Ceylan Essays On Prices

Essays on
markets, prices, and consumption

in the Ottoman Empire
(late-seventeenth to mid-nineteenth centuries)
By
Ceylan, Pinar
A thesis submitted to
the Department of Economic History
of the London School of Economics
for the degree of Doctor of Philosophy
London, December 2016
_________________________ _____________________________________________ _______ ___1
DECLARATION
I certify that the thesis I have presented for examination for the PhD degree of the London
School of Economies and Political Science is solely my own work other than where I have
clearly indicated that it is the work of others (in which case the extent of any work carried out
jointly by me and any other person is clearly identified in it).
The copyright of this thesis rests with the author. Quotation from it is permitted, provided that
full acknowledgement is made. This thesis may not be reproduced without the prior written
consent of the author. I warrant that this authorisation does not, to the best of my belief,
infringe the rights of any third party. I declare that my thesis consists of 84,500 words.
_________________________ _____________________________________________ _______ ___2
ABSTRACT
This thesis consists of separate papers that examine markets, prices, and consumption in the
Ottoman Empire between the late seventeenth and mid-nineteenth centuries. Recent
scholarship has posited that market development, new consumption patterns, and
productivity gains in non-agricultural sectors that were marked by changing price-product
structures are among the structural alterations that paved the way for industrialisation at the
turn of the nineteenth century. This research investigates whether these phenomena were
particular to the West or can be expanded to other parts of the world. As such, the study
contributes to the literature seeking to understand where the “distinctive advantage” of
Northwestern Europe lay.
The findings reveal that on the eve of the first wave of globalisation, domestic wheat markets
in the Ottoman Empire were no better integrated than they were two centuries previously.
Nevertheless, Europe and the Ottoman Empire shared several characteristics of early-modern
consumerism. This research demonstrated that the interiors of Ottoman houses grew richer
and more varied throughout this period. From the second half of the eighteenth century
onwards, Ottomans who were not richer and who were not better-positioned in the social
hierarchy than their counterparts in 1700 owned a greater quantity and variety of domestic
goods. In both regions, a decline in the real prices of consumer goods was a major factor, if
not the only one, that triggered this change. Moreover, the analysis on prices and inventory
valuations refutes the argument that the decline in prices of non-food items was a distinctive
pattern in Northwestern Europe in the pre-industrial era; instead, this was mirrored in the
Ottoman Empire.
Overall, the findings of this research point to long-term market development (and its absence),
rather than changing consumption patterns, as well as productivity gains in non-agricultural
sectors as a major source of divergence prior to the Industrial Revolution between parts of
Europe and the Ottoman Empire.
_________________________ _____________________________________________ _______ ___3
ACKNOWLEDGEMENTS
First and foremost, I am grateful to my supervisors Patrick Wallis and Sevket Pamuk, who never
failed to be patient and provide support and encouragement over the course of this project.
This work could not have been completed without their invaluable guidance. I would like to
express my special appreciation to Patrick, who has been a tremendous mentor for me and
who kept me on track during good and bad times alike. I could not have imagined having a
better supervisor.
Oliver Volckart and David Chilosi substantially contributed to this thesis with their many helpful
comments and advice. I also cannot forget Stefan Martens’ sincere efforts and accurate
interventions in making this thesis “legible.” The staff at the Center for Islamic Studies (ISAM)
Library helped me without hesitation during the long days and nights I worked in their
archives. I acknowledge them all with gratitude.
I am also grateful to the Turkish Cultural Foundation, the Istanbul Research Institute, the
TurkPetrol Foundation, the Economic History Society, the Middle East Centre at LSE, and the
Economic History Department for the financial support that they provided.
I thank Erik Thoen and all the team members of the GINI project at Ghent University for the
wonderful work environment they offered me, as well as their patience during the “very, very
final stage of my PhD” that took about two years.
I also extend my admiration to Ahmet Kuyas and Hulya Canbakal, who taught me the value of
intellectual honesty, academic curiosity, and simple thinking, which might be rare assets in
certain academic environments.
Many thanks to my comrades, whose own struggles in the world kept mine going: Olga,
Yasemin, Zeynep, Selcuk, Gizdo, Selen, Kuru, Ozgur, Ulas, Aytul, Alp, and Ali. They witnessed
my academic journey and made it joyful – even if they could never make sense of my
enthusiasm for eighteenth-century wheat prices.
And special thanks to all beautiful and courageous people of the Human Rights Foundation
of Turkey, who have devoted their lives to making Turkey a country where human dignity is
respected. They shared the darkest moments of summer 2016 with me and gave me hope for
better days.
And finally, my grandmother, Fatma Irfan and my mother, Ayse Tulin, the two extraordinary
and brilliant women who inspired and supported me in all my pursuits… I dedicate this
dissertation to them.
_________________________ _____________________________________________ _______ ___4
TABLE OF CONTENTS
List of tables…………………………………………………………………………………………………………….…………….....5
List of figures…………………………………………………………………………………………………………………………….8
List of maps…………………………………………………………………………………………………………………………….10
CHAPTER 1- Introduction………………………………………………………………….……………………………...…….11
CHAPTER 2- Ottoman inheritance inventory as a source for price history…………………………..…..60
CHAPTER 3- Measuring market integration in the Eastern Mediterranean……………………………..86
CHAPTER 4- Was there a consumer revolution in the Ottoman Empire?........................................164
CHAPTER 5- Relative prices and the rise of household goods………………………………………..……..220
CHAPTER 6- Gender, religious status, and ownership of household goods…………………………..246
CHAPTER 7- Conclusion…………………………………………………………………………………………………...…..271
Primary sources………………………………………………………………………………………………………………….…282
Bibliography………………………………………………………………………………………………………………………...285
Appendix……………………………………………………………………………………………...........................................311
_________________________ _____________________________________________ _______ ___5
LIST OF TABLES
2.1. Goods reported in units in Ottoman inheritance inventories

2.2 Household goods in inventories
2.3. Capital goods reported in inventories
2.4. Mean unit valuations of silk and cotton sheets in Üsküdar
2.5. OLS regression results-Inventory valuations and wealth in Üsküdar
2.6. Mean valuations of copper kitchenware by quartiles of total wealth in Manisa, Trabzon and
Istanbul
2.7. Mean valuations of wheat by quartiles of total wealth in Manisa and Ayntab
3.1. OLS regression results- Transportation costs and distance
3.2. Journey durations by horse
3.3. Price data: Sources (Ottoman Empire)
3.4. Price data: Sources (Adriatic)
3.5. Metric equivalents of local measurement units (Ottoman Empire)
3.6. Moneys and local measurement units (Adriatic)
3.7. Descriptive statistics of the price series
3.8. Fixed Effects Results (Time trend), Ottoman Empire (Complete sample), 1675-1800
3.9. Fixed Effects Results (Aggregated price relatives), Ottoman Empire (Complete sample),
1675-1800
3.10. Fixed Effects Results (Time trend), Entire Eastern Mediterranean, Ottoman Mediterranean,
and Adriatic regions, and Adriatic-Ottoman Med. market pairs, 1660-1840
3.11. Fixed Effects Results (Aggregated price relatives), Entire Eastern Mediterranean, Ottoman
Mediterranean, and Adriatic regions, and Adriatic-Ottoman Med. market pairs, 1660-1840
3.12. Descriptive statistics of the regression variables
3.13. OLS regression results-Bilateral relative prices and distance between markets
3.14. Fixed Effects Results for 30-year intervals (Time trend)
3.15. Fixed Effects Results for 30-year intervals (Time trend), Summary
3.17. Fixed Effects Regression Results, Grain-trading network
3.18. ADF and OLS results
4.1. Wealth brackets
_________________________ _____________________________________________ _______ ___6
4.2. Wealth, Descriptive statistics

4.3. Ownership of real estate according to periods
4.4. Distribution of the sample according to gender and religious status
4.5. Estate owner with titles
4.6. Index of amenities
4.7. Descriptive statistics of regression variables
4.8. Goods concerned with bedding
4.9. Floor coverings
4.10. Sitting implements
4.11. Goods concerned with storage
4.12. Towels
4.13. Household linen
4.14. Lightening implements
4.15. Heating
4.16. Mirrors
4.17. Clocks
4.18. New furniture
4.19. Crockery, plates, and cutlery
4.20. Average amount of copperware possessed
4.21. Correlations between the (pooled) stock of household durables and (pooled) wealth
4.22. OLS Regression results
4.23. Exponential coefficients of time dummies
4.25. Estimated index scores for 949 constant akçe (Muslim male)
5.3. Comparison of prices, early seventeenth to mid-nineteenth centuries
5.4. Estimated time trend coefficients
5.5. Unit valuations of selected household goods from Üsküdar inventories
5.6. Household durables contained in basket
5.7. Value of household durable basket from 1700 to 1850
6.1. Mean total estate value by gender and religious status groups
_________________________ _____________________________________________ _______ ___7

6.3. OLS regression results
6.5. Ownership of real estate
6.6. Descriptive statistics of the regression variables according to gender and religious status
6.7. Coefficients of determination from OLS regressions
6.8. Descriptive statistics of the regression variables (Sub-samples)
6.11. Logistic regression results predicting ownership of selected goods from total wealth,
gender, religious status and period
_________________________ _____________________________________________ _______ ___8
LIST OF FIGURES
2.1. Edirne wheat prices from inheritance inventories and waqf account books
2.2. Edirne wheat prices from the account books of the Selim II Waqf and the Mevlana
Celaleddin-i Rumi Waqf
2.3. Istanbul wheat prices from waqf account books and inventories
2.4. Trabzon and Amsterdam copper prices
2.2. Beledi mattress and beledi duvet index
3.1. Number of war years per decade, 1660-1840 (European and Eastern fronts)
3.2. Number of years of naval hostility per 25 years in the Eastern Mediterranean
3.3. Number of drought years per decade
3.4. Number of epidemic years per decade
3.5. Wheat prices, Istanbul and Western Anatolia
3.6. Wheat prices, Balkans
3.7. Wheat prices, Damascus, Ayntab, Konya, Cairo
3.8. Wheat prices, Adriatic
3.9. Aggregated relative prices, Ottoman Empire (Complete sample)
3.10. Aggregated relative prices, Eastern Mediterranean (Entire)
3.11. Aggregated relative prices, Adriatic and Ottoman Mediterranean
3.12. Aggregated relative prices, Adriatic – Ottoman Med. market pairs
3.13. Coefficient of variation, Ottoman Empire
3.14. Coefficient of variation, Entire Eastern Mediterranean region
3.15. Coefficient of variation, Ottoman Mediterranean and Adriatic regions
3.16. Aggregated relative prices, Ottoman Empire with and without Istanbul
3.17. Aggregated relative prices, Ottoman Mediterranean with and without Istanbul
3.18. Aggregated relative prices, Ottoman Mediterranean and Manisa, Istanbul, Edirne
4.1. Wealth and wages
4.2. Distribution of the sample according to gender and religious status
4.3. Mean index and mean wealth
4.4. Mean amount of copperware owned and mean wealth
4.5. Average amount of copperware possessed, Lower and upper wealth groups
_________________________ _____________________________________________ _______ ___9
4.6. Index A, Total estate value, and total estate value less the value of household durables
stock
4.7. Index B, Total estate value, and total estate value less the value of household durables
stock
4.8. Copperware, Total estate value, and total estate value less the value of household durables
stock
4.9. Estimated index scores for an estate of 949 constant akçe (Muslim male)
4.10. Estimated amount of copperware owned for an estate of 949 constant akçe (Muslim
male)
5.1. Manufactured and traded goods’ prices
5.2. Unit valuations of sheets, mattresses, duvets, cushions and pillows
5.3. Price of beledi mattresses and yemeni duvets
5.4. Unit valuations of beledi mattresses from Üsküdar and Manisa
5.5. Unit valuations of felts and chests
5.6. Valuation of copperware
5.7. Value of household durable basket
5.8. Value of household durable basket, and the estimated index A and B for an estate of 949
constant akçe (Muslim male)
5.9. Price of copperware, and the estimated amount of copperware owned for an estate of 949
constant akçe (Muslim male)
6.1. Value of household durables against wealth (Men and Women)
_________________________ _____________________________________________ _______ ___10
LIST OF MAPS
1.1. The Ottoman Empire c. 1683-1800
1.2. Ottoman towns of Üsküdar, Manisa, Ayntab, and Trabzon
3.1. Ottoman maritime network
3.2. The Eastern Mediterranean

_________________________ _____________________________________________ _______ ___11
CHAPTER 1
INTRODUCTION
This thesis consists of separate papers that examine markets, prices, and consumption
in the Ottoman Empire between the late seventeenth and mid-nineteenth centuries. Recent
scholarship has posited that market development, rising consumer demand, and productivity
gains in non-agricultural sectors that were marked by changing price-product structures are
among the structural alterations that paved the way for the industrialisation of the late
eighteenth and early nineteenth centuries, or at the very least, made quicker growth in Europe
possible during the era of the Industrial Revolution. These developments, which stand out as
three important phenomena associated with the growth path of pre-industrial Europe, occupy
a central place in the current debates in economic history about the nature of pre-industrial
economies and the causes of global economic divergence and growth. The present study is
an attempt to contribute to these debates by providing insights from a non-Western context.
Since Pomeranz’s (2000) widely cited but equally controversial book, The Great
Divergence: China, Europe and the Making of the Modern World Economy , was first
published,1 when and why Europe diverged from the rest of the world in terms of productivity
and living standards has been discussed in the light of ever-burgeoning evidence and new
insights into regional variations both in Europe and in Asia.
As the California School’s argument, namely, that the most advanced parts of Europe
and Asia were on the same development level with “shared constraints” before 1800
(Pomeranz 2000: 107), has been attacked on several grounds, early-modern history has
attracted unprecedented attention. Seeking the roots of modern economic growth in the pre-
modern era, recent research has replaced the picture of the early-modern economy as a
stagnant economy incapable of generating long-term growth with a more optimistic tone.
Today, the preponderant view is that the North Sea Area, if not the whole continent, underwent
a series of structural and institutional changes that led to the rise of living standards, making
1 Wong’s (1997) China Transformed; Frank’s (1998) ReORIENT; Parthasarathi’s (1998) “Rethinking wages
and competitiveness in the eighteenth century: Britain and South India”; Lee and Feng’s (1999) One
Quarter of Humanity; and Goldstone’s (2000) “The Rise of the West- or not? A Revision to Socio-
economic History” are the other essential works representing the California School’s position.
Chapter 1-Introduction 12
this region “more dynamic, competitive, and creative than the rest of the world” (van Zanden
2009) even before industrialisation took place, while also preparing the ground for the
Industrial Revolution.
In an effort to understand where pre-modern Northwestern Europe’s ‘distinctive

advantage’ lay, economic historians have directed their attention towards institutions, markets,
technical innovation, and demand-side changes as possible candidates. While institutional
improvements and innovations (security of property, representative institutions, efficient
systems of taxation and government debt management, effective national banks, etc.) have
been proposed as the ultimate or deeper explanation, the rise of efficient and integrated
markets, productivity gains in non-agricultural sectors, and upward demand shifts are cited as
closer and more direct sources of pre-modern growth.
Despite several controversies in the growing body of empirical research on these

topics, today it is widely accepted that during the early-modern era, most advanced parts of
Europe gradually – if not without setbacks – became more integrated within themselves and
with other parts of the continent; that there was a marked increase in the consumption of
durable goods and luxuries by individuals from different economic and social backgrounds;
and that manufactured and traded goods became increasingly inexpensive relative to
agricultural products in these regions. Drawing upon trade-led, demand-led, and innovation-
led growth theories, economic historians have assigned varying roles to these developments
in explaining pre-modern economic growth.
With the aim of understanding whether these phenomena were particular to the West
or can be expanded to other parts of the world, this thesis first looks at the evolution of
Ottoman commodity markets from the mid-seventeenth to mid-nineteenth centuries; second,
it examines the existence of an ‘Ottoman consumer revolution’; and third, it explores whether
consumer goods in the Ottoman realm dropped in price, relative to agricultural products
during the same period. In so doing, the study employs new evidence from inheritance
inventories.
The Ottoman Empire has been cited alongside Qing China and Mughal India as a site
where divergence can be identified and analysed. It was “one of the greatest, most extensive,
and longest-lasting empires in the history of the world,” (Quataert 2005: 3). Together with the
Venetian Republic, it represents a region that stood at the centre of the pre-Columbian world
economy but which gradually lost ground throughout the seventeenth and eighteenth
centuries, in parallel to the rise of the North Sea area. Although as a ‘waning’ power (Tabak
2010), the Ottoman Empire provides fertile ground to explore the sources of early divergence,
it has assumed little significance within the Great Divergence literature compared to early-
modern China and India, a lacuna that can be partly explained by the lack of long-term and
systematic quantitative research. The present study addresses this geographical bias in the
existing literature.
The period under study here begins in the second half of the seventeenth century,
when economic and social life in the Ottoman realm showed signs of recovery from the
seventeenth-century crisis and ends in the mid-nineteenth century, shortly before the first
wave of globalisation. The 1840s have been considered a threshold in the economic and
political modernisation of the empire. From this date onwards, new dynamics, institutions, and
actors shaped the Ottoman economy.
This periodization also corresponds to De Vries’ reinterpretation of the “long-

eighteenth century” that stretches across the period 1650-1850. It should also be noted that
the short eighteenth century that stands in the middle of the time span of the study has been
identified by recent research on markets, prices, and consumption in Europe as an era of
substantial developments. There is a general tendency to locate the consumer revolution in
the eighteenth century, while several studies have detected integration in European markets
in the first half of the century. In contrast, the second half of the eighteenth century appears
as a period of decline in the prices of manufactured and traded goods.
In what follows, I present an outline of the Ottoman Empire and provide an overview
of the political and economic background of the period, together with a discussion of the
empire’s place in the divergence debate. Subsequently, the chapter surveys the literature on
early-modern market integration, the consumer revolution, and changes in relative prices of
manufactured and agricultural goods, briefly discussing how each of these phenomena were
linked to pre-industrial growth in Europe. Next, I introduce the primary sources that are
employed in the analysis. The chapter concludes with an outline of the following chapters,
while also highlighting the empirical and methodological contribution of this new research.
1. The Ottoman Empire: Political and Economic Environment
The Ottoman state emerged in western Asia Minor around 1300, expanding in the
following centuries from a small chiefdom to an empire with vast territories. Stretching from
the Balkans and the Black Sea region through Anatolia to Syria and Mesopotamia, and from
the Gulf to Egypt and most of the North African coast, the Ottoman Empire stood at the
crossroads of intercontinental trade until its dissolution after World War I (See Map 1). In the
seventeenth and eighteenth centuries, its population exceeded 30 million.
Map 1- The Ottoman Empire c.1683-1800
Source: Quataert (2000: 39)
1.1. The Ottoman Empire before the seventeenth century
Throughout its entire history, the Ottoman Empire remained an agrarian economy.
Before the Land Code of 1858, the Ottoman legal and political system did not develop a
category of alienable property rights over agricultural land. Arable lands were registered as
state lands (miri) and as a rule, they could not be bought, sold, or inherited. 2 On this legal
basis, the organisation of agricultural production traditionally relied on the çift-hane system,
where peasant households, acquiring hereditary usufruct rights, cultivated the land with a pair
of oxen and family labour and paid taxes to the local agents from the centre (Keyder 1991).
These local agents were cavalrymen and other members of the ruling class who were
granted the right to collect the tax revenue from arable lands in certain localities in return for
performing military services during wartime. They were responsible for supervising their
territory and its method of cultivation by peasants. Fiefs could not be inherited, although it
was not uncommon for these to be reassigned to a son provided he performed military service.
This tax collection and revenue-sharing system complementing the Ottoman land regime was
called the tımar system, and lasted from roughly the fourteenth century through to the late
sixteenth century (İnalcık 1994). The Ottoman central administration did not attempt to impose
the tımar regime in all of its conquered territories, however. In many of the more distant areas
such as Eastern Anatolia, Iraq, Egypt, Yemen, Wallachia, Moldavia, and the Maghreb, the
Ottomans were eager to collect taxes but only altered the existing land regimes to either a
limited extent or not at all3 (İnalcık 1954).
Ottoman society consisted of the tax-exempt ruling class (askeri) and the large subject
class of reaya. The ruling class was comprised of people who were in the sultan’s service, such
as bureaucrats, army troops, and religious functionaries. Merchants, artisans, and peasants
constituted the ruled, and were organised into religiously based communities called
cemaʿât, taife or millet, as well as into guilds ( esnaf), mystic orders of dervishes ( tariqât) and
other groupings that formed a substratum of Ottoman society. The Ottoman social structure
did allow upward mobility, that is, the penetration of the lower classes into the ruling class
2 There were, however, important exceptions to this rule. Among others, small plots of arable land
(orchards, vineyards, and vegetable gardens) in proximity to urban areas were registered and recognized
by the state as freely held property and could be bought, sold, and inherited. Land under this status is
frequently encountered in the property transaction documents in court registers and in probate
inventories (Keyder 1991). Although it has not attracted the attention of Ottoman historians until now,
sources point to the existence of a lively, if not large, land market, particularly near large urban
agglomerations.
3 For an analysis of regional variation in Ottoman tax practices, see Coşgel (2015). Coşgel stresses the
significance of political economy constrains, alongside economic factors, in explaining the tax bases and
rate structures that varied across different parts of the empire.
(Karababa 2006). Membership in the ruling class was open to all who declared and manifested
loyalty to the sultan, his dynasty, and his empire, and who accepted the religion of Islam. The
Ottoman state was unique compared to its European counterparts in that it lacked an
aristocracy but was run by men chosen by merit and loyalty to the sultan (Karababa 2006).
Over the centuries, the empire developed a complex system of administration with
the sultan as the supreme ruler. The central administration consisted of three main aspects:
the sultan’s household; the departments of government grouped under the control of the
grand vizier who was the sultan’s deputy in all state matters; and the Muslim religious
institution, which consisted of Muslim functionaries concerned with education and law. The
administration employed a language (Ottoman Turkish) which was Turkish in grammar but
largely Arabic and Persian in vocabulary, and written in an Arabic-based script (Karpat 1972).
Administratively, the empire was first subdivided into provinces ( eyalet), ruled by
governors appointed by the centre. Sancaks (sub-provinces) were governed by sancak beys,
selected from the high military ranks by the central government. The kaza was a subdivision
of the sancak, and the fundamental division in the Ottoman administrative and judicial system,
ruled by a kadı, who looked after local administration and criminal and civil law (Karpat 1972).
In the late-sixteenth century, when the empire reached its greatest extent, it
comprised central Hungary, the Balkan Peninsula, Anatolia, Mesopotamia, Syria and Palestine,
western Arabia, Egypt, and lands in the Caucasus and western Iran. In Europe, Transylvania,
Moldavia, and the Crimea were tributary principalities, while in North Africa, Tripoli, Tunis, and
Algiers were semiautonomous provinces. The Balkans, together with western and central
Anatolia, including the capital city and its environs, constituted the core regions of the
Ottoman Empire. These were administered by the capital with institutions closely resembling
those in the Istanbul region, and were integrated into the larger Ottoman system of taxation,
provisioning, and trade. In contrast to these core lands, in the more distant regions, institutions
and administrative structure were shaped by local practices and power relations with the
centre (İnalcık 1954).
1.2. Political environment and decentralisation
In his monumental work, Freedom and Growth, Epstein (2000) stresses the political
conditions of pre-industrial economic growth, and posits jurisdictional centralisation under the
late Medieval and early-modern rules as the ultimate source of the rise of efficient and
integrated markets in Europe. In new institutionalist economics, political regimes are defined
as ways to facilitate cooperation for mutual advantage. In line with this literature, Epstein
suggests that pre-modern centralised states were more efficient than decentralised ones
because they suffered from fewer multiple coordination failures.
Although decentralisation, centralisation, and the changing relationship between the

imperial centre and the periphery in 1700-1850 have been subject to ample research in
Ottoman historiography (Inalcik 1977; Piterberg 1990; Salzmann 1999, 2004; Hathaway 2002;
Khoury 1990, 2002; Smiley 2008), these processes have rarely been addressed in relation to
the general functioning of the economy and its underlying incentive structure. Between the
seventeenth and mid-nineteenth centuries, the empire witnessed major socio-political
transformations which reshaped the institutional environment within which economic actors
operated. Without examining what occurred in the political sphere, we cannot acquire a full
understanding of the changing economic environment.
Building upon Byzantine and Seljuk military agrarian organisation, the Ottomans
established a strong centralised regime in the fourteenth and fifteenth centuries with state
control of agricultural lands, a central military force, and a fiscal organisation oriented towards
the central treasury (Pamuk 2009). As Salzmann (1993: 396) states, “the Ottomans… achieved
one of the first examples of an early-modern state structure with centralised judicial, military,
and administrative powers.” During the classical period, the rise of the Ottoman Empire was
closely associated with territorial expansion. Military success, in turn, depended on the land
tenure and fiscal regimes that supported a large cavalry-based army (Pamuk 2009). Up until
the second half of the sixteenth century, the fief-holding sipahis as the traditional cavalry
armed with conventional weapons of bow and arrow, lance, and sword, constituted the
backbone of the Ottoman army. In this system, 30 to 40 percent of military expenses were
covered by revenues collected in rural areas by the sipahis (Kunt 2000).
However, from the late sixteenth century onwards, warfare increasingly began to
constitute a drain upon Ottoman state finances. At the end of the sixteenth century, the
Ottoman Empire was engaged in a series of long and exhaustive wars on the Iranian and
Habsburg frontiers. Unlike the successful campaigns of the earlier period, which had led to the
inflow of annual remittances from the newly incorporated provinces, the limited conquests
during this period were no longer sufficient to render warfare profitable (Pamuk 1984).
At the same time, military costs began to soar as technology altered warfare. The ever-
more widespread use of firearms by the Habsburg armies made obsolete the traditional
cavalry, which had proven ineffective against Austrian musketeers (Kunt 2000). In order to keep
up with the transformation in warfare, the Ottoman government was forced to increase the
number of janissaries, the standing infantry corps. Their numbers rose from 13,000 in the 1550s
to 38,000 in the 1600s. Another source of soldiers using muskets was the numerous vagrants
in the countryside (Pamuk 2001). Armed young men from reaya or from nomads were
recruited as mercenaries by the central government (Kunt 2000).
The additional costs resulting from the expansion of the central army and the
recruitment of mercenaries fell upon the central treasury, since their wages were paid in cash.
The costs of warfare were also enhanced by the price revolution (Pamuk 2001). With the
outbreak of war with the Persians in 1578, the treasury began to experience shortages of silver
for payments to soldiers. The debasement of 1584-86 was a symptom of financial crisis (Pamuk
2001).
The military transformation brought about changes in Ottoman fiscal practices, and
the dissolution of the classical institution of the tımar. The central government tried to increase
its income by implementing a new revenue collection system, the iltizam system, where an
entrepreneur was delegated the right to collect taxes from a tax source by the state for a
limited period of time in return for an annual lump-sum. The tax contractor was endowed with
several privileges and immunities. This new tax-farming system was not limited to the tithes
and taxes on villages and fields, but also expanded to cover a variety of state revenues from
market taxes to custom revenues (Salzmann 1993).
It is often stated that during the early-modern period, the Ottoman Empire followed
a trajectory of state-formation that is different than that of Western Europe (Barkey 1994). The
institutional centralism of the Ottoman state in the classical period was gradually succeeded
by a fragmented political structure over the course of the seventeenth and eighteenth
centuries. These two centuries witnessed the rise of provincial elites who held local power and
who operated autonomously from the capital. In almost all parts of the empire, the central
state became visibly less important (Khoury 2006). The provincial power holders, ayan, came
from two different groups: prominent notables whose families had been among the local elites
of an area before the Ottoman period, and centrally appointed officials who subsequently put
down local roots (Quataert 2005).
In both cases, their economic and political power was rooted in the life-term tax-
farming system. The Habsburg-Ottoman war of 1683-1699 exerted considerable pressure on
Ottoman finances in a period of extreme financial instability. Attempts to create funds to
finance the war and to balance budgets could not ease the government's need for cash. Under
these conditions, the iltizam system was expanded in 1695 into the Malikane system which
depended on contracts on state revenues held for life, with the contractor determined by
public auctions (Genç 1987).
During the first half of the eighteenth century, the redistribution of state wealth and
titles did not subvert state authority, as provincial elites did not rebel against the centre. Until
the late 1760s, the relation between provincial elites and the centre was one of mutual
recognition and interest, and it depended on negotiation to establish a balance of power. Local
dynasties recognised the sultan and central authority in general, sent troops for imperial wars
and, more importantly, responded to the increasing cash demand of the state, in exchange for
legitimacy from the central state, the right to collect state revenues, and several other
privileges and immunities (Quataert 2005).
The coalition between the centre and periphery lasted for as long as it was lucrative
for both sides. The 1768-1774 Russo-Ottoman war marked the breakdown of the tacit pact
between local elites and the central government. In the post-1770 period, as economic
conditions worsened, the rates of profit in the tax-farming sector declined, leading provincial
elites to shift their portfolio towards more lucrative areas of the economy, such as money
lending and urban real estate. Besides, the wartime chaos gave the notables greater latitude
of action, which weakened the provincial ties to the centre. The pattern of negotiation, mutual
recognition, and control lost its place to an unavoidable conflict between the centre and
periphery. A struggle for power between the imperial centre and the provincial elites marked
the last decades of the eighteenth century (Salzmann 1993).
Although the practice of the malikane was maintained until the 1840s, institutional
centralisation began at the turn of the century. The central government took an important step
in 1826 by destroying the Janissary corps, which had resisted centralisation efforts. These
efforts culminated in the imperial reform edicts of 1839 and 1856. Ending political
fragmentation and rivalry, the edicts ushered a centralised bureaucracy into power.
Administrative centralisation and a vast expansion in the numbers and responsibilities of
bureaucrats accompanied military changes that resulted in a state apparatus vastly larger and
more powerful than that of the previous era (Quataert 2005).
For a long time, Ottomanists interpreted decentralisation as a sign of the dissolution

of the classical order of the golden age of the empire and associated it with the “decline” of
state power. This view has been challenged in recent decades by a growing body of revisionist
literature, which suggests that the decentralisation in the seventeenth and eighteenth
centuries should be interpreted as a way of adapting to changing circumstances in Eurasia in
the early-modern era (Barkey 1994; Salzmann 1993; Khoury 2006).
In this line of research, Salzmann re-conceptualised the tax-farming system, the

backbone of Ottoman decentralisation, as the “privatisation of the fiscal agency,” (1993: 395)
and claimed that this was a transitory phase between the “precocious imperial centralisation”
(1993: 396) of the fifteenth and sixteenth centuries, and the institutional centralisation of the
early nineteenth century. According to Salzmann (1993: 399), during the seventeenth and
eighteenth centuries, Ottoman rulers chose to implement a strategy of “selective inclusion of
certain strata of provincial society within the political hierarchy” to amplify the state’s political
capacity, rather than attempting to directly eliminate these alternative sources of power, as
their European counterparts did. The pragmatic attitude of the rulers and the institutional
flexibility substantially contributed to the longevity of the empire, preventing the resistance of
peripheral sources, and enabling their incorporation into the political system.
While thoroughly examining the Ottoman institutional regime of the seventeenth and
eighteenth centuries from the perspective of political efficiency, the revisionist view did not
assess this “alternative path” in terms of its fiscal outcomes. Recent findings by Karaman and
Pamuk (2010, 2013) suggest that in comparison to their counterparts in the centralised states
of Western Europe, Ottoman fiscal institutions were relatively less efficient. Examining long-
term trends in the revenues of the Ottoman central administration based on the evidence
provided by Ottoman budgets, Karaman and Pamuk’s (2010, 2013) studies reveal that between
the sixteenth and nineteenth centuries, the Ottoman central administration did not experience
a growth in its extractive capacity, unlike its European counterparts. It remained immune from
the strong rising pattern in per capita tax revenue and in total revenues across Europe during
this period. As a result, by the eighteenth century, a large gap had emerged between the
revenues of most European states and those of the Ottomans – a development that was
accompanied by a discrepancy in military performances.
Karaman and Pamuk (2010, 2013) suggest that the decentralised nature of the tax-
collection system that involved numerous intermediaries was the underlying reason for the
Ottoman state’s failure to expand its resources. “The intermediaries influenced the size and
incidence of the tax burden and captured a significant portion of tax revenue at the cost of
the central treasury through temporary assignments, long-term alienation, or embezzlement,”
(2010: 595). These studies confirm the association between the enhanced fiscal capacities of
the early-modern European states and the political centralisation and state consolidation
processes they experienced, effectively providing support to Epstein’s argument regarding the
benefits of jurisdictional centralisation.
1.3. Economic policy
According to Genç (2000), until the middle of the nineteenth century, the Ottoman
economic mentality was shaped by three principles: provisionalism, fiscalism, and
traditionalism. Provisionalism entailed making goods and services accessible, ample, and
affordable for the empire's subjects via state intervention in production and trade. Fiscalism
was predicated on maintaining or increasing the revenues of the treasury and decreasing
expenditures with the aim of bolstering the state's financial power. In line with this principle,
all economic activity was considered above all as a basis of potential tax income. Traditionalism
refers to the concern of preserving a social and economic balance that was believed to be
immutable. Ottoman authorities were uneasy with economic changes and innovations and
consistently tried to achieve a return to the economic status quo ante through reform.
Genç (2000) states that out of these three principles, provisionalism was the most
important principle of Ottoman economic policy. This policy undergirded state monopolies,
internal tariffs, export/import prohibitions, and price control mechanisms. Like many local and
central governors in pre-industrial Europe, Ottoman rulers perceived the availability of
essential goods at affordable prices in the urban markets as a prerequisite for social order and
political stability (Yıldırım 2003). Supplying the capital city, the armed forces, and to a lesser
extent, other urban areas with abundant foodstuffs, was always a major concern for them.
From an early period, Ottoman internal and international trade policy was shaped by this
priority (Genç 2000). In general, imports were fostered and facilitated, while exports were
curtailed by prohibitions, quotas, and taxes. When the possibility of shortages emerged, the
government prohibited the exportation of basic necessities, especially foodstuffs and raw
materials (Pamuk 2009).
The state established a complex provisioning network that encompassed the Ottoman
territories, with the capital at its centre. A large portion of the surplus production in grain,
sheep, cattle, horses, butter, wax, and timber was shipped from the Black Sea coast of Bulgaria,
the Romanian principalities, Thrace, Macedonia, Thessaly, Morea, Egypt and several districts of
Anatolia to Istanbul. The organisation of this network included many institutions and policies
(Yıldırım 2003). Although the major priority was to supply the capital, the provisionist policy
also concerned other parts of the empire, particularly during times of supply interruption. In
these situations, measures taken by the state included waiving tax obligations and obliging
prosperous regions to make temporary loans of specified amounts of grain to areas hit by
hardship (Murphey 1987).
Alongside provisioning the cities, the requirements of warfare and providing raw
materials for public construction works and manufactured goods for members of the palace
and military class were the principal reasons for state interference (Quataert 2005). As in the
past, the Ottoman state continued to intervene in markets to varying degrees in the eighteenth
century, claiming the right to control economic resources when deemed necessary. According
to Genç (1984), warfare, which put enormous pressure on the state’s finances, was the most
important determinant of the degree of intervention. This explains why during the period of
relative peace that lasted until the 1760s, intervention remained limited, whereas in the
following period marked by wars, state involvement in the economy substantially increased.
This intervention took the form of price ceilings, trade regulations, internal tariffs, forced
purchases, and state command over resources.
Quataert (2005: 42) underlines the doubly disruptive and negative effects of these
interventions: “The state often paid below-market prices for the goods and, often, drained
away all or most of a commodity, thus creating scarcities. Crops of entire areas or the
manufacturing output of certain guilds were commandeered for particular purposes, for
example, to supply the royal household or marching armies.” These pejorative effects on
producers and consumers, as well as their political consequences, were not unknown to state
officials, who were actually very eager to avoid or at least limit these effects. Therefore,
particularly after the seventeenth century, state intervention in domestic markets was done
selectively (Pamuk 2009). For instance, for price ceilings ( narh), Pamuk (2004: 19) states: “The
narh came to be considered not as permanent policy but as an instrument reserved for
extraordinary conditions such as wars, exceptional difficulties in the provisioning of the capital
city, or periods of monetary instability.” The centre always sought a balance between its
military and political priorities, the welfare of the producers and consumers, and the potential
effects of its actions on the overall economy. It also took measures to avoid negative effects
or redressed the losses of producers and consumers whenever possible. Its limited
technological capacities also restricted the state’s ability to exercise strict control over the
markets. Thus, it is often suggested that one should not exaggerate the extent of the state
interference in pre-modern period (Quataert 2005).
1.4. General trends in the Ottoman economy
In the 1580s, the Ottoman Empire entered an era of political difficulties, social unrest,
economic turmoil, and financial crises. The first half of the seventeenth century was
characterised by military revolts and conflicts among political factions in the capital; bandit
campaigns and rebellious governors in Anatolia and in the Arab lands; and long and
exhausting wars on the eastern and western fronts – all of which had destructive effects on the
Ottoman economy and its finances. The crisis manifested itself in the form of a dramatic fall in
agricultural and craft production, while also resulting in a drop in tax revenues, vast population
movements, and urban decline (Faroqhi 1994). Attacks, invasions and plundering by bandits,
mercenaries, landless peasants, and religious school students resulted in the large-scale
depopulation of Anatolian villages and towns in the first decade of the seventeenth century.
An important aspect of the seventeenth-century crisis was the decline and even
interruption of domestic trade. Besides the fall in production, the insecurity of routes was an
important impediment to interregional exchanges. Even the main routes in Anatolia were
partly controlled by bandits, and traveling merchants had no surety for their lives or their
goods. Celali bands established themselves in fortresses, assembled small armies, laid siege to
commercial centres and towns of considerable importance and even briefly occupied Bursa,
an imperial town and industrial centre. Under these conditions, mid-sized and large towns
sought to protect themselves by constructing city walls and by developing self-sufficiency.
Infrastructure and investments previously made by pious foundations in order to facilitate
trade, such as shops, storage spaces, covered markets, caravanserais, and roads, were
damaged and often totally destroyed (Faroqhi 1994).
As relative peace and order was restored in Anatolia by the mid-seventeenth century,
agricultural and craft production was resuscitated, precipitating a decline in food prices and a
revival of trade. Official attempts to prevent peasants from abandoning their land and to
resettle refugees in their former places of residence had born fruit by this time. The extent of
the recovery remains unknown, but the long wars on the European frontier and the fiscal
difficulties that resulted accordingly were likely to have restricted it (Faroqhi 1994).
Our knowledge of the Ottoman economy in the eighteenth century is more ample
than that of the late seventeenth century. Ottoman economic and political history in this
period is divided into two parts. The first period, which runs from the turn of the century until
the 1760s, was an era of peace and stability for the empire. Despite the on-and-off conflict on
the Iranian border between 1722 and 1747, the Ottoman state was not engaged in large-scale
fighting during the three decades between the Treaty of Belgrade (1739) and the Russo-
Ottoman war of 1768-1774. These decades were also characterised by a general trend in
expansion and development in almost all sectors of the economy. The growth in industrial
production and rise of new industrial centres in Anatolia and in the Balkans in the first half of
the eighteenth century is well-documented (Genç 1994). Like industrial output, agricultural
production and exports also significantly rose during this period. The state did not confront
major fiscal difficulties, and even experienced an increase in its revenues, albeit an
unspectacular one (see Figure 1). The purchasing power of the Ottoman akçe remained largely
stable (Pamuk 1994), while the emergence of new trade nodes in the Ottoman urban network,
the maintenance of security on the trade routes, the spread of commercial agriculture, and the
growth of agricultural output created conditions favourable to economic growth (Faroqhi
1979).
This period of expansion came to an end with decades of war, fiscal difficulty, inflation,
and rising insecurity in the countryside after the late 1760s (Faroqhi 1994). Declining tax
revenues due to a decline in agricultural and industrial output, together with the increasing
exigencies of warfare provoked massive state interference in markets, which took the form of
price ceilings, trade regulations, internal tariffs and state's command over resources. The akçe
lost half its value by the end of the century (Pamuk 1994).
Figure 1- Revenues of the Ottoman Central Administration (in tons of silver)
Source: Karaman and Pamuk (2010: 604)
The early nineteenth century was marked by treasury shortfalls, frequent

debasements, and monetary instability. This era of long and exhaustive wars with European
powers was also one of national awakening in the Balkans that began with the Serbian
Revolution (1804-1815). Napoleon's invasion of Egypt was succeeded by the Russo-Turkish
War of 1827-28, the Greek War of Independence, and Mehmed Ali's invasions of Syria and
Anatolia. Despite the wars and high inflation of the early 1800s, the Ottoman economy
experienced a second wave of expansion towards the mid-nineteenth century in parallel to its
increasing integration into the world economy. As with many regions of the world, following
the Napoleonic wars, the Ottoman Empire was incorporated into the world market at an
unprecedented pace due to its agricultural exports (Quataert 1994).
Three events around 1840 – the 1838 Anglo-Ottoman convention, the 1839 reform
edict, and the introduction of the first railway line in 1856 – represent milestones in the
economic and political history of the Ottoman Empire and marked the opening of a new era.
In the several decades that followed this date, the explosion of international trade; the rise of
commercial agriculture; economic liberalisation; the centralisation of the state apparatus; and
extensive improvements in transportation and communication facilities changed the face of
the economy. Although most of these transformations began in the late eighteenth century,
they accelerated during this period, leaving substantial legacies from the mid-nineteenth
century onwards.
The urban real wage series constructed by Pamuk (2000a) is the best evidence
available to observe the evolution of standards of living in the Ottoman realm. Taken together
with the changes in population, these series confirm the general trends in the Ottoman
economy outlined above. His estimations suggest that from their relatively high levels at the
end of the fifteenth century, construction workers’ wages in Istanbul experienced a steady and
large decline during the sixteenth century by as much as 40 percent (see Figure 2). By the end
of the sixteenth century, real wages stood close to their levels before the Black Death in the
mid-fourteenth century. From this date onwards until the turn of the eighteenth century, urban
real wages remained largely unchanged. Between 1700 and 1750, real wages continuously
increased, albeit remaining under the level from the early sixteenth century. The decline in
daily real wages in the third quarter of the eighteenth century was followed by a recovery and
then increase until the mid-nineteenth century. From 1775 to 1850, wages almost doubled,
equalling the levels from the early sixteenth century. Overall, the post-Black Death peak in
wages was not surpassed during the early-modern era. At the end of the eighteenth century,
wages in Istanbul were no higher than their levels at the end of the fifteenth century (Pamuk
2014). In short, the wage evidence suggests limited growth in the Ottoman realm prior to the
nineteenth century.
Figure 2- Real Daily Wages of Construction Workers: Istanbul, 1489-1914

(Wages in Grams of Silver/CPI)
Notes: 1489/90 Unskilled=1

Source: Özmucur and Pamuk (2002: 306)
1.5. The Ottoman Empire and the divergence debate
Although it offers a promising ground for comparative work, the Ottoman Empire has
found little place within the recent literature on the Great Divergence, unlike China and India.
There are only rare studies that deal with the Ottoman economic history in the pre-industrial
era in a comparative perspective, and that explore when and why the region diverged from
the Western Europe in terms of economic performance and standards of living. This limited
body of literature is mostly comprised of Pamuk and Kuran’s works, which focus on
institutional environment in the Ottoman Empire as the ultimate source of absence of intensive
growth in this region.
Both Pamuk and Kuran claim that institutional improvements and the evolution of the
legal systems paved the way to modern economies in the Western countries in the early-
modern era, whilst institutional stagnation in the Ottoman Empire impeded economic
development. Nevertheless, Kuran and Pamuk represent the two strands of explanation of the
institutional trajectories. Kuran stresses the role of Islam as a legal and cultural system in
shaping this institutional setting, while Pamuk highlights vested interests, social structure, and
power relations.
Pamuk’s (2005) comparative study on the evolution of the urban real wages in several
European cities and in Ottoman Istanbul, shows that the gap in urban real wages between
Northwestern Europe and the Ottoman Empire existed as early as the first half of the sixteenth
century. This gap persisted until the Industrial Revolution and further widened thereafter.
However, such a gap occurred only after industrialisation between the Ottoman Empire and
other parts of Europe. Moreover, urban wages remained close to each other in most parts of
Europe, with the exception of Great Britain and the Low Countries, where real wages were
distinctly higher than the rest of the continent during the eighteenth century (See Figure 3).
While on the one hand, points to the regional differentiation within pre-industrial Europe itself,
Pamuk’s findings go partly against the revisionist accounts of the California School, in that
they demonstrated the existence of a gap in labour productivity between Northwestern Europe
and the rest of the continent, as well as the Ottoman Empire, prior to the Industrial Revolution.
Figure 3- Real Wages of Unskilled Construction Workers in European Cities, 1450–1913

(Wages in Grams of Silver/CPI)
Source: Pamuk (2005: 221)
Pamuk (2007) suggests that the wage gap originated in the Black Death, which created
a high-wage, labour-scarce environment. According to him, the Low Countries and England
responded more successfully to the changing environment via a number of long-term
structural changes that occurred in demographic behaviour, agriculture, manufacturing, trade,

technology, and, most importantly, the institutional setting.
While Pamuk’s study (2007) offers an explanation for Northwestern Europe’s success,
Kuran concentrates on Ottoman “failure.” Kuran’s 4 explanation for the economic backwardness
of the Middle East focuses on the legal infrastructure of the region, as shaped by Islamic law.
He suggests three reasons for these economic shortcomings that had become evident by the
eighteenth century, namely, the Islamic law of inheritance, which inhibited capital
accumulation; the absence of the concept of a corporation in Islamic law, which kept civil
society weak; and the waqf institution, which locked vast resources into unproductive
organisations for the delivery of social services. Although these Islamic institutions,
incompatible with modern economic life, were radically reformed during the nineteenth
century, traditional Islamic law remained a factor in the Middle East’s on-going economic
disappointments (Kuran 1997, 2000, 2003, 2005, 2012) 5.
Pamuk’s (2009) analysis of Ottoman institutions and factor markets provides a more
nuanced, balanced, and realistic account that escapes Orientalist narratives. While on the one
hand, he emphasises that the flexibility of certain Ottoman fiscal and economic institutions
4 For a comprehensive overview of Kuran’s views, see Kuran (1997, 2004) For other essential works on
the same subject by the author, see Kuran (2000, 2003, 2005, 2012).
5 Kuran’s views represent a revival of an old paradigm that is coloured by modernisation theory and the
association of Islam with underdevelopment. His views have been severely criticised by revisionist
scholars on a number of fronts. First, Kuran overestimates the role of legal norms. Recent research
(Peirce 2003; Çizakça 2004; Ergene 2010) showed that in several cases, Ottoman judicial practice
significantly diverged from the sharia and that local power relations and individual strategies played an
important role in shaping legal outcomes. Second, in Kuran’s account, the “West” and the “Middle East”
appear as monolithic entities mainly defined by religion, and completely isolated from each other. One
needs hardly emphasize the regional diversities within both geographical areas, as well as the several
channels of exchange and flows between them. Furthermore, Kuran focuses only on religion and ignores
vested interests, social structure and power relations, distributional coalitions, and factor endowments
as determinants of institutional environment and its evolution.
For a political economy approach to the impact of Islamic law on economic development, see Coşgel
(2011). Coşgel (2011) explains institutional and technological change and stagnation in the Ottoman
Empire focusing on the relationship between the rulers and the Islamic legal community, which was an
influential group due to its power in conferring legitimacy on rulers. According to him, change and
stagnation were an outcome of the interaction between these two parties, rather than a reflection of an
intrinsic quality of Islamic law.
contributed to the longevity of the empire, on the other, he argues that many of the key
institutions of the Ottoman order, including state ownership of land and urban guilds,
remained intact until the nineteenth century due to the resistance of ruling elites who had
vested interests in the traditional order. In his account, the fact that economically productive
classes were never in a position to influence state policies and push for institutional changes
appears as the main explanation for the relative institutional stagnation prior to the
modernising reforms of the nineteenth century, and in turn was responsible for the absence
of intensive economic growth before this date.
Albeit limited, this body of literature on the long-term trajectory of the Ottoman
economy, and the role of institutional factors in its divergence from that of Europe, provides
a basis for the present research. Assessing the institutional environment, explaining its
evolution and stagnation, and investigating how it shaped the economy is outside the scope
of this study. Nevertheless, by attempting to identify differences and commonalities in the
developmental prospects of pre-modern Europe and the Ottoman Empire, this research will
draw on and contribute to this literature linking the Ottoman experience in pre-industrial era
to the Great Divergence debate.
2. Literature review: pre-industrial growth, market integration, consumption, and

relative prices
After this brief introduction to the political and economic environment of the Ottoman
Empire, and an overview of the two strands of explanations concerning the causes of its
divergence from the European path of growth, we now turn our attention to the literature on
markets, prices, and consumption. Market integration, the rise of new consumer patterns, and
changes in the relative prices of industrial goods are three phenomena associated with the
growth path of pre-industrial Europe. As such, they stand firmly entrenched at the centre of
the debates on divergence across different parts of the world prior to the Industrial Revolution.
Yet, these three potential drivers of divergence have not been explored in the Ottoman
context. In what follows, this paper discusses trade-led, demand-led, and innovation-led
growth theories with reference to the existing empirical literature on Western and non-
Western economies.
2.1. Market integration and trade-led growth
The “trade-led growth” theory is perhaps the best-established and most-widely

recognised explanation of pre-industrial growth since Adam Smith. It argues that the rise of
efficient and integrated markets, and the intensification of trade across regions were the key
driving forces for economic growth in early-modern Europe, and set the stage for
industrialisation (De Vries 1994; Persson 1988; van Zanden 1999; 2002). According to this view,
since the contribution of technological innovation remained limited, improvements in
productivity in pre-industrial era were generated by the development of trade between
previously distinct markets, which facilitated specialization and a more efficient allocation of
resources on an inter-regional scale (Özmucur and Pamuk 2007). Accordingly, market
integration is defined as the “opening and development of trade between heretofore
autonomous markets and their integration into a single operative entity” (Jacks 2004: 286).
Integration in commodity markets occurs when trading costs (inclusive of

transportation and transaction costs) decline, making trade over longer distances profitable.
The drivers of reduction in trading costs can be categorised into two headings, technological
and institutional. Technological progress fosters integration by curtailing transportation and
information costs, as well as costs concerning storage and spoilage. Among the policy-
related/institutional causes of market integration, a particular emphasis is placed on the
elimination/lowering of policy-promoted barriers to trade such as internal and external tariffs,
or import and export prohibitions. Political and economic liberalisation, as well as the
emergence of custom unions, is often cited as a crucial development that acted as an impetus
for market expansion (Pamuk and Özmucur 2007). Innovations which spread risks (i.e., marine
insurance), increased the mobility of capital (i.e., bills of exchange, improvements in the
banking system), and reorganised commercial activity (i.e., new firm models), are also
mentioned among the dynamics that brought about a decline in the transaction costs in the
long-run (Jacks 2004). On the other hand, the integration of markets is susceptible to sporadic
external shocks (i.e., wars, epidemic diseases, urban unrest, and others), which result in the
interruption of interregional trade. These lead to a temporary disintegration in markets without
permanently altering shipping costs.
In recent decades, a wide-ranging corpus of literature on market efficiency and market

integration has emerged, as focus has shifted to institutions and markets in explaining long-
term economic growth. 6 Scholars have largely focused on questions of determining when
European and transatlantic markets began to integrate, and the relative importance of
technological and institutional factors in fostering integration. The spectacular increase in
trade volumes in the second half of the nineteenth century, and the strong trend towards
integration in intra-European and intercontinental commodity markets during the same
period, are among the best-documented phenomena in economic history. However, historians
disagree as to whether this was a sudden and unprecedented phenomenon triggered by the
transport revolution, or a continuation of a trend that started much earlier with improvements
in the institutional environment.
The traditional account of nineteenth-century globalisation as established by

O’Rourke and Williamson (1999, 2002, 2004) addresses the drop in freight charges due to
technological change as the major cause of market integration in the long run, stating that a
large boost towards integration occurred only after the proliferation of steam ships and
railroads. According to them, in the absence of improvements in overland and sea transport
technologies –hence, of important reductions in freight charges 7 - a continuous trend of
integration was not possible in the early-modern era.8
6 For the integration of national markets see the following: Austria–Hungary- Good (1981); Szabad
(1961), the Benelux countries- Buyst, Dercon, and Campenhout (2000); Dejongh, Campenhout, and
Ceusters (2000); Griffiths (1982), France- Chevet and Saint-Amour (1991, 1992); Ejrnas and Persson
(2000); Roehner (1994), Germany- Fremdling and Hohorst (1979); Gerhard and Engel (2000); Kopsidis
(2002); Kuczynski (1960), Norway- Hodne and Gjolberg (1981), Italy- Sereni (1947, 1966); Zamagni
(1983); Federico (2007), Russia- Metzer (1974); Milov (1995), Spain- Pena and Sanchez-Albornoz (1984),
the United Kingdom- Chartres (1995); Gourvish (1970); Paterson and Shearer (2001), and the United
States- Coelho and Shepherd (1974); Jue (1999); Slaughter (2001). For the integration of international
markets see Chilosi et al. (2011); Federico (2011); Findlay and O’Rourke (2003); Gonzales, Garcia-
Hiernaux, and Guerrero (2012); Goodwin and Grennes (1998); Hatton, O’Rourke, and Williamson (1994);
Jacks (2005); Latham and Neal (1983); Özmucur and Pamuk (2007); Persson (1999); Rönnbäck (2009);
Sharp and Weisdorf (2013).
7 Although the introduction of three-mast, larger trading vessels in the latter part of the fifteenth
century partially reduced costs, such changes remained limited after this date (Rosenberg, Birdzell, and
Mitchell 1986). Freight costs in the mid-eighteenth century were only slightly lower than they were in
the high Middle Ages (Menard 1991).
8 These early works by O’Rourke and Williamson (1999, 2000, 2002) are in contradiction to their more
recent study (2009) on inter-continental and intra-European spice markets before and after voyages of
Contrary to accounts emphasising the technological limits to integration, scholars

such as Epstein (2000) and North (1958, 1968) have developed a new institutionalist
perspective. Their argument is based on the idea that pre-modern societies operated
significantly below their technological and productive potential, meaning the fundamental
constraints to market size – and to growth – were in fact institutional, rather than technological.
In other words, coordination failures that increase transaction costs (information, negotiation,
enforcement, exaction costs, and the like) were, to a large extent, a result of inefficient
institutions. Thus, the removal of direct institutional impediments and the improvement of the
institutional environment could and did foster integration in parts of late Medieval and early
modern Europe despite stagnant transport technology. In addition, it played a more
substantial role in nineteenth-century globalisation than technological improvements.9
Within this literature emphasizing the political conditions of pre-industrial growth,

Epstein (2000) argued that this early market development in several late medieval European
states was facilitated by jurisdictional centralisation, which reduced pre-existing seigniorial
dues; helped overcome prisoners’ dilemmas between rival feudal and urban monopolies; gave
rise to better coordinated systems of road maintenance and systematization of legal codes,
weights and measures; and which decreased the rulers’ incentives to act autocratically as a
‘stationary bandit.’
What do the empirical findings show? First, it is necessary to look at transatlantic trade.
Relying on a set of prices of commodities subject to intercontinental trade, O’Rourke and
discovery. In this study, they conclude that the advantage in terms of transport efficiency in using the
Cape route, and the changing structure of trade, which led to a more competitive Euro-Asian trade,
resulted in lower costs in the inter-continental spice trade, while at the same time stimulating the
integration of intra-European spice markets.
9 Among others, Uebele (2011) and Jacks (2009) suggested that economic and political liberalisation
and wars (or, more accurately, their absence) were the real drivers of global integration in the nineteenth
century. In a similar vein, North (1958, 1968) repeatedly argues that organisational improvements played
a more important role in lowering transport costs and spurring change in international market
integration in the first half the 1800s. More recent studies have also questioned the extent of the drop
in freight costs. Transportation costs in Atlantic trade fell after the 1870s, although the fall was modest
rather than dramatic. Jacks (2009) estimates that trade costs fell by an average of 10-16 percent between
1870 and 1913. According to Persson (2004), when evaluated against a more accurate deflator that
reflects the general fall in the international price level in the last quarter of the nineteenth century, the
case for a sharp decline in freight costs disappears.
Williamson (1999, 2002, 2004) discovered no price convergence in intercontinental markets

between 1500 and 1800. However, more recent studies employing larger data sets have
revealed episodes of transatlantic integration preceding the nineteenth-century transport
revolution. Uebele (2011) suggests that important steps towards integration in transatlantic
wheat markets occurred in the first half of the nineteenth century before steamships and
railroads could make a substantial impact and that the speed of globalisation slowed down in
the second half of the century in contrast to conventional wisdom. Rönnbäck (2009) argues
that the early-modern period featured a succession of waves of integration and disintegration,
with huge variations across trade routes and commodities. Gonzales, Garcia-Hiernaux, and
Guerrero (2012) detected a general trend towards closer market integration that started in the
eighteenth century. According to their results, globalisation progressed, not without setbacks,
gradually into the nineteenth century instead of suddenly appearing at some point after the
1820s. To sum up, the results are contradictory 10 and provide evidence to support both theses.
What is certain is that the case for early globalisation cannot be easily dismissed at the present
state of research.
Empirical findings on early-modern European markets are no less ambiguous with

regards to long-term and continent-wide integration. While some economic historians claim
that the process of market integration began under the rule of absolutists in the late Medieval
and early-modern era (Achilles 1957; Abel 1980; Unger 1983; Allen and Unger 1990; Persson
1999; Clark 2015; Jacks 2004; Chilosi et al. 2011; Gonzales, Garcia-Hiernaux and Guerrero 2012),
others suggest that there was little or no grain market integration in Europe between the late
fifteenth and early nineteenth centuries, emphasizing that price differentials persisted
(Özmucur and Pamuk 2007; Bateman 2011; Federico 2008).
Despite inconsistencies, existing research gives us important clues, provided a

nuanced approach is adopted. First, if Europe as a whole did not become engrossed in a
complete and overarching system of markets, some regions within the continent were
becoming increasingly more integrated within themselves and among one another during the
early-modern era (Özmucur and Pamuk 2007; Chilosi et al. 2011; Studer 2008). Studies on
10 These contradictory results are rooted in differences in geographical coverage and quality of data,
as well as in the multiplicity of the methodologies employed in measuring integration (Federico 2012).
Northern and Northwestern Europe unambiguously depict a growing integration of

commodity markets and an increasing synchronisation of prices during these centuries (Jacks
2004; van Tielhof 2002; van Bochove 2008; Gonzales, Garcia-Hiernaux and Guerrero 2012). By
the early eighteenth century, a national market in wheat emerged in England (Granger and
Elliot 1967). Alongside this regional differentiation, Studer (2008) introduces a further
distinction, based upon access to waterways. Integration among “lowland markets,” clustering
around the North Sea, was already very high at the beginning of the eighteenth century, while
integration among “landlocked” markets caught up throughout the century.
Second, the study of price convergence has nearly always focused on long-distance
trade. However, in order to obtain a fuller picture of the process of European market
integration, one needs to also look at the process of regional convergence, which could have
followed a different trajectory than inter-regional or international markets (Studer 2008). While
transportation costs were more important in the intercontinental and inter-regional trade, and
these depended largely on technological capacities, short- and medium-range trade might
have been more responsive to institutional pressures, and the costs could have been
significantly curtailed by policy-promoted changes, despite no changes in transportation
technology. Studying integration between Swiss towns that are 100 to 200 km apart, Studer
(2008) identified a pattern very different to that of long-distance markets. His results suggest
a much more continuous process of convergence that started in the 1730s and gradually led
to a near-equalisation of prices by the 1880s. Thus, distance was a determinant of the extent
of integration prior to the nineteenth century. It is reasonable to assume that market expansion
started from a regional level, then moved to an interregional, and finally to a Europe-wide (and
global) level (Studer 2008). From this perspective, the eighteenth century appears as an era of
integration at the regional and national levels (Agelan and Mendez 2001).
Finally, the late Middle Ages and the early-modern period featured a succession of
waves of integration and disintegration within Europe, if not a clear long-term trend. For this
reason, it would be more accurate to speak of the absence of permanent and sustainable
integration, rather than a total absence thereof. Hence, inquiring as to what impeded
sustainable market expansion is as important as considering what drove it.
Although there is no consensus as to whether trade-led growth can help explain the
Industrial Revolution, regional patterns of specialisation facilitated by expanding markets are
cited by many economic historians as factors that contributed to pre-industrial growth.
Therefore, whether there was a discrepancy between the Western and the non-Western worlds
in terms of the extent of market development, and whether it led to divergence in economic
performances across different parts of the world; are highly relevant questions with regards to
the Great Divergence debates.
Studer’s (2008) comparative study on Indian and European commodity markets, is

the most comprehensive work that links the issue of market integration to the pre-industrial
economic divergence. In this study, Studer (2008) detected a significant discrepancy between
Europe and India in terms of the extent of market development between 1700 and 1914. His
results suggest that the level of integration was higher across European wheat markets
throughout the entire period and that the region experienced a gradual expansion of markets,
even before the nineteenth-century transportation revolution had substantial impacts. Indian
wheat markets in contrast, remained largely isolated due to high transportation costs and
political fragmentation, and there was no observable development in the market before the
mid-nineteenth century.
Studer (2008) also looks at several economic indicators to understand whether the
discrepancy between the two regions in terms of market integration was accompanied by a
divergence in economic performances. He concludes that alongside having larger markets,
Europe was economically much more productive and richer than India already in the
eighteenth century. The findings supported the hypothesis that divergence in economic
performances across different parts of the world was well underway prior to industrialisation
and that unequal market development was an important factor in divergence.
Shiue and Keller’s (2007a, 2007b) comparison of markets in China and Europe,
however, reveals a different picture. Their analysis shows that before the Industrial Revolution
spread to Continental Europe, markets in China were as efficient and well-integrated as those
in Europe. Britain, on the other hand, constituted an exception with a significantly higher level
of market development around 1770 than those in the most advanced parts of Continental
Europe, as well as China. The divergence in terms of market integration and efficiency between
China and Europe in general occurred suddenly in 1780-1830 with industrialisation. During
this period, markets in the Continental Europe rapidly caught up to British markets, while no
such improvement occurred in China. They conclude that since no Industrial Revolution
ensued in China, markets cannot be the explanation of the rise of Europe. They may be a
necessary condition, but not a sufficient one for economic development.
Bassino’s (2007) findings on Japanese rice markets are similar to Shiue and Keller’s
findings on China. He suggests that during the long period of peace under the rule of the
Tokugawa (1603-1867), Japan experienced a remarkable degree of market integration, as well
as the development of regional specialisation. A very high degree of integration was already
achieved in western Japan in the early eighteenth century, and throughout the century, eastern
Japan rice markets showed a strong trend towards integration within themselves and with the
western part of the country.
The question of whether unequal market development across different parts of the
world is a key factor in explaining the divergence in economic performances before
industrialisation remains an open matter of debate, as studies on different geographical areas
reveal different stories. As Studer (2008) has emphasised, in contrast to Europe and the
Western world, which are front and centre in the extant literature, empirical evidence from
non-Western markets is extremely rare. Therefore, bringing insights from Eastern
Mediterranean wheat markets, which operated within a political and institutional environment
that was substantially different from that of contemporary Europe, will provide an invaluable
contribution to this literature. Unlike China, Japan, and India, commodity markets and trade
conditions in the early-modern Ottoman realm have not been the subject of research until
today.
One prominent issue in the study of Ottoman commodity markets in the existing
literature is the effect of regional differentiation and the importance of examining markets at
a regional, rather than national, level. Between 1660 and 1840, the Ottoman Empire
encompassed a vast territory stretching from the North African coast to the Balkans and
comprised different geographical and economic zones. Taking into account the findings of
previous research on European markets, which revealed that long-distance and short-distance
markets, as well as coastal and landlocked regions, followed different trajectories in terms of
market integration, the necessity of narrowing the geographical scope and separately
analysing regional/inter-regional integration becomes evident.
Another important point concerns the political and institutional conditions underlying
early-modern integration. Since the rise of efficient and integrated commodity markets in parts
of Europe was linked to institutional improvements, policy-promoted changes, and processes
of jurisdictional centralisation in the pre-modern era, the Ottoman Empire, which followed an
“alternative path to a modern state” characterised by administrative decentralisation and
political fragmentation, and which lacked a powerful merchant class to push for institutional
changes, offers an excellent case study to highlight the similarities and contrasts with Europe
in terms of market conditions.
Finally, previous research has highlighted the importance of analysing short-term

trends, as well as long-term integration. By identifying the periods for which dynamics towards
integration were in play, as well as the periods in which these were impeded by counteracting
forces, we can arrive at a better understanding of the processes triggering and hindering the
integration of commodity markets and, hence, pre-industrial growth.
2.2. Early-modern consumerism and demand-led growth
An alternative explanation for pre-industrial growth concentrates on demand-side

changes rooted in sociocultural transformations. This second strand of literature argues that
changes in demand in the early-modern era preceded and caused production shifts. According
to this account, during the long eighteenth century, a growing desire for consumer goods in
the Northwestern Europe led to higher levels of market participation and an increase in the
labour supply, triggering the agricultural revolution and proto-industrialization, and paving
the way for the Industrial Revolution.
Research on England, France, Spain, Italy, the Netherlands and the American colonies
has revealed a rise in the consumption of durable goods in the early-modern period. De Vries
(1975), Weatherill (1988), and Overton et al. (2004) relying on large samples of probate
inventories, demonstrated that throughout the early modern era, the number and kinds of
domestic chattels possessed by British and Dutch households significantly increased, making
the domestic environment richer, more comfortable and more specialised. The evidence
suggests that the rest of the continent was not immune from this trend. Among others, Roche
(1987) showed that France, too, experienced the same increase in the amount and variety of
clothing and furniture pioneered by the urban middle classes in the second half of the
eighteenth century. Simultaneously, studies on American probate inventories (Carr and Walsh
1980, 1988; Perkins 1991; Main 1983a, 1983b, 1988; Main and Main 1988) displayed
eighteenth-century colonials’ participation “in what became a transatlantic revolution in
consumer tastes” (Main and Main 1988: 44), through a mass adoption of consumer durables
(including new types of furniture and decorative accessories). This “revolution” not only
concerned durables, but also goods such as sugar, tea, coffee, tobacco and new kinds of
textiles, all of which began to be consumed not only by the rich, but also by the lower strata
of society.
When McKendrick first introduced the idea of an early modern revolution in

consumption in 1982, he defined it as an independent cultural revolution associated with the
abolition of the estate-based society. According to him, the willingness to consume in the
eighteenth century England stemmed from a social system characterised by increased social
mobility and class competition. The replacement of given social boundaries by a new social
hierarchy that was ordered through emulation generated the major driving force behind the
proliferation of consumer goods. The middling segment’s emulation of the aristocracy, and
the consequent efforts to effect social distinction on the part of the aristocratic elite, especially
laid the foundations for an early-modern consumer society in which consumer goods were
conceived as markers of status, rather than utilities. Fashion was generated by the taste of the
elite and “trickled down” the social echelon through a process of aping. In the second half of
the eighteenth century, consumption spread rapidly among the middling sorts, especially
those living in urban and industrial areas and, most notably, in London. Although not to the
extent of the middling groups, the poor in many regions also became more and more
integrated into consumer society, turning away from the traditional moral economy. The result
was an increasingly rich, more colourful, and more diversified material culture that was shared
by all social classes (McKendrick 1982).
Since McKendrick first coined the term, the spread of consumer goods among the
middling ranges and the lower classes, was posited as the characteristic feature of the
eighteenth century consumer revolution. Nevertheless, this postulation is in contradiction with
previous and more contemporary findings on long-term trends in real wages, which reveal a
decline or stagnation rather than an increase in wage rates and labour productivity in Europe
as a whole, especially during the eighteenth century. In Britain, for instance, the rise in both
wages and income between 1600 and 1750 was very modest – less than 20 percent (Clark
2008). This contradiction has led several economic historians to attempt to bring an
explanation to higher levels of consumption by the middling and lower stratum in the face of
declining wage rates in the eighteenth century.
The most popular answer to the puzzle is De Vries’ ‘industrious revolution’ thesis
(1994, 2008). De Vries defines a process of the reallocation of time between work and leisure,
and the higher involvement of women and children in labour that characterized Northwestern
Europe between 1650 and 1850. In the main, an increase in daily wage rates is conventionally
considered to indicate an increase in standards of living. However, the ability to spend depends
on disposable yearly incomes, which are a function of both wage rates and the number of
workdays per year. Therefore, earnings can be raised by increasing the labour supplied, even
when real day wages are stagnant or declining. De Vries (1994) claims that this was what
occurred in the Netherlands starting in the seventeenth century. Driven by an overwhelming
desire for more goods, which found its source in changing tastes and commercial incentives
(also related to changes in relative prices and reduced transaction costs), consumers began to
work more hours per year and to substitute consumer goods with leisure. This increased both
the supply of marketed commodities and labour and the demand for market-supplied goods,
which eventually led to improved levels of material well-being –contrary to the picture
depicted by real wage rates.
While for both McKendrick (1982) and De Vries (1994), the early modern consumerism
was an exceptional and unique phenomenon that was strongly associated with sociocultural
and economic trajectory of the Northwestern Europe, recent research (Belk, Ger and
Askegaard, 2003; Ger and Belk 1999; Howes 1996; Zhou and Belk 2004; Karababa 2006, 2012;
McCabe 2014) has suggested that the boundaries of early modern consumerism can be
broadened to include the non-Western world. By showing that multiple consumer cultures
began to develop across the globe during the early modern era, these studies have challenged
a euro-centric account of the consumer revolution.
Consumption is one of the fields, in which Pomeranz (2000: 111) refuted a Western
superiority that manifested itself prior to the nineteenth century, claiming that there were no
significant differences between Europe and China in the “ability and inclination of a minority
of well-to-do households … to stimulate economic change through changes in what they
wished to buy”. He suggests that changes in consumption, at least in elite consumption in
parts of Asia, were “roughly” comparable to Europe between 1400 and 1800, although Asian
consumption did not show the on-going acceleration experienced by Europe during the same
period. In China, Japan, and to a more limited extent India, the change in attitude towards
goods revealed itself through the display of material possessions as determinants of status,
proliferation of different kinds of status goods for different levels of people, and discussions
about the “proper,” “tasteful” way to use various commodities.
According to Pomeranz’ estimates, Chinese consumption of sugar, was higher in 1750

than that in continental Europe, even in 1800, and the consumption of tea and tobacco stacked
up quite well against European consumption. China and Japan also witnessed a striking
increase in the quantity and variety of home furnishings, elaborate clothes, and eating utensils
among the wealthiest people in the sixteenth through eighteenth centuries. There was also a
huge boom in the printing of religious texts, medical manuals, and almanacs using simplified
language and aimed at a popular audience, at least in Yangzi Delta, during the same period.
In the case of Tokugawa Japan, these improvements were enabled by the substantial share of
peasant incomes (at least 20 percent) available for savings or discretionary non-subsistence
spending (Hanley and Yamamura 1977).
Whether the findings on European and North American probate inventories about
changing patterns of consumption are equally valid for other parts of the world, or whether
they were specifically a characteristic of the Western economic shift, are central questions in
understanding the early roots of consumerism, as well as explaining long-term economic
growth. Still, evidence from non-Western contexts is fragmentary and relatively rare. In the
absence of Asian counterparts to European probate inventories (Pomeranz 2000), the
divergence in terms of consumption of durables across different parts of Eurasia prior to the
nineteenth century has been primarily discussed on the basis of qualitative and anecdotal
evidence. This lack of data restricts our ability to make sound comparisons and to grasp the
real nature of changes that occurred in the domestic environment and in the possession of
personal goods outside Europe and America. In this regard, Ottoman inheritance inventories,
which report all movable and immovable assets of the deceased, including personal and
domestic goods, provide us with a unique opportunity to achieve quantitative insights into
pre-industrial consumerism in a non-Western context.
Beyond this advantage provided by inheritance inventories, studying consumption in

the Ottoman realm is illustrative in terms of how consumerism was embedded in sociocultural
processes. If changing tastes were the major driver of the early-modern consumer revolution
in Europe, and these were rooted in the transformations in the social and cultural spheres as
McKendrick (1982) and several others have argued, then we would expect that a similar
process would have occurred in the Ottoman Empire, which experienced a profound change
in its social structure and power relations from the seventeenth century onwards. In parallel to
the dissolution of the classical order, established boundaries between the tax-exempt ruling
class (askeri) and the large subject class of reaya began to disintegrate as mobility between
the classes increased. With the monetization of the economy, Ottomans began to gain access
to goods, positions and social standing on the basis of their ability to purchase. Throughout
the eighteenth century, the new class of local elites that emerged in the provinces struggled
not only for political power and economic interests, but also for social influence (Karababa
2006; Hamadeh 2004). This should have created favourable conditions for consumption to
gain a social role and began to be used for social differentiation.
Was early-modern consumerism primarily a demand-side phenomenon that was

driven by sociocultural transformations and changing tastes and which, in turn, triggered
higher industrial output levels? Or was it a consequence of the already changing nature of
production and distribution? Even after three decades of research on the roots of modern
consumerism, answering these questions and explaining the increased acquisition of
consumer goods and luxuries remain controversial tasks.
De Vries’ argument of an ‘industrious revolution’ has been severely criticized by Clark

and Werf (1998) on the grounds that there is no evidence in favour of a universal trend
displaying an increase in work habits in early-modern Europe. According to them, labour input
per person in England was already high in the Medieval Era. As for women and children, even
if De Vries is correct, their participation in the workforce would have exerted only a small
impact on the total labour supply. Therefore, the “seeming paradox” (Clark 2004: 21) of rising
consumption in the face of stagnant or declining real wages requires a different sort of
explanation. Among others for Clark (2004), the proliferation of consumer goods among
European societies and the rising demand for these were a result, rather than a cause of,
growth. These accounts suggest that the explanation for early-modern consumerism lies in
the changing price-product structures indicating productivity gains in particular sectors of the
economy. The following section is concerned with the literature that takes the declining
relative prices of industrial goods as a sign of innovation-led growth.
2.3. Changes in price-product structures and innovation-led growth
While the trade-led and demand-led growth theories postulate technical stagnation
and seek the dynamics underlying pre-modern growth elsewhere; a third strand of literature
suggest that as early as the sixteenth century Northwestern Europe experienced significant
productivity gains in the non-agricultural sectors, which occurred due to technical and
institutional innovation, and which were manifested by the declining prices of manufactured
and traded goods in comparison to agricultural prices. With the increasing productivity in the
manufacturing and transportation, it is argued, the region became more productive than the
southern parts of the continent and the developed parts of Southeast Asia, long before the
technological advances of the late-eighteenth century (Broadberry and Gupta 2006).
Almost 20 years after Shammas’ (1994) pioneering article on the decline of textile
prices in England and British America between the mid-sixteenth and late-eighteenth
centuries, early-modern changes in price structures has once again attracted the attention of
historians, this time in relation to the Great Divergence debate. Recent research has
demonstrated that between 1500 and 1800, the prices of a wide range of manufactured and
traded goods in Europe rose much less than the prices of agricultural products, bringing about
a rise in the consumption of everyday luxuries and in the acquisition of consumer goods
(Hoffman et al. 2002; Allen et al. 2004; Clark 2004; Malanima and Pinchera 2012).
For Allen et al. (2004) and Broadberry and Gupta (2006), during this period,
Northwestern Europe particularly stood out as a region with the world’s most expensive grains
vis-à-vis the world’s cheapest non-food industrial goods, reflecting gains in productivity,
specific to this part of Europe, and was not experienced elsewhere. According to them, this
product-price pattern refutes the arguments in favour of a Northwestern European superiority

in growing food and discredits higher agricultural productivity as “distinctive advantage”, and
thus, as a potential explanation of early divergence. Rather, it points to productivity gains in
the non-agricultural sectors and to improvements linked to urban and non-agricultural
developments (Broadberry and Gupta 2006).
An important implication of these findings on prices concerned the previous

measurements of real incomes, on which Pomeranz has based his arguments. Broadberry and
Gupta (2006: 6) suggest that across-region comparisons of living standards based on grain
wages underestimated the divergence between Northwestern Europe and the rest of the world
prior to the nineteenth century, since real consumption wages in this region rose “through the
increased consumption of non-agricultural goods and services, the prices of which were falling
relative to the price of grain,” (Broadberry and Gupta 2006), while high silver wages in
Northwestern Europe did not translate into high grain wages before the nineteenth century.
However, in the absence of research on the evolution of the relative prices of traded
and industrial goods in the non-Western world throughout the early-modern period, it is not
possible to realistically assess these arguments. Although manufactured goods were cheaper
relative to grains in Northwest Europe or other parts of the continent and presumably other
parts of the world, this does not necessarily mean that the decline in relative prices was not a
global trend during the same period.
While Clark (2004) identifies technological discoveries from 1200 to 1700 (i.e., the
introduction of the knitting frame, printing press, windmills, sailing ships, and new optical
instruments – spectacles, telescopes, and microscopes – the mechanisation of silk spinning, et
cetera) and the enhanced ability to trade overseas as the prime reasons for the drop in prices
of consumer goods and luxuries, Broadberry and Gupta (2006) attribute a crucial role to
improvements in the service sector (distribution and finance), alongside improvements in
production. However, the change in relative prices might also be interpreted in demographic
terms (Hoffman et al. 2002). Under constant productivity, one would expect the relative prices
of agricultural goods to rise in epochs of rising population and the relative prices of industrial
goods to decline (Malanima and Pinchera 2012). At this juncture, looking at how the prices of
manufactured and traded goods evolved relative to agricultural prices in different economic
and demographic settings will contribute to our understanding of the source of changing price
patterns.
Ottoman manufacturing in the eighteenth century might not have been a flourishing,
dynamic sector competing on international markets, but it was not stagnant either. As Genç
(1994) asserts, it displayed a complex pattern, with subsequent phases of expansion and
retraction. During this century, Ottoman manufacturing competed first with Asian, and then
European, imports but managed to survive. Despite its marginality in foreign markets, the
Ottoman textile industry continued to produce for a vast domestic market. Even in the last
third of the eighteenth century, when the cotton industry in England was being transformed
by the Industrial Revolution, the empire remained self-sufficient in cotton textiles. Around
1820, the share of imported cotton yarn and cloth in total domestic consumption was just 4
percent (Pamuk 1986).
This notwithstanding, throughout the period under study, no substantial

improvements seem to have been implemented in the organisation or methods of production
in the Ottoman manufacturing sector. Until the mid-nineteenth century, artisans were
organised around small-scale manufacturing establishments (Pamuk 1986). At the same time,
the Ottoman industry remained under the control of the guilds, which acted as interest groups
protecting their members by restricting production and overseeing quality and prices
(Quataert 2005), as they sought and obtained the support of the government whenever
merchants attempted to organise alternative forms of production (Pamuk 2004). Under these
conditions, any significant productivity increases that might have occured due to technical and
institutional innovation is highly unlikely. Therefore, if the Ottoman realm experienced a
declining trend in the relative prices of manufactured goods – as was the case for several other
European countries – then we should turn our attention to factors other than innovation to
explain this drop.
This section provided a brief overview of the literature on three mechanisms of pre-
industrial growth. Until today, all three mechanisms of growth have been investigated primarily
with evidence from European economies. Nevertheless, the explanatory power of market
development, consumer demand, and technical innovation depend not only on their existence
in Northwestern Europe, but also their absence elsewhere. This is a key issue in determining
whether the Industrial Revolution was a phenomenon that occurred suddenly and at the
intersection of several contingencies, or whether it was the culmination of a long-run
evolution. If those regions that were growing slowly also experienced these phenomena, then
it would cast doubt on the connection between them and growth and the Industrial
Revolution. In this respect, it is relevant to ask whether there was a discrepancy between the
Western and the non-Western worlds in terms of the extent of market development,
consumption patterns, and trends in relative prices, and whether it led to divergences in
economic performances across different parts of the world. Despite this, the non-Western
world has attracted scant attention within the existing literature on markets, prices, and
consumption, in contrast to the abundant empirical research on pre-modern Europe.
This is where this study provides an important contribution to the debate, improving
on the empirical evidence from a non-Western context that has so far often defied sound
assessment by presenting new quantitative evidence for the Ottoman Empire. The present
study investigates whether these three phenomena were restricted to Europe or could be
experienced to other parts of the world. In a series of five papers, the following questions are
asked: (1) Did Ottoman commodity markets experience long-term and sustainable integration
so as to produce regional patterns of specialisation? (2) Did the Ottomans enjoy a greater
variety and quantity of goods despite stagnant real wages, as was the case in eighteenth-
century Europe? (3) Did manufactured goods become cheaper relative to agricultural goods
between 1700 and 1840?
3. Sources
3.1. The Ottoman inheritance inventories
This study relies on evidence provided by Ottoman inheritance inventories.

Inheritance inventories are lists of the possessions of a deceased individual recorded by a
judge to distribute the inheritance among the heirs. These sources are included in the şer’iye
registers (judicial court registers), which were compiled in all major kazas of the empire.
Inventories were registered either together with other judicial court records or in separate
books known as tereke defterleri (inheritance books). Sometimes inheritance inventories for
the ruling and ruled classes were registered separately. 11
Thousands of these sources are available for several Ottoman towns and survive for
long, continuous periods from mid-fifteenth century to the early twentieth century. Inventories
are obligatory in Islamic law when there are minor heirs, a pregnant wife or missing heirs. The
court also intervened upon the request of the emin-i beytülmal (the local representative of the
state treasury) or (1) if no heir existed other than the spouse(s), (2) if no heirs were known, (3)
if the heirs were far away, (4) if the and Pascual 1992).
Preparing an inheritance inventory 12 involved the following process: the kassam, a

judicial official and expert in inheritance matters, would arrive at the home of the deceased
and ascertain the particulars of any property possessed by the deceased by referring to copies
of court warrants (hüccets) in the hands of the family, and possibly, to the testimony of
witnesses. He would then make a record of the description. Each of the legal heirs would be
summoned, as required by law to act as eyewitnesses themselves (Matthews 2001).
The inheritance inventory constitutes a summary of four main operations

subsequently executed by the kadı (judge): the identification of the deceased and heirs, the
listing of assets, the enumeration and deduction of debts, and the apportioning of shares. In
the initial section, “introductory protocol,” the deceased would be identified by given name
and father’s name and place of residence (by neighbourhood or village and the city).
Subsequently, the names and degree of affinity of the legatees, the title of the treasury agent
(emin-i beytülmal) and the date of portioning would be recorded. Occupation and the cause
of death were also occasionally mentioned. (Matthews 2001) The second section, “inventory”
describes the deceased's patrimony in detail: Buildings (houses, shops, watermills), vineyards,
trees and crops, livestock, personal and household goods, stores, commercial goods, as well
11 Inheritance inventories belonging to the askeri class were recorded by the kadiasker (top
administrative official), while the inheritance inventories of ordinary people were recorded by the kadı
(Karababa 2006).
12 The terminology used in this study is that of Matthews. She suggests that the term “probate” is
unsuitable for Ottoman terekes, because unlike in Western societies whose legal system relied upon
Roman heritage, in Ottoman society, the partitioning of the estate among legatees was executed in a
prescribed manner by the court in accordance with Islamic law. (Matthews 2001: 19)
as outstanding loans13 (der zimmet) and the name of the borrower. All items were recorded
with their worth (Establet and Pascual 1992). The values assigned may have reflected a price
estimate or the actual amount for which the item was sold at auction. The third section,
“personal liabilities,” constituted the claims against the inheritance: debts incurred by the
deceased (Düyun), an outstanding bride price to the wife, claims on the estate, bequests, and
sundry expenses (medical expenses, funeral costs, the cost of the inheritance registry process,
and taxes). In the fourth section, the net amount of the assets was divided among the heirs
(Matthews 2001). (See Appendix for a sample inheritance inventory from the late eighteenth
century).
3.2. Comparison with British probate inventories
Although Ottoman inheritance inventories resemble their European and American

counterparts in many respects, some significant differences exist in their content. Here, I will
focus on English probate inventories for a comparison with Ottoman inventories. The Statue
of 1529 and other contemporary legal texts explicitly regulated what a probate inventory
should include: all the movable property of the person deceased -goods, chattels, wares,
merchandise - found in the houses, yards, barns, stables, shops and workshops - money,
jewellery, livestock and other domesticated animals, growing crops on the fields, and debts
owed to the deceased. Land and buildings, and all non-movable property including fixtures of
a building such as ovens, furnaces and window glasses, wild animals, self-produced crops and
debts owned by the deceased were excluded (Overton et al. 2004).
Secondary sources are silent as to whether the process and rules for preparing an
Ottoman inheritance inventory was regulated by any particular code besides religious law,
confining themselves to asserting that all the movable and non-movable properties of the
deceased were fixed, appraised and distributed according to Islamic inheritance rules (İnalcık
1953; Barkan 1966; Öztürk 1995). Hence, we need to turn our attention to the probate
inventories themselves to determine what was truly included and what was not.
Unlike English probate inventories, Ottoman inventories included real estate for
residence (beyt, menzil, hane, konak, oda, dam), shops, workshops, mills, stores, barns, and
13 In the case of married women, this included the unpaid instalment of the dowry, which the husband
was liable for.
rural real estates (çiftlik, tarla, bağ, bahçe, bostan, arsa, yoncalık, çehrilik, ağaçlık, otluk, yurd).
However, it should be noted that only freely held lands (mülk) were subject to the inventory
process, and Islamic inheritance law did not apply to state lands ( miri), which constituted the
majority of the agricultural lands until the mid-nineteenth century. Peasants possessed only
usufruct rights over state lands, the transfer of which was regulated by customary law (İnalcık
1953). Owning shares of a piece of real estate was a common practice in Ottoman society, and
such shares were recorded and inherited. Like English inventories, Ottoman inheritance
inventories also included cash money, debts owned by and owned to the deceased, as well as
slaves, livestock, commercial goods (raw material and manufactured goods), means of
production and personal and household belongings.
At a preliminary glance, the most important advantage Ottoman inheritance

inventories have over English probate inventories was the inclusion of real estate. If the
question of omissions is ignored, they appear to present a more complete picture of an estate
owner's wealth. However, houses or other buildings were never described in detail in Ottoman
inventories: They contain no information about the numbers of the rooms or how they were
used. On this later point, English inventories are more extensive. In the process of probate,
English appraisers often listed items room by room, even though they were not required to
do so (Overton et al. 2004). Although the risk of underrepresentation of room numbers is
mentioned by historians, the inventories still provide crucial insights into early-modern English
houses. Room use is part of material culture. Furthermore, from the perspective of
consumption studies, the location of objects in the houses gives clues as to the meanings
attributed to them (Overton et al. 2004).
In comparison to English inventories, Ottoman inventories are far less useful sources
for the study of agricultural history. Rural underrepresentation is the crucial but not sole factor.
Like English inventories, which cannot tell us directly about the extent of land ownership
(Overton et al. 2004), Ottoman inventories do not shed light on small family farms, which were
the basic form of early-modern Ottoman agriculture (Pamuk 2005). Since small farms were not
private property but state land, they left no traces in the sources. For the same reason,
information about crops in the fields in England, which have been utilized to infer agricultural
productivity in a number of path-breaking works (Campbell 1983; Overton 1979, 1990; Allen
1988), are absent in the field of Ottoman studies.
Ottoman inventories contain information about both the debts owed by and owed to
the deceased person, whereas English inventories only contain information about the latter.
On this matter, the evidence provided by inheritance inventories can be used to shed light on
the extent that credit relations were common in Ottoman society, or to identify moneylenders
in a particular town, in order to gain an understanding of them through their social status and
economic means. However, Ottoman inventories do not provide any insights about interest
rates or terms of debt, which are of central importance for studying credit transactions.
The differences in the amount of detail inserted into the description of certain goods
is also noteworthy. In Ottoman inventories, clothing items and textiles used for home furniture
were described in a very detailed manner. The description of a cloth always involved its type –
often the type and origin of the fabric – and in many cases, its colour and ornaments. In
contrast, English inventories often defined clothing items as simply “apparel,” lumping them
together instead of itemising they separately (Overton et al. 2004).
Finally, an important differentiation resulting from inheritance laws needs to be

underlined. According to English common law, the possessions of a married woman belonged
to her husband, as long as he was alive (Overton et al. 2004). Hence, only unmarried women
(primarily widows) made wills or had inventories. Unlike English common law, Islamic property
law recognized married women’s property rights. A substantial number of inventories
belonging to married and unmarried women exist, allowing comparisons across genders. This
implies that total estate values reported in married men’s inventories were not representative
of overall household wealth, as men’s estates did not include assets of wives. The same applies
to household goods. The goods in the house were a combination of goods possessed by
husband and wife. For a comprehensive understanding of the inner Ottoman house, male and
female possessions need to be evaluated together.
3.3. Limitations
With all the detailed information they provide, inheritance inventories constitute a
unique data source for reconstructing the economic and social history of the pre-modern
Ottoman Empire (Ergene and Berker 2008). Since Barkan’s (1966) pioneering work, “Edirne
Asker Kassamına Ait Tereke Defterleri (1545-1659),” 14 inheritance registers have been
employed, particularly in the context of urban and provincial history, to shed light on several
aspects of Ottoman society and economy. These included family size and structure, 15 annual
and seasonal mortality rates,16 economic trends, credit relations, 17 money, real estate prices,18
polarization of wealth, 19
social stratification, 20
factors of wealth accumulation, 21
and
consumption.22
Despite their usefulness, inheritance inventories suffer from critical shortcomings in

depicting a realistic picture of the society. Here, I focus on three main problems often
associated with employing these records for quantitative study: representativeness, omissions,
and accuracy of valuations.
The most recognized shortcoming articulated regarding probate/inheritance

inventories concerns their ability to represent the living population of the time. Not all estates
were recorded, and those that were registered did not usually constitute a random sample of
the individuals in the society23. The frequency of estates registered increases as we ascend the
14 In this article, Barkan (1966) presents tabulations for more than 3,000 inheritance inventories,
demonstrating the distribution of the estates according to sex, social origin, marital status, number of
wives and children, as well as the different ranks within the military class, following a short introduction
about the sources and Islamic rules of inheritance.
15 See Barkan (1966); Demirel (1990); Establet and Pascual (1994); Öztürk (1995); Rafeq (1994).
16 See Establet and Pascual (1994)
17 See Barkan (1966); Cezar (1977); Establet and Pascual (1994); Gerber (1988); (1995); Rafeq (1994);
Seng (1991); Todorov (1983).
18 See Establet and Pascual (1994); Gerber (1988); Öztürk (1995); Seng (1991); Todorov (1983).
19 See Coşgel and Ergene (2011); Establet and Pascual (1994); Raymond (1974).
20 See Canbakal (2007); Raymond (1974); Establet and Pascual (1992), Todorov (1983a)
21 See Ergene and Berker (2008); Canbakal (2007); Establet and Pascual (2004); Öztürk (1995); Coşgel
and Ergene (2011).
22 See Karababa (2006, 2012); Göçek (1996); Ceylan (2010).
23 Coşgel and Ergene (2014) extensively discuss the selection bias in the Ottoman court records in the
context of eighteenth-century Kastamonu.
economic ladder, since there was little point in making an inventory when a deceased person
left an estate of little value in relation to the costs of obtaining probate (Overton et al. 2004).
Besides, the population of the deceased was naturally generally older and hence wealthier
than the living population. Therefore, such estate records underrepresent the poor and over-
represent the rich. In other words, there is a systematic upward economic bias in these
registers.
Another important question is the degree to which probate inventories were

comprehensive and accurate in listing the possessions of the deceased. Many historians draw
attention to the fact that, for a variety of reasons, such inventories were incomplete accounts
of the movable and, in the case of Ottoman inventories, non-movable assets. Missing items
might have been present in the household but ignored by the appraisers, or been concealed
by one or more of the heirs before the inventory was made (Overton et al. 2004). The most
liquid items, cash money and jewellery, were sometimes removed between the death of the
estate owner and the intervention of the court. Everyday objects of little value, whose existence
or absence had few economic implications, such as ceramics, sewing gear, or children's toys,
were often overlooked by the appraisers (Bedell 2000).
A further doubt regarding probate/inheritance inventories concerns the accuracy of

valuations. Like the wealth bias in the sources, or the likelihood of omissions from estates,
unrealistic or erratic valuations of inventory items can lead the historian to distorted
conclusions, as the total estate values analysed in such studies were calculated from a sum of
the total values of the deceased's movable and immovable possessions, the cash they owned
and the debts owed to him. To what degree valuations in Ottoman inheritance inventories
reflected conventional prices will be discussed in detail later in this study in chapter 2.
3.4. Geographical coverage
This research employs inventories belonging to the Ottoman towns of Üsküdar,

Manisa, Ayntab, and Trabzon, which are available at the Centre for Islamic Studies (Istanbul).
Map 2- Ottoman towns of Üsküdar, Manisa, Ayntab, and Trabzon
Located on the Asian shore and a gateway to Istanbul, Üsküdar was one of four official
divisions of the capital in the eighteenth and nineteenth centuries. It was a commercial and
military hub, and a site for small farming and agricultural industry. The town was the first
station of the relay system on the trade routes connecting Istanbul to Anatolia, Arab lands and
to Iran, and as such, it was the departure point for the pilgrimage and campaigns to the East.
In the eighteenth century, the town’s population was estimated to be around 60,000
inhabitants (Bostan 2012).
A residence for imperial princes until the end of the sixteenth century, Manisa was one
of the most important administrative centres of the empire and was richly endowed with pious
foundations. Surrounded by gardens and vineyards, the city was a supplier of Istanbul markets
from an early date, meaning it had strong trading links with the capital (Faroqhi 1994). Manisa
benefited from its advantageous location in the vicinity of İzmir and increasingly specialised in
cotton production during the eighteenth century. Compared to its symbolic significance, the
city's population was modest by the late sixteenth century, with approximately 2,000 taxpayers
(Erder and Faroqhi 1980). The substantial population size of the city in the early nineteenth
century (an urban population of around 35,000 (Emecen 2003)) suggests that Manisa
experienced a growth spurt during the eighteenth century. It is widely assumed that the city
owes its prosperity and its size in the later period to commercial agriculture and large, cotton-
growing farms established to meet increasing European demand (Teoman and Kaymak 2008).
Ayntab (modern-day Gaziantep) was a medium-sized town in Southeastern Anatolia

with no particular administrative or economic significance for the centre. Due to its insular
position, the city did not engage in export-oriented agricultural production during the
eighteenth and nineteenth centuries. Nevertheless, references in local sources regarding the
local specialty fabrics from Trabzon and Tokat in the north, Ruha in the east, and Hama in the
south indicates, according to Canbakal (2007: 43), the town's participation “in a wider
commercial network which was primarily domestic, and stretched across Eastern Anatolia,
northern Syria, and Iraq.” Therefore, inventories from Ayntab can be expected to provide
insights into the emergence and development of the regional economy in this part of the
empire.
Trabzon, located on the eastern Black Sea coast, was a middle-sized town that
contained 2,122 tax-payers in the late sixteenth century (Jennings 1976). It was the seat of a
sancakbey, an administrative centre, established immediately after the conquest by Mehmed
II in 1463. After suffering immensely from the seventeenth-century crisis, the city benefited
substantially from Iranian trade and turned into a great international trading entrepôt on the
Istanbul-Tabriz trade route and became an important port for the re-export of Iranian silk and
goods from the Caspian coast to Europe from the late seventeenth century onwards (Jennings
1976). By the nineteenth century, Trabzon was among the few Anatolian cities with a
population exceeding 40,000 (Erder and Faroqhi 1980). In this regard, Trabzon is a cogent
setting to observe the impact of trade routes on early-modern urbanisation.
4. Empirical and methodological contributions
Alongside the overall motivation to relate Ottoman economic history to the Great
Divergence debate, each paper possesses its own objectives that cut across several current
themes in European and Ottoman historiography. The thesis consists of three sections. Section
1 is devoted to a methodological discussion regarding the use of inventories as historical
sources. From the 1960s onwards, parallel to the discovery and widespread use of probate
inventories by the historians of early-modern Europe and America, substantial effort has been
directed towards testing the reliability of the sources, identifying the problems, and
developing methodologies to correct them. In the case of Ottoman inheritance inventories,

however, almost no attempt has been made in this direction, despite their increasing use in
the last decades. The paper “Ottoman inheritance inventory as a source for price history”
addresses historians’ doubts concerning the accuracy of valuations in inheritance inventories.
It provides a detailed examination of the quality of the valuations and discusses to what degree
these reflect conventional commodity prices of the time. It is the first study to test the reliability
of the information provided by inheritance inventories, and as such, provides a reference point
for future research employing these records, including the remainder of this thesis.
This paper also explores the potential that the inheritance inventories might possess
in expanding Ottoman price history. Until today, most research based on inheritance
inventories concentrated on wealth and wealth-related issues (polarisation of wealth, factors
of wealth accumulation, etc.). With this paper, I aim to introduce a new area of investigation
to inheritance inventory studies. The results strongly suggest that inventory valuations are
generally consistent and were closely related to the conventional prices of the time. Building
upon these findings, it is reasonable to assume that the prices contained in Ottoman
inheritance inventories can be reliably employed for historical research.
Employing a price dataset compiled from inheritance inventories and secondary

sources, Section 2 examines regional, interregional, and international wheat markets in the
Eastern Mediterranean from the late seventeenth to mid-nineteenth centuries in an effort to
understand whether these experienced sustainable and long-term integration, and to identify
integration and disintegration episodes in the shorter term. Besides contributing to the market
integration literature by providing insights from the Eastern Mediterranean, a geographical
area neglected by the existing scholarship, this paper sheds light on an under-explored topic
in the Ottoman economic history. Trade within and between Ottoman provinces and domestic
commodity markets in the seventeenth and eighteenth centuries is far less understood
compared to foreign trade during the same period (McGowan 1994). Here, using price data,
this paper enhances our knowledge on the conditions of domestic as well as international
exchange.
This paper reveals that unlike the North Sea Area and some other parts of Europe, the
Eastern Mediterranean despite hosting two leading powers of the pre-Colombian world did
not move towards a regional existence during this period, although in the Adriatic basin a
strong trend of integration was observed, particularly in the early nineteenth century.
Furthermore, on the eve of the first wave of globalisation, domestic wheat markets in the
Ottoman Empire were no better integrated than they had been two centuries previously.
Section 3 is concerned with the evolution of domestic comfort and conveniences in

the Ottoman Empire in a long-term perspective. The first paper in this section, “Was there a
consumer revolution in the Ottoman Empire? (1700-1850)” looks at the types and quantity of
goods possessed by households in Üsküdar as they appeared in the inventories. It explores
how the ownership of domestic chattels evolved from 1700 to 1850, and examines the
existence of an Ottoman consumer revolution during this period. This paper is the first long-
term study on Ottoman consumerism, and the first to quantify the change in consumption
patterns over time. It reveals new insights into Ottoman living standards, which cannot be
captured by real wage series alone. Doing so, it contributes new quantitative evidence to the
broader literature on living standards and consumption from a non-Western context.
Focusing on how the material environment inhabited by ordinary Ottomans evolved,

this paper relates itself to the ‘decline debate.’ For a long time, traditional historiography
assumed that following its golden age in the sixteenth century, the empire entered an era of
decadence enveloping the military, political, and economic spheres, ultimately leading to its
inevitable dissolution. In the last decades, the “Ottoman decline” during the seventeenth and
eighteenth centuries as a “totalising historiographical device” has been severely criticised and
has been replaced by a more nuanced and non-linear view of the trajectory that the empire
followed in these centuries. This study tests the revisionist argument that “in the eighteenth
century, Ottomans lived better than their ancestors of the Suleimanic era, in terms of material
culture and means, [although] their lot was not improving relative to the lot of those living in
other parts of the world (Kafadar 1999: 68).”
The results on Üsküdar inventories depict a picture that is similar to those depicted by
most studies on European probate inventories, which show an increase in the ownership of
household durables in the ‘long eighteenth century’; hence, I reject the postulate of a
regressing economy as put forward by the traditional historiography. The interiors of Ottoman
houses grew richer and more comfortable throughout this period, with an increase in the
quantity and variety of the household goods possessed.
The second chapter of Section 3 expands on the issue by exploring the changing price
structure of durable goods and investigating whether change over time in the types and
quantities of domestic chattels possessed by deceased Ottomans was associated with a
change in the overall value of domestic durables stock in their estates, and/or with the
cheapening of manufactured goods. The results are similar to those for early-modern Europe
and America, which have shown that both the absolute value of probated consumer goods
and their share within the overall estate remained stable or declined, while people owned
increasingly more goods. Throughout the same period, however, the valuations of selected
household goods, as well as the prices of several manufactured and traded goods, declined
relative to agricultural products’ prices. This decline was particularly visible in the 1730-1790
period, when the ownership of goods rose most rapidly.
Until today, Ottomanists have viewed early-modern consumerism as a symbolic field

enjoying a broad autonomy and detached from economic constraints. Consumer behaviour
and new consumption patterns have intrigued them with their social and cultural aspects.
Without ignoring the significance of sociocultural processes in determining consumer choices,
and in shaping the manner in which the Ottomans imagined a comfortable and desirable life,
this study approaches the phenomenon from a different angle, focusing on the Ottomans’
ability to possess consumer goods. Approaching Ottoman consumerism from an economic
perspective, this paper fills a significant gap in the literature, by showing that the inhabitants
of Üsküdar could own more consumer durables without re-allocating their resources in favour
of durables.
The third chapter of Section 3, “Gender, religious status, and household durables”
focuses on how gender and religious status was related to ownership of domestic goods. I
compare male and female and Muslim and non-Muslim estate owners’ possessions and the
share of consumer durables within their overall estates and examine whether the ownership
of selected categories of goods varied across these groups in order to understand how
material culture varied between them. Unlike British and American probate inventories,
Ottoman inheritance inventories offer a favourable ground to explore differences across
genders and religious groups. These sources allow us to identify the social segments that
pioneered the introduction of new consumption patterns in the Ottoman realm, understand
whether a gender-based differentiation existed in terms of ownership of domestic chattels,
and ascertain whether Ottoman women were more eager consumers than men.
The quantitative analysis demonstrates that Muslim men in the town, rather than the
non-Muslim community or Muslim women, appeared to be inclined to acquire novel consumer
goods associated with westernization, and a western lifestyle, such as clocks, mirrors, and
modern furniture. The ostensible supremacy of women – and particularly Muslim women – in
terms of the ownership of household effects resulted from the possession of greater quantities
of more established ‘traditional’ goods that were always part of the Ottoman inner house, such
as household linen, chests, and towels. Rather than a sign of greater interest in the new
consumer regime, the higher shares of household goods in the estates of Muslim women
compared to Muslim men at the same wealth level reflects both the limited control over
resources and the fewer investment opportunities available to women in Ottoman society, and
the gender roles that relegated women to the home and made the creation of a comfortable
domestic environment a prime female responsibility.
Overall, the findings of this research point to long-term market development (and its
absence), rather than a change in attitudes towards consumption, and productivity gains in
the non-agricultural sectors as a major source of divergence prior to the Industrial Revolution
between parts of Europe and the Ottoman Empire; and provide support to the literature
stressing the significance of institutional improvements and political conditions for the rise of
efficient and integrated markets and thus, for pre-modern growth.
59
SECTION 1
SOURCES
60
CHAPTER 2
OTTOMAN INHERITANCE INVENTORY AS A SOURCE FOR PRICE
HISTORY24
Ottoman historians have been severely limited by the poor availability and low
quality of primary sources from which historical prices can be compiled. The lack of reliable
information about local prices is a key impediment to our understanding about a range of
significant issues, including real wages, wealth inequality, consumption, and market
integration. As is the case in many fields of history, historical records of prices are less readily
available and accessible for earlier periods, as well as for medium-sized and small towns, than
they are for larger centres and more recent periods. The capacity of inheritance inventories –
which are available for several Ottoman towns and survive for long, continuous periods – to
bridge this significant gap in the historical record has failed to attract sufficient attention. By
presenting new quantitative evidence on the potential and limitations of the prices recorded
in inheritance inventories, this chapter demonstrates that these documents do, in fact, offer a
reliable alternative source for Ottoman prices.
The relation between inventory valuations in the early modern European and
American inventories and market prices has been subject to much speculation in the relevant
scholarly literature. Recent studies have shown that these are generally consistent and closely
related to the conventional prices of the time (Overton 2000). The use of these sources in the
field of price history facilitated the discovery of completely new findings and enabled
historians to shed further light on the fundamental changes that emerged in the nature of the
early modern economies. Shammas’ (1994) path-breaking work revealing the decline in textile
prices in pre-industrial England and America and Overton’s (2000) extensive study providing
long-term agricultural, industrial, and consumer price indices for England are only some of the
works based on these sources.
24 This paper has been published in Historical Methods: A Journal of Quantitative and Interdisciplinary
History 49(3): 132-144, 2016.
Chapter 2-Ottoman inheritance inventory as a source for price history 61
Like European and American probate inventories, Ottoman inheritance inventories list
the items composing a portion of the deceased’s estate along with their unit valuations. The
idea of using these valuations for constructing price series is not novel. Some studies in the
past (Barkan 1966; Çizakça 1978; Establet and Pascual 1994; Öztürk 1987, 1988, 1992) have
employed the valuations of certain items in the inventories to offer an insight into conventional
prices for specific places, usually for short periods of time. However, the majority of work on
Ottoman prices has focused on other types of sources. Moreover, none of the existing studies
have offered quantitative evidence on the relationship between the inventory valuations they
utilise and market prices even though the significance of the issue was recognized in one of
the earliest studies on Ottoman inheritance inventories (Barkan 1966).
Whether inventory valuations reflect market prices and whether they are reliable are
also decisive questions for inventory studies that explore wealth and wealth-related
questions25. Doubts regarding the accuracy of valuations are often cited among the limitations
and problems associated with deploying inventories for quantitative analysis. Like the wealth
bias in these sources, or the likelihood of omissions from estates, unrealistic or erratic
valuations of inventory items can lead the historian to distorted conclusions, as the total estate
values analysed in such studies are computed from a sum of the total values of the deceased's
movable and immovable possessions, the cash they owned and the debts owing to him. If
valuations did not reflect market or auction prices, or if they were substantially higher or lower
than the conventional prices of the time, then the total estate value would not reflect the real
wealth of the deceased. In such a situation, research on the evolution or distribution of wealth
in a certain region, or wealth accumulation, can be seriously affected.
The question is also critical for studies on the differences in the composition of the
wealth of estate owners from divergent social groups and from different periods, as inaccurate
or inconsistent valuations might distort the share of the value of consumption and capital
25 For some inventory studies on wealth, wealth accumulation, wealth distribution and inequality in the
Ottoman Empire see Berker and Ergene (2008); Canbakal (2007, 2009, 2008-2012); Coşgel and Ergene
(2011); Establet, Pascual and Raymond (1994); Özdeğer (1988); Todorov (1983).
goods within the whole estate 26 . Hence, an evaluation of the consistency of inventory
valuations will contribute extensively to the growing range of research that relies on these
sources.
The basic process through which inventory items were valued was first described
clearly by Barkan (1966), who showed that possessions were either sold at public auction by
an auctioneer, with the prices of the transactions being recorded in the inventory, or that they
were appraised by experts appointed by the court ( ehl-i hibre). However, recent work has
revealed that the processes were more complicated, as in certain cases, heirs purchased the
items from the estate or shared them among themselves without publicly attending auctions
(Gradeva 2005). Perhaps the same applied for creditors.
Two contradicting views exist in the literature on the credibility of the valuations in
Ottoman inventories. The first view argues that court officials generally tended to overvalue
an estate, since the court's own fee was indexed to this valuation (Barkan 1966; Cezar 1977;
Ergene 2002; Seng 1991). As Barkan (1966) showed, decrees were sent by the administrative
centre to local judges warning them against artificially inflating the values of estates.
Nevertheless, the existence of decrees does not, ipso facto, prove that the manipulation of the
valuations was a common practice. The historical documents tend to over-represent the
violations of the rules, as there would be no need to take action in cases where the rule is
obeyed. According to Gradeva (2005) and Establet and Pascual (1994), any desire of the court
to manipulate the valuations would have been constrained by the presence of local people,
heirs and creditors, as it would probably end in an unfair assignation of shares, whilst the heirs
and creditors would resist biased valuations. The discussion remains open, however, and
empirical studies are required for us to acquire a precise idea about the accuracy of the
valuations.
Recently, Bozkurt (2011) offered a new argument in favour of the accuracy of inventory
valuations. He suggests that the valuations in the inventories could be manipulated only when
they were appraised prices, since it was the transaction prices that were directly recorded in
26 For studies on the composition of estates, see Establet and Pascual (2000, 2002); Rafeq (2012); Wilkins
(2010).
the inventory when part or all of the estate was sold at an auction. Considering that individuals
typically came to the court to register a deceased's estate due to situations that involved either
conflicts (among heirs and creditors/debtors, or among the heirs themselves), or the existence
of minors among the heirs, he argues that it is likely that most of the estates were sold at an
auction and hence reflected actual market prices. Bozkurt (2011) supports his argument by
demonstrating that in 90 percent of the 792 estates in his sample, the estates were sold at an
auction27. However, he also acknowledges that this situation might have been particular to
Istanbul and/or to the period under study. Especially in small towns where the market economy
was underdeveloped, estates could have been shared directly by the heirs and creditors
instead of being sold.
This paper provides further evidence about the accuracy of inventory valuations by
testing their reliability in the light of empirical evidence from primary and secondary sources.
Further, this study undertakes a comparison between inventory valuations and price data from
independent sources, investigating whether the valuations are internally consistent.
The paper is organized as follows: In the first section, the potential of inheritance
inventories for expanding the field of Ottoman price history and the advantages of employing
these alternative sources are highlighted. This is followed by an investigation of various types
of items reported in the inventories. Agricultural, consumer and capital goods, for which prices
can be compiled from these documents, are listed. Subsequently, some alternative uses of
inventory valuations are presented and the accuracy of the relationship between market prices
and inventory valuations are tested.
1. Advantages of inheritance inventories as a source for price history
The scarcity of price data is the main reason why most Ottoman price studies are
restricted in their chronological and geographical scope. Aside from Pamuk's (2000a) study
covering 500 years of prices in Istanbul and in some other cities of the empire, the general
tendency of price studies has been to focus on prices over short periods (not exceeding 50
27 It is possible to identify estates that were sold at the auction, as these included auctioneers’ fees
(dellaliye) or porters’ fees (hammaliye) that were deducted from the total value before it was shared
among the heirs.
years) in the big cities and important administrative centres.
To date, early modern and modern Ottoman price studies have generally relied on the
following four groups of sources:
 Waqf palace account books (vakıf muhasebe defterleri), which show prices paid
by pious foundations and their soup kitchens ( imaret) (Barkan 1975; Faroqhi
1984; Güran 2006; Pamuk 2000a; Orbay 2001, 2007a, 2007b);
 Palace kitchen account books, which reflect the food prices in the palace
purchases (Barkan 1964; Pamuk 2000a);
 Ceiling-price listings (narh), which include state-controlled prices for basic items.
These listings were promulgated as separate registers or were placed among
court records (Kütükoğlu 1978, 1983; Özlü 2006; Öztürk 1987, 1988, 1992;
Sahillioğlu 1967);
 Wholesale prices of the Commodity Exchange (for the nineteenth century)

(Pamuk 2000a).
Among these sources, Waqf account books have a particular prominence. In Islamic
law, these institutions are designed to be permanent institutions whose modi operandi are
described in detail in the deeds that establish them. These sources allow us to continuously
observe current market prices over time. The use of books that belong to a number of different
pious foundations established in the same city and in the same period allows us to check their
consistency and reliability (Orbay 2008), and also to substitute these sources with each other
when data is partially missing. Barkan (1975) presented the first price index relying on waqf
account books to elucidate the sixteenth century “price revolution in the Ottoman Empire.”
Subsequently, Faroqhi (1984) published sixteenth- and seventeenth-century price data for
Konya, a mid-sized Anatolian city, in her study on Konya waqfs. Waqf account books also
constitute the main source of Pamuk's (2000a) study.
Two other sources have also been important: tax registers and private expenditure
books. For proportionally taxed products like grains, enumerators had to specify a price to
convert physical quantities to nominal values in order to calculate the total tax revenue in each
village (Coşgel 2006). In his study, Coşgel (2006) makes use of these sources for exploring
agricultural productivity in the sixteenth-century Ottoman Empire and gives grain prices for a
number of localities. The extent to which these prices reflect market prices remains a critical
question, although this is beyond the scope of this paper. However, considering that tax
registers are among the most available and accessible historical sources for Ottoman history,
the opportunity they provide for price history should be addressed by further studies. The
prices reported in private expenditure books recording the purchases for the households of
Ottoman statesmen have been studied by Pamuk (2000a), who concluded that data drawn
from these sources are not particularly helpful, since they include both wholesale and retail
prices and, hence, are not standard.
Compared to these widely used sources, inheritance inventories possess several

advantages as an alternative source for Ottoman price studies. One of these is related to the
availability and accessibility of these documents. These legal documents were prepared by
Ottoman courts through an effectively standard procedure across the empire, yet they were
recorded and compiled on an urban basis, reflecting the local economy and society. The easy
availability and accessibility of inheritance inventories for numerous medium-sized and small
towns located in the more remote areas of the empire, as well as for big towns in the core
regions, makes them a unique source of comprehensive local price data. For most of the cities
where they survive, inventories can offer regular and continuous information over centuries,
an important requirement for long-term price studies.
In comparison to the other main sources used in Ottoman price studies, inheritance
inventories have two other advantages. First, the prices indicated in the inventories can be
assumed to reflect consumer prices, whereas most other surviving price data comes from
institutions. Institutions are likely to have paid lower prices in the market than those most
individual consumers, as their purchases were conducted regularly and in large quantities.
Overton (2000) argues that in the British case, the difference between prices in the institutional
records and general retail prices could be quite large. Second, and more importantly,
inventories cover a wide range of commodities that were excluded from the existing series
(Overton 2000). Allen et al. (2004) suggest that price histories have been biased towards over-
using food prices and under-using other prices for consumer and capital goods. A pertinent
reason for this is a lack of sources, including information about these latter items. In this regard,
inventories that encapsulated a wide range of manufactured goods can be of crucial benefit
for an effective study of the topic.
2. Goods in the inventories
Inventories describes the deceased's patrimony in detail: Buildings (houses, shops,

watermills), vineyards, trees and crops, livestock, personal and household goods, stores,
commercial goods as well as debts owed to the deceased. Each item is usually recorded with
its worth.
Estate owners’ movable goods can be grouped under three categories: agricultural,
consumer, and capital goods. In a significant number of estates, agricultural goods that were
stored for consumption or commercial purposes, including growing crops on the field, were
reported in units of measurement, allowing a unit valuation to be calculated (see Table 1). The
types of agricultural goods appearing in the estates, and their frequency, substantially varied
from one region to another and over time. However, for a considerable number of
commodities in Table 1, continuous and long-term yearly price series can be constructed
based on inventory valuations.
The second group of commodities, consumption goods, can be categorized into two
sub-groups: household and personal consumer goods. Personal consumer goods included
mainly clothing, watches and leisure goods (books, tobacco pouches, hookas, and the like).
Since the clothing items that were reported were usually described as 'old' or 'torn' and had
little to no market value, their valuations are less easily deployed for price studies. Although
the adjectives 'old' and 'torn' were also applied to household goods (see Table 2), it is possible
to identify 'new' or 'unused,' as well as relatively more standard goods among this second
category. Furthermore, despite being second-hand, household goods possessed a visibly
higher market value than the clothing items. The third group, capital goods (see Table 3),
mainly concerned agricultural production. For this group, the differences between unused and
second-hand goods were probably less marked. In regions where sericulture and cotton
production was common, it is possible to find raw cotton and silk given in weight units among
inventories. Similarly, processed cotton and silk were frequently reported in certain locations.
Table 1- Goods reported in units in Ottoman inheritance inventories
Most frequently appearing Other food items Non-food items

food items
Wheat (bugday, hınta, kamh, Cooking oil (sadeyag) Fodder (Saman, alef)
kendüm) Olive oil (zeytinyağı) Firewood (odun, hateb)
Barley (arpa, sair) Coffee (kahve) Coal (kömür)
Cracked wheat (bulgur) Salt (tuz) Boll (Penbe kozası)
Bitter vetch (burçak) Sugar (seker) Cottonseed (penbe çekirdeği)
Chick-pea (nohud) Roasted chick-pea (leblebi) Tobacco (Tütün, duhan)
Lentil (mercimek) Almond (badem) Soap (sabun)
Corn, white corn (darı, ak Olive (zeytin) Wax, beeswax (mum,
darı) Onion (soğan) balmumu)
Rice (pirinç) Vinegar (sirke)
Broad bean (bakla) Grapes (üzüm)
Cowpea (börülce)
Flour (un, dakik)
Molasse (pekmez)
Honey (bal, asel)
Table 2- Household goods in inventories
Household linen Sheet, pillow case, cushion case, table cloth, curtain, towel
Mattress Mattress, thin mattress
Interior lighting implements Candlestick, cresset, oil lamp
Floor coverings Carpets, rugs, felts, rush mats, prayer rugs
Sanitation utensils Bath jug, bath bowl, bath tub, washtub, dipper, loincloth, etc.
Coffee utensils Coffee pot, coffee cup, saucer
Cookware Pot, cauldron, pan, pastry tray, etc.,
Cutlers and serviettes Plate, vessel, knives, forks, spoons, tray, compote bowl etc.
Heating utensil Brazier
Table 3- Capital goods reported in inventories
Plough (saban, çift alatı)

Plough iron (saban temuru)
Millstone (değirmen tasi)
Scales (kantar)
Balance (terazi)
Oxen and horse (öküz, at)
Sickle (orak)
Hatchet (balta)
Adze (keser)
Windlass (çıkrık)
Raw silk and cotton
Processed silk and cotton
3. Alternative uses of inventory valuations
To facilitate comparisons across time and space, price studies focus on intermediary
and standard commodities, rather than goods whose quality variations cannot be ascertained
(Allen et al. 2004). This is, however, a criterion of inclusion which is often quite hard to satisfy,
particularly when studying historical prices. For instance, one of the reasons historians avoid
positing prices for a wider range of manufactured goods, particularly those of textiles, is
because it is not possible to control for the changes in their quality (Pamuk 2000a). If we are
focused on absolute price levels, as is the case when comparing the prices of a particular
commodity across different localities, or if we are investigating the relationship of a price series
with another variable series expressed in absolute terms, as when constructing a standard of
living index, then ensuring the goods remain as homogenous as possible is of crucial
importance. In this regard, most of the items for which we can find inventory valuations,
particularly consumer goods, are not of great use for conventional price analysis.
However, price analysis does not always concern absolute levels, and when appraising
relative prices and price trends, valuations which were not in perfect accord with the market
prices can be utilized by applying alternative methodologies. As Overton (2000: 123) states, "It
is important to realize that absolute price levels are less important in historical analysis than
relative prices. What is of most significance, therefore, is not whether inventory valuations are
identical to the actual prices the items concerned would fetch when sold, but whether the
trend of inventory prices follow the sale prices.”
Price trends and relative prices can be constructed for many inventory items that
cannot be included in conventional price analysis. Inheritance inventory valuations can be
employed for constructing price indices for consumer and capital goods to investigate price
movements or the evolution of relative prices. For instance, the question of whether the
relative prices of consumer goods declined, leading to a change in consumer preferences in
the seventeenth- and eighteenth-centuries Ottoman Empire, can be explored by producing a
basket of consumer goods from the inventories and comparing this basket to wheat and barley
prices. In this vein, Overton (2000: 21-2) uses British inventories to confirm a rise in the price
of capital goods relative to consumption goods, arguing that “the increased prices of such
items as ploughs, harrows and spinning wheels suggest improvements in their production and
possibly, therefore, technological change leading to increased production efficiency.” Similarly,
price relatives can be imputed for domestic and foreign goods to investigate their relative
competitiveness in the domestic market. Çizakça (1978) makes use of valuations in inheritance
inventories to follow the evolution of the price ratio of raw silk to silk cloth between 1550 and
1650 in Bursa, a major centre of silk cloth production, concluding that the sharp increase in
the relative price of raw silk demonstrates the decline of the Ottoman silk industry vis-à-vis its
European counterpart.
Inventory valuations can equally be used to follow inflation in a certain locality or to

create a local deflator. Rothenberg (1979) relies mainly on prices from probate inventories in
constructing a price index for rural Massachusetts for the period 1750-1855 and uses the
frequency of the appearance of goods in inventories in weighing the consumer basket.
Similarly, Harris (1996) observes inflation in early America through valuations in these sources.
On the other hand, the lack of a deflator that can be used to observe the evolution of the real
wealth in eighteenth-century Maryland led Carr and Walsh (1980) to impute a deflator based
on these valuations.
In studies on Ottoman inheritance inventories and more generally in the field of

Ottoman economic history, a deflator based on Pamuk's (2000a) consumer price index is now
often used to transform nominal values into real terms. However, the widespread utilization of
this deflator can present problems. First, Pamuk’s index belongs to Istanbul, the capital city.
Nevertheless, available evidence suggests that prices moved together in the Ottoman realm,
except for short-term fluctuations that were particular to a town or a region. So, this first point
is probably unimportant in the middle and long term. More seriously, however, the deflator is
based on a commodity basket dominated by agricultural goods (80 percent) rather than
manufactured goods (20 percent). This might be deceptive when applied to deflate the value
of consumer and capital goods, real estate and cash, rather than agricultural goods. The
distortion would be even greater if the relative prices changed substantially over the period
under study.
4. Testing inventory valuations
The best method of evaluating the accuracy of inventory valuations is to compare

them with prices from independently created sources. This methodology has been extensively
applied to British inventories. For example, Cox and Cox (1985) and Overton (2000) compare
unit valuations in British inventories for wheat, barley, pewter and horses with historical price
data from other sources. They demonstrate that in all cases, prices and valuations fluctuated
in concert, even though they did not perfectly overlap. In this paper, I explore four versions of
this type of consistency test. I first evaluate Ottoman inventories by carrying out a series of
comparisons with several independent price series for wheat and copper, drawn from other
sources. I then examine the internal consistency of prices for linked goods, in this case two
textile products. I also look at whether inventory valuations reflected quality differences. In the
final section, I apply a separate method, developed by Overton, to test the internal consistency
of inventory prices by using the status of the deceased, for goods where external series are
not available.
4.1. Comparing inventory price series and independent series: Wheat
I begin with a comparison of wheat prices from Barkan’s (1966) sample of inheritance
inventories in the Ottoman town of Edirne and from waqf account books gathered by Orbay
(2012). Edirne, located in the European portion of the empire, was a medium-sized town with
a population of around 30,000 inhabitants in the late sixteenth century (Gökbilgin 1994).
Barkan (1966) uses inheritance inventories belonging to the Ottoman state elite in Edirne
between 1545 and 1659, whilst Orbay (2012) extracts his data from the account books of the
Beyazıd II waqf for the period 1597-1640.
Figure 1- Edirne wheat prices from inheritance inventories and waqf account books
(per kile28 in akçe)
Sources: Waqf account book prices: Orbay (2012)

Inventory prices: Barkan (1966)
As can be seen in Figure 1, although waqf account book prices were, in general, slightly
higher than inventory prices, they were quite close in absolute terms. The product moment
correlation coefficient between inventory valuations and waqf account book prices is +0.90,
which suggests that they moved in concert – that is, relative changes in inventory valuations
followed relative changes in waqf prices. Another way to analyse the closeness between two
series is to look at the yearly price differentials as a percentage of the average price. The
average price difference for the period 1597-1640 was 21 percent of the mean price, which
remains within a reasonable margin of error in any study on historical prices.
To offer a guide to what this level of variance means, we can compare two wheat price
series (1596-1620) based on the account books of different waqfs established in the city of
Konya (see Figure 2). The first series comes from the Selim II waqf (Orbay 2007a) and the
second from the Mevlana Celaleddin-i Rumi waqf (Orbay 2007a).
28 A kile is a local unit of volume used in measuring grains. One kile of Istanbul and Edirne wheat were
equal to, respectively, 35 L (Pamuk 2000a) and 28.8 L (Barkan 1964) of wheat.
Figure 2- Edirne wheat prices from the account books of the Selim II Waqf and the
Mevlana Celaleddin-i Rumi waqf (per kile in akçe)
Sources: Orbay (2007a)
Despite originating from identical types of sources, the two series produce a lower
correlation coefficient (+0.58), and the average yearly price difference between them is higher
in terms of the average wheat price (0.32 percent). This comparison points to two conclusions.
First, it can be assumed that wheat prices in inheritance inventories reflected market prices, at
least as much as waqf account books do. Second, the account book prices and inventory prices
are comparable.
The same exercise can be conducted for Istanbul wheat prices over a longer period
(Figure 3). The inventory prices come from Üsküdar court registers belonging to the period
1656-1812, whereas the waqf prices are drawn from Pamuk’s study (2000a), which incorporates
data from more than 6,000 account books belonging to several waqfs in the capital.
Figure 3- Istanbul wheat prices from waqf account books and inventories
(per kile in akçe)
Sources: Waqf account book prices: Pamuk (2000a)

Inventory prices: Data collected by the author. See Primary Sources section
for details.
In the case of Istanbul, the correlation between the two series is stronger, with a
correlation coefficient of +0.96. On average, the yearly differences equalled 21 percent of the
mean wheat price. The absolute values overlapped closely until the beginning of the
nineteenth century. From this date until the end of the period, the waqf account book prices
were largely higher than the inventory valuations. The gap might be a consequence of the fact
that the grain prices in Üsküdar, located on the Anatolian side of the capital city, were slightly
different to those on the European side.
4.2. Comparing inventory price series and independent series: Copper
Figure 4 depicts the varying average price of copper for Trabzon, a port town in the
Black Sea region, and Amsterdam over a period of 15 years 29. The Trabzon copper price series
is collected from inventories and computed based on the valuations of copper kitchenware
29 Amsterdam is chosen for comparison due to practical reasons. To the best of the author’s knowledge,
Amsterdam copper price series are the only continuous and long-term series available. A wide market
for copper existed between European states as early as the late Middle Ages. This essential war material
was also a long-established item of European trade with the Levant (Ayduz 2006). Prior to the sixteenth
century, copper came largely overland from Central Europe and was exported in Venetian ships (Ayduz
2006). During the eighteenth century, Ottomans imported copper from Europe (Zorlu 2008).
reported in weight units ( okka) in the inventories. These valuations reflect the second-hand
prices of manufactured copper products. Still, it can be safely assumed that these prices were
linked to the contemporaneous price of processed copper in the market. The Amsterdam
copper prices are obtained from van Zanden and van Tielhof (2009) 30, who acquired them from
Amsterdam exchange records. Despite the fact that these mirror the prices of processed
copper, and not that of manufactured products, the co-movement between two series is
evident with a correlation coefficient of +0.68.
Figure 4- Trabzon and Amsterdam copper prices

(silver gr/kg for 15-year moving averages)
Sources: Amsterdam: van Zanden and van Tielhof (2009)

Trabzon: Data collected by the author. See Primary Sources section for details.
This co-movement in prices might derive from a number of reasons. Shared

improvements in copper mining and processing technology, or simultaneously changing
demand conditions in both markets, might explain the situation. Such co-movement has also
been posited as an indicator of market integration by several historians. Regardless of the
reason, the simultaneous ups and downs in copper prices in both markets indicate the
credibility of the copper price series extracted from the inventories.
30 Amsterdam copper prices are reported in guilders per pounds in van Zanden and van Tielhof (2009).
For conversion into silver grams, I used data supplied by de Vries, Smiths, and van Riel (see
http://www.iisg.nl/hpw/).
The close relation between inventory prices and the price of copper in the market is
also attested by the match between the price series and the reported consequences of political
intervention in the Ottoman market. In a letter to the central government dated 1793, a
revenue farmer at the Tokat smelter complains that the copper prices dramatically dropped
for that year due to the state's refusal to purchase the yearly expected amount (Beşirli 2004).
As this would suggest, the copper price that year, which is computed from inventory
valuations, was the lowest (4.76 silver gr/kg) observed from 1700 to 1840.
4.3. Testing the internal consistency of inventory prices for cognate goods
One of the advantages of inventories as a source – their inclusion of products for

which we have few other records of prices – has the unfortunate effect that we cannot conduct
comparisons with independent sources for many of the commodities we wish to analyse. We
can still test the reliability of inventory prices for consumer and manufactured goods in other
ways, however. Here, I compare the evolution of the prices of two distinct goods which both
utilise the same raw material. If inventory prices are close to market prices, we would expect
to see the prices of these goods move in correlation.
Figure 5 displays the index constructed based on valuations of beledi31 mattresses and
duvets in Manisa inventories. The index reflects the average unit valuations of these items,
collected from inventories at five-year intervals from 1700 to 1820. 1720 is used as the base
period. The co-movement of the two series is confirmed by the high correlation coefficient
(+0.95).
31 Beledi is a local cotton textile which occasionally included silk. Since the early sixteenth century,
textiles were produced in Urla and Tire on the Aegean coast of Anatolia. Until the late eighteenth
century, individuals from all segments of society used mattresses and duvets covered with beledi.
Figure 5- Beledi mattress and beledi duvet index (in silver gr)
Sources: Data collected by the author. See Primary Sources section for details.
Notes: Base period is 1720
Although we lack detailed price data for beledi in particular, and for cotton textiles in
general, we can compare these general trends with some observations in the existing literature.
For instance, we know that the raw cotton prices doubled in the course of the eighteenth
century, whilst the price of yarn more than tripled (Faroqhi 2011). This is also the case depicted
by inventory valuations. In 1805, the average of two indices (mattresses and duvets) was 2.99,
3.46 times the average figure from 1700 (0.87). Historical sources (Faroqhi 2011) also suggest
that from the late eighteenth century onwards, cotton textile prices skyrocketed, which would
confirm the trends demonstrated by the index.
4.4. Testing the internal consistency of inventory prices with quality differences and the
wealth of the testator
But even the method just outlined for comparing connected goods is only viable in a
few cases. As such, Overton (2000) proposes a further alternative method: determining
whether inventory valuations are consistent with respect to the quality of the items being
valued. This can be achieved either by exploring whether they reflected the adjective used to
describe them and whether the value of goods in the inventories was correlated with wealth.
Initially, I compare the unit valuations of cotton and silk sheets in Üsküdar inventories.
It is a well-known fact that silk was an expensive cloth, associated with luxury and reserved for
the affluent, whereas cotton was a more affordable and ordinary textile, accessible by Ottoman
individuals with limited means. Table 4, which looks at the mean unit valuations of sheets,
demonstrates that both in the early and the late eighteenth centuries, sheets made of cotton
as recorded by the appraiser were substantially cheaper than those made of silk. As such, we
can assume that the higher the quality, the higher the valuation.
Table 4- Mean unit valuations of silk and cotton sheets in Üsküdar (in silver gr)
1700 1790
Silk 35.64 67.59
Cotton 8.84 12.63
Sources: Data collected by the author. See Primary Sources section for details.
Second, I test the inventory valuations against the wealth of the testator. Individuals
of higher economic status were expected to own goods of better quality, and thus of higher
value, than poorer individuals. A positive correlation between wealth and the average value of
specific goods is thus evidence of the dependability of inventory valuations 32. To look at this
relation, I examined Üsküdar inventories from the early (1695-1705) and late eighteenth
century (1785-1795). I identified several categories of basic household goods that frequently
appeared in the estates of all wealth groups: cushions, sheets, duvets, mattresses, and rugs.
For each period, the unit valuations of these household goods are separately
regressed against the wealth of the estate owner.
𝑉𝑎𝑙𝑢𝑎𝑡𝑖𝑜𝑛𝑐𝑢𝑠ℎ𝑖𝑜𝑛 =∝ + 𝛽𝐶𝑃𝐼 + 𝛿𝐿𝑁𝑊𝐸𝐴𝐿𝑇𝐻 + 𝑢 (1)
Unit valuations are expressed in silver grams. I controlled for changes in the general
level of prices using Pamuk’s (2000a) consumer price index. This was required because each
period covered a 10-year interval. LNWEALTH is the natural logarithm of wealth in constant
32 It might be suggested that total estate value and item valuations are endogenous, as estates
involving more highly valued items would be of higher value themselves. However, as shown by several
studies on the composition of wealth, real estate and debts possess the greatest share within the estate,
and the weight of consumer or agricultural goods is too small to affect the total estate value.
akçe, obtained by deflating the nominal total estate value by the CPI.
Table 5- Inventory valuations and wealth in Üsküdar – OLS regression results
MATTRES
1695-1705 CUSHION SHEET DUVET RUG
S
CONSTANT -14.89 -63.71 -55.96*** -89.35 -34.09*
(17.08) (28.72) (14.91) (43.14) (18.67)
CPI -9.99 -2.35 31.90*** 63.5 22.22*
(13.28) (19.95) (11.92) (30.92) (13.08)
LNWEALTH 6.80*** 10.83*** 6.72*** 4.93** 5.15***
(1.35) (2.86) (1.18) (2.78) (1.41)
N 71 43 91 43 41
R2 0.25 0.23 0.3 0.18 0.23
1785-1795
CONSTANT -22.5 -17.5 -62.56 -89.35** -10.83
(36.42) (24.78) (41.56) (43.14) (58.44)
CPI 5.91 -3.62 28.92 63.50** -12.99
(28.45) (15.41) (28.86) (30.92) (35.16)
LNWEALTH 5.28*** 5.03*** 6.30*** 4.93* 7.88***
(2.4) (1.22) (1.65) (2.78) (2.8)
N 48 29 47 43 20
R2 0.1 0.35 0.25 0.18 0.11
Sources: Analysis is based on data collected by the author. See Primary Sources section for
details.
Notes: *, **, *** indicates significance at the 10 percent, 5 percent and 1 percent levels
respectively.
As shown in Table 5, the results again accord neatly with our expectations about
consumption and wealth. Household items owned by wealthier individuals were generally
valued more highly than those owned by less wealthy people. In all cases, the value of the
household goods was positively associated with the total wealth. The coefficient of LNWEALTH
is positive and significant at the 1 percent level, except in the case of mattresses.
Table 6- Mean valuations of copper kitchenware (in silver gr/kg)

by quartiles of total wealth in Manisa, Trabzon and Istanbul
Manisa Trabzon Istanbul

1700-20 1740-60 1760-80 1700-20 1740-60 1760-80 1700-20 1740-60 1760-80
Min-LQ 117.98 124.07 164.21 79.19 93.41 148.12 98.24 121.44 156.33
LQ-Median 121.60 126.86 160.66 79.71 94.52 146.15 101.00 120.00 159.04
Median-UQ 121.31 125.17 163.52 79.40 91.65 148.54 99.38 118.63 160.03
UQ-Max 118.54 123.33 167.58 77.68 90.98 150.00 97.05 124.95 159.40
N 42 52 49 41 45 55 48 51 55
details.
Notes: LQ: Lower quartile UQ: Upper quartile
Table 7- Mean valuations of wheat (in silver gr/HL)

by quartiles of total wealth in Manisa and Ayntab
Ayntab Manisa
1680-1700 1700-1720 1720-1740 1700-1720 1720-1740

Min-LQ 205.05 227.91 477.165 223.56 548.22
LQ-Median 194.51 220.25 520.52 227.45 533.235
Median-UQ 217.11 300.82 529.13 205.085 501.28
UQ-Max 269.62 320.86 482.2 202.735 496.145
N 54 58 56 43 47
details.
Notes: LQ: Lower quartile UQ: Upper quartile
We can take further reassurance about the credibility of inventory prices from the
finding that for wheat and copper, no such association between unit price and wealth is
observed in inheritance inventories from Manisa, Trabzon, Istanbul and Ayntab (See Table 6
and 7). This reflects the fact that these are more or less standard goods, for which the variations
in the quality are limited. Furthermore, the agricultural goods we encounter in the inventories
were not primarily food stored for consumption purposes, but mainly commercial goods
owned by individuals of various occupations related to agriculture or trade 33. Therefore, quality
variances were more or less equally distributed within the whole sample. It is also likely that in
valuing copper kitchenware reported in weight units, a standard price reflecting the market
value was applied.
In addition to providing evidence of the accuracy of the valuations in the inventories,

these results also imply that the total estate values recorded in the inventories were likely a
true reflection of the deceased's wealth. Of course, these results do not rule out the possibility
of omissions in the inventories. But the positive correlation between consumer good prices
and estate values suggests that these provide us with a valid proxy for distinguishing the
wealthy from the poor. The results of studies demonstrating an association between total
estate values and official and social titles (status) also support this argument 34.
Furthermore, given that the quality of goods varies with wealth, it is necessary to take
the distribution of wealth groups into account when constructing price series and making
comparisons across regions and over time (Overton 2000). A possible and easy way to control
for sample bias and to minimize the variance of quality of goods is to exclude outliers and
include goods only from middling group's estates, which are assumed to be of 'average'
quality. The middling group can be defined on the basis of wealth brackets previously
established35 or can be identified as a cross-section of the distribution of the whole sample.
5. Problems and limitations
An important challenge that any price historian interested in pre-modern prices must
confront is a method of determining local units and converting them into grams of silver per
metric unit. This conversion process requires two sets of information: the silver content of
moneys in circulation and the metric equivalents of local weights and measures. Silver content
33 The fact that grains were rarely observable in Trabzon and Istanbul inventories also confirms this
point. These two cities were not producers but consumers of grain.
34 This does not imply that the distribution of the estates is representative of the whole population.
35 For an example, see Todorov (1983). The author divides the populations of eighteenth-century Vidin,
Sofia and Ruse into groups of 'property owners.' Those who had assets of 500 qurush and below
belonged to the lower stratum.
and exchange rates of various currencies in circulation in the Ottoman realm from the early
fourteenth century to the early twentieth century are provided in detail by Pamuk (1994, 2000a,
2001).
However, converting local measurement units into metric units is often a complex and
tedious task, since local units showed a great variety, and varied not only from region to region
but also from date to date and according to the commodity measured. Whenever available,
information from secondary sources can be used 36 or the conversion rates can be derived
directly from the document because, for their own accounting purposes, Ottoman officials
sometimes made a note of the equivalent of the local measurement unit in Istanbul units,
which can be converted into metric units with ease. In a case where no information on the
local units is accessible, Coşgel (2006) computed his own methods to compute the conversion
rates37.
Periodicals, dictionaries, encyclopaedias, guides and travel books that are available in
European archives are also beneficial sources for Ottoman metrology. These publications
aimed at providing information to European merchants traveling and trading all around the
world and embraced a wide range of local knowledge from climate to language or customs in
a particular country. Local units of measurement and exchange rates of local currencies were
often included. This notwithstanding, the information available in these sources merely
concerns port cities, trade centres or towns that were integrated into the world economy
through commercial agriculture.
Another problem in calculating agricultural prices –particularly grain prices – derives
36 For conversion rates for some local Ottoman measurement units, see Hinz (1955); İnalcık (1983);
Kürkman (1991); Kürkman and Işın (2003).
37 Coşgel (2006: 12) compares local prices of a particular commodity with the known standard price, in
order to calculate a conversion rate: “For proportionally taxed products like grains, enumerators had to
specify a price to convert physical quantities to nominal values in order to calculate the total tax revenue
in each village. In cases of unknown conversion rates for a district, (the enumerators) compared the
price of wheat specified in the registers of this district with the (standard) prices used in the registers of
other districts for the same time period to determine whether the enumerators were likely to have used
a standard kile for measurement. If the price appeared too low or too high compared to known standard
prices, I relied on comparable prices and conversion rates observed in the region to specify a rate of
conversion for this district.”
from agricultural price movements. In the early-modern period, food prices fluctuated
substantially even within periods as short as a couple of months. These fluctuations occurred
in parallel with those of supply, which were due to harvest conditions, wars and other causes
that could affect the agricultural production and transportation of products from the
countryside to urban areas (Orbay 2008). As Orbay's (2008) observations of prices from waqf
account books in the early seventeenth century suggested, grain prices might vary by up to
70 percent within a period as short as six months (Orbay 2008).
Hence, to be able to arrive at accurate yearly prices for cereals, we need a large
number of observations that are more or less continuously available for the period under
study. Also, these observations need to be distributed equally among the months to reflect
the seasonal ups and downs of the prices, which is quite difficult to achieve when working with
inheritance inventories. Although the month of death can be established in some regional
traditions of records keeping, in several cases this valuable information is missing in the
inventories. Although this might affect the accuracy of short-term data, it can be argued that
this is trivial for medium- and long-term analysis. It is reasonable to assume that the
distribution of inventories according to the months is random, meaning there is no
introduction of systematic biases that might substantially distort the price levels of trends in
the long run and at an aggregate level.
It should be also recalled that valuations in the inventories for consumer and capital
goods reflect the prices of second-hand goods and hence are lower than the market prices of
their unused equivalents. For Overton (2000: 141) this is a benefit of these sources rather than
a weakness: “... the Rogers and Beveridge series are for new goods, whereas most transactions
in early modern England probably involved the sale of second-hand goods, which the
inventory prices represent.” As evidence of the commonality of the use of second-hand goods
in the early-modern Ottoman Empire, Bozkurt (2011) shows the frequency of deeds of sale
among court records. These documents feature information about transactions of a range of
second-hand items among people of different social status, including the elites in a significant
number of cases. Moreover, it was again a common practice to confiscate valuable items from
among the estates of the Ottoman state elite for rendering to the sultan himself or for use in
the imperial palace.
As we have seen, care needs to be taken when using inventory prices to ensure that
the characteristics of each type of good are recognised and addressed. By shaping the quality
of commodities that individuals possessed, the wealth of the deceased had different effects
on the reported prices of different groups of possessions. In the case of agricultural goods,
quality variation (and with it price variation) is a negligible problem. At the very least, this
problem does not affect inheritance inventories more than it concerns other types of historical
sources. However, for consumer and capital goods, quality variation is a serious issue. Most of
the goods that were included in the estates were final goods whose quality varied substantially.
Goods reported in the inventories were of differing sizes, were made of various materials and
possessed varying degrees of obsolescence. This imposes a serious limitation on the utility of
valuations of most goods included in the inventories. These valuations cannot be posited to
reflect absolute price levels, although they remain useful for following price trends if
approached with caution.
There are a number of solutions that can be suggested for controlling for the
differences in quality of these goods. First, a selection criterion can be imposed. Some types
of consumer and capital goods show greater heterogeneity. For instance, most clothing items
are less suitable for inclusion in a price study than some household goods, such as sheets,
which are relatively more homogenous. This second type of goods can be allowed to stand for
all of a particular type. In a similar vein, goods that were described as 'old,' 'torn,' and the like
can be excluded, whilst consumer and capital goods can be categorized into groups according
to the terms describing the material they consist of and their size. Second, the value of these
goods is highly correlated with wealth, indicating a quality differentiation of the goods based
on the economic status of their owner. Therefore, including goods merely from middling
group's inventories will significantly decrease the variations in quality. Last but not least,
utilizing a large sample size will help to eliminate the unsystematic biases in the sample.
6. Conclusion
The accuracy of the relationship between the valuations in Ottoman inheritance

inventories and market prices has long been a source of uncertainty for historians, both in
terms of the reliability of the quantitative analysis based on these sources, and in terms of the
potential they possess for expanding the field of Ottoman price history. Two arguments have
generally been advanced to support the argument that these valuations reflect the market
prices of the time: first, the appraisal process was strictly regulated by Islamic law and
conducted in the presence of witnesses; and second, in most cases, inventory valuations
represented the sale prices for the goods constituting a portion of the estate. However, the
accuracy of the prices themselves has never been examined directly.
As we have seen, a detailed examination of the quality of inventory prices suggests

strongly that they were based on market prices. When comparing inventory valuations with
independent sources of prices, we find that the price series were closely related. For products
where no external source of prices is available, the prices of cognate goods were closely
connected, as we would anticipate if they were driven by market prices, and inventory
valuations are consistent with our expectations of the quality of the items being valued, based
on the wealth of their owners. Based on these results, it can now safely be assumed that the
prices contained in Ottoman inheritance inventories can be reliably employed for historical
research.
85
SECTION 2
MARKETS
86
CHAPTER 3
MEASURING MARKET INTEGRATION IN THE EASTERN
MEDITERRANEAN (1660-1840)
“The profoundest distances are never geographical”

The Magus, John Fowles
This paper investigates, the trends in the regional, national, and inter-national
integration across the wheat markets of the Eastern Mediterranean from 1660 to 1840, a region
that stood at the centre of the pre-Colombian world economy.
Regional patterns of specialisation facilitated by integrating commodity markets are

often associated with the growth path of pre-industrial Europe. Alongside rising agricultural
productivity, urbanization, and the shift in consumption patterns, the emergence and
development of trading networks and the rise in domestic and international trade are cited
among the structural changes that paved the way for the Industrial Revolution (De Vries and
van der Woude 1997; Epstein 2000; Persson 1988, 1999; Studer 2008; Pamuk and Özmucur
2007).
The empirical research that has been carried out on early-modern European markets
reveals contradictory findings in regards to a long-term and continent-wide integration prior
to the nineteenth century 38 . Still, most economic historians agree that parts of Europe
gradually – despite the occasional setback – became more integrated within themselves and
with other parts of the continent during this period, and that this allowed the establishment
of a geographical division of labour, and a more efficient organization of production (Özmucur
38 Özmucur and Pamuk (2007), Bateman (2011), and Federico (2008) found no evidence of a permanent
trend of market integration across Europe prior to the nineteenth century. However, relying on the most
extensive price data employed until now, Chilosi et al. (2011) detected widespread European integration
between 1620 and 1789, although they recognised that a European grain market did not truly emerge
before the nineteenth century. According to Gonzales, and Garcia-Hiernaux, and Guerrero (2012) too,
from the early eighteenth century onwards, price dispersion in wheat gradually decreased within Europe.
Chapter 3-Measuring market integration in the Eastern Mediterranean 87
and Pamuk 2007; Chilosi et al. 2011; Studer 2008). There was a particular trend towards greater
integration between the sixteenth and nineteenth centuries in Northern and Northwestern
Europe (Granger and Elliot 1967; Jacks 2004; van Tielhof 2002; van Bochove 2008; Gonzales,
Garcia-Hiernaux, and Guerrero 2012)– a tendency that has also been confirmed by studies on
urban trading networks. These suggest that a number of regionally based urban networks
emerged during the sixteenth century without a continent-wide structure and were integrated
into a single European network and hierarchy during the course of the seventeenth century,
in parallel with the rise of the Dutch trading system (Ringrose 1998). Did Eastern Mediterranean
commodity markets experience a similar trend towards integration during the same period, or
was this a cause and a sign of the rise of Northwestern European?
A zone of intensified commercial and cultural exchange, this part of the Mediterranean
was shared by the Ottoman Empire and the Venetian Republic until the fall of the latter at the
end of the eighteenth century. The pair were the two most important protagonists in long-
distance trade between East and West before they lost ground following the great discoveries
culminating in Columbus’ voyage to America and the new route to Asia. During the
seventeenth and eighteenth centuries, the emergence of the Atlantic powers (the Netherlands,
France, and England) gradually pushed them into a peripheral position (Pezzolo 2013).
Recent research has highlighted that this was far from a dramatic and irreversible crisis
that emerged immediately. Rather, the “waning” of the Ottomans and the Venetians was
gradual, slow and with reversals, occurring only from the late sixteenth century onwards (Tabak
2010). Indeed, Venice, thanks to foreign shipping, remained a very busy port, perhaps the
busiest in the whole Mediterranean even as late as the seventeenth century (Pezzolo 2013).
According to Braunstein (1971), the Adriatic ports, despite their discords, were very closely
bound by the extant business, navigation, and monetary networks in the eighteenth century.
Similarly, ample evidence from the sixteenth century points to the existence of a relatively well-
developed urban network, encompassing the coastal regions of the Ottoman Mediterranean
and their hinterlands (Panzac 1992). In the seventeenth and eighteenth centuries, the liveliness
of the Ottoman maritime world was maintained, as indicated by the rise of two important port
cities, İzmir and Salonika (Eldem, Goffman, and Masters 1999).
At the outset, this study poses once more, in the context of the Eastern Mediterranean,
a central question that has occupied economic historians in the last few decades: did pre-
modern commodity markets witness a long-term trend towards integration in spite of
stagnant transportation technology? This paper adopts a nuanced view by hypothesizing a
possible regional differentiation in terms of market expansion. In the light of previous research,
one would expect that early-modern integration processes fostered by institutional change
primarily concerned short-distance markets and coastal regions, whereas long-term markets
and landlocked regions witnessed little impact from this trend (Studer 2008). This hypothesis
will be tested here.
A second question concerns the relative “decline” of the Mediterranean powers.

Although by the last quarter of the seventeenth century, this part of the Mediterranean had
lost its centrality in world economic flows (Tabak 2010), international and regional trade
continued to be a major source of revenue for both the Venetians and the Ottomans (Pezzolo
2013). Merchant ships were still operating intensely in the Adriatic, Ionian, and Aegean seas. I
inquire as to whether the initial resistance and subsequent decline of the two protagonists in
the Oriental commerce was accompanied by the emergence of a regional economy in the
waters of the Eastern Mediterranean during the early-modern era. Putting it differently, I aim
to understand whether the Eastern Mediterranean gained a regional existence as the Venetian
Republic and the Ottoman Empire were downgraded to regional powers.
The empirical findings of this study demonstrate that Eastern Mediterranean did not
move towards an integrated regional market system during this period, a strong trend of
integration in the Adriatic basin notwithstanding. On the eve of the first wave of globalization,
domestic wheat markets in the Ottoman Empire were no better integrated than they had been
two centuries previously. Neither the extended sample that incorporates coastal and
landlocked regions and short-distance and long-distance markets, nor the one restricted to
the littorals of the Ottoman Mediterranean, produce evidence to support a continuous and
sustainable declining trend in trading costs prior to the nineteenth century. The results also
reveal that at equal distances, the cost of trading wheat within the Adriatic market was smaller
than that of trading within the Ottoman Mediterranean market, suggesting that this region
was better integrated compared to the latter in 1720-1840.
Finally, a multivariate regression analysis on price differentials across the Ottoman

grain-trading network centred on the capital posited wars, occupations, and plagues as causes
of fragmentation in the wheat markets and shows that, following controls for asymmetrical
shocks, there were no significant differences in the extent of market integration between 1748
and 1774, a period of intense regulation and state interference, and the post-1774 period,
which, according to the revisionist scholarship (Ağır 2013) witnessed the liberalisation of
Ottoman grain policy.
Although the Adriatic region also falls within the scope of this study, the primary focus
is on the Ottoman Empire. In what follows, I first briefly outline the trade environment in the
pre-nineteenth-century Ottoman realm, focusing on transportation infrastructure, economic
policy, wars, and asymmetrical shocks engendered by droughts and plague epidemics.
Subsequently, I introduce and discuss the data and the methodology employed in measuring
market integration in the Eastern Mediterranean. Finally, I present the results and the
robustness checks.
1. Trade environment in the Ottoman Empire
This section surveys the general trade conditions in the Ottoman Empire between
mid-seventeenth to mid-nineteenth centuries, focusing upon the central factors: political
structure, transportation network, economic policy, wars, droughts, and plague.
At its apogee, the Ottoman realm as a “vast domestic economic entity,” represented
one of the world economies as defined by Braudel (Panzac 1992: 202). The economic
interdependence of provinces was seen as essential not only for the welfare of the subjects
but also for the political cohesion of the empire (Panzac 1992). Therefore, the state played a
crucial role in the establishment and promotion of the complex network of production and
consumption that encompassed the empire’s territories on three continents.
Faroqhi (1979) speaks of a relatively well-developed Ottoman trading network

supported by ‘precocious imperial centralisation’ in the fifteenth and sixteenth centuries. The
proliferation of fairs and local markets in the second half of the sixteenth century, particularly
in settlements located on trade routes, indicates a lively internal trade in manufactured as well
as agricultural goods. However, the general trade environment in the empire substantially
changed with the turn of the seventeenth century, with political instability, rising insecurity,
falling agricultural output, and declining population. As a result, the first decades of the
century witnessed a sharp drop in the volume of domestic trade, the disappearance of local
markets, and a significant level of market fragmentation. Our account of Ottoman commodity
markets starts in the 1660s, following an era of recovery from this crisis in production and
trade (Faroqhi 1994).
As McGowan (1994) states, trade within and between Ottoman provinces in the
seventeenth and eighteenth centuries has been far less well documented and understood
compared to foreign trade during the same period. Still, some general traits can be
highlighted. First, from 1700 to 1850, in volume and value, domestic trade vastly exceeded
international trade, despite the expansion of the latter, especially after 1750 (Quataert 2005;
McGowan 1994; Panzac 1992). According to Panzac’s (1992: 202) estimates, at the end of the
eighteenth century, the total value of the transactions between Europe and the Ottoman
Empire was 110 million to 120 million livres tournois, whereas the estimated value of maritime
trade between Ottoman provinces was almost twice this figure, at 180 million to 200 million
livres tournois.
Second, the provisioning network established by the state in the fifteenth century, and
encompassing the Ottoman territories with the capital at its centre, was successfully
maintained until territorial disintegration occurred in the nineteenth century (Yıldırım 2003).
As far as grain policy was concerned, a shift in the traditional supply patterns occurred only
after the loss of Ottoman dominance over the seaborne traffic in the Black Sea in the late
eighteenth century. The effective loss of the Romanian principalities, the main source of grain
for Istanbul, imposed the need to develop greater self-sufficiency based on domestic
Anatolian grain production (Panzac 1992; Yıldırım 2003). Before this date, however, the
“imperial division of labour,” (Yıldırım 2003: 266) as more or less defined during the apogee of
the empire, continued to function.
Third, maritime transportation remained essential to the domestic relations of the

empire. The Mediterranean not only brought the Ottomans into direct contact with Venice and
Western Europe, but also connected the empire’s lands on three continents. The maritime
route was “the most direct, the quickest, and the cheapest route for the capital and for a good
number of Ottoman provinces,” (Panzac 1992: 195). To a large extent, geographical position,
and access to the sea accounted for regional differences in terms of participation in inter-
regional trade. In fact, close commercial relations and economic interdependence mostly
concerned the coastal regions (the Black Sea coast of Bulgaria, the Romanian principalities,
Thrace, Macedonia, Thessaly, Morea, Western Anatolia, the Egyptian delta, and the Arabian
coasts of the Red Sea), whereas landlocked areas of Anatolia were dependent on their close
vicinity or even forced to rely on self-sufficiency and isolation until well into the nineteenth
century.
1.1 Political structure
In his monumental work, Freedom and Growth, Epstein (2000) stresses the political
conditions of pre-industrial economic growth, and posits jurisdictional centralisation under the
late Medieval and early-modern rulers as the ultimate source of the rise of more efficient and
better-integrated markets in Europe. In new institutionalist economics, political regimes are
defined as ways to facilitate cooperation for mutual advantage. In line with this literature,
Epstein suggests that pre-modern centralised states were more efficient than decentralised
ones because they suffered from fewer multiple coordination failures (Epstein 2000). Although
decentralisation, centralisation, and the changing relationship between the imperial centre and
the periphery in 1700-1850 have been subject to ample research in Ottoman historiography
(Inalcik 1977; Piterberg 1990; Salzmann 1999, 2004; Hathaway 2002; Khoury 1990, 2002; Smiley
2008), these processes have rarely been addressed in relation to the evolution of the overall
economy, and more particularly, the trade environment.
The institutional centralism of the Ottoman state in the classical period was gradually
succeeded by a fragmented political structure over the course of the seventeenth and
eighteenth centuries. These two centuries witnessed fiscal and administrative decentralization
and the rise of provincial elites who held local power and who operated autonomously from
the capital. In almost all parts of the empire, the central state became visibly less important
(Khoury 2006).
The prosperity brought about by the expansion of regional economies during the first
half of the eighteenth century allowed for the development of a modus vivendi between local
elites in the provinces and the Ottoman government. However, by the second half of the
century these relationships, which depended on mutual recognition and common interests
between the centre and the provincial elite, began to be undermined. In the power vacuum
engendered by the Russo-Ottoman war of 1768-1774, the centre found itself fighting a
number of rebellions by semi-autonomous provincial power-holders in the Balkans and the
Middle East (McGowan 1994).
Following an era of political crisis and of rising insecurity in the countryside in the late-
eighteenth century, the central Ottoman state began to transform itself into a more powerful,
more rational, and more specialized structure, with the modernization and centralization
efforts of the nineteenth century. The capital employed its expanding bureaucracy and military
to weaken and destroy its domestic rivals, and battled against diverse groups such as the
Janissaries, guilds, tribes, religious authorities, and provincial notables to gain political control.
The imperial reform edicts of 1839 and 1856 finally ended the political fragmentation and
rivalry of the last two centuries and ushered a centralized bureaucracy into power (Quataert
1994).
Recent scholarship stresses that the seventeenth- and eighteenth-century process of

Ottoman decentralization should be considered as a viable political strategy developed in
response to the changing circumstances of the early modern world rather than as institutional
decay (Barkey 1994; Salzmann 1993; Khoury 2006). From the perspective of trade however, the
fragmented polity of multiple provincial dynasts and notables should have made commerce
more difficult during the eighteenth century. Particularly, the power struggles between the
centre and the periphery, and among rival local notables in the second half of the eighteenth
century are likely to have acted as a factor that multiplied coordination failures, and hence,
increased the transaction costs of trade.
Despite their interest in trade, it was tax-farming and usury that remained the most
lucrative enterprises for the local power holders (Pamuk 2000b). Prior to the nineteenth
century, the fiscal regime and the decentralized Ottoman political structure did not create an
impetus for improvements to the institutional environment to foster trade, as merchants and
domestic producers never became powerful enough to exert sufficient pressure on the central
government to change or even modify its policies, as was the case in Europe (Pamuk 2000b).
Measures to promote domestic and international trade, such as attempts to standardize local
measures, weights and legal codes, and the establishment of gendarmerie units to provide
safety on trade routes were only implemented around the mid-nineteenth century under the
impact of Western influence.
1.2. Transportation network
A large portion of the regional and inter-regional trade in Ottoman Anatolia took
place via land routes, as most cities – including manufacturing towns such as Kayseri, Tokat,
Ankara, and Aleppo – were inland centres, while waterways, except for the Euphrates-Tigris
system, were not navigable on a regular basis. Transportation by camel caravans remained a
dominant feature in Anatolia even after the introduction of railways (Faroqhi 1982, 1994).
Three major caravan routes connected Istanbul to Damascus and Aleppo, as well as to Iran
(see Map 1). The most famous one was the diagonal route passing through Akşehir, Konya,
Tarsus and Adana, and from there to Damascus. Iranian trade followed the Tokat-Erzincan-
Erzurum itinerary. Diyarbekir, a major entrepôt for Iranian goods located at the south of this
route, was linked to the major itinerary by circuitous routes. Two alternative caravan routes
existed between Istanbul and Tokat. The first passed through Eskişehir and Ankara, while the
second traversed the hill chains of Northern Anatolia, with Amasya being an important stop
(Faroqhi 1994).
While sea transportation remained of limited importance with respect to Anatolia, it

played a major role in connecting the Balkans, Western Anatolia, the western shores of the
Black Sea, the Egyptian delta, and the Arabian coasts of the Red Sea. As Map 1 shows, the
Istanbul-Smyrna-Alexandria route was the main axis of the network that intersected with the
other routes to Macedonia, Crete, Syria, and the Maghreb. Beyond Chios and the Aegean Sea,
an east-west route that linked the Maghribian ports to those of Syria ran along the African
coast and crossed the north-south axis at Alexandria (Panzac 1992).
Map 1- Ottoman maritime network
Source: Panzac (1992: 136)
Seaborne traffic was particularly important for food supply. Due to the high costs of
overland transportation, long travel periods, and inadequate food preservation techniques,
only durable and high-cost low-bulk goods, such as textiles and other manufactured wares,
were traded inter-regionally. The transportation of foodstuffs over long distances was quite
rare (Quataert 2005). The price of wheat almost doubled over 100 kilometres (Grehan 2007).
Nevertheless, relatively lower costs in maritime transportation and shorter journey times
allowed several foodstuffs, particularly grains, to be exchanged between coastal areas (Faroqhi
1994). In this respect, it would be reasonable to expect a higher degree of integration in
markets with easy access to sea.
We can obtain an indication of the much lower burden of transport costs from data
given by Aynural (2001: 25-26), which details the freight costs between Istanbul and several
Ottoman ports in the early nineteenth century. Based on this information, the transportation
cost of a hectolitre of wheat is estimated per 100 nautical miles with an OLS regression,
employing transportation cost in silver grams (TC) and sea distance in nautical miles
(DISTANCE) between indicated ports. The results are presented in Table 1.
Table 1- OLS regression results- Transportation costs and distance
Dependent variable TC
CONSTANT 1.953***
(-0.355)
DISTANCE 0.994***
(-0.078)
N 28
Adjusted R2 0.86
Source: Freight costs are taken from Aynural (2001).

Notes: *, **, *** indicates significance at the 10 percent, 5 percent, and 1 percent levels
respectively.
The estimated transportation costs are calculated with the following formula:
𝑇𝐶 = 1.953 + 0.994 ∗ 𝐷𝐼𝑆𝑇𝐴𝑁𝐶𝐸
The scale of the difference between sea and land travel becomes visible when we
compare the cost of transportation. The estimated cost of transporting by sea, a hectolitre of
wheat from, for instance, Galatz port (located at a distance of 320 miles from Istanbul) in the
initial decades of the nineteenth century was 5.13 silver grams, a figure equalling 10 percent
of the market price of wheat in Istanbul and 21 percent of the price gap between the two
locations.39
As in other parts of the world, the Ottoman Empire did not experience any significant
improvements in transportation and communication technologies in the early-modern period.
Still, even in the absence of technological innovation, improvements in transportation
39 It should be noted that grain was transported from other Ottoman provinces to the imperial capital
by a combination of state-owned and leased ships (Murphey 1987). In the latter case, the state played
a dominant role in price arbitration and in determining a fair freight charge. For this reason, evidence
of freight charges in shipments to the capital might be a misleading indicator of charges elsewhere, as
these were probably higher in shipments between other parts of the empire.
conditions could occur due to changes in the institutional framework supporting the
transportation system (state attempts to maintain road paving, blaze new trails, construct
bridges and channels, etc.). Public works could help reduce costs by shortening the average
journey times between two locations. Despite the importance that the Ottoman state gave to
facilitating the transportation of both pilgrims and merchants (Halaçoğlu 2002), guard station
registers demonstrate that such attempts had little impact on the duration times of journeys.
Three guard station registers dating to around the mid-seventeenth, mid-eighteenth and mid-
nineteenth centuries show that journey durations by horse for the three main itineraries in
Anatolia were only slightly shorter at the end of the period (see Table 2).
Table 2- Journey durations by horse
1643 1766 1843
Istanbul-Aleppo 257h 253h 248h

(Right branch)
Istanbul-Mosul 372h 369h 349h

(Middle branch)
Istanbul-Erzurum 276h 276h 266h

(Left branch)
Source: Halaçoğlu 2002; Çetin 2013; Bozkurt 1966
Thus, it would be reasonable to conclude that before the introduction of steamships

(1828) and railways (1858), there was little or no change in overland transportation costs, while
only limited improvements were recorded in sea transport (Faroqhi 1994).
1.3. Provisionalism, interventionism, and trade policy
Besides high transportation costs, the most substantial impediments to the rise of
efficient and integrated markets in the pre-modern world were state monopolies, internal
tariffs, export/import prohibitions, and price control mechanisms (Özmucur and Pamuk 2007).
The market for grain was particularly heavily regulated, and was the last to be liberalized due
to its strategic importance in the Ottoman realm, as was the case elsewhere in the pre-modern
world (Chilosi et al. 2011).
Throughout the period under study, grain was subject to close state monitoring and
control because of its status as the essential foodstuff for the Ottoman population. The
Ottoman state regulated and was involved in several phases of the production, procurement,
transportation, storage, and distribution of grain in order to provide an adequate and
continuous supply of grain to the capital and other major urban centres, to provision the army
during times of fighting, and to transfer grain to regions struggling with shortages that
occurred due to climatic factors and other external shocks (Yıldırım 2003).
The principal grain-growing regions for the imperial centre included the plains of
Thrace; the Danubian basin; the steppe region from the Dobruja to the Don river; Macedonia;
the plain of Thessaly; Western Anatolia; and the Egyptian delta (İnalcik 1994). The western
Black Sea coasts (Romanian and Bulgarian lands) were traditionally the most important grain
supplier for Istanbul (İnalcık 1994).
The organisation of this network included many institutions, policies, and local and
central actors. Export bans, price controls, licence requirements to purchase, transport, and sell
grains, as well as forced purchases, were the main tools of the Ottomans’ grain policy during
this period (Güçer 1949). The form, tools, and the degree of state intervention in grain markets
varied throughout time in response to several factors affecting demand and supply conditions,
and reflected the evolution of the political, administrative, and financial structure of the
Ottoman state.
Until the beginning of the eighteenth century, exporting grain was prohibited in
principle, although smuggling could never be totally suppressed (Yıldırım 2003). During the
eighteenth century, the Ottoman state adopted a more tolerant attitude towards wheat
exports (and the export of several other agricultural goods), and export inhibitions began to
be temporary and exceptional. Yet, under the unfavourable conditions of the second half of
the eighteenth century, export inhibitions again multiplied (Pamuk 2004).
In the course of the eighteenth and nineteenth centuries, two major changes occurred
in supply patterns. With the expansion of the Russian presence following the Treaty of Küçük
Kaynarca in 1774, the Black Sea was opened to foreign trade, and the Romanian provinces
gradually escaped from direct Ottoman control. As an increasing share of the grain produced
in the region was oriented towards western Mediterranean markets, the Western Anatolia and
Mediterranean coasts became more important for provisioning Istanbul in the last quarter of
the eighteenth century. Still, the region continued to supply grain for the capital, albeit at lower
volumes than previously (Güran 1986). The second shift in provisioning patterns took place in
the second half of the nineteenth century with the introduction of railways. Within a short
period of time, Anatolia became an important factor in Istanbul’s grain supply. Istanbul millers
began buying rail-shipped wheat from inner Anatolia instead of waterborne grain from Russia,
Bulgaria and Romania. Thereafter, under more favourable circumstances, the Anatolian railway
provided over 90 percent of all wheat delivered to the capital for local consumption (Quataert
2005).
Changes in the patterns of supply were one aspect of the transformations that
Ottoman grain policy underwent during the eighteenth and nineteenth centuries. Questions
as to when Ottoman trade and grain policy began to liberalise, and whether there was any
radical change in the provisionist policy prior to the nineteenth century remain crucial, yet
contentious, issues (Ağır 2013; Yıldırım 2003; Quataert 1994; Güran 1986).
In traditional accounts of Ottoman economic policy, the 1838 Anglo-Ottoman

convention imposing the elimination of state monopolies and barriers to domestic and foreign
trade marked a radical change, and the government’s efforts to direct the economy diminished
gradually after this date (Toprak 1992; Owen 1992; Özveren 2001). However, revisionist
scholars have rejected 1838 as a break point, claiming that the convention did not radically
change the state’s existing approach to the economy, but continued the government’s earlier
commitment to liberalise grain and trade policies, which had appeared at the end of the
eighteenth century (Quataert 1994, 2005; Ağır 2013).
As for grain policy, two opposing views exist. The traditional view holds that the state’s
role in provisioning Istanbul with grain did not show any sign of significant deviation from
traditional patterns until well into the nineteenth century (Yıldırım 2003; Güran 1986), while
the revisionist approach suggests that starting with the 1780s, Ottoman policy-makers
adopted a more liberal attitude towards price-formation in grain markets and deliberated on
removing pre-emptive privileges (Ağır 2013). According to Ağır (2013), the relaxation of price
controls on wholesale grain and the consideration of a system based on a network of private
merchants as a new alternative signalled a shift in grain policy that was rooted in a new concern
for the state of agricultural production.
The establishment of the Grain Administration ( Zahire Nezareti) in 1793 is interpreted

in two different ways in these opposing analyses. For Yıldırım (2003: 253), by establishing a
central institution to control and finance a greater share of the grain trade in the empire,
governors aimed at “bringing the commodity chain of grain under the full sway of the state,”
with this testifying that the realm of grain provisioning remained “the most tightly regulated
aspect of the Ottoman economy until the dawn of modern times.” However, Ağır (2011: 3) sees
the new institution as a reflection of the “policy shift towards a more-centralized-yet-flexible

use of regulatory tools in the Ottoman grain trade.” She highlights that an imperial decree in
the same year ordered the purchase of grain for the capital to be set at the market price. By
terminating price controls and the quota system, this decision was an indicator of the
liberalisation of Ottoman grain policy in the late eighteenth century.
The issue of the transformation in grain policy has not been discussed with reference
to quantitative evidence until now. Most of the evidence that both sides cite is qualitative in
nature and does not allow us to grasp how the grain policy (and its change) impacted on the
Ottoman markets for grain. Without a shift in focus from the economic mindset of the
government and state policy to the question of what occurred in grain markets in actuality, it
is difficult to comment on the degree of state interference, its effectiveness, as well as the real
impact of changes over time.
In terms of internal tariffs and other taxes on domestic grain trade, 1760 appears to
be a turning point. In this year, in addition to regular internal tariffs, a new tax on the domestic
trade of grains (3 percent ad valorem) was introduced. Furthermore, between 1760 and 1821,
the state increasingly expanded internal tariff zones and established new ones (Genç 2000). 40
40 The internal Ottoman customs system was not a network that covered all Ottoman territory but was
rather organised in the form of circular areas around certain big trade nodes, such as Istanbul, İzmir,
and Salonika.
The number of internal tariff zones exceeded 100 by 1801. Reaching its zenith in 1821, the
internal customs regime gradually shrank in the following period. In 1843, newly established
internal custom zones were abolished, although older ones continued to exist (Genç 2000;
Kütükoğlu 1996). Last but not least, the government implemented several trade regulations
for the provisioning of the capital and other big cities during the war years of the late
eighteenth century (Quataert 2005). All these factors can be considered as institutional barriers
to trade that considerably increased trade costs.
1.4. Wars
Several studies on domestic and international market integration in the early-modern

period have emphasised the negative impact of wars (Federico 208; Chilosi et al. 2011; Jacks
2006; Bateman 2011). Wars were disruptive to commerce, making it dangerous to move goods
across and sometimes within borders. It should be noted that in the Ottoman case, wars had
a number of other direct and indirect impacts on commodity markets. On the one hand,
Ottoman soldiers marching to and returning from the fronts were a source of insecurity for
the villages close to the campaign area. Furthermore, the vacuum of authority due to warfare
led to a significant increase in bandit activity in the provinces (Quataert 2005).
Warfare also counteracted the forces of market integration indirectly through

government policies. The requirements of warfare were an important motive for state
intervention in markets. By putting immense pressure on state finances and seriously
disturbing productive activities, wars forced the Ottoman government to give weight to the
priorities of the state over those of the society and economy (Genç 1984). It can be assumed
that state purchases at official and/or market prices ( mübayaa), the imposition of extraordinary
taxes in kind or in cash, the inhibition of trade in strategic goods and regulations for
provisioning the big cities such as price ceilings ( narh) – all of which were implemented more
frequently during times of war – had important effects on Ottoman interregional markets
(Genç 2000). Given the direct and indirect effects of the wars, one would expect prices to have
diverged across markets during times of conflict.
Figure 1- Number of war years per decade, 1660-1840 (European and Eastern fronts)
10
9
8
7
6
5
4
3
2
1
0
Number of war years on the European front

Number of war years on the Eastern front
Figure 1 depicts the number of years per decade from 1660 to 1840 in which the
Ottoman state fought on the European and Eastern fronts. As can be seen from the figure, the
long conflicts of the late seventeenth and early eighteenth century were followed by a period
of relative peace. The approximate quarter century between 1746 and 1768 – between the end
of the Ottoman-Iranian war and the outbreak of the Russo-Ottoman war of 1768-1774 – is
generally accepted as both an era of peace and order and a period of rising agricultural and
industrial output, urban wealth, and population growth. The outbreak of the Russo-Ottoman
war, however, ushered in a new era of long and exhausting armed clashes. The early century
also witnessed the Serbian and Bosnian Revolts and the Greek Independence War.
Given the primacy of maritime links in long-distance trade, naval conflicts that
threatened the security and continuity of commerce in the Eastern Mediterranean had a
particularly decisive effect on commodity market integration in the region, both at national
and international levels. In order to understand the general trade environment, it is useful to
identify the periods of peace and hostility in these waters specifically, and they are summarized
in Figure 2. It appears that from the last quarter of the sixteenth century to the mid-
seventeenth century, peaceful relations between Venice and the Ottoman Empire provided a
favourable ground for the development of trade within the Eastern Mediterranean. According
to several diplomatic sources, the coexistence between the two polities that began in 1573
allowed closer collaboration between authorities and merchants from both sides, particularly
in the frontier areas such as Dalmatia (Fusaro 2015). It seems likely that higher levels of security
in the Eastern Mediterranean and lower transaction costs contributed to market extension at
domestic and international levels. In 1646, the outbreak of the Cretan War terminated the era
of peace. The war turned the Eastern Mediterranean, particularly the Adriatic, Ionian, and
Aegean Seas into a theatre of battle. The Cretan war lasted almost a quarter century, and
included numerous naval engagements and raids around the Aegean Sea, with Dalmatia also
becoming the scene of sporadic naval operations. This war was followed by the Morean War
(1684-1699), and the Ottoman-Venetian War of 1714-1718.
Figure 2- Number of years of naval hostility per 25 years in the Eastern Mediterranean
25
20
15
10
Years of naval hostility/25 years
The 1714-1718 war was not only the last conflict between the Ottoman State and the
Venetian Republic, but also the beginning of a second period of peace in the Eastern
Mediterranean that would last until the infiltration of the Russian navy into the Aegean in 1768.
Within this period, the Aegean, Ionian, and Adriatic Seas witnessed no naval hostility of
significant scale involving either regional or external powers. Even in the last quarter of the
nineteenth century, which was marked by intensified fighting for the Ottoman Empire, the
Eastern Mediterranean was largely spared from naval clashes, as most of the Ottomans’ wars
were fought in the European lands of the empire and the Black Sea. But the Ottoman
Mediterranean once again became a zone of military clashes at the turn of the nineteenth
century when Napoleon’s fleet entered the area and the Greek War of Independence erupted.
In the first four decades of the century, the Ottoman armada fought several naval wars against
British, French, and Russian vessels in its domestic waters.
1.5. Drought
It is often assumed that under the conditions of pre-modern agricultural technology,

harvests were strongly affected by extreme weather conditions (Orbay 2008). While not always
resulting in crop failures, droughts led to bad harvests and, thus, higher agricultural prices.
Even when several markets were simultaneously hit by the same climatic shock, droughts
tended to produce a divergence in prices, while the level of price rises depended on several
factors related to conditions of supply and demand.
Drought and famine appear in several Ottoman narrative sources and are addressed
in numerous urban historical studies (Kılıç 2002; Aydıner 2006; Baer 1977). However, these
studies examine individual incidents in particular towns and regions, and no scholar has
produced a study of drought and famine from a broad temporal and spatial perspective.
Though limited in number, recent studies on the climatic history of the Eastern Mediterranean
provide a more comprehensive understanding of the linkages between non-human and
human histories.
Figure 3-Number of drought years per decade
10
9
8
7
6
5
4
3
2
1
0
Number of drought years per decade
Sources: Akkemik et al. (2006) (Anatolia); Xoplaki, Maheras, and Lutherbacher (2001), Gounaris
(2009) (South-eastern Europe); Grehan (2007) (Damascus); Baer (1977) (Egypt)
Figure 3 depicts the number of years per decade in which severe drought occurred in
one or more regions included in this study (Anatolia, Ottoman Europe, Egypt, and Damascus).
During the eighteenth century, droughts occurred in one or more of these regions on an
average of 3.4 years every decade. The 1690s and 1710s particularly witnessed severe episodes
of drought, while the figure was also above average for the 1740s, 1750s, and 1780s.
1.6. Plague
The Black Death, a common cause of mass deaths in Europe for centuries, disappeared
at the end of the seventeenth century in the west of the continent before soon also
disappearing from Central Europe. Nevertheless, it remained a major scourge in the Balkans,
Anatolia and the Middle East until the late nineteenth century (Panzac 1985). In the
seventeenth and eighteenth centuries, it continued to be one of the most frequent and most
serious natural calamities to affect the Ottoman Empire. The port cities of Istanbul, Salonika,
Izmir, and Alexandria, as well as big trade centres such as Aleppo and Cairo, were among the
towns that were repeatedly and severely affected by the epidemic (Quataert 2005). Studying
waves of plague in the six Ottoman cities mentioned above, Panzac (1985) suggests that the
cities witnessed a severe plague epidemic on an average of every 23 years from 1700 to 1850.
Panzac estimates that each serious wave caused a death toll of about 20 percent of the
population in each town it affected (Panzac 1985).
Plague epidemics were important factors behind the interruption of intercity relations.
Recognising that human interaction was responsible for the spread of plague, European
polities took measures from the late fourteenth century on to prevent the disease by isolating
infected individuals and limiting the mobility of people, animals and goods across regions
(Boerner and Severgnini 2011). It is highly likely that the efforts to restrict the spread of plague
through the interruption of trade and travel and the suspension of intercity relations had
resulted in the disintegration of the European markets. Studying the case of the port of Izmir,
the major export centre of the empire during the eighteenth century, Panzac (1985) argues
that the city experienced economic difficulties due to the cessation of foreign trade during
times of serious plague epidemics. European merchant ships avoided the port, and the city
significantly suffered from the suspension of relations with the wider world. Still, the degree
to which plague isolated Ottoman towns from each other and led to the disintegration of
domestic markets remains unclear. Did plague epidemics exert a similar impact on domestic
trade? Did seasonal caravans continue to operate across Anatolia? How was sea-based trade
in the Mediterranean affected?
Plague should have caused a disintegration in commodity markets not only when it
led to the suspension of interregional relations, but also when it created asymmetrical shocks
in local markets which influenced local supply and demand conditions and, thus, prices. The
decline in population or the deceleration of population growth rates due to rising mortality
can be identified as the main source of plummeting prices in European markets in the late
fourteenth century. On the supply side, however, plague generated a counteracting force
precipitating labour scarcity and higher wages. The total impact of plague on prices can be
said to be an outcome of the interaction between the demographic impact of the disease and
local conditions.
Figure 4-Number of epidemic years per decade
10
9
8
7
6
5
4
3
2
1
0
Number of plague years per decade
Sources: Ayalon (2008), Panzac (1985); Varlık (2011); Kostis (1995); Grehan (2007)
Figure 4 depicts per decade, the number of years in which plague was seen in 10 or
more Ottoman cities, from 1660 to 1840. Throughout the period, plague was seen in 10 or
more Ottoman cities at least one year in every decade, and geographically widespread plague
epidemics occurred most frequently in the late seventeenth and late eighteenth centuries.
This section has overviewed the general trade environment in the Ottoman Empire
with reference to the political structure, transportation costs, state policy, wars, climatic
conditions, and plague epidemics. Like elsewhere in the world, substantial reductions in
transportation costs were not observable in the Ottoman realm prior to the introduction of
steamships and railroads. In the first decades of the nineteenth century, transporting grain
between Ottoman ports was still tremendously expensive. As for the political conditions, from
1660 to 1800, the empire experienced decentralisation and political fragmentation and
diverged from the political trajectory of its European counterparts, which were evolving
towards centralised states. Political rivalry between the centre and local elites and the rising
insecurity in the countryside in the late eighteenth century, should have led to increased
transaction costs in regional and inter-regional wheat trade.
On the other hand, in the absence of a powerful merchant class, the institutional
improvements and policy-promoted changes that undergirded the rise of efficient and
integrated markets in pre-modern Europe did not appear in the Ottoman realm before the
political and economic reforms of the mid-nineteenth century, as impediments such as internal
tariffs and state interference in commodity markets persisted throughout the era. As for
asymmetrical shocks that might have affected price divergence, severe and frequent droughts
marked the late seventeenth and early eighteenth centuries, while plague epidemics affected
vast swathes of the empire in the late eighteenth century.
Overall, the picture of trade conditions depicted here is in accordance with the
previous accounts of the overall Ottoman economy from the mid-seventeenth to the mid-
nineteenth centuries. Although it is difficult to detect the direction of the causality, a positive
association is visible between economic expansion and a favourable trade environment. The
second third of the eighteenth century, an era of relative peace, stability, and economic growth
for the empire, featured conditions favourable to market development. In contrast, the late
seventeenth and early eighteenth centuries were marked by frequent wars, while the late
eighteenth century was an era of economic retraction and frequent state intervention in
markets; both periods were highly likely to have witnessed market fragmentation.
2. Data
Before assessing the levels and processes of market integration, we first need to look
at the price data at hand, which forms the quantitative base for the analyses in the following
section. As Studer (2008) has shown, in Europe, short-distance and long-distance markets and
coastal and landlocked regions followed a different trajectory in terms of market integration.
Therefore, at a first instance, I look at the geographical coverage of the price series, and briefly
present each of the markets with their basic geographical and commercial characteristics.
2.1. Geographical coverage
The price data employed to measure market integration in the Eastern Mediterranean
embraces Ottoman and Adriatic wheat markets.
The Ottoman Empire
Wheat prices used for the analysis come from 11 Ottoman towns (Manisa, Istanbul,
Bursa, Edirne, Cairo, Salonika, Patmos, Candia, Konya, Ayntab, and Damascus), and from three
regions in the Ottoman Europe, Wallachia, as well as the Eastern and Western Balkans. During
this period, the Ottomans reigned over a vast area stretching from the Balkans to Yemen, from
Crimea to the Mediterranean coasts of Africa. As can be seen in Map 2, the data provides a
good representation of the realm under Ottoman rule in the eighteenth century. The sample
includes small, medium-sized and big towns, coastal and landlocked regions, agricultural and
urban areas, self-sufficient towns and those that were dependent on their hinterland for a
supply of grain.
Map 2- The Eastern Mediterranean
The regions covered by the data can be divided into two groups. The first group
includes Istanbul, Manisa, Bursa, Edirne, Patmos, Candia, Wallachia, Salonika, the Eastern and
Western Balkans, and Cairo. All these locations were part of the trading network centred on
the capital, which connected the coastal regions of the Mediterranean, Aegean, and Black seas.
They had easy access by sea to the imperial capital and to other markets in the network. In
addition, these areas played a role in the grain provisioning for Istanbul, while they generally
also engaged in the trading of grain among themselves. The second group consists of
Damascus, Ayntab, and Konya, which are located in landlocked regions, and which had no
direct ties with the trading network incorporating the coastal areas and the capital city 41.
Group 1: Centres in the Maritime Trading Network
Manisa, the central town of the province of Saruhan, was located in the Gediz river
valley in the hinterland of İzmir, which developed into the primary port for Ottoman
agricultural exports from the seventeenth century. Since its conquest, the town was part of the
core provinces of the empire, and as such, it was integrated into the larger Ottoman system of
taxation, provisioning, and trade (Emecen 1989). Barkey and Rossem (1997) state that as early
as the fourteenth and fifteenth centuries, the Ottoman state had furthered the development
of a coherent trading network between Istanbul and Manisa. The town supplied the capital
with grain, and several other foodstuffs, shipped from Foçalar and İzmir. In parallel to its
incorporation into the world economy in the seventeenth and eighteenth centuries, Manisa
became one of the first Ottoman towns to engage in commercial agriculture, with the rise of
big farms oriented towards export products. How this affected the town’s crucial role within
the provisioning system is not certain. Goffman (2002) argues that as Europe made forays
along the Western Anatolian coast for grains, Istanbul increasingly struggled to control its own
delivery routes.
The first capital of the Ottoman Empire (1326-1402), and the most prominent
commercial centre in Western Anatolia, Bursa, was among the largest Ottoman cities, boasting
a population that varied between c. 30,000-75,000 from the late fifteenth century until the
middle of the nineteenth century (Canbakal 2012). It was a major centre of silk trade and
manufacturing and the last entrepôt on the Silk Road, where an interchange of commodities
from the East and the West took place (İnalcık 1994). The town’s position on the southern
coast of the Marmara Sea made it a keystone in the trading network that linked the southern
41 Unfortunately, our knowledge of the Ottoman internal tariff system is extremely limited. No
information available as to which of the markets included in this study were located in the same customs
region, and how the geographical coverage of these latter changed over time. Therefore, it was not
possible to group the markets according to their customs regions.
Marmara region and Istanbul. Bursa was easily accessible by sea from the imperial capital.
From its numerous docks, wheat, barley, rice, fruits, and vegetables produced in the region
were transported to the capital (Çiftçi 2004).
Candia was the capital city of Crete, the biggest Aegean island. Crete was conquered
by the Ottomans in 1669 after a long period of clashes, and was the last Ottoman conquest.
Balta (1997) defines Candia, the biggest town on the island, as one “that verged on self-
sufficiency, at least until the middle of the eighteenth century.” Combined with its hinterland,
the town supplied the western part of Crete and in good years it exported small amounts.
Patmos is a small island positioned very close to mainland Anatolia. The island was
controlled by the Ottoman Empire for many years, but it enjoyed certain privileges, mostly
related to tax-free trade by the monastery.
Located in the heart of Thrace on a fertile plain, Edirne, another ancient capital of the
Ottoman state, was an actor not only in the provisioning of Istanbul with grains, but also in
the provisioning of Ottoman troops during expeditions to Europe (Aksan 2013). Positioned on
Via Egnatia, the major trade route that linked the capital to the European lands of the empire,
the town was a densely populated central market by the standards of the day, and according
to several accounts, the city enjoyed a lively commercial life (Gökbilgin 1994). Grain from the
Edirne region was transported to Istanbul from the closest ports, Enez, Tekirdağ, and Ereğli
(Gökbilgin 1994; İnalcik 1994).
Istanbul received the largest portion of its provisions and raw materials from the
western Black Sea region (Wallachia, Moldavia, and Bulgarian coasts). This region remained
the major grain supplier of the capital until the nineteenth century. The Principality of
Wallachia, located to the north of the Danube and south of the Southern Carpathians, was a
tributary state of the Ottoman Empire. The voyvodes of Wallachia and Moldovia had to provide
a predetermined amount of grain for the Istanbul market on an annual basis (İnalcik 1994). In
the Danubian delta, Braila, Issacea, and Constanta developed as the main transit ports between
the Romanian lands and Istanbul. With the opening of the Black Sea to international trade and
shipping in the last quarter of the eighteenth century, the principality gradually escaped from
the Porte’s monopoly over its foreign trade, and oriented itself towards world markets.
Likewise, after the Treaty of Adrianople (1829), Wallachia and Moldavia gained “the full liberty
of trade for all the productions of their soil and of their industry.” This also led to a shift in the
grain policy of the Ottoman government, which turned towards its core lands out of
provisionist concerns.
In the Balkan peninsula, two price series are included in the database, from the east
and west of the Vidin-Matapan42 line. To the east of this line are Bulgarian lands. Grain and
other manufactured and agricultural goods from Bulgaria were shipped to the capital from the
ports of Varna and Burgas on the Western Black sea coasts. The region to the west of the line
encompasses Macedonia, Thessaly, and Illyria, which were connected to Western Anatolia and
Istanbul by the Aegean Sea. While the Black Sea was closed to non-Ottoman traders, and
legitimate or contraband trade with Europe was scarce until the late eighteenth century,
agricultural exports to Europe through smuggling in the western part of the peninsula could
never be fully inhibited by the government. As a result, the region established a commercial
relationship with Europe as early as the seventeenth century (Faroqhi 1994). Large commercial
farms emerged in the western Black Sea coast in the early seventeenth century, but it was only
in the later seventeenth and eighteenth centuries that they became visible in Macedonia and
Thessaly (Faroqhi 1994).
Salonika was one of the flourishing Ottoman port towns of the seventeenth and
eighteenth century, which according to Faroqhi (2001: 99), were “privileged points of entry
into the Ottoman commercial system.” Salonika’s hinterland produced grain, dried fruit and
some cotton, all of which were traded inter-regionally. Furthermore, Istanbul relied heavily on
Macedonia for grain supply, and one would have expected that most of these supplies were
shipped through Salonika (Faroqhi 2001). As its industries declined in the second half of the
seventeenth century onwards, domestic and foreign trade became increasingly important to
the economy of the town (Svoronos 1956). Faroqhi (2001) highlights a difference between the
Ottoman officials’ attitude towards eastern and western shores of the Aegean. She argues that
the trade in foodstuffs on the Balkan coast of the Aegean was somewhat less strictly controlled
by Ottoman officialdom than its Anatolian counterpart.
42 Cape Matapan is the southernmost point of the Peloponnese peninsula.

Soon after its conquest (1517), Cairo was organized into a regional centre for grain
supply, as an integral part of the Ottoman trading network. During the seventeenth century,
Cairo’s commerce with Anatolia, the Balkans and the Maghreb far outweighed its trade with
Europe, although the town remained a centre for international trade (particularly the transit
trade in coffee) even after the Portuguese opened up the ocean route to India (Faroqhi 1994).
Egypt’s supply radius stretched eastwards to the holy cities of the Hedjaz and northward to
the Anatolian ports of the Mediterranean coast and Istanbul. Egypt sent wheat and rice to the
Ottoman capital, thanks to the Nile and the Mediterranean, which facilitated the transport of
Egyptian grain and rice, which were exported from Damietta and Alexandria (İnalcik 1994).
According to figures cited by Murphey (1987), in the course of the seventeenth and eighteenth
centuries, annual grain shipments of 48,000 ardabs (3,341 tons) were sent to Mecca and
Medina and 20,000 ardabs (1,392 tons) to Istanbul for the Imperial Household. In later periods,
Egypt functioned as an emergency source of grains when the flow from Macedonia and the
Danubian region was interrupted, either due to natural causes or as the result of wartime
blockade (Murphey 1987).
Group 2: Landlocked Centres
Unlike the towns in the Balkans, the Danubian region, Western Anatolia, and the
Egyptian delta; Damascus, Ayntab, and Konya had no direct economic ties with the capital and
the trading network incorporating the coastal areas. The most important characteristic of these
three towns was their limited market access due to their geographical position in somewhat
landlocked regions. To access the wider world, they overwhelmingly relied on overland
transportation. The high cost of overland transportation dictated by the conditions of early-
modern technologies restricted the tradability of wheat – a bulky, low-value product – and
isolated these towns to a serious extent from other Ottoman grain markets.
Damascus was located in a large oasis fed by the Barada River in the Arab lands of the
Ottoman Empire. Despite the short distance to the Mediterranean shore, the town could not
easily reach the sea. Grehan (2007: 43-44) writes,
“The main roads fanned out along a north-south axis, threading between the desert
to the east and the Lebanese highlands to the west. Though the sea was in reality less
than fifty miles away, it seemed far more distant – about three to four days on the
road through difficult and hostile terrain where the authority of Ottoman governors
was no more than nominal…Communications with the coast did not begin to improve
until the mid-nineteenth century, as Syria was progressively pulled into the European
world economy. In the meantime, Damascus seldom looked towards the sea.
Operating within its network of overland routes, Damascus made contact with the
wider world primarily through its numerous caravans, which carried people, products,
supplies, and that most precious commodity – news – into and out of the region.”
The grain output of the Damascene oasis itself, which was perceived as inferior, was
not sufficient in quantity to feed the town’s population. Damascus was dependent on its
hinterland for agricultural products. The town received a good portion of its grain supplies
from the Biqa' Valley, east of Mount Lebanon, and the Hawran plain to the south. Difficulties
in supplying grain due to the isolated position of Damascus and the general limits on overland
transportation were combined with unfavourable climatic conditions during most of the
eighteenth century, which resulted in grain shortages, high prices, and, in several instances,
“bread riots” in the city (Grehan 2007).
Ayntab was a medium-sized town in Southeastern Anatolia that was home to

approximately 14,000 inhabitants in the late seventeenth century (Canbakal 2007). Some 50
kilometres to the west of the Euphrates and surrounded by extensions of the Taurus Mountains
in the west and the north, Ayntab was positioned on a fertile plain with abundant water
supplies. The town was self-sufficient in terms of agricultural products, and most of the grain
output of the region was consumed locally.
The town was not on any of the long-distance trade routes although it stood very
close to them. The journey from Aleppo, the main entrepôt for Indian trade during sixteenth
and seventeenth centuries, to Anatolia followed a route closer to the Mediterranean, bypassing
Ayntab (Canbakal 2007). Canbakal (2007: 4) defines Ayntab, as a town of “no particular
importance” to the imperial centre: “Ayntab did not stand out from a strategic point of view
or in terms of the resources it contributed to the well-being of the ‘well-protected domains;’
geographically, it stood outside the core lands of the empire. Therefore, the imperial centre
had no reason to pay special attention to the control or well-being of the city; nor did it have
the means to be part of the day-to-day business of the town due to the distance involved.”
Konya, located in the south of the dry Central Anatolian plateau, was a sizeable
Ottoman town in the sixteenth century that was severely hit by the seventeenth century crisis.
In the first decades of the century, Konya experienced depopulation on a large scale and a
significant decline in agricultural output. Although the town was a major stop on the diagonal
route traversing Anatolia and connecting Istanbul and Aleppo, Konya did not become a
strategic grain producer for the imperial centre and other Ottoman markets until the late
nineteenth century. Pamuk (1984: 112) states that “potentially the most important wheat-
growing areas of the Empire, Konya and Ankara provinces of Central Anatolia, continued to
remain outside the reach of Istanbul and other domestic and export markets until after the
arrival of the Anatolian Railway in the early 1890s.”
The Adriatic region
Despite sharing a weakening position in the world economy, the two adjacent sub-
regions of the Eastern Mediterranean, the Adriatic region under Venetian influence and the
Aegean, Ionian, and Levantine basins controlled by the Ottoman Empire, were different from
each other in several aspects. Most importantly, the Adriatic region was characterized by
political fragmentation, although it was embedded in the Venetian economic space until the
late eighteenth century. Venetian Dalmatia was a narrow and long strip of land along the
western coast of the Adriatic Sea. It was interrupted only by the Republic of Ragusa, which was
under Ottoman influence. In the southern part, there were the Ottoman eyalets of Albania,
Bosna, and Rumeli. On the western shores lay the Papal States under Venetian influence and
the Kingdom of Napoli, another rival of the Serenissima. The Venetian Republic, the dominant
force in the Adriatic region as late as the seventeenth century, experienced a political decline
in the second half of the eighteenth century, which culminated in its invasion and fall in 1797.
By contrast, the Aegean and the Ionian Sea basins were unified under Ottoman rule until the
end of the period under study.
Eight Adriatic markets are included in the sample: Ljubljana, Udine, Padua, Rovigo,
Ferrara, Ancona, and Senigallia. During the period under study, Udine, Padua, and Rovigo were
under the rule of the Venetian Republic; Ferrara, Ancona, and Senigallia were under the control
of the Papal States; and Ljubljana was controlled by the Habsburg Empire. After the
establishment of the free ports of Trieste (1719), Fiume (1723), and Ancona (1732), Venice
faced new challenges to its Adriatic commerce, which began to be diverted towards the free
ports (Arbel 2013). Accordingly, Venetian dominance in regional trade gradually declined. First,
Ancona became the port of attraction for raw materials from the Balkans and the Levant. From
the second quarter of the eighteenth century onwards, the Austrian port of Trieste began to
rise. From 1770, the Trieste port sidelined the port of Venice (Arbel 2013).
2.2. Sources
The sources for the Ottoman and Adriatic price series used in this paper are reported
in Tables 3 and 4. Annual wheat price series for Manisa and Ayntab (1660 to 1840) are new
series using data extracted from inheritance inventories. 43 For other Ottoman towns, wheat
prices are taken from secondary sources, which report waqf account book prices (Pamuk
2000a), inventory prices (Öztürk 1992; Sahillioğlu 1999), and prices in other court registers
(Balta 1992; Raymond 1974; Svoronos 1956). In the previous chapter, I demonstrated that
inventory prices reflect the conventional prices of the time and are comparable with prices
from other primary sources.
The Ottoman database includes three regional series (Wallachia, Eastern and Western
Balkans). Berov (1976) states that the series are collected from towns to the east and west of
the Vidin-Matapan line, dividing the peninsula into two sub-regions. The origin of his regional
series and how prices from different towns were weighted to construct regional averages are,
however, unknown. In the paper in which Berov presents the wheat series for the first time, he
mentions that the data was “found and systematized from several hundred Turkish, Bulgarian,
Yugoslavian and Greek historical sources,” (1974: 170). In his extended work on price
movements in the Balkans, he states that 5,371 references to 114 commodities in various years
were taken from hundreds of historical publications and archives related to the Balkan
countries (1976: 317). Similarly, Jacks (2004, 2005) reports a Wallachia series from Popa (1978)
43 For each calendar year, two to three observations are collected from inventories. The annual price
given is the average of all observations recorded for the particular year.
without mentioning the primary sources used to extract prices.
The Adriatic series are taken from Jacks (2004, 2005) at

http://www.sfu.ca/~djacks/data/prices/prices.html and the Allen-Unger Global Commodity
Prices Dataset at http://www.gcpdb.info/. Except for the Ljubljana series, these rely on
statistical data published by governmental bodies.
Table 3- Price data: Sources (Ottoman Empire)
Location Source Primary source
Manisa Collected by the author44 Inheritance inventories
Istanbul Pamuk (2000a) Waqf account books
Pamuk (2000a); Sahillioğlu (1999) Waqf account books (Pamuk 2000a);

Edirne inheritance inventories (Sahillioğlu 1999)
Cairo Raymond (1974) Court registers
Pamuk (2000a); Öztürk (1992) Waqf account books (Pamuk 2000a);

Bursa inheritance inventories (Öztürk 1992)
Svoronos (1956); Balta (1992) (taken
Salonika
from Vasdeavellis 1952) Court registers
Patmos Asdrachas (1972) ?

Balta (1992) (taken from Stavrinidis Court registers
Candia
1975; 1976; 1978; 1984; 1985)
Jacks (2004; 2005) (taken from Popa ?
Wallachia
1978)
Balkans (East) Berov (1976) ?
Balkans (West) Berov (1976) ?
Ayntab Collected by the author45 Inheritance inventories
Konya Pamuk (2000a) Waqf account books
Damascus Grehan (2007) Court registers
44 See Primary Sources section for details.

45 See Primary Sources section for details.
Table 4- Price data: Sources (Adriatic)
Ljubljana Jacks (2004; 2005) Valenčič (1977);
Ferrara Jacks (2004; 2005) Ministero di Agricoltura, Industria e Commercio (1886)
Udine Jacks (2004; 2005) Ministero di Agricoltura, Industria e Commercio (1886)
Padua Jacks (2004; 2005) Ministero di Agricoltura, Industria e Commercio (1886)
Rovigo Jacks (2004; 2005) Ministero di Agricoltura, Industria e Commercio (1886)
Pesaro Allen-Unger Database Government of the United Kingdom (1826-1827)
Ancona Allen-Unger Database Government of the United Kingdom (1826-1827)
Senigallia Allen-Unger Database Government of the United Kingdom (1826-1827)
2.3. Conversion to metric units and silver grams
The prices employed for the analysis are annual silver prices per hectolitre of wheat.
To obtain these prices (1) all currencies are converted to akçe; (2) unit prices in akçe are
calculated by dividing the total price by the quantity reported in local unit ( kile, ölçek, ardab,
kıyye); (3) average prices per local unit are multiplied by a conversion rate to calculate prices
per metric unit; and finally (4) nominal prices per metric unit are converted to silver grams.
Different currencies are converted to akçe in accordance with exchange rates

calculated by Pamuk (1994) (1 qurush = 150 akçe in the period 1660-1680; 1 qurush = 120
akçe in 1680-1840; 1 qurush = 40 para in 1660-1840; 1 sülüs = 80 to 90 akçe in 1660-1840).
Nominal prices calculated on a yearly basis are fixed in silver grams to remove monetary effects
and to permit across-space comparisons. The silver content of akçe and para is taken from
Pamuk (2000a: 36-41).
In measuring grain, a variety of local units (kile, ölçek, ardab, muzur, ghirara) were
used in the Ottoman realm. Among these, the kile was the most common measure of volume.
However, it was not identical in all parts of the empire and for all kinds of goods. Its metric
equivalent could vary even in the same area according to the commodity measured (e.g. The
kile for rice was half the kile used for wheat) or could change over time in the same region
(İnalcık 1983). The data collected for this study mentions five different kile: the kile of Istanbul,
the kile of Edirne, the kile of Salonika, the kile of Konya, and the kile of Romania. The kile of
Istanbul was the standard kile, which was equal to 37 litres or 0.97 Winchester bushels (Pamuk
2000a). The conversion rates for other local units have been taken from a number of sources.
(For metric equivalents and sources, see Table 5). These are first converted to the weight
measure of okka, which was standard across the Ottoman lands. It equalled 1.283 grams and
1.805 litres of wheat.
Table 5- Metric equivalents of local measurement units (Ottoman Empire)
Standard Metric
Location Local unit equivalent equivalent Source
Manisa ölçek 15 okka 27.07L Eldem (1970)
Ayntab kile 80 okka 144.39L Öztürk (1989)
Istanbul kile of Istanbul 20 okka 35.27L Pamuk (2000a)
Edirne kile of Edirne 28.79L Barkan (1964)

French merchant
Kahire ardab46 175.55L magazines
Konya kile of Konya 24 okka 43.32L Inalcik (1994)
Bursa kile of Bursa 12 okka 21.66L Inalcik (1994)
Salonika kile of Salonika 84 okka 151.61L Svoronos (1956)
Patmos kile of Istanbul 20 okka 151.61L Svoronos (1956)
Candia muzur 15 okka 27.07L Balta 1992
Wallachia kile roumaine 10 kile of Istanbul 352.7L Jacks 2004; 2005

Balkan (East) *Reported in silver gr/kg (Berov 1976)
Balkan (West) *Reported in silver gr/kg in (Berov 1976)
Damascus ghirara 265L Hinz 1955
46 Ardab is a unit of volume for grains often used in Egypt. There are different estimates of its metric
equivalent which vary between 75lt and 184lt (Pamuk 2000a). Yet, the variance of equivalents of “ ardab
for wheat” given by a number of guides and magazines for French merchants (Voyage de la Turquie
d'Europe, La Decade Egyptienne Journal Literaire et d'Economie Politique 1798-9, Correspondance de
Napoleon 1er, etc.) is quite narrow (between 172 and 182lt). Here, an average equivalent is taken.
For the Adriatic region, in converting local monetary units to silver grams, Jack’s (2004;
2005) table of “Silver content of currencies 1258-1979” available at
http://www.sfu.ca/~djacks/data/prices/Metals/prices.html was utilised. Table 6 reports the
moneys and measurement units in which the prices are reported.
Table 6- Moneys and local measurement units (Adriatic)
Money Measure Measure

1 mernik=26.5L,
Ljubljana Krajcarjih Mernik; vagan 1 vagan=61.49L
Ferrara Italian Lira Hectolitre
Udine Italian Lira Hectolitre
Padua Italian Lira Hectolitre
Rovigo Italian Lira Hectolitre
Pesaro English Pence English Grain Quarter 285.79L
Ancona English Pence English Grain Quarter 285.79L
Senigallia English Pence English Grain Quarter 285.79L
Tables 7 presents the descriptive statistics of the price series for each location. The
unit of observation is a series of annual wheat prices in a specific market. While some authors
prefer interpolation/extrapolation techniques to obtain complete price series (Özmucur and
Pamuk 2007; Federico 2012), most avoid doing so because this may distort the results. In this
study, the price series are not interpolated/extrapolated. Instead, the database features a high
proportion of missing observations particularly in the Ottoman series. In only three of the
series in this sample are annual prices available for more than half of the 180 years under
study.
Table 7- Descriptive statistics of the price series
Coverage
N Mean S.D. Min. Max
as %
OTTOMAN EMPIRE
Manisa 129 0.72 37.66 17.29 9.5 111.6
Istanbul 93 0.52 40.34 17.46 15.7 98.4
Edirne 81 0.45 31.66 16.57 7.5 97.2
Cairo 45 0.25 36.93 21.34 9.7 100.9
Bursa 42 0.23 46.8 30.13 14.6 132.7
Salonika 49 0.27 78.05 83.44 20.3 426.8
Patmos 23 0.13 59.05 9.7 39.6 76.9
Candia 21 0.12 32.44 12.97 13.3 62.6
Wallachia 44 0.24 36.28 25 10 116.8
Eastern Balkans 56 0.31 82.67 88.15 10.8 395.2
Western Balkans 51 0.28 32.55 27.41 13 167.9
Ayntab 93 0.52 29.63 13.52 8.4 60.2
Konya 22 0.12 72.01 42.85 18.7 137.2
Damascus 58 0.32 146.19 110.45 44.8 650.9
ADRIATIC
Ljubljana 131 0.73 69.95 22.68 37.7 160.4
Ferrera 54 0.3 73.85 25.31 34.7 153.7
Udine 180 1 58.86 26.96 23.9 189
Padua 40 0.22 74.32 30.4 37 179.6
Rovigo 67 0.37 63.65 22.04 37.6 159.8
Pesaro 118 0.66 53.86 25.53 19 166.9
Ancona 117 0.65 48.2 20.52 17.9 132.1
Senigallia 95 0.53 57.98 23.33 26.7 148.6
As has been discussed in the previous chapter, the availability and quality of primary
sources from which historical prices can be compiled pose a serious limit to the study of price-
related subjects. Prior to this research, the only study on long-term commodity price
movements (Pamuk 2000a) concerned the imperial capital, and used a series that only
possessed a coverage of 52 percent. Despite the serious number of missing observations,
considering the state of research in the field of Ottoman price history, I contend that the
present study presents a valuable contribution, bringing together the available local series
published in various sources and adding two new series constructed based on inheritance
inventories. To check its robustness, the findings based on the complete sample with its high
numbers of missing observations are compared with results using a restricted yet more
complete sample. The consistency of results between the extended and restricted samples
suggests that despite the missing observations, general trends in price differentials can be
safely drawn from this data set.
Before moving on to present the methodology used to measure trading costs, I briefly
discuss some basic characteristics of the wheat price series included in the study. As can be
seen in Figures 5 to 8, while wheat prices in other towns moved together and the mean values
were close to each other in Damascus, Konya, and the Balkans, the average prices were
remarkably high compared to the rest of the sample, and experienced sharp fluctuations,
particularly during periods of shortage such as the 1720s or 1740s. Trends differed. In Konya
and Damascus, the silver price of wheat continuously and dramatically rose throughout the
period. In the Eastern and Western Balkans, it declined around the mid-eighteenth century,
approaching the level of other Ottoman towns. Price volatility was lowest in Western Anatolia,
Cairo, Ayntab, and Istanbul. In general, wheat prices in the Adriatic region were higher than
prices in the Ottoman realm. The price of wheat in silver grams in the Adriatic shows a pattern
similar to that in Western Anatolia, steadily rising in the nineteenth century and peaking in the
early nineteenth century before declining towards the mid-century.
Figure 5- Wheat prices (silver gr/HL) - 15-year moving averages

Istanbul and Western Anatolia
Figure 6- Wheat prices (silver gr/HL) – 15-year moving averages

Balkans

Damascus, Ayntab, Konya, Cairo

Adriatic
3. Methodology
Market integration is divided into two sub-processes: price convergence (the

diminution of price differentials over time) and price adjustment (the speed at which profitable
price differential are arbitraged away) (Jacks 2005). In measuring the degree of market
integration, historians employ a variety of methodologies, testing either of the following two
complementary, yet distinct, conditions: the equilibrium level of prices being equal (or the law
of one price); and prices returning to this level with ease and rapidity after any shock (Federico
2012).
The law of one price implies that in a competitive equilibrium, price differences are
equal to trading costs between trading markets. The difference in autarky prices between two
locations are compared with the total costs for shipping the product from the lower-priced
market to the higher-priced one. Whenever the former exceeds the latter, there is an
opportunity for profitable arbitrage, which profit-seeking traders can exploit. If, however, the
shipping costs are greater than the difference in autarky prices, traders incur a loss that
encourages them not to trade (Federico 2012).
Here, converging prices rather than measures of market efficiency are taken as the
most crucial indicator of integrating commodity markets. Federico (2011, 2012) argues that
although market efficiency, the second of these conditions, is important in terms of market
integration, mildly inefficient markets may experience market integration, making, in practice,
converging prices a more robust indicator of integrating markets. Similarly to Federico, Findlay
and O’Rourke (2003: 15) also suggest that price convergence is the best measure: “Price gaps
reflect all relevant costs of doing trade between markets: not just transport costs, but also
trade barriers, and those costs associated with wars, monopolies, pirates, and so on.” Since the
law of one price is almost never met due to transaction costs, and estimating these latter
factors in an effort to gauge their relation to the price differentials between different locations
is difficult, Federico recommends the examination of price trends rather than price levels, and
a measurement of the extent to which prices of the same commodities converged over time
in different locations (Federico 2012: 477). To test for price convergence (divergence), I look at
two alternative indicators: aggregated relative prices and yearly coefficients of variation.
Scholars have frequently noted that Istanbul’s grain prices were held artificially low by
provisionist practices. Berov (1974: 168-9) writes, “A comparison of the data found about the
market price of wheat in Istanbul and in other towns in the Balkans for separate years in the
sixteenth-eighteenth century showed that in about 65 percent of the cases, prices in Istanbul
were up to twice lower, and in 35 percent of the cases – equal or a little higher than those in
the provincial towns in the Balkans.”
A crucial question here is whether viable market integration analysis can be conducted
under these circumstances based on price data. It could be argued that artificially low prices
in Istanbul markets could cause us to underestimate the actual price gaps. Bateman (2010: 28)
for instance, states, “in terms of the analysis, it must be remembered that Ottoman authorities
were not afraid of interfering with the market, setting and adjusting prices where they wished.
Hence, any examination of price data must be treated with some caution.” Having said that,
the alternative sample excluding Istanbul, and used for robustness check produced the same
trends with the original sample incorporating the city, suggesting that inclusion of the capital
city into the sample did not alter the general trends.
Relative prices
The first indicator of market integration that I use in looking at convergence in prices
is relative prices. Convergence (or divergence) in prices between two locations can be
estimated with a log-linear regression of trends in relative prices or price gaps. However, as
the number of possible pairs increases, interpreting results for a large number of markets can
be difficult (Federico 2012). For these reasons, individual relative prices between each of the
city-pairs are calculated and compiled into a panel data set, where the cross-sectional
component is the city pair and the time series is years. Price relatives are aggregated through
a city pairs fixed effects panel regression model. A non-interpolated price series is used for
this analysis.
In panel data, each entity (in our case, each city pair) has its own individual
characteristics that may influence the predictor variables, like the size of the markets under
comparison, the distance between city pairs, whether or not the cities are connected by sea,
and the like. The fixed effects model allows us to control for variables that account for
individual heterogeneity. It removes the effect of time-invariant characteristics, so that we can
assess the net effect of the predictors on the outcome variable. It also has the advantage of
avoiding some of the problems that arise from missing data.
In the first regression, the time effect (TIME) is employed as the explanatory variable
for the whole period and separately for sub-periods. A second fixed-effects panel regression
incorporating a series of dummies for each decade between 1680 and 1835 is applied to the
data in order to extract the average relative price at different points in time, using the
estimated coefficients in the time dummies. The dummy for the initial period (1675-1685 for
the complete Ottoman sample, and 1660-1675 for the rest), is omitted as the reference period.
(1)
𝑃1𝑡
|𝐿𝑛 ( ) | = 𝛼 + 𝛽𝑇𝐼𝑀𝐸 + 𝑢
𝑃2𝑡
(2)
𝑃1𝑡
|𝐿𝑛 ( ) | = 𝛼 + Σ𝛽𝑘 𝐷𝐸𝐶𝐴𝐷𝐸𝑘 + 𝑢
𝑃2𝑡
As bilateral price differences decline, relative prices move towards 1. Pit is the wheat
price in city i in year t, Pjt is the wheat price in city j in year t. Since ln(Pit/Pjt) = lnPit −lnPjt for all
i ≠ j; if Pit/Pjt = 1 then ln(Pit/Pjt) = 0. Thus, as markets integrate |ln(Pit/Pjt)| moves towards 0, and
as markets disintegrate the price relatives move away from 1. If markets were integrating, we
would expect the coefficient of TIME to be negative and significant.
However, a significant time trend in the aggregated price gaps might not always mean
that the price differentials at the end of our period were different than those in the initial
period. Cycles of divergence and integration can produce an impression of change, even while
the long-run trend remains static. A statistically significant negative trend in price dispersion
might be observed, for instance, due to a sharp price divergence in the early eighteenth
century, which had been compensated for in the following period even though the relative
prices in the nineteenth century were no higher than those in the late seventeenth century.
For this reason, the presence/absence of a long-term integration trend is evaluated by first
looking at the coefficient of the TIME variable in equation (1) before an examination of the
coefficients of the last two period dummies ( 1825-1834, 1835-1840) in equation (2). Only if
both the coefficient of the time trend variable and the period dummies for the last two decades
are statistically significant and negative can we conclude that the prices converged.
An insignificant coefficient for a time dummy may reflect a lack of changes across all
sub-regions, but could also be due to opposite and offsetting trends between sub-regions,
which imply integration in some areas and disintegration in others (Federico 2012). For this
reason, the analysis is conducted separately on five different samples (entire Eastern
Mediterranean, entire Ottoman Empire, Ottoman Mediterranean, Adriatic, and Ottoman
Mediterranean-Adriatic samples) that cover geographical areas with distinctive features.
Coefficient of variation
When markets that are not in a direct exchange relation are considered, price
differences might not be equal to the trading costs. In such cases, estimates of trading costs
based on price differentials/relative prices are negatively biased (Chilosi et al. 2011). As has
been discussed above, most of the markets included in the study are connected to each other
and the imperial capital, although we do not have direct evidence of on-going trade in each
case. In such cases, coefficient of variation analysis gives us more reliable results. Therefore,
the robustness of the fixed effects results is also checked against the results from the
coefficient of variation analysis.
The second measure of market integration that I use is the coefficient of variation.
First, I look at the dispersion of all markets together by imputing the coefficient of variation
(the standard deviation normalized by the mean) in each period of time. Sigma-convergence
– a statistically significant decrease in the standard deviation of prices over the period – is
considered to indicate convergence in prices. On the advantages of looking at sigma-
convergence, Federico (2011: 95) states, “It needs no a priori assumptions about the process
of adjustment of prices; it is more robust to errors in data; it is highly flexible; and, above all, it
focuses on changes in prices which oriented the decisions of millions of producers and
consumers rather than on gains and losses for a few specialized traders.”
As a first step, I use the Augmented-Dickey Fuller (ADF) test to determine whether the
time series of the coefficient of variation of prices is stationary. 47 The null-hypothesis of ADF
47 A constant is used in the test regression. The results from two different forms – with no lagged
difference terms, and one lagged difference term – are reported.
is non-stationary. Therefore, if the null can be rejected at 10 percent, the series is taken to be
stationary. In other words, if the series has a constant variance over time – signifying the lack
of any trend – then the hypothesis of convergence (divergence) can be ruled out. When the
results are in favour of a non-stationary hypothesis, I estimate the rate of convergence
(divergence) using the following regression:
(3)
𝐿𝑛(𝐶𝑂𝑉𝑡 ) = 𝛼 + 𝛽𝑇𝐼𝑀𝐸 + 𝑢
If the prices converge, a negative and significant coefficient of TIME is expected. Since
the coefficient of variation for each commodity is based on the number of cities for which data
is available, the missing annual observation may create difficulties in interpreting the results.
Therefore, I constructed a second series by weighting the coefficient of variation by the
number of cities for which data is available.
4. Results
Here, the integration and disintegration in the wheat markets are evaluated by looking
at the changes in the aggregated relative prices. The findings from the coefficient of variation
analyses given in section 6 are used for robustness checks. As will be discussed in more detail
subsequently in this study, both methods employed to measure the extent of market
integration produce consistent results.
For each sample, results from the fixed effects regression analysis are presented in
two separate tables. In each case, the first table reports the results of the fixed effects
regression (1) employing the time trend as the explanatory variable, while the second table
displays results from the regression (2) employing period dummies. The figure showing the
estimated average relative prices uses the coefficients on the period dummies from regression
(2), regardless of whether or not they are significant.
Three criteria need to be met to conclude that wheat markets experienced long-term
integration/disintegration: A statistically significant time trend should be detected in the
regression (1); the coefficient of the last period dummy in regression (2) should be significantly
different than that of the initial period; and the sign of the TIME and the last two period
dummies should be the same.
4.1. Long-term trends
Initially, I examine the complete set of Ottoman markets covering Anatolia and Syria
in the east, the Balkans in the west, and the Egyptian delta in the south. This complete set
incorporating 14 markets (Istanbul, Cairo, Manisa, Bursa, Candia, Patmos, Edirne, Wallachia,
Eastern and Western Balkans, Salonika, Ayntab, Konya, and Damascus) provides a
representation of the wider Ottoman region during the period under study. By looking at this
extended sample of markets separated by large distances, I trace the trends of market
integration across long-distance Ottoman markets. Since this data is most complete for the
1675-1800 period, the analysis is restricted to these years.
Alongside the markets incorporated into the imperial grain-trading network, this
complete sample includes three landlocked and isolated markets -Damascus, Ayntab, and
Konya-, which had no direct ties with this network, and presumably with each other. In the
light of the previous empirical findings, and considering that the restricted tradability of wheat
overland prior to the introduction of railroads, a long-term integration trend is not expected
in this complete sample.
The results presented in Tables 8 and 9 are in accordance with our expectations. At
the beginning of the nineteenth century, the wider Ottoman region, encompassing the Arab
lands, the Egyptian delta, Anatolia, the Balkans, and Istanbul, was no more integrated than it
was in the last quarter of the seventeenth century. At the 10 percent level, the regression (2)
detects no time trend in the aggregated relative prices for the whole period, and in regression
(3), the estimated relative price for 1795-1804 is not significantly different from that for 1675-
1685.
Table 8- Fixed Effects Results -Time trend

Ottoman Empire (Complete sample), 1675-1804
Dependent variable |Ln(Pi/Pj|
Coef. Std.Err.
CONSTANT 0.821*** -0.041
TIME -0.002 -0.001
N 1412
ADJUSTED-R2 0.32
respectively.
Table 9- Fixed Effects Results - Aggregated price relatives

Ottoman Empire (Complete sample), 1675-1804
Dependent variable |Ln(Pi/Pj)|

Coef. Std.Err.
CONSTANT 0.674*** -0.081
1685-1694 0.053 -0.098
1695-1704 0.064 -0.097
1705-1714 0.282*** -0.098
1715-1724 0.339*** -0.093
1725-1734 0.094 -0.101
1735-1744 0.241*** -0.095
1745-1754 -0.089 -0.096
1755-1764 -0.096 -0.100
1765-1774 0.136 -0.098
1775-1784 0.064 -0.095
1785-1794 0.097 -0.097
1795-1804 0.048 -0.099
N 1412
ADJUSTED-R2 0.35
respectively. The period 1675-1684 is the omitted category.
It is often stressed that in the pre-modern world, economic activities were regionally
rather than nationally organised. Accordingly, I take as a second step, the Eastern
Mediterranean region (the Levant, the North African shore, the Aegean and Ionian basins, and
the Adriatic) as an economic entity that brings together different political units. The sample
incorporates the eight wheat price series from the Adriatic (Ljubljana, Udine, Padua, Rovigo,
Ferrara, Pesaro, Senigallia, and Ancona) and nine series from the Aegean and the Ionian area
(the Western Balkans, Salonika, Edirne, Istanbul, Bursa, Manisa, Candia, Patmos, and Cairo), all
of which are situated at most at 100 kilometres from the Mediterranean coast.
As shown in Table 10, from around the mid-seventeenth to the mid-nineteenth

century, the fixed effects regression detected a trend of decline in the aggregated price
differentials in the entire Eastern Mediterranean region. However, the coefficients of the last
two period dummies are positive and significant at the 1 percent level (see Table 11), implying
that the estimated price relatives for 1825-1834 and 1834-1840 were significantly higher than
the estimated price relatives for the initial period of 1660-1675. This implies that in the early
nineteenth century, trading costs in the Eastern Mediterranean as a whole were higher than
they had been in the last quarter of the seventeenth century and that the trend of decline
detected derives from the high values in the early eighteenth century.
Table 10- Fixed Effects Results, 1660-1840 - Time trend

Entire Eastern Mediterranean, Ottoman Mediterranean, and Adriatic regions, and
Ottoman Mediterranean-Adriatic market pairs
Dependent |Ln(Pi/Pj)|
variable
EASTERN MED. OTTOMAN
ADRIATIC OTT. MED-ADR.
(ENTIRE) MED.
CONSTANT 1.997*** 0.58*** 0.442*** 0.575***
(-0.264) (-0.045) (-0.016) (-0.028)
TIME -0.001*** -0.001** -0.002*** 0.000
(0.000) (0.000) (0.000) (0.000)
N 4759 624 1683 2452
ADJUSTED-R2 0.29 0.22 0.39 0.12
respectively.
Table 11- Fixed Effects Results - Aggregated price relatives

Entire Eastern Mediterranean, Ottoman Mediterranean, and Adriatic regions, and
Dependent variable |Ln(Pi/Pj|
EASTERN MED. OTTOMAN OTT. MED-

ADRIATIC
(ENTIRE) MED. ADRIATIC
CONSTANT 0.191*** 0.406*** 0.341*** 0.164***

(-0.037) (-0.099) (-0.041) (-0.053)
1675-1684 0.06 0.229* -0.194*** 0.037
(-0.051) (-0.127) (-0.061) (-0.075)
1685-1694 0.265*** 0.221* 0.218*** 0.273***
(-0.048) (-0.117) (-0.061) (-0.07)
1695-1704 0.447*** 0.056 0.078 0.689***
(-0.043) (-0.111) (-0.048) (-0.063)
1705-1714 0.419*** 0.086 0.038 0.635***
(-0.042) (-0.111) (-0.047) (-0.062)
1715-1724 0.432*** 0.236** 0.094** 0.582***
(-0.042) (-0.111) (-0.046) (-0.061)
1725-1734 0.231*** 0.098 -0.076* 0.348***
(-0.042) (-0.119) (-0.044) (-0.062)
1735-1744 0.254*** -0.019 -0.073* 0.414***
(-0.041) (-0.115) (-0.044) (-0.06)
1745-1754 0.124*** -0.035 -0.129*** 0.218***
(-0.041) (-0.121) (-0.045) (-0.063)
1755-1764 0.096** -0.104 -0.125*** 0.169***
(-0.045) (-0.127) (-0.047) (-0.067)
1765-1774 0.328*** 0.338*** -0.076* 0.471***
(-0.043) (-0.122) (-0.046) (-0.064)
1775-1784 0.225*** 0.197* -0.135*** 0.346***
(-0.042) (-0.116) (-0.045) (-0.062)
1785-1794 0.158*** -0.03 -0.151*** 0.293***
(-0.041) (-0.113) (-0.044) (-0.06)
1795-1804 0.296*** -0.004 -0.120*** 0.529***
(-0.041) (-0.115) (-0.043) (-0.06)
1805-1814 0.142*** -0.051 -0.163*** 0.249***
(-0.04) (-0.125) (-0.043) (-0.06)
1815-1824 0.108*** -0.036 -0.215*** 0.228***
(-0.04) (-0.127) (-0.043) (-0.06)
1825-1834 0.274*** 0.028 -0.246*** 0.573***

(-0.042) (-0.129) (-0.044) (-0.063)
1834-1840 0.237*** 0.069 -0.227*** 0.491***
(-0.049) (-0.173) (-0.047) (-0.076)
N 4759 624 1683 2452
ADJUSTED-R2 0.36 0.26 0.43 0.56
As the third step in this analysis, I compare trends across the two main sub-regions
within the Eastern Mediterranean economic zone: the Adriatic and the Ottoman
Mediterranean. In doing so, I investigate whether Ottoman Mediterranean markets followed a
trajectory that diverged from those of the wheat markets in other proximate Mediterranean
regions. This comparative approach will allow us to differentiate between trends in market
integration that were experienced by the Mediterranean world as a whole (or commonly by
adjacent sub-regions) and those that were specific to the Ottoman realm. Furthermore,
comparing the trends in market integration in the Northwestern part of the Ottoman
Mediterranean against trends in the Adriatic region will help us to understand how factors that
cut across both areas – such as wars, climatic fluctuations, and epidemics – affected the market
development in different institutional settings and policy environments.
As can be observed in Table 10 and 11, in the long term, wheat market integration
only appears to have occurred in the Adriatic region. Both the coefficient of the time trend
variable and the coefficient of the last two period dummies are negative significant at the 1
percent level in the Adriatic sample. In the Ottoman Mediterranean, although a negative time
trend is observed, this appears to reflect the region’s recovery from the early eighteenth-
century growth in price dispersion, rather than a long-term integration trend, as shown by the
statistically insignificant and positive period dummies for 1825-1834 and 1834-1840.
Finally, I look at the evolution of estimated price relatives in the Ottoman

Mediterranean-Adriatic market pairs, which indicate the price differentials in foreign trade. The
findings show no long-term trend of decline/increase in the costs of international trade within
the Eastern Mediterranean region. It can be concluded that as with region as a whole,
international markets in the region did not experience long-term integration/disintegration.

Having said that, it should also be noted that the coefficients of the last two period dummies
are positive and highly significant, showing that estimated relative prices were higher at the
end of the period under study in comparison to the initial period.
4.2. Levels
What did market conditions within these different trading regions look like? Did the
lack of integration in Ottoman markets reflect a relatively well-integrated market system, or
did it occur against the backdrop of a fragmented and costly exchange network? We can shed
some light on this by looking at aggregated relative prices, estimated by the fixed effects
regressions. These capture the level of price difference within each trading area. Figures 9 to
12 depict the evolution of aggregated relative prices within the Ottoman Empire, the Eastern
Mediterranean, the Ottoman Mediterranean, the Adriatic region, and in the Ottoman
Mediterranean-Adriatic market pairs. Among the five different samples, the first sample
incorporating the landlocked and coastal, as well as short-distance and long-distance Ottoman
wheat markets, produced the highest estimated price relatives. Throughout the period, these
ranged between 1.8 and 2.8. It should be underlined that even under the favourable conditions
of the mid-eighteenth century, the aggregated relative prices in the Ottoman markets was 1.8,
while this figure was 1.4 for the Ottoman Mediterranean, and just 1.2 for the Adriatic.
The wide price gaps across the wheat markets suggest that prior to the introduction
of steamships and railroads, landlocked markets in the wider Ottoman region remained largely
fragmented, meaning a ‘national wheat market’ did not emerge. These results are hardly
surprising, since under the conditions of pre-modern transportation and communication
technologies, an empire extending to three continents could not be expected to operate as a
single economic entity.
Figure 9- Aggregated relative prices

Ottoman Empire (complete sample)
Figure 10-Aggregated relative prices

Entire Eastern Mediterranean

Adriatic and Ottoman Mediterranean

Given the high costs of transportation prior to the nineteenth century, it is crucial to
take distance into account when comparing different samples in terms of trading costs. Here,
I compare trading costs in the Ottoman Mediterranean and Adriatic regions, controlling for
distance between market pairs, in order to understand whether the difference in the level of
price gaps across the two regions was merely the result of the longer distances separating
Ottoman wheat markets, or whether it indicates a discrepancy in terms of the extent of market
integration.
The logged relative prices in Adriatic and Ottoman Mediterranean market pairs are
regressed against LNDISTANCE and OTTOMANMED, using the OLS model. LNDISTANCE is the
logged flight distance (in 100 km) between markets in each pair. OTTOMANMED is a dummy
variable indicating the market pairs from the Ottoman Mediterranean. The reference category
is the Adriatic market pairs. If there was a significant difference in trading costs per 100km
between the two regions of the Eastern Mediterranean, we would expect the coefficient of
OTTOMANMED to be significant.
(4)
𝑃1𝑖
|𝐿𝑛 ( ) | = 𝛽0 + 𝛽1 𝐿𝑁𝐷𝐼𝑆𝑇𝐴𝑁𝐶𝐸𝑖 + 𝛽2 𝑂𝑇𝑇𝑂𝑀𝐴𝑁𝑀𝐸𝐷𝑖 + 𝑢
𝑃2𝑖
Table 12 reports the descriptive statistics of the regression variables.
Table 12-Descriptive statistics of the regression variables
N Mean S. D.
|Ln(P1/P2)| 2322 0.294 0.343
LNDISTANCE 2322 0.624 0.946
OTTOMANMED 2322 0.275 0.447
The regression results are given in Table 13. As expected, price gaps across market
pairs are strongly and positively associated with distance between the markets, throughout
the 1660-1840 period. The coefficient of LNDISTANCE is positive and significant at the 1
percent level in all three regressions. In every additional 100 km, the relative prices increased
by 12, 8, and 4 percent, respectively, in the 1660-1720, 1720-1780, and 1780-1840 periods.
Table 13-OLS regression results

Bilateral relative prices and distance between markets
Dependent variable |Ln(P1/P2)|
1660-1720 1720-1780 1780-1840

CONSTANT 0.416*** 0.240*** 0.141***
-0.039 -0.018 -0.006
LNDISTANCE 0.117*** 0.080*** 0.036***
-0.035 -0.017 -0.007
OTTOMANMED -0.043 0.282*** 0.198***
-0.062 -0.041 -0.014
N 326 676 1320
ADJUSTED-R2 0.04 0.2 0.21
respectively. Adriatic market pairs are the omitted category.
Once controlled for distance, no significant difference is detected between trading

costs in the Ottoman Mediterranean and Adriatic regions prior to 1720. This shows us that in
the late seventeenth and early eighteenth centuries, the extent of wheat market integration in
both regions was comparable. However, in the following two periods, the level of price gaps
in the Ottoman Mediterranean was significantly higher than the price gaps in the Adriatic
region. The coefficient of the OTTOMANMED is positive and significant at the 1 percent level.
In 1720-1780, the price gap between Ottoman Mediterranean wheat markets was 28 percent
higher than the price gap between Adriatic markets separated by equal distance. In 1780-1840,
this figure was 20 percent. These results suggest that from the second quarter of the
eighteenth century, Adriatic wheat markets became better integrated than Ottoman
Mediterranean markets.
4.3. Short-term trends
One of the stylised facts that can be derived from previous research is that market
integration was not a linear process. Despite the absence of consensus on the early roots of
commodity market integration, most studies reveal that the late Middle Ages and the early-
modern period featured a succession of waves of integration and disintegration within Europe.
The findings of Chilosi et al. (2011) for instance, show that following the Thirty Years’ War,
Europe-wide price dispersion decreased until the mid-eighteenth century, remained stationary
from this date onwards until the French Revolution, and increased during the Napoleonic Wars.
This leads them to conclude that European market integration occurred in a step-by-step
fashion with occasional periods of regression.
In order to identify short-term integration and fragmentation episodes, I conducted

fixed effects regressions separately for each 30-year interval between 1660 and 1840,
employing TIME as the explanatory variable. While in Table 11, the estimated coefficients of
the decade dummies estimate the difference in aggregated relative prices between the
omitted (initial) decade and the decade represented by the particular period dummy; the
coefficients of TIME variable presented in Table 14 tell us whether there was a positive or
negative time trend in bilateral relative prices within each of the sub-periods.
Table 14- Fixed Effects Results for 30- year intervals (Time trend)
Dependent variable |Ln(Pi/Pj)|
OTTOMAN EMPIRE
1675-1690 1690-1720 1720-1750 1750-1780 1780-1800
CONSTANT 0.537*** 0.391*** 1.874*** -0.536* 1.927***
(-0.125) (-0.113) (-0.246) (-0.311) (-0.588)
TIME 0.020* 0.011*** -0.017*** 0.014*** -0.010**
(-0.011) (-0.003) (-0.004) (-0.003) (-0.005)
N 126 292 332 312 343
ADJUSTED-R2 0.38 0.69 0.31 0.54 0.5
EASTERN MEDITERRANEAN (ENTIRE)

1660-1690 1690-1720 1720-1750 1750-1780 1780-1810 1810-1840
CONSTANT -4.615*** 0.436 15.342*** -15.414 -4.429** -11.164***
(4.037) (3.232) (2.859) (2.676) (1.897) (2.131)
TIME 0.003 0.001 -0.009*** 0.009*** 0.003*** 0.006***
(0.002) (0.002) (0.002) (0.002) (0.001) (0.001)
N 239 545 540 651 1336 959
ADJUSTED-R2 0.31 0.55 0.32 0.49 0.31 0.37
OTTOMAN MEDITERRANEAN
Ottoman Med. 1660-1690 1690-1720 1720-1750 1750-1780 1780-1810 1810-1840
CONSTANT 0.384*** 0.301* 1.471*** -1.297** 1.235*** -0.549
(0.131) (0.183) (0.419) (0.605) (0.458) (0.586)
TIME 0.01 0.004 -0.013** 0.019*** -0.006* 0.005
(0.007) (0.004) (0.006) (0.006) (0.004) (0.004)
N 64 124 93 94 166 76
ADJUSTED-R2 0.33 0.41 0.35 0.3 0.24 0.01
ADRIATIC
1660-1690 1690-1720 1720-1750 1750-1780 1780-1810 1810-1840
CONSTANT 0.349*** 0.562*** 0.836*** -0.037 -0.007 0.665***
(0.081) (0.144) (0.114) (0.133) (0.111) (0.11)
TIME 0.004 -0.002 -0.007*** 0.002* 0.001* -0.003***
(0.004) (0.003) (0.001) (0.001) (0.001) (0.001)
N 31 105 268 202 483 560
ADJUSTED-R2 0.01 0.51 0.44 0.17 0.17 0.15
OTTOMAN MEDITERRANEAN-ADRIATIC
1660-1690 1690-1720 1720-1750 1750-1780 1780-1810 1810-1840
CONSTANT 0.392*** 1.007*** 1.212*** -0.648*** -0.178 -2.102***
(0.059) (0.137) (0.207) (0.244) (0.241) (0.38)
TIME 0.001 -0.003 -0.009*** 0.011*** 0.005*** 0.016***
(0.003) (0.003) (0.003) (0.002) (0.002) (0.002)
N 144 304 400 355 680 509
ADJUSTED-R2 0.26 0.5 0.2 0.43 0.14 18
respectively.
To facilitate the interpretation of findings, Table 15 summarises the results by showing

the sign of the coefficient of the time trend variable and whether it is significant at the 10
percent level.
Table 15- Fixed Effects Results for 30 year intervals (Time trend)
Summary
1660-1690 1690-1720 1720-1750 1750-1780 1780-1810 1810-1840

Ottoman Empire +* +* -* +* -* N/A
Eastern Med. (Entire) + + -* +* +* +*
Ottoman Med. + + -* +* -* +
Adriatic + - -* +* +* -*
Ott. Med.-Adriatic + - -* +* +* +*
Although for the late seventeenth century, the coefficient of TIME is positive in all the
samples – except in the regression based on the extended Ottoman sample – none of the
coefficients are significant at the 10 percent level. Thus, it can be concluded that the Ottoman
Empire as a whole suffered fragmentation from 1660 to 1690, while trading costs remained
stable in the other four samples. Within the Ottoman realm, market deterioration on an inter-
regional level continued between 1690 and 1720.
Between 1720 and 1750, following the end of the long Venetian-Ottoman naval wars
in the region, a period of integration in domestic and international markets emerged in the
Eastern Mediterranean. We can see a strong declining trend in trade costs within the Ottoman
Mediterranean, the Adriatic, and across Ottoman Mediterranean-Adriatic wheat markets
during this period of peace in the Eastern Mediterranean waters. Similarly, the entire Eastern
Mediterranean region, as well as the wider Ottoman geography, became more integrated
during this period. In all the five samples, the coefficient of TIME is negative and significant at
the 1 percent level. The coefficients of the explanatory variable, signifying the magnitude of
the annual change in the aggregated price relatives are, respectively, -0.017, -0.009, -0.013, -
0.007, and -0.009, for the Ottoman Empire, Eastern Mediterranean, Ottoman Mediterranean,
Adriatic, and the Ottoman Mediterranean-Adriatic wheat markets. Thus, the decline in the price
gaps was largest in the entire Ottoman sample, and the Ottoman Mediterranean.
However, in the absence of adequate price data for the period prior to the 1660s, we
cannot be sure what peace-time levels were during the periods of long maritime peace that
occurred during the Ottoman-Venetian coexistence between the last quarter of the sixteenth
century and the mid-seventeenth century. Therefore, it is impossible to ascertain whether

wheat markets across the Eastern Mediterranean around the mid-eighteenth century attained
higher levels of integration than they had 150 years previously.
The 1750-1780 period, was marked by a general trend of disintegration. Regional and
inter-regional wheat markets in the Ottoman Empire, as well as international markets in the
Eastern Mediterranean, became significantly fragmented, as is demonstrated by the positive
coefficients, all significant at the 1 percent level. Adriatic regional markets also experienced
disintegration during the same period, yet the magnitude and the statistical significance of the
coefficient of the time trend variable in this sample was lower.
Particularly during the Russo-Ottoman war of 1768-1774, price dispersion reached

high levels in Ottoman wheat markets. Several historical sources report unprecedentedly high
grain prices, shortages and famine in the capital during the war years, a phenomenon that is
indicative of the dissolution of the provisioning network centred on Istanbul. The ground
battles between the two powers were fought on the Danubian front, inflicting heavy damage
on the region’s production and markets. The Russian occupation of the Romanian provinces
during the war years severed the already weak ties between the Balkans and Central Europe.
In the Romanian provinces, occupation and fear of servitude drove many peasants into
Habsburg territory.
A series of naval battles in the Aegean and the presence of Russian ships that had
sailed from the Baltic Sea, through Gibraltar, and across the Western Mediterranean to the
Ottoman Mediterranean appears to have affected trade in the Eastern Mediterranean and to
have precipitated the fragmentation of inter-regional markets in the Eastern Mediterranean
and regional markets in the Ottoman Mediterranean.
By roughly 1780, Adriatic and the Ottoman domestic markets began to follow
diverging trajectories. While Ottoman wheat markets experienced a relative recovery in the 30
years following this date, Adriatic regional markets were impacted by the political decline of
the Venetian Republic and experienced fragmentation.
At the same time, the trend towards recovery is not visible in international wheat
markets. After 1750, international trading costs continuously rose until the mid-nineteenth
century. Multiplication of export bans on essential foodstuff, grains being in the first place, is
a commonly accepted fact that explains this episode of price dispersion on the international
scene during this period.
It should also be noted that the extent of the recovery in Ottoman regional and
interregional markets was more restricted compared to the extent of the fragmentation in the
previous period. While between 1750 and 1780, the annual rates of increase in the price gaps
in the Ottoman Empire and the Ottoman Mediterranean were 1.4 and 1.9 percent, respectively,
the annual rates of decrease in the following period were 1 and 0.6 percent.
Adriatic regional markets became more integrated in the early nineteenth century.
Venice’s occupation at the hands of the Napoleonic armies, before the city’s fall in 1797,
ushered in a new era in the Adriatic. Austria seized Venetian possessions in the Balkans and to
the east of the Adige rivers in accordance with the peace with France, while France annexed
the Lombard part of the state. In the Adriatic, the rise of the Austrian port of Trieste had already
begun in the 1760s. With the fall of Venice, Adriatic markets began to move towards further
cohesion under Austrian influence. During the same period, price gaps in the Ottoman
Mediterranean remained stable, yet the region was cut off from Adriatic markets.
5. Impact of grain policy on market integration
In this section, I estimate the difference in the level of price gaps across four sub-
periods, separated by benchmarks in terms of Ottoman grain policy, in order to understand
whether the extent of integration varied in compliance with the degree and method of
regulation in the grain markets. To do this, I examine the Ottoman grain-trading network that
covers 11 markets (the Eastern Western Balkans, Wallachia, Salonika, Edirne, Istanbul, Bursa,
Manisa, Candia, Patmos, and Cairo) in the database during the 1675-1815 period, for which
the price series are most complete.
The 1675-1815 period can be divided into four sub-periods according to the grain
policy tools applied and the degree of regulation in the grain markets.
Our first benchmark year is 1711. In 1711 and 1714, the Moldavian and Wallachian
princes elected by the boyars were replaced by the Phanar-dependent Greeks of the Porte.
The establishment of the Turkish-Phanariot regime represented a milestone in terms of the

Ottoman control over the two Danubian principalities. With the incorporation of the
principalities into the empire’s political and economic system, the Ottoman centre reinforced
its monopoly over the main products of the principalities (Alexandrescu-Dersca Bulgaru 1992).
This political dominance over the region allowed a substantial portion of its grain surplus to
be transferred to the capital via non-market extraction methods. In the following years, the
Danubian provinces and the Eastern Balkans turned into the empire’s “cellar,” satisfying the
predominant share of the capital’s need for grains, while the role of the Mediterranean shores
in the provisioning system became increasingly marginalised (Alexandrescu-Dersca Bulgaru
1992).
As such, our first period, 1675-1711, signifies an era preceding the full incorporation
of the Danubian Principalities into the Ottoman political and economic orbit, during which
Western Anatolia, the Western Balkans, and Egypt played a relatively more important role in
the provisioning of Istanbul with grain. As Ağır (2013) highlights, due to geographical factors,
the state’s capacity to enforce export bans and internal barriers on grain trade on the
Mediterranean coasts was limited. Since smuggling grain from the Mediterranean coasts
through the archipelago was easy, the central government had to take market forces into
account in order to procure necessary quantities. Therefore, it can be claimed that before the
eighteenth century, the central government’s efforts to supply the capital was not “conducted
according to very strict regulations (Salakides 2002: 131),” and a less coercive price policy was
applied.
This situation, however, began to change when the quantities of grain brought to
Istanbul failed to satiate the capital’s rising demand for grain (Salakides 2002). At the turn of
the seventeenth century, Istanbul faced a wave of immigration with the abandonment of land
in the neighbouring vilayets of Anatolia and Roumelia. The population of the capital rapidly
grew, and as the difficulties in provisioning the capital’s inhabitants increased, political
tensions surfaced and several bread riots erupted. The pressure of meeting the needs of the
city’s inhabitants led the Ottoman state to apply a more coercive grain policy (Aktepe 1958;
Alexandrescu-Dersca Bulgaru 1958).
As revealed by several imperial decrees imposing measures to control Istanbul’s

population and ordering grain and other essential foodstuffs to be transferred from the
European provinces to the city, the problem became exacerbated during the initial decades of
the eighteenth century. In parallel to the rising pressure, forced purchases at fixed prices, which
were implemented in previous centuries only during times of war, began to be applied for the
purpose of provisioning Istanbul with essential foodstuffs, particularly grain (Aktepe 1958;
Alexandrescu-Dersca Bulgaru 1958). The practice was known as the miri mübayaa regime,
which imposed a tax-like levy to facilitate the provision of goods and services for the state at
a price significantly lower than the market price. The policy particularly concerned the
European provinces (Ağır 2011). Hence, our second period, 1711-1748, witnessed the stiffening
of the Ottoman grain policy, which was imposed by the growing problems of the capital and
facilitated by the direct control over the Danubian provinces.
Our third period is 1748-1773. In 1748, a new system of grain deliveries, known as a
comparative quota system (mukayese nizamı), was introduced in the zones traditionally
designated as the hinterland of the capital, more particularly in Ottoman Europe. With this
compulsory system of deliveries of wheat, the transfer of grain surpluses from the hinterland
to Istanbul was systematised. Through an investigation of Istanbul’s central grain market
registers, which featured data about the amount of grain sent to the capital in previous years,
the authorities determined how much surplus each district could produce in normal years.
According to these estimates, each district was assigned a quantity to be delivered to Istanbul
(Ağır 2013).
In this system, differing amounts of grain were purchased at the miri (pre-determined
fixed price, which was substantially lower than the market price), and rayiç prices (determined
through negotiations between state agents and producers and set at a level lower than the
market price but higher than the fixed price), depending on conditions of supply and demand
(Ağır 2011).
From an early date onwards, the obligation to obtain state permission to purchase,
transport, and sell grain was always a key component of the provisioning policy. Until roughly
the mid-eighteenth century, such permission was granted to merchants by the central
government on a case-by-case basis following the application of the concerned merchant via
local authorities of the region where the grain was to be sold (Güçer 1952). Such a practice
accorded considerable power to these authorities and prominent members of the local
community, particularly the judge who was responsible for regulating grain prices and exports.
Concurrently with the introduction of the quota system, the state permission system
was transformed into a licence requirement, restricting the privilege to provision Istanbul with
grain to a group of officially authorised private merchants and government agents (Ağır 2011).
According to Alexandrescu-Dresca Bulgaru (1992), this new regulation was a response by the
central state to numerous complaints from inhabitants of the rural Eastern Balkans regarding
abuses by the merchants and state officials that ultimately resulted in several limitations to
their conduct.
Our last period is 1774-1815. 1774 marks the end of the Russo-Ottoman war of 1768.
With the Treaty of Küçük Kaynarca terminating the conflict, the Black Sea was opened to
foreign trade, and the Romanian provinces, the traditional zone of grain supply for Istanbul,
gradually escaped from direct Ottoman control. As an increasingly greater share of the grain
produced in the region was oriented towards the Western Mediterranean markets, Western
Anatolia and the Mediterranean coasts became more important for provisioning Istanbul in
the last quarter of the eighteenth century. Still, the region continued to supply grain for the
capital, albeit at lower volumes than previously (Güran 1986).
According to Ağır (2011), this shift in the geographical patterns of supply came
together with a liberalisation in the Ottoman grain policy in the last quarter of the eighteenth
century. In 1783, the comparative quota system and the practice of purchase at controlled
prices were terminated. In 1793, the Ottoman Grain Administration – considered by Ağır (2011:
3) as the sign of a “policy shift towards a more-centralised-yet-flexible use of regulatory tools”

– was established, and the licence requirement to sell grain in Istanbul was lifted.
The fixed effects regression (5) estimates the level of aggregated price relatives for
these four sub-periods. The period 1748-1774, considered an era of intense regulation in the
Ottoman grain markets and characterised by the practice of a comparative quota system and
licence requirements, functions as the reference category and has been omitted from the
regression. In order to distinguish the alterations in trading costs due to institutional/structural

factors from aleatory and temporary shifts caused by external shocks, control variables (wars,
debasements, droughts, and plague years) have been incorporated.
(5)
𝑃1𝑡
|𝐿𝑛 ( )| = 𝛽0 + 𝛽1 𝑊𝐴𝑅_𝑁𝐴𝑉𝐴𝐿 + 𝛽2 𝑊𝐴𝑅_𝐼𝑁𝑇𝐸𝑁𝑆𝐼𝑇𝑌 + 𝛽3 𝐷𝑅𝑂𝑈𝐺𝐻𝑇 + 𝛽4 𝑃𝐿𝐴𝐺𝑈𝐸_𝐿𝐴𝐺
𝑃2𝑡
+ 𝛽5 𝑂𝐶𝐶𝑈𝑃𝐴𝑇𝐼𝑂𝑁 + 𝛽6 𝐷𝐸𝐵𝐴𝑆𝐸𝑀𝐸𝑁𝑇 + 𝛽7 1660 − 1710 + 𝛽8 1711 − 1747
+ 𝛽9 1774 − 1815 + 𝑢
Two separate variables are constructed for Ottoman wars. WAR_NAVAL is a dummy
variable assuming 1 for the years during which the Ottoman state fought in Mediterranean or
Black Sea waters. Given the centrality of sea transportation within the grain-trading network,
naval wars in the Mediterranean and Black Sea, which led to an interruption of maritime trade,
are expected to have a significant impact on the trading costs. Our second control variable,
WAR_INTENSITY, is the number of years per decade during which the Ottoman state was at
war on the European or eastern front. This variable is employed as a measure of intensity of
wars.
DROUGHT is a dummy variable for years in which extreme aridity occurred in the
Balkans or Anatolia according to climatological studies (Akkemik et al. 2007; Xoplaki, Maheras,
and Lutherbacher 2001). It assumes 1 if the year was a drought year, and 0 otherwise.
In constructing a dummy variable in order to assess the impact of local sporadic

shocks generated by plague, the ideal situation would be to detect the years in which the
epidemic was seen in each of the towns in the sample. However, determining when and to
what degree individual towns were affected is not possible due to the lack of detailed historical
information.
Here, I take years in which plague was observed in at least 10 towns in the Ottoman
Empire, according to the data compiled by Panzac (1985). In that sense, the variable
constructed here is a measure of the geographical spread of the epidemic, rather than the
severity or appearance in the towns in the sample. For the period 1700-1840, the dummy
assumes 1 if plague was detected in more than 10 cities in the Ottoman Empire in the particular
year, and 0 otherwise. For the period 1675-1700, which is not covered by Panzac’s (1985) study,
I included three episodes of the epidemic, 1677-80, 1687-90, and 1697-99. Even though it is
unknown how many Ottoman towns were hit by the disease during these years, it is known
that it occurred in Southeastern Europe, Anatolia, and Damascus concurrently (Ayalon 2008;
Kostis 1995; Varlık 1991). Thus, it is assumed that these three episodes affected large areas
around the Eastern Mediterranean.
It behoves us to note, however, that the geographical spread of the epidemic, which
the variable proxies for, risks reflecting intensified interregional trade since the plague spread
more rapidly across space during periods of lively commercial activity. Cipolla (1974) identified
travelling merchants and population density as the main transmitters of the disease among
cities. Boerner and Severgnini (2011), for instance, use the spread of the Black Death from
1347-1351 as a proxy to study trade among different cities. If plague spread during periods in
which trade intensified, the plague variable could emerge to be negatively correlated with
price differentials. Assuming that the interruption of trade due to plague took place sometime
after the appearance of the disease in a particular locality, I constructed PLAGUELAG by
delaying the PLAGUE variable by one year so as to avoid this problem.
Debasement was another instrument of the state to extract resources and finance
warfare. The late eighteenth and early nineteenth centuries were marked by frequent
debasements and high inflation. How quickly local markets responded to the changes in the
silver or gold content of the money is unclear in the absence of relevant research. However,
we can assume that under the conditions of limited communication technology in the early-
modern period, the adjustment procedure was not immediate despite the monetary
unification in the late seventeenth century. This likely resulted in discrepancies at the local
money markets, which in turn were reflected in profitable price differentials across regional
commodity markets. DEBASEMENT is a dummy variable assuming 1 for debasement years, as
indicated in Pamuk (2000a).
The provisioning system broke down when the flow of deliveries was interrupted in
periods of territorial occupations in the regions supplying Istanbul. We know that during the
Austrian occupation of Western Wallachia (1718-1739), the Russian occupations of Moldavia
and Wallachia (1711, 1768-1774; 1787-1792; 1828-1834), and Napoleon’s occupation of Egypt
(1798-1801), inhabitants of the capital were affected by grain shortages and high grain prices.
Therefore, a dummy variable has been introduced for occupations. OCCUPATION assumes 1
for years in which one or more regions within the grain supplying network were under foreign
occupation, and 0 otherwise.
Table 16 reports the descriptive statistics of the regression variables.
Table 16- Descriptive statistics of the regression variables
Variable N Mean S.D.

|Ln(P1/P2)| 992 0.602 0.548
WAR_NAVAL 992 0.264 0.441
WAR_INTENSITY 992 4.513 2.836
DROUGHT 992 0.339 0.474
PLAGUELAG 992 0.396 0.489
DEBASEMENT 992 0.130 0.337
OCCUPATION 992 0.398 0.490
1660-1710 992 0.162 0.369
1710-1748 992 0.192 0.394
1774-1815 992 0.539 0.499
The fixed effects regression results are presented in Table 17. At the 10 percent level,
the coefficients of NAVALWAR, WARINTENSITY, DROUGHT, PLAGUELAG, and OCCUPATION
are statistically significant.
Table 17- Fixed Effects Regression Results

Grain-trading network, 1675-1815
Dependent variable |Ln(P1/P2)|

Coef. Std.Err.
CONSTANT 0.405*** 0.056
NAVAL_WAR 0.108** 0.052
WAR_INTENSITY 0.017** 0.008
DROUGHT -0.072** 0.037
PLAGUELAG 0.075** 0.037
OCCUPATION 0.131*** 0.039
DEBASEMENT 0.081 0.054
1675-1710 0.146* 0.078
1711-1747 0.301*** 0.068
1774-1815 -0.065 0.067
N 885
ADJUSTED-R2 0.26
As expected, war intensity and naval wars in Ottoman waters were positively
associated with price gaps in the Ottoman grain-trading network. During years of Ottoman
naval wars, trading costs increased by 11 percent. The results confirm the significance of sea
transportation within the Ottoman grain network and the disruptive effect of wars on markets
and posit the interruption of maritime trade due to naval clashes on the Mediterranean and
Black seas as an important cause of market fragmentation in the region.
Interestingly, droughts exerted a negative impact on the price differentials. DROUGHT

is significant at the 5 percent level. During drought years, estimated bilateral price relatives
were reduced by about 7 percent. By concurrently affecting local markets and influencing local
supply conditions, climatic shocks might have caused grain prices to converge across local
markets. On the other hand, provisionist policies that were implemented particularly during
times of supply difficulties might also explain this result. Kütükoğlu (1996) suggests that when
an essential foodstuff was scarce in a local market to a degree that it negatively impacted
everyday life, the import of that good from abroad as well as its internal trade was exempt
from both internal and external customs. These findings support Kütükoğlu’s argument relying
on narrative sources.
Again in accordance with our expectations, plague epidemics and territorial

occupations created an upward shift in trading costs. Following the years of the beginning of
plague epidemics and during territorial occupations, trading costs rose respectively, by 7.5
percent and 13 percent, respectively. Finally, although the coefficient of the DEBASEMENT
dummy is positive as expected, it is not significant at the 10 percent level, suggesting that
debasements did not have a substantial effect on the price gaps. This might be indicative of
relative well-integrated monetary markets within the Ottoman trading network centred on
Istanbul.
One of the purposes of this exercise was to detect whether wheat markets in the
Ottoman grain-trading network became more integrated in the post-1774 era, marked by
changes in provisioning policies. After controlling for the impact of wars and external shocks,
no statistically significant difference can be detected in the aggregated relative prices between
the 1748-1773 period, during which a comparative quota assessment system and licence
requirement were in practice, and the post-1774 period, which witnessed a shift in
geographical patterns of supply, and arguably, the relaxation of the grain policy. The
coefficient of the 1774-1815 dummy is not significant at the 10 percent level. These findings
do not provide support to Ağır’s (2013) argument that authorities began to liberalise Ottoman
grain policy in the late eighteenth century with the lifting of several restrictions to free trade.
If policy changes occurred in this direction, they were not sufficient enough, along with several
other influences, to lower trading costs and lead to better-integrated wheat markets across
the Ottoman grain-trading network centred on Istanbul.
Another striking finding concerns the 1711-1748 period. The coefficient of the
dummies for this period is significant at the 1 percent level. For this period, the estimated
relative price within the grain-trading network was 30 percent higher than the estimated
relative price for the 1748-1774 period. The substantially wide gaps in wheat prices between
the Balkans and the Mediterranean coasts in the early eighteenth century appear to be
responsible for the difference.
A closer inspection of the price series shows that during this period, wheat prices in
the Balkans reached sky-high levels, condemning the masses to famine (see Figure 4). In fact,
the subsistence crisis in the Balkans that marked the first half of the eighteenth century has
been documented by several studies. While the traditional explanations posit climatic factors
and over-exploitation by the Ottoman state as the source of the crisis, Kostis (1993) and
Gounaris (2009) claim that the shortage of grain and the high prices were ultimately linked to
the inability to attract random imports. This, they argued, was due to the relevant conflicting
interests of a network of local actors who played a part in the regulation and operation of the
grain trade in a local setting (including the local judge, who was the official regulator of prices
and exports; local janissary aghas, who controlled local grain trade; local notables, who
possessed high quantities of surplus grain; and French consular and captains who performed
and organised grain trade operations across the Mediterranean). Given the large price
differentials between the Balkans and other parts of the Ottoman grain-trading network during
this period, the explanation offered by Kostis (1993) and Gounaris (2009) seems much more
plausible than the traditional view.
The involvement of local agents in the administration of grain trade between different
parts of the empire should be considered within the overall political and financial context. As
Yıldırım (2003) rightfully claims, “the provisioning of grain to Istanbul was not insulated from
the official policy of the state to turn its sources of revenue to tax farms during the eighteenth
century. Many offices involved with the reception, taxation, redistribution and supervision of
the grain supply were farmed out to private entrepreneurs for a short-term or on a life-long
basis.”
As Gounaris (2009) has presented shown through a detailed account of Salonika, the
interplay between conflicts of various local interest groups led to coordination failures and
high transaction costs impeding the flow of grain from where it was available to where it was
scarce. In this respect, the subsistence crisis in the Balkans in the first half of the eighteenth
century is an excellent demonstration of how coordination failures observed in decentralised
structures relying on a multiplicity of intermediaries led to inefficient and fragmented markets,
as Epstein (2000) argues.
In light of these new findings, the comparative quota system and the licence
requirement, the main tools in the Ottomans’ mid-eighteenth century grain policy, can be re-
assessed. Both of these tools can be considered as a step towards a more centralised
administration of Ottoman grain trade that culminated in the establishment of the Ottoman
grain administration in 1793. The comparative quota system, which implied that the amounts
to be sent by each district to the capital were centrally determined, and the licence system,
which created a privileged class of “trustworthy” private merchants centrally authorised to
provision Istanbul with grain, likely curtailed the power of the local authorities and
intermediaries in the provisioning system to some extent, while helping limit their arbitrary
conducts. This notwithstanding, local intermediaries continued to play a substantial role in the
regulation of the grain trade. As such, these tools likely helped reduce coordination failures
and lower transaction costs.
6. Robustness check
In order to check the robustness of the findings presented above, results from the
fixed effects regressions are compared with results from an analysis of the coefficient of
variation. In a further step, decennial aggregated relative prices are estimated for alternative
samples, excluding the capital, and restricted to the markets with the most complete price
series in the database to understand whether these produce consistent findings with the larger
samples included above.
6.1. Coefficient of variation analysis
Here, I look at the evolution of price dispersions in wheat markets by employing

coefficients of variation, an alternative indicator used to measure the extent of integration.
Initially, I apply an ADF test to determine whether the coefficient of variation series is
stationary. The null-hypothesis of the ADF test is non-stationarity. If the null-hypothesis of
non-stationarity cannot be rejected at the 10 percent level, the series is considered stationary,
and the logged coefficient of variation is regressed against a time trend variable to deduce
whether a statistically significant decreasing/increasing trend in price dispersion is observable
and to estimate the rate of convergence/divergence in wheat prices across different markets.
Table 18- ADF test and OLS regression results
ADF t-value TIME

Ottoman Emp. -3.210 Null hypothesis rejected
Ottoman Med. -2.492 Null hypothesis cannot be rejected -0.004***
Adriatic -1.016 Null hypothesis cannot be rejected -0.012***
Eastern Med. -1.874 Null hypothesis cannot be rejected -0.002***
Table 18 reports the ADF test and OLS regression results. Figures 13 to 15 depict 15-
year moving averages of yearly coefficients of variation of wheat prices in the Ottoman Empire,
Ottoman Mediterranean, Adriatic and entire Eastern Mediterranean regions.48 In the complete
Ottoman sample, the null-hypothesis of non-stationarity is rejected at the 10 percent level,
ruling out a long-term trend of integration/disintegration. In the Adriatic, Ottoman
Mediterranean, and the entire Eastern Mediterranean samples, the null hypothesis of non-
stationarity is not rejected at the 10 percent level, while the coefficient of the TIME variable is
negative and significant. However, as depicted in Figure 15, only in the Adriatic region is the
price dispersion in the early nineteenth century below its late-seventeenth century level. In the
case of the Ottoman Mediterranean and the Eastern Mediterranean as a whole, the coefficients
of variation of wheat prices in the first decades of the nineteenth century are slightly higher
than the initial level. The time trend detected in these series apparently stem from episodes of
fragmentation and integration in the early eighteenth century which the linear regression
model failed to capture. In sum, as in the fixed effects regression employing the bilateral price
relatives as the dependent variable, long-term integration in wheat markets is only observable
in the Adriatic region.
48 Since a coefficient of variation is a measure of dispersion within a sample, it cannot be separately

applied to the Adriatic-Ottoman Mediterranean market pairs.
Figure 13- Coefficient of variation (15-year moving averages)

Ottoman Empire

Entire Eastern Mediterranean

Ottoman Mediterranean and Adriatic
The short-term trends identified by employing two different methodologies also

overlap. According to the coefficient of variation analysis, the late seventeenth and early
eighteenth centuries appear as an era of fragmentation in the Ottoman realm and the Adriatic
region, as well as in the Eastern Mediterranean as a whole. From around the 1720s, wheat
prices converged until the 1760s. In the Adriatic region, from this date onwards, the price
dispersion remained stationary until the early nineteenth century before falling continuously
until 1840. The 1760s appear as a break point for Ottoman markets. In the complete Ottoman
sample, prices diverged in the last decades of the eighteenth century. In the Ottoman
Mediterranean, the impact of the shock caused by the 1768-1774 Russo-Ottoman war was
reversed in the following decades, and price dispersion remained stable in the nineteenth
century.
6.2. Alternative samples
2.1. Is Istanbul distorting results? Ottoman Empire and Ottoman Mediterranean without
Istanbul
The general framework assumes that trade flow runs from low-price to high-price
regions. Although this assumption appears valid for trade between most Ottoman towns,
Istanbul presents an important exception. Due to state policies aiming to keep the price of
grains affordable in the capital, the price of wheat in Istanbul was lower than the hinterland
that provisioned the town. Here, I look at the evolution of the price gaps in the Ottoman
Empire leaving Istanbul out in order to understand whether the region outside the capital
followed a different pattern in terms of market integration trends.
Figure 14 compares aggregated decennial price gaps between the two samples. The
findings suggest that market integration/disintegration trends in the Ottoman Empire
excluding Istanbul were nearly identical with the sample including the imperial centre.
Similarly, as can be seen in Figure 15, the Ottoman Mediterranean sample without Istanbul
produced similar results with the sample incorporating the capital.
2.2. Are missing observations a source of error? Manisa, Edirne, Istanbul
Finally, considering the high number of missing observations in the database of

Ottoman prices, the aggregated price gaps are estimated for a restricted group of markets for
which relatively more complete price data is available. The price series from the Manisa,
Istanbul, and Edirne markets have respective coverage rates of 72, 52, and 45 percent. As these
markets are located in the Ottoman Mediterranean region, we would expect the results based
on this restricted sample to be similar to the enlarged Ottoman Mediterranean sample.

Ottoman Empire with and without Istanbul

Ottoman Mediterranean with and without Istanbul

Ottoman Mediterranean and Manisa, Istanbul, Edirne
The overlap between the short-term trends in the two samples is visible in Figure 16,
which compares decennial aggregated relative prices from Manisa, Edirne, and Istanbul with
the complete Ottoman Mediterranean sample. Since the average distance in this restricted
sample is smaller than the enlarged Ottoman Mediterranean sample, the estimated price gaps
are smaller, but the movement over time was very similar, suggesting that the omitted data is
not distorting trends in the full sample.
In this section, the robustness of the findings based on the fixed effects regression
employing bilateral price relatives as the dependent variable have been checked against the
results from a coefficient of variation analysis and fixed effects regressions conducted on
alternative samples. The evolution of coefficients of variation of yearly prices revealed no long-
term integration trend in wheat markets in the Ottoman Empire, Ottoman Mediterranean,
Adriatic and the entire Eastern Mediterranean regions. Like the fixed effects regressions, the
coefficient of variation analysis detected integration only in the Adriatic wheat markets. The
aggregated bilateral relative prices estimated for the samples (1) excluding the capital, which
had a distinctive price pattern, and (2) restricted to three markets which largely had the most
complete price series also support the findings in the original samples.
7. Conclusion
This study investigated changes in the extent of integration in regional, national, and
international wheat markets in the Eastern Mediterranean from 1660 to 1840. The results
revealed that this geographical area hosting the two leading powers of the pre-Columbian
world did not move towards a regional existence during this period, a strong trend of
integration in the Adriatic basin notwithstanding.
On the eve of the first wave of globalisation, domestic wheat markets in the Ottoman
Empire were no better integrated than they were in the second half of the seventeenth century.
Neither an extended sample that incorporates coastal and landlocked regions and short-
distance and long-distance markets, nor one restricted to the littorals of the Ottoman
Mediterranean produced evidence of a continuous and sustainable trend of decline in trading
costs prior to the mid-nineteenth century. Likewise, there is no sign of an emerging Anatolian
wheat market. As highly fluctuating prices in Konya and Damascus also suggest, before the
construction of railways produced a substantial impact, wheat markets in inland Anatolia and
Arab territories remained largely disconnected. Studer (2008) has shown that unlike long-
distance markets, short-distance European markets experienced long-term and sustainable
integration throughout the early-modern era, leading to the emergence of regional
economies, even in the absence of technological improvement. The lack of regional integration
in the Ottoman Mediterranean suggests that this could be a peculiarity of Europe and that
failed market development in the non-Western world could help to account for diverging
growth paths.
Similar results are evident at the international level. Despite the absence of a long-
term trend in the costs of international trade within the Eastern Mediterranean region, it should
be noted that the average trading costs between 1825 and 1840 were higher than they were
in the initial period between 1660 and 1675.
Looking at short-term trends produces other intriguing findings. Until the second half
of the eighteenth century, significant parallels were visible between the trends in the Ottoman
Mediterranean and in the non-Ottoman Adriatic region. In the late seventeenth/early
eighteenth century, regional, inter-regional, and international wheat markets in this part of the
Mediterranean were highly fragmented, which can be at least partially explained by the long
and consecutive maritime wars between Venice and the Ottoman Empire. The period of non-
aggression in the waters of the Eastern Mediterranean that began in the second decade of the
eighteenth century facilitated integration in the entire Eastern Mediterranean region, as well
as in the wider Ottoman geography.
During the Russo-Ottoman war of 1768-1774, Ottoman regional and inter-regional

wheat markets experienced significant market deterioration. The trend of recovery in the
following decades brought the Ottoman Mediterranean back to antebellum levels of
integration around the turn of the nineteenth century. Despite this recovery, the entire
Ottoman region did not return to mid-century levels, and relatively high trading costs marked
the second half of the eighteenth century. From 1750 to 1840, international wheat markets
and the entire Eastern Mediterranean region continuously fragmented, suggesting that the
Ottoman coasts increasingly decoupled from the non-Ottoman Eastern Mediterranean. While
trading costs in the Ottoman Mediterranean remained stable in the first half of the nineteenth
century, Adriatic wheat markets became increasingly integrated during this period, with the
price gaps declining below early eighteenth-century levels.
The estimated long-term and short-term trends in trading costs, and the results from
the multivariate regression analysis on the relative prices in the Ottoman grain-trading
network tie in well with the relevant literature on trade conditions in the Ottoman realm. Wars,
occupations, and plagues appeared as source of fragmentation in the wheat markets, and once
we control for these asymmetrical shocks, no significant difference was detected in the extent
of market integration between the 1748-1774 period, which was seen as a period of intense
regulation and state interference, and the post-1774 period, which arguably witnessed the
liberalisation of Ottoman grain policy. These result provide support Yıldırım’s (2003) assertion
that the Ottoman economic policy regarding grain markets did not show signs of liberalisation
before 1839, when the state adopted a free market strategy under foreign pressure.
Finally, the overlap between the general trends in market integration and the ups and
downs of the overall Ottoman economy are noteworthy. As revealed by the findings of this
study, the period of expansion in the first half of the eighteenth century was an era of price
convergence, while the period of economic turmoil and retraction in the second half of the
eighteenth century was marked by price dispersion and high price gaps in the entire Ottoman
Empire. It should be recalled that only steady and long-term integration processes can bring
about growth, as sufficient time is required for structural changes to occur in the economy. It
is difficult to comment as to whether a quarter century is sufficient time for such changes and
whether market integration played a role in the episode of expansion.
Indeed, it seems entirely probable that the causality ran from output growth to
integration. Brautaset and Grafe (2005) highlights scale economies in market efficiency as an
explanation for market integration. That is, price convergence is a function of the volume of
trade, since under constant transport technology, the fixed costs –particularly those related to
transportation – tend to decline as the latter increases. In line with this thesis, Sharp (2008)
claims that the main reason for declining price gaps between the United Kingdom and the
United States was the increase of American wheat supply. Similarly, it is quite likely that the
rise in agricultural and manufacturing output, the commercialisation of agriculture,
urbanisation, and population increase, which triggered an upward shift in aggregate supply
and demand and led to higher trade volumes, fostered integration across Ottoman wheat
markets between 1735 and 1765, while the reversal of this general economic trend in the later
part of the century led to disintegration.
162
SECTION 3
CONSUMPTION & PRICES
163
CHAPTER 4
WAS THERE A CONSUMER REVOLUTION IN THE OTTOMAN EMPIRE?
“Rüyası ömrümüzün çünkü eşyaya siner”

[Cause permeates into things our life’s dream]
Ahmet Hamdi Tanpinar
This paper offers a quantitative assessment of Ottoman consumption in the

eighteenth and nineteenth centuries in an attempt to contribute to the debates on early-
modern consumerism by exploring whether progress in material culture during this period can
be observed outside the Western world.
In the last three decades, a rich body of consumption literature based on probate
records has revealed that in the long eighteenth century, the ownership of consumer goods
increased substantially among European and North American societies despite the stagnation
and decline in real wage rates in the same period 49. This literature has posited early-modern
consumerism as an indicator of the social and economic transformations and a precursor of
the West’s modern society and economy. According to some historians, the changing patterns
of consumption and growing consumer demand in early-modern Northwestern
Europe prepared the ground for industrialisation by generating a strong impetus for the more
efficient production and distribution of goods; by leading to increased demand for cash
49 A large degree of inventory-based scholarship has focused on the American colonies and England.
For some examples on early-modern colonial America, see Carr and Walsh (1980, 1988); Horn (1988);
Lemon (1967); Main (1983a, 1983b, 1988); Main and Main (1988); Perkins (1991); Shammas (1980, 1990,
1997); Sweeney (1984); and Walsh (1983).
For a selection of works on material culture and consumption in early-modern England based on
evidence from probate records, see McKendrick, Brewer, and Plumb (1982); Overton et al. (2004); Earle
(1989); King (1997); McCants (2006); Pennell (1999); Shammas (1990, 1997); Thirsk (1978); Weatherill,
(1986, 1988).
For continental Europe, see De Vries (1975); Garnot (1995); Pardailhe-Galabrun (1991); Ramos (2004);
Roche (2000); Schuurman and Woude (1980).
Chapter 4-Was there a consumer revolution in the Ottoman Empire? 164
among ordinary people; and by stimulating them to work more (Gilboy 1932; Sombart 1967;
De Vries 1994, 2008; Horrell 1996). In several of these accounts, consumerism appeared as a
Western creation that was only exported to other regions of the world with the development
of modern economies and the export of industrialisation (Braudel 1985; Mukerji 1983; De Vries
1994, 2008).
More recent research, however, has attacked the idea of Western exceptionalism in
rising consumerism and pointed to the possibility of a worldwide pattern of increasing
consumption in the long eighteenth century. Burke (1993); Belk, Ger and Askegaard (2003);
Hanley (1997); Hanley and Yamamura (1977); Karababa (2006); and Pomeranz (2000) have
suggested that increasing concern with material culture was not particular to Western societies
and that the boundaries of early-modern consumerism can be broadened to include the non-
Western world. By showing that multiple consumer cultures began to develop across the globe
during the early-modern era, these studies have challenged a euro-centric account of the
“consumer revolution.”
Still, while the spread of consumer goods throughout Western societies has been well-
documented, evidence from other parts of the world is fragmentary and relatively rare. In the
absence of Asian counterparts to European probate inventories, the divergence in terms of
consumption of durables across different parts of Eurasia prior to the nineteenth century has
been primarily discussed on the basis of qualitative and anecdotal evidence (Pomeranz 2000).
In this regard, the Ottoman inheritance inventories provides us with a unique opportunity to
bring quantitative insights into pre-industrial consumerism in a non-Western context.
Based on quantitative evidence from probate inventories, this paper investigates the
evolution of household comfort and conveniences enjoyed by the inhabitants of the Ottoman
town of Üsküdar from 1700 to 1850. From the seventeenth century onwards, the Ottoman
Empire witnessed major sociopolitical and cultural transformations that many historians argue
underpinned the rise of a consumerist society in Europe. As the system of hierarchies that
characterized the Ottoman classical order was eroding, greater mobility among social and
professional groups led to new social and financial aspirations, and novel recreational and
cultural practices (Hamadeh 2004). On the other hand, Westernisation efforts that began in
the military and administrative spheres in early eighteenth century went hand in hand with the
deepening of Western economic and cultural influence on urban Ottoman society, which is
often associated with the introduction of a new attitude towards home and social life (Göçek
1993). Were these transformations in the sociocultural sphere accompanied by an increased
ownership of household goods? Was there a consumer revolution in the Ottoman Empire?
Ottoman historiography generally identifies the eighteenth century as the period

when Ottoman consumer culture started to develop. Quataert (2000) suggests that novel
consumption patterns were first introduced into the Ottoman Empire in the seventeenth
century with mounting coffee and tobacco consumption; continued to broaden and deepen
in the eighteenth century as ownership of consumer durables, particularly textiles, became
widespread; and developed further in the nineteenth century with the explosion of European
imports. Karababa (2006) demonstrates that in the seventeenth century, the Ottoman Empire
experienced the trickling down of certain personal goods that were restricted to the ruling
class in the mid-sixteenth century50. According to Artan (1998), the infamous Tulip Era 51 (1718-
1730), which is associated with the extravagant lifestyles of high-state officials, was also a
turning point for ordinary Ottomans, who began to be driven by a growing consumerist desire.
By looking at the types and quantity of goods possessed by the deceased, as reported
in the inheritance inventories, I ask whether the Ottomans in the mid-nineteenth century were
better off in terms of domestic comfort compared to their eighteenth-century counterparts at
the same level of wealth.
50 However, two weaknesses of this study should be underlined. First, illustrating the trickling down of
certain consumer durables is not enough to support the thesis of an early modern Ottoman consumer
revolution, which would require further evidence of an increase in the variety and quantity of the
personal and household goods consumed. That ordinary Ottomans in the mid-seventeenth century
possessed items that were reserved for the elites does not necessarily mean that there was an increase
in the overall number and variety of consumer durables owned by this group. Second, based on
Karababa’s (2006) findings, it can be suggested that the trickling-down process only concerned
personal goods, particularly clothing. As for the household goods, her results do not support the
argument.
51 The Tulip Era is a period in Ottoman history that lasted from the Treaty of Passarowitz on 21 July
1718 to the Patrona Halil Revolt on 28 September 1730. It was a relatively peaceful period, during which
one could argue that the Ottoman Empire began to orient itself towards Europe.
Evidence from probate inventories provides the empirical basis for the idea of a rise
in material consumption in the long eighteenth century. However, these sources have several
limitations, which might jeopardize the quantitative analysis. Notably, Clark (2010) suggests
that the growth of consumer goods as observed in the inventories in England might be a
reflection of the upward bias in the inventory samples, rather than a real indicator of progress.
As the frequency of probates declined between 1600 and 1750, he claims the characteristics
of the average testator in England were changing markedly in favour of men of higher status
and wealth. The average testator in 1750 was much better placed in the distribution of wealth
within society than the average testator of 1600. This implies that inventories in this period
might have exaggerated the society-wide changes in wealth and consumption.
In order to overcome similar doubts regarding the representativeness of Ottoman

inventories, this study pursues a different line of interrogation to those conventionally followed
in research on consumption. It investigates whether Ottomans in the mid-nineteenth century
who were no wealthier than their ancestors in the early eighteenth century were nonetheless
living in a richer domestic environment. By examining the ownership of household durables at
constant wealth levels, we can explore the change in the quantity and variety of the household
goods in the inventories that is not derived from rising wealth.
By comparing the rates of ownership of goods between two counties for individuals
in the same wealth categories, Overton et al. (2004) showed that wealth was not the major
explanation for the differences in material culture that they observed between Kent and
Cornwall. Even when the effect of difference in wealth between the two counties is removed,
the adoption of new commodities was still slower in Cornwall. This, they concluded, was due
to a “location effect,” which captures regional differences in the level of interest and
information about commodities, and in participation in an urban commercial culture.
This study applies the intuition underlying the analysis of Overton et al. (2004).
However, instead of making an across-space comparison to uncover a “location effect,” I
conduct an across-period analysis to grasp the “time effect,” which cannot be explained by
variations over time in wealth – variations in the wealth of the living population, as well as
possible variations in the degree of wealth bias in the sample. Put differently, by controlling
for the changes in wealth and some other characteristics of testators over time, I attempt to
distinguish the changes in the amount and variety of household durables in inventories that
occurred due to rising wealth from that which occurred due to other time-dependent social
and economic factors. Price reductions, product innovations, enhanced distribution and retail
networks, increased interest in consumer goods and changing consumer behaviour are some
of the possible candidates for why an increase in consumption occurred. 52 All of these
processes point to structural changes on the demand and supply sides of the economy and
are often closely associated with rising consumerism in the early-modern era.
According to the results obtained here, the growth of consumer goods in the long
eighteenth century, which De Vries identifies as a feature of Northwestern Europe, also
occurred in the Ottoman realm. The interiors of Ottoman houses grew richer and more varied
throughout this period. More importantly, from the second half of the eighteenth century
onwards, Ottomans who were not richer than their counterparts in 1700 owned a greater
quantity and variety of domestic goods, and thus, enjoyed higher levels of domestic comfort.
If increasing acquisition of consumer goods by all segments of the society independently of
wealth is considered as the trademark of the early modern consumer revolution, then the
Ottoman Empire experienced it in the second half of the eighteenth century.
When evaluating the results of this study, it should be recalled that the acquisition of
personal items might have followed a different pattern than the acquisition of domestic goods.
Karababa (2006) argues that due to the privacy of the innerhouse, which the Ottomans were
keen at protecting, social differentiation was sought through arrangements in clothing, rather
than in the household environment. If Karababa (2006) is correct, then it is quite possible that
the adoption of new consumer habits occurred in the field of personal goods even earlier.
In the subsequent sections of this paper, I provide a brief overview of the Ottoman
town of Üsküdar and its social and economic structure, and introduce the sample and
methodology used in the study before presenting the results.
52 Having said that, it should be emphasized that the wealth here is used as a control variable, rather
than an explanatory variable. Although the stock of consumer goods possessed on an estate is positively
associated with the total estate value, consumption is a ‘flow’ concept that relates to income rather
than wealth.
1. Setting the scene: Üsküdar
The interest in consumer durables and the ability to possess them are strongly related
to social hierarchy and class structure, urban commercial culture, proximity to networks of
trade and the availability of goods, and participation in the market economy, all of which vary
across localities and regions. Consumption patterns show significant discrepancies from one
town to another. Among others, Overton et al. (2004) detected a striking difference in terms
of the adoption of new material goods based on location. According to their results, whereas
more and more varied material goods reveal higher standards of comfort, new methods of
heating and cooking, and new ways of eating and drinking in Kent, the domestic environment
in Cornwall contrastingly became more impoverished between the early sixteenth and mid-
seventeenth centuries. Although our study focuses on a single region, we need to briefly look
at the socioeconomic landscape of Üsküdar in the interests of acquiring a comprehensive
understanding of the evolution of domestic comfort.
Located on the Asian side of the Bosporus and functioning as a gateway to Istanbul,
Üsküdar was one of four official divisions of the capital in the eighteenth and nineteenth
centuries, together with Galata, Eyüp, and the city of Istanbul proper. It was a commercial and
military hub, and a site for small-scale farming and the agricultural industry. The town was the
first link in the relay system on the trade routes connecting Istanbul to Anatolia, the Arab lands
and to Iran and, as such, it was the departure point for the pilgrimage to Mecca and the
campaigns to the east (Halaçoğlu 2002).
Only rough estimates exist regarding the demographic structure of the town during
the period. The conventional view suggests that Üsküdar followed the pattern that is common
to most Anatolian towns: urban growth in the sixteenth century, depopulation and recovery in
the seventeenth century, and resurgence in the first half of the eighteenth century with an
increasing population and new neighbourhoods (Faroqhi 1979). We do not know whether the
town lost its significance in the second half of the eighteenth century with the decline of trade
to the east and the expansion of the European economy. The Ottoman population census of
1885 records a total population of 95,667 in the town. In light of this figure, Bostan’s (2012)
estimate of 60,000 in the eighteenth century seems a little high. It is suggested that Muslims
constituted about two thirds of the overall population, while the other third consisted of
Greeks, Jews, Armenians and Turkish-speaking non-Muslims settled in specific areas. Half of
this population lived in the town centre, and the other half in the villages nearby (Bostan 2012).
It is reasonable to assume that the town’s socioeconomic structure was determined

by its strategic position for trade and its role in the supply chain around the capital. With
numerous caravanserais and inns constructed in the sixteenth century, Üsküdar was a centre
of regional trade. Stressing the importance of Üsküdar in channelling supplies into the capital,
Seng (1991: 27) suggests that the town served as a supply depot for Istanbul and that it hosted
a merchant community, which was specialized in the eastern trade: “The commercial class of
Üsküdar organized and dealt with internal or the eastern international trade destined for
Ottoman consumption, or more specifically, Istanbul consumption, and was distinguished
from the European merchants whose venue was trade to and with Europe and who settled in
Galata-Pera”. How this commercial class was affected by the decline in eastern trade from the
second half of the eighteenth century onwards is a question that remains to be investigated.
Üsküdar was at the same time a good example of urban agriculture oriented towards
urban consumption. The town provided Istanbul with fruits and vegetables, and it was
particularly known for its sweet grapes. Covered with vineyards, gardens, and orchards,
Üsküdar kept its semi-rural nature well into the early twentieth century. According to Bostan
(2012), in the seventeenth century, more than half of the total population made their living
from agriculture. Evidence from inheritance inventories suggests that this figure is also viable
for the eighteenth and early nineteenth centuries. Vineyards and gardens, and agricultural
means of production, such as ploughs, sickles, hoes, and oxen frequently appeared in the
estates, particularly of the poor and the middling groups. Owning agricultural land also seems
to have been an investment strategy for the wealthy. Evidence also suggests a developed
transportation sector in Üsküdar, in line with the town’s role as the supplier of the capital
(Mazak 2005). Bostan (2012) points to the high number of boat owners who also appear in the
estates. These point to high levels of participation in the market economy in the town.
Despite its semi-rural character, the town had a vibrant urban life, with numerous
shops and covered bazaars (bedestans), mosques, charity kitchens, religious schools, and
dervish lodges constructed and maintained by large pious foundations (Mantran 1986). The
town was the site of principal markets for bread, clotted cream, yogurt, grapes, fruit and fish,
as well as slaughterhouses. Tanners, lumber workers, arrow makers, pipe makers, candle
makers and suppliers of yogurt frequently appear in seventeenth-century court registers for
the town, suggesting that these were the most important crafts in Üsküdar (Seng 1991). During
the nineteenth century, a small-scale textile industry developed. Bostan (2012: 367) reports
that in 1867, 2,750 looms and around 3,500 textile workers were recorded. This picture is
complemented by the existence of köşks and palaces for the use of the sultan during his
eastern campaigns.
The town’s position on the outskirts of the capital and at the crossroads of the major
eastern trade routes was favourable for the supply of most of consumer goods. It would not
be incorrect to assume that many consumer durables, including imported goods that were
available in Istanbul markets, were accessible to the inhabitants of Üsküdar. The town
provisioned Istanbul, and the returning shipping could have been used to carry consumer
items imported from Europe (Solmaz 1979). Its proximity to the largest city in the empire, with
a lively urban commercial culture, likely also facilitated the transmission of information about
material goods.
As a middle-sized town, Üsküdar was likely to have offered its inhabitants the relative
anonymity of an urban environment, which according to Overton et al. (2004) provided more
opportunity to fashion an identity through consumption. Furthermore, the coexistence of
several religious communities and the presence of a commercial class alongside religious and
administrative bureaucracy likely resulted in a dynamic social life in which consumption could
act as a means of distinction.
But in discussing the consumer behaviour of the inhabitants, the pastoral character of
Üsküdar should be taken into account as well, since this might have affected the pace at which
new tastes and goods were introduced. It is often argued that urbanity encouraged the
emergence of a consumer culture. Nevertheless, several empirical studies (Overton et al. 2004;
Weatherill 1988) indicate the absence of a clear-cut distinction in the ownership of goods
between the rural and urban population. However, in England changes in income and its
distribution among the rural and urban classes might have been a significant factor
determining the variety and quantity of goods owned, particularly if the relative prices of
agricultural and manufactured goods changed significantly. Additionally, increased demand
for market-supplied goods and services is associated with the transition from self-sufficiency
to market dependence for households. A high level of market participation would be expected
among the rural population of Üsküdar, as most of the peasants were engaged in commercial
agriculture oriented towards the capital and, as such, they assumed a place in the market
economy.
With its advantageous geographical location with regard to the supply of goods and
high levels of market participation, Üsküdar cannot be considered a typical Ottoman town.
Thus, one needs to be cautious when expanding the results of an investigation on the adoption
of material goods in eighteenth and nineteenth-century Üsküdar to other parts of the empire.
It is highly likely that rural households in remote regions of the empire remained self-sufficient
for much longer and their mode of consumption shifted slowly and only as the result of
increasing involvement in commercialized agriculture, which allowed households to earn
money to spend on textile yarn or ready-made cloth.
2. Sample
2.1. Sampling strategy
For this study, I selected six periods each of ten years’ duration from 1700 to 1850.
Each period marks the beginning of a 30-year phase: 1695-1705; 1725-1735; 1755-1765; 1785-
1795; 1815-1825 and 1845-1855. In each period, I constructed a sample consisting of about
80 inheritance inventories from the town of Üsküdar. Initially, I surveyed all the available
inventories in the court registers belonging to the years 1700, 1730, 1760, 1790, 1820 and
1850, as well as the two years preceding and following these dates, recording the total estate
values of each along with relevant information on gender and religious status (For the details
of the inventories employed see Primary Sources section). The frequency distribution of these
inventories (for two years preceding and following the benchmark years) was divided into four
wealth groups according to the total estate values in constant akçe53. The boundary values
that defined each group are noted, and the resulting wealth brackets are given in Table 1.
53 The monetary values are deflated by Pamuk’s (2000a) price index, which takes 1489-1490 as the
base period.
Table 1- Wealth brackets (in constant akçe)
1700 1730 1760 1790 1820 1850

Q1-Q2 339 952 567 455 1,095 585
Q2-Q3 968 2,619 1,345 1,155 2,360 1,700
Q3-Q4 2,856 7,143 5,670 4,330 7,403 5,574
N 99 17 98 120 78 89
Subsequently, for each wealth quartile in each period, I randomly selected 20 inventories
from the court registers for the five years preceding and following the benchmark years. The
method applied in the selection of the inventories is designed to yield a stratified sample that
reflects the frequency distribution of available inventories in the court registers as accurately
as possible.
For the 1730 period, the number of available inheritance inventories in the registers is
less than 80. In this case, I used all available inventories. Yet, this is only a partial solution. As
will be seen in the analysis later, this sample generates some unusual findings (such as
unrealistically high mean wealth). On the other hand, a number of inheritance inventories that
depict only real estate appear in the Üsküdar court registers. It seems that in several instances,
the relatives of deceased individuals applied to the court merely to share immovable
properties but did not make the deceased’s movable assets subject to probation. As estates
are randomly selected, such inventories were included in the initial sample but were omitted
in subsequent steps, reducing the sample size (aside for 1730) to around 70 inventories. The
final sample consists of 380 inventories.
2.2. Composition of the sample
Wealth
As defined by Overton et al. (2004: 138), the wealth of a household consists of

“anything that members of the household possess which can be sold for money or used to
acquire other commodities: real property, durable goods, financial assets and human capital.”
Ottoman inheritance inventories present a more complete picture of estate owners’ total
wealth than seventeenth- and eighteenth-century British and American inventories do. These
sources report all movable and immovable assets possessed at the moment of death, as well
as the slaves, debts owed to and debts held by the deceased.54 Owning shares of real estate
was a common practice in Ottoman society and these shares were also recorded and inherited.
However, since married women had the right to their own possessions according to Islamic
property law, the total value of the estates reflects personal rather than household wealth.
An important problem concerns the representation of the poorest segments of society

in inventories. In the case of Ottoman inheritance inventories, there are no independent
sources that we can use to check the social representativeness of these sources. Therefore, it
is not possible to calculate the proportion of the population who had few possessions, and
thus who did not leave any inventories, as Overton (2014) did for British probate records by
linking inventories to the list of taxpayers. However, considering that inventories were
obligatory in Islamic law – regardless of the amount to be inherited – in the event that were
minor heirs, a pregnant wife or a missing heir, it would not be incorrect to assume the poor
appeared in these records to some degree. Accordingly, some of the inventories included in
the sample belonged to individuals who did not have any material possessions, except a few
personal and household goods of insignificant value.
In what follows, I compare real wealth series constructed based on Üsküdar

inheritance inventories with Pamuk’s (2000a) real wage series for construction workers in
Istanbul. Real wages are the best indicator available for long-term trends in real incomes in
the Ottoman realm (Özmucur and Pamuk 2002). Table 2 looks at the descriptive statistics of
wealth in constant akçe in each period. The “wealth” refers here to the net wealth of the
deceased, obtained by subtracting the debts owed by the deceased from the total estate value.
Real wealth, or wealth in constant akçe, is calculated by dividing the net estate values by
Pamuk’s (2000a) consumer price index. Figure 1 compares mean and median real wealth values
and real wages. The real wages given for each of the 10-year intervals are the average daily
wage rates for skilled and unskilled construction workers in Istanbul (Pamuk 2000a). To
facilitate a comparison with wealth, the wages are reported as 1,000 days’ wages.
54 However, it should be noted that only freely held lands mülk were subject to the inventory process,
while Islamic inheritance law did not apply to state lands ( miri), which constituted the majority of
agricultural lands well until the mid-nineteenth century. The peasants possessed only usufruct rights to
state lands, while the transfer of such rights was regulated by customary ( örfi) law (İnalcık 1953).
Table 2- Wealth in constant akçe (Total estate value/CPI)

Descriptive statistics
N Mean Median S. D. Min Max

1695-1705 70 2,989 949 5,023.41 64 26,776
1725-1735 31 4,780 2,883 5,078.56 226 17,813
1755-1765 68 3,504 1,590 4,569.16 267 20,361
1785-1795 74 4,040 1,385 6,745.29 160 41,775
1815-1825 69 6,582 2,606 12,069.34 123 86,341
1845-1855 68 6,116 2,076 9,910.13 149 57,791
The distribution of wealth in eighteenth- and nineteenth-century Üsküdar as observed

in inheritance inventories was highly positively skewed, with means greater than medians and
the standard deviations large. The general trend in average wealth was upward and, to a large
extent, reflected the real wages. Still, it should be noted that the relationship between income
and wealth might have been significantly altered during this period.
Figure 1-Wealth and wages (in constant akçe)
Nevertheless, the growth rate of mean and median wealth values from 1700 to 1730
was unrealistically high, suggesting an upward bias in the inventories included. Taken together
with the restricted number of observations from this period, this reflects the limits to this
sample. Table 3, which shows the across-period variations in the ownership of real estate as a
percentage of all observations, also confirms this statement. While in the other periods,
between 16 percent and 21 percent of all inventories contained real estate, about half of estate
owners had immovable property in 1730. It is likely that in the second period, the primary
motive of inheritors who applied to the court was to share the immovable properties of the
deceased. Thus, the wealthy who owned immovable property were overrepresented in the
available inventories in the registers.
Table 3- Ownership of real estate according to periods
Real estate owners as % of all

observation
1700 20.9
1730 48.1
1760 20.6
1790 15.9
1820 17.4
1850 16.8
At this point, it should be underlined that the question of how representative our
sample is of the living population in terms of wealth distribution, is in fact of relatively little
significance for the purposes of this study. While the affluent may have been overrepresented,
the poor also appear. For the type of analysis being conducted here, it is the latter condition
that is more important. This is one of the advantages of this study over previous research on
the ownership of goods from probate records, whose reliability depends on the correctness of
the assumption that the samples of probates employed provide an accurate picture of the
population.
Studies of the adoption of material goods also emphasize identity (gender, social
status, and occupation of individuals) as a determinant of material culture and consumption
patterns. If the latter were related to personal attributes, variations over time in the stock of
household chattels could also reflect across-period changes in the composition of the sample
in terms of the identity of the estate owners. For instance, if women left significantly more
household goods than men, the inclusion of more female inventories in the later periods might
lead us to falsely assume that such goods proliferated over time. For this reason, the sample
needs to be controlled as much as possible for factors that play on the ownership of household
chattels. In what follows, I briefly discuss the relationship between different aspects of identity
and ownership of goods and provide a description of the sub-samples with regards to the
personal attributes of the estate owners.
Gender and religious status
The gender and religious status of the estate owners can be safely obtained from
Ottoman inheritance inventories. It is often noted that in order to study gender differences in
consumption and material culture, British and American probate records and wills are not ideal
sources. Only unmarried women (primarily widows) made wills or had inventories, because
according to the common law, the possessions of a married woman belonged to her husband,
as long as he was alive (Overton et al. 2004). As a result, the number of female inventories and
wills that exist is considerably lower than that of male inventories, and for several regions and
periods, there are not sufficient for gender-based comparisons to be made (Shammas 1980).
An equally important limitation concerns the upward bias in the female estates. Widows,
whose property was probated at a later stage of the life cycle than married women, constitute
the great majority of the female estate owners. The inventories of this group of women include
both what they accumulated themselves and what had been passed down to them by their
husbands and fathers (Berg 1996).
Ottoman inheritance records provide a more favourable ground to study gender

differences, as women’s property rights were guaranteed by Islamic law, and the estates of
adult women of all ages and social origins were brought to the court, like those of men. As for
non-Muslims, churches and synagogues were responsible for distributing the inheritances of
members of their communities. However, non-Muslims were free to apply to the court for the
distribution of the estates. They applied to the kadı to distribute their inheritances especially
when taxes paid to the kadı were lower than the ones paid to synagogues and churches, or
when the distribution of the estate according to Islamic law was more profitable for the heirs
(Karababa 2006: 75).
In Üsküdar during the 1700-1850 period, both women and non-Muslims were only
slightly underrepresented among the subjects of inventories, as can be seen in Table 4 and
Figure 2. The share of women and non-Muslims within the overall estate owners is 42.6 percent
and 30 percent, respectively. Muslim men were the focus of 40 percent of all the inventories
in the sample, followed by Muslim women (30 percent), non-Muslim men (17.4 percent), and
non-Muslim women (12.6 percent). Considering that non-Muslims constituted about one third
of the inhabitants of the town, it seems that our sample provides a good representation of the
population of Üsküdar in terms of gender and religious status.
Figure 2-Distribution of the sample according to gender and religious status
Table 4-Distribution of the sample according to gender and religious status
Muslim men Non-Muslim men Muslim women Non-Muslim women

N % N % N % N %
1700 28 40.0 5 7.1 33 47.1 4 5.7
1730 8 25.8 12 38.7 6 19.4 5 16.1
1760 26 38.2 14 20.6 24 35.3 4 5.9
1790 26 35.1 17 23.0 8 10.8 23 31.1
1820 34 49.3 8 11.6 23 33.3 4 5.8
1850 30 44.1 10 14.7 20 29.4 8 11.8
TOTAL 152 40.0 66 17.4 114 30.0 48 12.6
Social status and occupation
Overton et al. (2004) suggest that in England social status and occupational categories
appear to be more strongly associated with the acquisition of new kinds of material goods
than wealth was. The acquisition of new material goods was linked both to the status hierarchy
of gentlemen, yeomen, and husbandmen and to occupations associated with the service and
retail trades, reinforcing Weatherill’s (1988) findings. In the existing literature, there are no
studies investigating how consumption was related to social status in the Ottoman society.
Unfortunately, the information provided by Ottoman inheritance inventories do not

allow us to control for differences in the occupation of the estate owners. The occupation of
the deceased was not systematically indicated in the inventories. These sources recorded the
economic activity of the deceased only occasionally. In most other instances, we lack this
important information. On the other hand, religious and official titles, which signalled the
position of Ottoman individuals in their society, were always reported in the estates together
with the name of the deceased. Having said that, the meanings of these titles in terms of the
social status of the individuals who possessed them underwent a transformation in the
seventeenth and eighteenth centuries.
In the context of seventeenth-century Ayntab, Canbakal (2007: 63-4) speaks of an

“inflation of honours.” She demonstrates that the distribution of titles and ranks was used by
the Ottoman state as a strategy to achieve a stronger degree of integration between the centre
and the provinces. As more and more people acquired religious and official titles, which were
indicative of membership in the tax-exempt ruling class, 55 the social prestige these titles
represented for their owners eroded.
The evidence from inventories suggest that in Üsküdar, “inflation of honours” did not
occur in the second half of the seventeenth century, but was a phenomenon of the second
half of the eighteenth century. As can be seen from Table 5, very few individuals acquired
religious or official titles during the first half of the eighteenth century or even by 1760. In
contrast, from the second half of the eighteenth century onwards, a greater share of Muslim
men and women were granted these titles. The expansion of the military class in the town did
not indicate a change in the composition of the population, but a change of the social
55 During the second half of the seventeenth century, there were three points of entry to the military
class: acquisition of revenue grants with or without performing civil service, claiming descent from the
Prophet Muhammad or acquiring membership in a military corps (Canbakal 2007: 63). These three
points were associated with religious and official titles. Religious titles included şerif and seyid and could
be attributed to both Muslim men and women, whereas most common official titles consisted of bese,
ağa, and efendi, and were only bestowed upon Muslim men.
significance of the honorifics. Likewise, the religious titles of seyid or şerif did not refer to the
same position in the social hierarchy in 1760, when only 6 percent of Muslim estate owners
had these titles, and in 1850, when almost one third of the inventories of the Muslims claimed
to be descendants of the Prophet.
Table 5- Estate owners with titles
Official title* Religious title**
N % N %
1700 4 14 0 0
1730 1 13 1 7
1760 14 54 3 6
1790 22 85 5 15
1820 28 82 10 18
1850 27 90 15 30
ALL 96 63 34 13
Notes: *As percentage of Muslim males

**As percentage of Muslim males and females
3. Ownership of household goods according to categories
3.1. Measuring household amenities
In order to observe how household comfort evolved over time, I examined the variety
and quantity of household chattels reported in the inventories. In the first stage of the analysis,
I look at the change in the possession of different categories of household goods by using
frequency counts, to offer a detailed picture of the Ottoman material culture in 1700-1850.
Inventories enable material culture to be quantified in a number of ways. Frequency counts,
based on the presence or absence of objects in the estates, are the most straightforward and
common way to arrive at such a quanitification (Overton et al. 2004). This is also a particularly
effective method in showing the spread of new goods throughout the society.
In the second stage of the analysis, I employ an index of amenities, an item-by-item

counting method that identifies the presence or absence of consumption goods. This method
helps reduce the data to a manageable size that generates easy comparisons. This method is
employed by Carr and Walsh (1980; 1988), Main (1988), and Göçek (1996), who sought to
measure the overall progress in the acquisition of amenities. I included 12 groups of goods in
this index: bedding items, sitting implements, storage equipment, floor coverings, lighting
and heating utensils, household linen, mirrors, clocks, modern furniture, and serving utensils.56
The particular items that each group includes are presented in Table 6. The first six groups
(bedding, sitting implements, linen, storage, floor coverings, and linen) on the list are
established goods that made part of an ordinary Ottoman household in the beginning of the
eighteenth century. These can be assumed to be minimal equipment for comfort and
cleanliness. The next six groups (lightening and heating utensils, mirrors, clocks, modern
furniture, and crockery, plates, and cutlery) appear rarely and only in the estates belonging to
the rich in the initial period. These goods were associated by the changing lifestyles, and
attitudes towards home.
56 In a number of cases, more than one word is used to indicate the same item or the items that fulfil
the same function but differ slightly in a particular feature. I took two words to indicate the same item
only in instances where they were used interchangeably (e.g. minder yüzü and minder kılıfı, both of
which mean cushion case). In order to remain loyal to the function attributed to the goods by the
Ottomans of the eighteenth and nineteenth centuries, I treated the items that are indicated by different
words as different goods. For instance, makreme, peşkir, havlu, and silecek, all refer to different types of
towel. Because they appeared next to each other in the same inventory on several occasions, I assume
that at the time there was a meaningful distinction between them. However, in some cases, a good
fulfilling a particular function was replaced over time by another fulfilling the same function. In these
cases, I considered both as a single item in order to keep the continuity over time. For example, kahve
ibriği (coffee ewer), which was used to make and serve coffee, gave way to cezve (coffee pot) towards
the end of the eighteenth century. I did not include cezve on the list as a new item but treated it as the
same good as a coffee ewer.
Table 6-Index of amenities
ITEMS
GROUP SCORE ITEMS
(OTTOMAN TURKISH)
Established goods
Mattresses, thin
döşek, şilte, yorgan, baş/yüz
Bedding 4 mattresses, quilts,
yasdığı
face/head pillows
Sitting
2 minder, makad Cushions, couches
implements
Sheets, curtains, table
çarşaf, perde, sofra örtüsü, döşek
Linen 10 cloths, bedclothes,
yüzü, yasdık kılıfı,
pillowcases/bedticks
Storage 2 sandık, boğça Chests, cloth bundles
halı, kilim, kaliçe, keçe, hasır, Carpets, rugs, felts, rush

Floor coverings 6
seccade mats, prayer rugs
Towels, washcloths,
Towels 4 Makrama, peşkir, havlu, peştemal
towelettes, etc.
New goods
Lightening 2 şamdan, fener Candlesticks, cressets
Heating 1 mangal Braziers
Mirrors 1 ayna/mirat Mirrors
Clocks 1 saat Clocks
Stools, chairs, tables,

Modern iskemle, sandalye, masa, dolab,
5 cupboards, chest of
furniture çekmece, koltuk
drawers, armchairs
Plates, forks, spoons,
Crockery, plates tabak, çatal, bıçak, kaşık, kase, knives, bowls, drinking
8
and cutlery bardak, fincan, zarf glasses, coffee cups, cup
holders
48
MAX. SCORE
I constructed two index scores for each inventory. Index score A indicates the presence
or absence of the selected goods in the inventories. The presence in an individual inventory of
an item results in one point. Each category of goods is attributed a total score corresponding
to the number of types of items it includes (i.e. the total score for bedding equals to four, as
this category includes four different types of household goods: mattresses, thin mattresses,
quilts, and pillows.) In calculating the index score B, an additional point is given if one of the
selected goods appeared more than once in an inventory. Accordingly, each inventory is given
a score A of between 0 and 48 and a score B of between 0 and 96. For instance, if two carpets,
three sheets and a mattress were reported, the index score A of such an inventory would be 3
(1 for each of the selected goods appearing in the inventory); and its index score B would be
5 (3 for the presence of three selected items, 1 for the presence of more than one piece of
kitchenware and 1 for the presence of more than one sheet). As such, score A reflects changes
in variety, whereas score B reflects changes both in the variety and quantity of goods.
It should be noted that the index scores constructed in an attempt to provide a

comprehensive measure of ownership of goods fail to fully capture the progress in the
ownership of household chattels. These only partially reflect the rise in quantity, as they only
look at (a) whether selected goods appeared in a certain inventory, and (b) whether they
appeared more than once. As such, the index scores will underestimate the multiplication of
goods in number, an important aspect of the new consumer regime.
A separate third series was necessary to measure copperware, which appeared

prominently among the household goods in the estates. Most of the kitchenware in the estates
consisted of objects linked to the preparation and consumption of food (cooking pots, frying
pans, cauldrons, shallow pans with two handles, round trays used as a table for serving meals,
trays, dishes, ewers, churns, vessels, strainers, and milk buckets). Besides kitchenware, copper
buckets, washtubs, and mugs frequently appear in the inventories. In the first three periods
(1700, 1730 and 1760), copperware was recorded as a separate item with unit valuations.
However, in the last three periods, copperware was generally reported in weight units and with
lump-sum valuations, probably because there were too many items to be listed individually.
For this reason, copperware cannot be included on the index of amenities alongside other
household items. Instead, I have calculated the amount of copperware included in each estate
in kilograms.
For the last three periods, the value of copperware is obtained by simply adding up
the lump-sum amounts indicated in an inventory. Since the weight of the copperware items
was recorded unsystematically in the earlier periods, I computed the monetary value of the
overall stock of copperware for each inventory before dividing the resulting value by the yearly
copper prices to obtain the amount of copperware possessed by estate owners.
To look at the evolution of the ownership of household goods according to

categories, I subdivide index A, into different categories of household goods (bedding, sitting
implements, floor coverings, storage, etc.). For each category, a score showing the number of
types of goods contained in the inventory is calculated. For instance, the bedding category
includes four different items: mattresses, thin mattresses, duvets, and pillows. The bedding
score for an inventory displays how many items this inventory contains out of the four items.
If the inventory records a mattress and a duvet, the bedding score equals 2. If none of these
items are listed, the inventory is attributed a score of 0 for bedding. Quantities are not taken
into account. This serves the purpose of making clear developments in each category while
elucidating their contribution to the index of amenities.
As a first pass at controlling for the effect of wealth during this discussion, I have
restricted the sample to inventories falling within two constant wealth brackets. Several studies
that analyse material culture based on frequency counts either use a sample of all available
inventories for selected dates, or focus on the middling range in each period. As has been
discussed, this casts doubt on the robustness of the results if evidence cannot be provided to
test how representative the sample is of the living population in terms of wealth distribution.
Since we are interested in quantifying change in the ownership of goods for Ottoman subjects
with comparable wealth levels, we look at the average scores of inventories that stand within
constant wealth brackets from 1700 to 1850. The lower and upper thresholds are 250 and
18,000 constant akçe. In determining the thresholds, I aim to select a range that can be traced
in all periods. The highest possible upper boundary and the lowest possible lower boundary
are taken. The upper threshold, 18,000 constant akçe, is the highest total estate value in the
first period when the outlier value of 26,776 constant akçe is omitted. The lower threshold, 250
constant akçe, is the lowest estate value in the 1760 period. Estates falling outside of these
thresholds have been left out.
According to the frequency distribution of inventories that fall between these two
thresholds in 1700, two wealth groups are defined. 1,650 constant akçe, the median value in
the initial period, is taken as the middle wealth bracket that separates lower and upper wealth
groups. Although in each period the frequency distribution of inventories according to total
estate value is not identical, samples that are reasonably comparable are constructed for an
across-period comparison.
The descriptive statistics of the regression variables are presented in Table 7.
Table 7- Descriptive statistics of regression variables
N Mean Median Std. Dev. Min Max

LNINDEXA 380 2.19 2.30 0.68 0.00 3.74
LNINDEXB 380 2.55 2.64 0.77 0.00 4.11
LNCOPPERKG 319 2.85 2.95 1.253 -1.63 5.58
LNWEALTH 380 7.55 7.51 1.39 4.16 11.37
MEN_NONMUSLIM 380 0.17 0.00 0.38 0.00 1.00
WOMEN_MUSLIM 380 0.30 0.00 0.46 0.00 1.00
WOMEN_NONMUSLIM 380 0.13 0.00 0.33 0.00 1.00
TITLE_MEN 380 0.26 0.00 0.44 0.00 1.00
TITLE_WOMEN 380 0.13 0.00 0.33 0.00 1.00
ELHAC 380 0.07 0.00 0.26 0.00 1.00
It is useful to begin by asking what the index of consumption tells us about changes
in households over this time period. Figure 3 and 4 display average index scores and average
copperware owned in kilograms in each period. The average index scores A and B steadily
increased between 1700 and 1820. Index A rose from 7.7 to 13.4, and index B from 11.5 to 20.7
during this period. From this date to the mid-nineteenth century, the scores remained
constant. The average amount of copperware per estate increased from 15 kilograms in 1700
to 41.5 kilograms in 1820, and then declined to 31 kilograms during the same period.
Figure 3- Mean index and mean wealth (in constant akçe)
Figure 4- Mean amount of copperware owned (kg) and mean wealth

(in constant akçe)
The growth in household possessions visible in the index could, of course, be due to
rising wealth or some other changes in the composition of the sample, and this issue will be
addressed fully in section 3.2. First, in order to better understand the nature of the changes in
household possessions that occurred over this period, and through this the underlying
foundations of the index, in this section I explore developments in the ownership of goods
among two slices of the sample whose wealth falls within two brackets.
Bedding
Over the entire period, items related to bedding were the essential household goods
that appeared in all Ottoman houses. These items were mattresses ( döşek), thin mattresses
(şilte), duvets (yorgan), and pillows (yasdık). The lower wealth group recorded an average of
two of the four items as early as the beginning of the eighteenth century. Some 150 years
later, the average number of types of bedding items possesed by estates owners from the
same wealth group was 2.6. In 1700, the mean score of this category of household goods for
the upper wealth group was 2.3, only slightly more than the first group, rising to 3.5 by around
1850 (see Table 8).
Table 8- Goods concerned with bedding
LOWER WEALTH GROUP 1700 1730 1760 1790 1820 1850

Mattresses (%) 44 56 40 46 63 76
Thin mattresses (%) 0 0 9 19 6 48
Duvets (%) 85 78 80 78 100 90
Pillows (%) 70 78 71 81 100 90
SCORE 2 2.1 2 2.2 2.6 2.6
UPPER WEALTH GROUP

Mattresses (%) 57 57 65 77 76 79
Thin mattresses (%) 7 14 16 36 60 68
Duvets (%) 89 86 84 87 82 97
Pillows (%) 82 91 94 90 87 94
SCORE 2.3 2.5 2.6 2.9 3.1 3.5
At the beginning of the eighteenth century, bedsteads were not in use. Instead, thick
matresses (döşek) were layed out on the floor at night, to be removed in the morning. Some
44 percent of the estates in the lower wealth group and 57 percent of the estates in the upper
wealth group contained matresses in 1700. By 1850, almost 80 percent of all estate owners
had matresses. These mattresses were usually stuffed with cotton and wool. Matresses stuffed
with feather appeared in the inventories of the rich very occasionally. Until the late eighteenth
century, mattresses used by all social segments were covered with beledi, a cotton textile,
which occassionally included silk. Production of the textile began in Urla and Tire, on the
Aegean coast of Anatolia, in the early sixteenth century. Before 1820, we come across
mattresses covered by a fabric called alaca, a striped cotton-silk mixture with a substantial
share of cotton, in just a few inventories of the rich. In the nineteenth century, the use of beledi
for mattresses dropped by half. Instead, çit, a printed cotton fabric produced in the Central
Anatolian towns of Sivas and Tokat, were increasingly used for covering mattresses. Thin
mattresses (şilte), the ownership of which was restricted to a small minority among the
wealthier group around 1700, trickled down to the lower middle class by around the mid-
eighteenth century. These were probably placed on mattresses to provide more comfort. They
were quite common among all estate owners in Üsküdar in the first half of the nineteenth
century.
In 1700, duvets were among the items that appeared most frequently in the
inventories. Over 80 percent of the estates owners from both groups possessed at least one
duvet throughout the period, while 50 percent owned more than one. From 1790 onwards,
most estate owners had at least two duvets. Until the late eighteenth century, duvets were
commonly covered with yemeni, an Ottoman adaptation of Indian prints. Though exceptional,
duvets covered with sandal, a cotton-silk mixture, were a marker of wealth. Around 1790, çit
and basma, local cotton prints replaced yemeni as the common material of duvets. After this
date, a large number of rich inventories contained sakızkari57 and kıbrıskari58 quilts.
By around 1700, 70 percent of estates from the lower wealth group and 82 percent
from the upper wealth group contained pillows, while an important share reported more than
one. In 1820, almost all estates included at least two of these items. In the second half of the
eighteenth century, pillows began to be differentiated according to their functions. Face
pillows (yüz yasdığı) and head pillows (baş yasdığı) were increasingly reported over time.
Floor coverings
Üsküdar estates contained a variety of floor coverings in various colours and

decorated with different designs: halı (carpets), kilim (rugs), kaliçe (small carpets), keçe (felts),
hasır (rush mats), and seccade (prayer rugs). These constituted important elements of the
57 A cotton textile produced in Chios.
58 A cotton textile produced in Cyprus.

eighteenth- and nineteenth-century Ottoman house. Keçe (felts) and kilim (rugs) appear to
have been the most popular types of floor coverings among the inhabitants of Üsküdar in this
period. During the second half of the eighteenth century, felts coming from Yambol ( Yanbolu),
located in the Western Black Sea region, appear frequently in the estates. Felts said to be
produced in Salonika were reported only a couple of times. As for the rugs, these were
described with their size, colour and state of wear rather than the place of origin.
Table 9-Floor coverings

Carpets (%) 15 22 6 3 6 3
Rugs (%) 52 11 29 22 6 21
Prayer rugs (%) 19 11 26 19 38 17
Rush mats (%) 4 11 11 19 13 0
Felts (%) 22 33 34 14 19 24
SCORE 1.1 0.9 1.1 0.8 0.8 0.5
UPPER WEALTH GROUP

Carpets (%) 46 10 10 13 0 0
Rugs (%) 54 10 19 42 27 38
Prayer rugs (%) 25 19 48 45 47 50
Rush mats (%) 29 14 26 10 20 18
Felts (%) 54 33 42 48 27 27
SCORE 2.1 0.9 1.5 1.8 1.2 1.1
The index scores present an interesting pattern for this group of items (see Table 9).
Unlike most other household goods, their ownership declined over time. The index score fell
from 1.1 in 1700 to 0.5 in 1850 for the lower wealth group, and from 2.1 to 1.1 for the upper
wealth group. Carpets totally disappeared in the mid-nineteenth century, while rugs and rush
mats diminished significantly. The felts lost popularity for the upper wealth group, but
continued to appear in the houses of the poorer. In the study of Kent and Cornwall estates
from the seventeenth and eighteenth centuries by Overton et al. (2004), the same situation is
observed. The authors cite improvements in the construction and the treatment of woods as
the reason for the development. According to them, as furniture became more decorative, the
necessity to cover it with carpets declined. Similarly, for Ottoman houses, it could be suggested
that with the introduction of new and more tasteful goods and improvements in construction
techniques and materials, floor coverings were less necessary for interior decoration.
Additionally, floor coverings also served as a measure against loss of heat (Üstündağ 2003). It
is possible that with the introduction of stoves to houses and the improvement of heating
facilities, the necessity for such goods disappeared.
The details given in the estates and the wide range of prices suggest that for the
inhabitants of Üsküdar, floor coverings were much more than a mere necessity. Rather, these
were means of social differentiation. For instance, rush mats ( hasır) were markers of wealth in
the eighteenth century. In the first three periods, they appeared only in the inventories of the
wealthy, and they were substantially more expensive than other types of floor coverings. After
their peak in the late eighteenth century, they gradually disappeared. In 1850, rush mats were
totally absent from Ottoman houses in Üsküdar.
Sitting implements
One of the main features of the eighteenth- and nineteenth-century Ottoman house
was the diwan known in Ottoman Turkish as sedir. This continuous and slightly raised platform
running along the walls of a room was furnished with long cushions and pillows. Tanyeli (2003)
suggests that in the sixteenth and seventeenth centuries, textiles were so expensive that even
the townspeople of upper-middle income groups did not use them in large quantities for
interior decoration. He argues that the interior organization of the Ottoman room, which
featured large quantities of textiles, was created from the seventeenth century onwards.
Charles White, a British traveller who spent three years in Istanbul in the 1840s,
described the sedir as follows:
“In Turkish habitations, the framework of divans generally consists of rough wooden
planks, over which is placed a long narrow mattress, filled with wool of straw. This is
covered with printed cotton, chintz, or cloth, bordered with fringe, frequently
festooned. In wealthy houses, these covers are of costly materials, such as silk or
velvet, embroidered with gold or silver, the frames of carved wood. At the back and
extremities are thick cushions of the same materials, and a long strip of white linen is
stretched over the seat and cushions from end to end to preserve them from being
soiled. These strips of linen, which can be removed and replaced in an instant, are kept
carefully stretched, clean and free from creases (1846: 170-1).”
Cushions (minder) were perceived as being essential for comfort in the view of
eighteenth-century Ottomans. The poor as well as the wealthy acquired a couple of cushions,
even in the early eighteenth century. For practical reasons, I focus on whether an inventory
contains one or more than one piece of this item. This methodological choice has the
disadvantage of neglecting an increase in the quantity beyond two. In the case of items which
were acquired in large quantities, such as pillows and cushions, we cannot accurately
determine change over time. Nevertheless, based on my reading of the inventories, it is clear
that the inhabitants of Üsküdar, with their different levels of wealth, acquired increasingly
greater numbers of pillows and cushions from the early eighteenth century until the mid-
nineteenth century.
Table 10-Sitting implements

Cushions (%) 74 67 63 68 75 66
Couches (%) 0 11 26 38 38 38
SCORE 0.8 0.8 0.9 1.1 1.1 0.9
UPPER WEALTH GROUP

Cushions (%) 79 76 74 71 67 68
Couches (%) 7 19 48 55 60 62
SCORE 0.9 1 1.2 1.3 1.3 1.4
Compared to pillows, the textiles used for cushions were less frequently recorded. The
available evidence suggests that from 1700 to 1760, alaca and yemeni were used for cushions,
whereas pillows were commonly covered by beledi or yemeni. An important marker of wealth
was kadife (velvet) and çatma59 pillows, which were substantially more expensive than ordinary
ones. Almost all members of the upper wealth group possessed one or several of these. Çatma
pillows were particularly very popular around 1760.
As reported in Table 10, in the early eighteenth century, couches (makad) were present
59 A silk-velvet mixture with embossed floral patterns

only in a small share (11 percent) of the inventories of the upper wealth group. Over time, they
trickled down and began to be adopted more widely. It was not before the late eighteenth
century that couches found their way into the houses of the poor. By this date, one third of
the members of lower wealth group, as well as more than half of the members of the upper
wealth group, had acquired them. Around the mid-nineteenth century, these figures reached
38 percent and 62 percent. From 1700 to 1760, makads were generally made of yemeni and
çuka. Around 1790, çit began to replace yemeni, and in 1850, these items were commonly
covered with çit and basma.
Storage
The Ottoman approach to collecting and keeping objects while facilitating their easy
transport is a legacy of the nomadic period. Nomads used to keep their objects in light, easily
collected and transportable bundles, saddlebags and chests. Turks living in permanent
settlements also used to wrap their valuable items, dresses, fabric products and dowry in
bundles (Usal 2010).
Chests (sandık) were among the most ordinary and indispensable items of domestic
furniture. It was a basic, modest and inexpensive necessity for the lower classes, while the
upper classes preferred more expensive pieces decorated with ornaments and made of better-
quality material. For instance, several wealthy inventories report chests featuring mother-of-
pearl in the nineteenth century. Similarly, towards the end of the eighteenth century, chests
made of cypress became fashionable among the upper classes in Üsküdar. Another item used
for storage was cloth bundles (boğca), which were used to wrap pieces of cloth, dresses, and
household linen.
Table 11- Goods concerned with storage

Chests (%) 22 33 51 57 56 62
Cloth bundles (%) 30 22 23 32 19 35
SCORE 0.8 0.6 0.7 0.9 1.1 0.8
UPPER WEALTH GROUP

Chests (%) 57 24 52 68 67 71
Cloth bundles (%) 29 29 16 16 51 53
SCORE 0.9 0.5 1 1.2 1.2 1.2
Table 11 shows that at the beginning of the eighteenth century, almost all estate
owners from both wealth groups had a chest or a cloth bundle, as the scores close to 1 indicate
(0.8 for the first wealth group, 0.9 for the second wealth group). As the items to be stored
increased, chests rose in popularity among the lower wealth group from 22 percent in 1700 to
62 percent in 1850. However, only a modest improvement (from 57 percent in 1700 to 71
percent in 1850) can be detected within the upper wealth group over the same period. This
might be explained by rising ownership of drawers and cupboards. However, these new items
did not totally replace chests, as high percentages at the end of the period indicate.
Household linen and towels
The interior organization of the eighteenth- and nineteenth-century Ottoman house

featured large quantities of textiles: towels, sheets, tablecloths, bedclothes,
cushions/pillowcases, and curtains.
Four different types of towels, probably with different functions, are mentioned in the
inventories: Makrama, peşkir, havlu, and peştemal. Among these, makrama seems to have lost
its popularity among the inhabitants of Üsküdar, while havlu became increasingly more
common throughout the eighteenth century. In the overall category of towels, there appears
to have been no progress in the lower wealth group. In the upper wealth group, the adoption
of towels rose from 1700 to 1820 by almost half before declining between 1820 and 1850 (see
Table 12).
Table 12- Towels

Makrama (%) 39 22 26 24 31 7
Peşkir (%) 11 0 9 14 6 7
Havlu (%) 7 22 14 19 38 24
Peştemal (%) 30 22 14 19 25 21
SCORE 0.7 0.9 0.8 0.8 0.9 0.7
UPPER WEALTH GROUP

Makrama (%) 39 29 48 55 47 12
Peşkir (%) 25 10 23 23 24 12
Havlu (%) 14 5 29 45 53 35
Peştemal (%) 18 10 19 19 22 18
SCORE 1 0.5 1.4 1.4 1.4 0.7
Interestingly, between 1700 and 1850, the inventories show only modest, if any,
improvements in the ownership of towels and household linen. The index scores for towels
passed from 0.7 in 1700 to 0.9 in 1820 in the lower wealth group, and from 1.1 to 1.4 in the
upper wealth group. The lower wealth group had on average of one type of household linen
per estate around the beginning of the eighteenth century, while this figure was 1.3 in the
early nineteenth century. The average score of household linen in upper wealth group also
rose slightly during the same period from 1.2 to 1.5. In the final period, the scores substantially
declined for both items in both wealth groups (see Table 13).
Table 13- Household linen

Curtains (%) 15 11 9 16 10 24
Table cloths (%) 0 0 9 8 13 3
Sheets (%) 56 33 49 70 73 35
Pillow/Cushion cases (%) 22 11 3 14 13 10
Bedclothes (%) 15 11 6 8 13 10
SCORE 1.1 0.8 0.8 1.2 1.3 0.8
UPPER WEALTH GROUP

Curtains (%) 50 0 26 19 13 47
Table cloths (%) 11 5 19 10 27 6
Sheets (%) 64 52 55 74 83 59
Pillow/Cushion cases (%) 29 14 10 26 13 9
Bedclothes (%) 15 5 16 23 24 9
SCORE 1.2 0.9 1.3 1.5 1.5 1.2
Nonetheless, these figures should not be taken as an indicator that household textiles
in Ottoman houses did not improve over time. Observations from these sources suggest the
acquisition of greater quantities of better-quality linen by many households appears to have
begun in the mid-eighteenth century. Qualitative evidence also confirms this observation.
Two factors probably lay behind the unlikely results for household linen. First, as will
subsequently be shown in this study, the possession of these items were closely linked to
gender. Differences in terms of the composition of the sample 60 across periods might be why
we cannot follow the real change in the ownership of these goods. Women, who were
predominant in terms of household textiles, constituted 58 percent of the estate owners in the
initial period, but only 35 percent in 1820 and 39 percent in the final period. However, a more
important limitation should be sought in the process of the registration of the inheritance
inventories. In the final period, a portion of the used textiles, available in the house –
particularly those worth insignificant amounts – were recorded as lump-sum quantities, while
newer, more valuable and ornamented pieces were mentioned item by item. This might be
masking the evolution of this category of goods in the inventories. Furthermore, it is also
possible that mattress and duvet covers, whose appearance diminished over time, began to
be considered as integral parts of mattresses and duvets, and thus, were valued and recorded
together with these items. Expressions such as “mattress covered with basma” support this
argument. Thus, the result seems very likely to reflect the increasing tendency to omit this
category of household good from the inventories.
Sheets were evident across all groups even at the beginning of the eighteenth century.
More than half of the estates in both wealth groups (56 percent in the lower wealth group,
and 63 percent in the upper wealth group) included sheets for this period. Around 1820, about
three quarters of estate owners had them (73 percent and 83 percent, respectively, for each
wealth group). Until 1790, sheets were prevalently made of alaca. A number of rich houses
contained “sheets of Egypt” ( Mısır çarşebi), which probably meant sheets made of Egyptian
cotton. In the nineteenth century, bürüncük, another local cotton fabric, began to be used for
sheets.
60 The sample refers to the restricted sample used in this exercise and not the extended sample
employed for the regressions in the earlier parts of this paper.
Curtains were recorded in different forms in the inventories. Door curtains (kapu
perdesi), which protected the inhabitants against cold draughts in winter, were the most
popular. Window curtains (pencere perdesi) and oven curtains (ocak perdesi/yasmağı) began
appearing in Ottoman houses over the course of the eighteenth century. Curtains were not
reserved for the richer group, even in the first half of the eighteenth century. However, like
bedclothes and duvets, the ownership of these items as recorded in the inheritance inventories
did not follow a meaningful pattern.
Unlike sheets and curtains, bedclothes and duvet covers were rarely part of a middle-
class house in 1700. They became widespread among these groups around the turn of the
nineteenth century. Their disappearance from the estates in 1850 is interesting and should be
linked to the factors mentioned above.
Lighting implements
The ownership of lighting elements is indicative of the use of time in early-modern

Ottoman society. The day began with the first light of day, preceding the morning prayer, and
ended early, soon after sunset. This pattern was not only imposed by technological limitations,
but was also a reflection of the daily routines of agricultural society. Lighting was a luxury by
the standards of the early eighteenth century, and was in limited use even among the rich.
Between the sixteenth and seventeenth centuries, şamdan (candlesticks) and fener
(cressets) replaced çerağ (oil lamps). In the lower wealth group, candlesticks made of copper,
brass, and iron doubled (19 percent to 38 percent) from 1700 to 1850 (see Table 14). Cressets,
which were absent from the inventories in this group in the initial period, appeared in just a
few instances (7 percent) around 1850. In the upper wealth group, less than one third of estate
owners had candlesticks in the initial period, but almost half had these items one-and-a-half
centuries later. The ownership of cressets also grew, passing from 7 percent to 15 percent.
Table 14- Lightening implements

Candlesticks (%) 19 22 27 19 19 38
Cressets (%) 0 0 9 3 13 7
SCORE 0.2 0.2 0.3 0.3 0.3 0.4
UPPER WEALTH GROUP

Candlesticks (%) 31 10 26 23 37 49
Cressets (%) 7 5 10 3 13 15
SCORE 0.3 0.2 0.4 0.4 0.4 0.5
These figures suggest that by the mid-nineteenth century, interior-lighting equipment

was no longer confined to a minority. However, it should also be noted that one almost never
comes across expensive lighting implements. Adopting elaborate lighting arrangements, such
as extravagant chandeliers, seems to be a phenomenon that emerged in the later part of the
nineteenth century.
Heating
Early-modern Ottoman houses were generally heated by ovens and fireplaces, which
also served for cooking purposes. Even imperial palaces did not use stoves before the Tanzimat
era (1839) (Üstündağ 2003). Mangals (braziers) made of copper, brass, metal or baked clay
vessels, were introduced to Ottoman houses after the invention of charcoal in the Middle Ages.
Braziers were used to heat rooms which did not contain ovens or fireplaces. Thus, it
can be expected that their usage became widespread with the increase in the number of rooms
in houses and with the functional differentiation of space. From the first half of the
nineteenth century, the covered stove began to infiltrate Turkish daily life, gradually
superseding the open brazier as a major form of heating, but even after the introduction of
the stove, the brazier was still used as an auxiliary heating source to heat other rooms of the
house with the embers from the stove. Today, in many parts of Anatolia, the brazier is still the
major form of heating.
Table 15-Heating

Braziers (%) 0 0 17 16 19 38
UPPER WEALTH GROUP

Braziers (%) 0 5 16 19 42 56
At the beginning of the eighteenth century, estates in Üsküdar did not report any
braziers (see Table 15). By 1730, there were still only two instances of the object. The share of
inventories from the lower wealth group containing braziers rose from 15 percent in the early
eighteenth century to 31 percent in the mid-nineteenth century. In 1850, 54 percent of the
estates of the upper wealth group included more than one brazier.
Mirrors and clocks
Mirrors and clocks, the consumption of which was associated with Westernization,
entered Ottoman houses in the late seventeenth century. In the first half of the eighteenth
century, only a small share (about 10 percent in the case of both items) of estates from lower
and upper wealth groups contained them (see Tables 16 and 17). The middle of the century
appears to have been a turning point in terms of the consumption of these items due to
increased European influence. About 40 percent of the upper wealth group and 15 percent of
the lower wealth group owned clocks around 1760, whereas 30 percent of the upper wealth
group and 10 percent of lower wealth group possessed mirrors. By this date, both of these
items had appeared only once in the estates of the lower wealth group. Yet, in 1820, 39 percent
of inventories from the poor mentioned clocks, whereas mirrors remained a luxury for this
group, even in the middle of the century, with only 15 percent of estates including this item.
As for the rich, both mirrors (57 percent) and clocks (43 percent) were quite popular among
this class.
Table 16- Mirrors

Mirrors (%) 0 0 6 5 6 28
UPPER WEALTH GROUP

Mirrors (%) 7 5 19 13 22 38
Table 17- Clocks

Clocks (%) 0 0 11 8 38 7
UPPER WEALTH GROUP

Clocks (%) 11 5 23 39 44 44
Western furniture
With the incorporation of influences from Western culture into the daily lives of
Ottoman subjects, European furniture entered Ottoman houses in the second half of the
eighteenth century, proliferating in the course of the nineteenth century. The adoption of new
furniture and changes in the interior decoration and inner arrangement of rooms began in
elite urban houses (Esenbel 1994). White’s observation of mid-nineteenth century Ottoman
society suggests that by this date, elites had already adapted themselves to the Western
standards of domestic comfort:
“In proportion as intercourse with Europeans extends, fashions and customs vary, so
that an important change is rapidly taking place in the furniture of houses. Thus, in
those of wealthy persons, chairs, sofas, tables, consoles, mirrors, wardrobes,
chandeliers, and a variety of Western essentials may be seen. Indeed, the Sultan’s
private day-apartments, at Tcheraghan and Beshiktash, are furnished more in the
European than Oriental style. … The middling classes are also making some progress,
but in general they retain their ancient simplicity. Their furniture is limited to divans,
mats, carpets, and a few pieces of glass or porcelain placed in wooden niches,” (White
1846: 174-5).
Table 18- New furniture

Chairs (%) 0 0 0 0 0 10
Stools (%) 4 0 11 14 31 24
Armchairs (%) 0 0 0 0 0 10
Drawers (%) 0 0 6 0 31 28
Cupboards (%) 4 0 3 24 6 7
Tables (%) 0 0 0 0 0 3
SCORE 0.1 0 0.2 0.5 0.6 0.6
UPPER WEALTH GROUP

Chairs (%) 0 0 0 10 0 29
Stools (%) 11 0 16 3 22 32
Armchairs (%) 0 0 0 0 0 32
Drawers (%) 0 0 16 19 27 44
Cupboards (%) 0 0 3 26 18 26
Tables (%) 0 0 0 0 0 9
SCORE 0.1 0 0.4 0.7 0.7 1.8
Üsküdar inventories record six different types of new furniture: chairs, stools,
armchairs, tables, cupboards, and drawers. In 1700, the average score of this category of goods
was 0.1 for both wealth groups (see Table 18). As Table 15 shows, particularly in the first half
of the nineteenth century, there was a significant shift in the ownership of Western-style
furniture. In 1850, the average score was 0.6 for the lower and 1.8 for the upper group. In terms
of modern furniture, stools and drawers were the first to enter Ottoman house and were
present in several estates as early as 1760. In both wealth groups, drawers and stools were by
far the most popular type of modern furniture in mid-nineteenth-century Üsküdar. Some 28
percent and 44 percent of estate owners from the lower and upper wealth groups had drawers,
and 24 percent and 32 percent had stools. Cupboards and chairs followed them. Tables, on
the other hand, were rare even in mid-nineteenth-century inventories, with less than 10
percent of the upper group owning them. Armchairs only appeared in the estates of the upper
wealth group in 1850. Furniture made of walnut tree or decorated with mother-of pearl
appeared in several instances among the higher levels of wealth in the mid-nineteenth century.
Crockery, plates, and cutlery
Changes in eating and drinking rituals and the introduction of modern habits in the
Ottoman realm was a phenomenon of the late nineteenth century. The custom of using tables,
chairs, forks, plates and glasses for everyone began in elite households and slowly filtered
down to moderate families. Before this date, the meal was placed in the middle of the sini, a
round copper tray, which was set up at meal times and taken away afterwards, and everyone
ate from the common dish using their spoons. Forks are almost completely absent from the
inventories throughout the whole period, only appearing in a couple of instances for wealthy
estate owners. Although common, knives were probably used for cooking rather than as an
eating utensil.
Table 19- Crockery, plates, and cutlery

Plates (%) 0 11 6 11 19 24
Cups (%) 4 11 17 19 13 14
Cup holders (%) 0 0 9 16 6 7
Forks, spoons, knives (%) 11 0 11 24 31 7
Drinking glasses (%) 0 0 6 11 6 7
Bowls (%) 0 0 3 8 13 14
SCORE 0.1 0.2 0.3 0.6 0.6 0.6
UPPER WEALTH GROUP

Plates (%) 7 5 13 16 38 35
Forks, spoons, knives (%) 14 5 11 32 33 24
Drinking glasses (%) 7 0 0 7 27 24
Cups (%) 4 5 23 26 44 24
Cup holders (%) 0 5 16 23 36 21
Bowls (%) 0 5 10 7 38 24
SCORE 0.3 0.1 0.4 0.7 0.9 0.9
Progress in the ownership of plates is quite striking (see Table 19). While in 1700 plates
were present in only 7 percent of the estates belonging to the upper wealth group, one fourth
of the lower wealth group and one third of the upper wealth group owned plates by around
the mid-nineteenth century. Although the increase is relatively less pronounced when
compared to British inventories, in which 85 percent of the deceased had plates reported in
their estates by the mid-eighteenth century (Overton et al. 2004: 99), it might be indicative of
the practice of eating from separate dishes.
Estates from 1700 do not report any bowls ( kase), but this item was introduced into
Ottoman households and proliferated in both wealth groups throughout the eighteenth
century. More than one tenth of the lower wealth group and about one fourth of the upper
segment had bowls by 1850. As bowls were used for soups and compotes, which have an
important place in Ottoman cuisine, their absence in the initial period is intriguing. Kase
probably referred to vessels made of glass or pottery, and replaced the copper tas, which was
used for drinking water and consuming stews and similar foods in previous periods.
The proliferation of drinking glasses, on the other hand, appears to be associated

with wealthier segments of society. As in the case of kase, bardak seems to refer to glass or
pottery, which began to replace tas in Ottoman houses. One fourth of the estates from the
second wealth group report this item around the mid-nineteenth century.
Coffee cups (fincan) were present in the estates of the lower wealth group throughout
the period, showing that drinking coffee was not a privilege completely reserved for the elite
even as early as the first half of the eighteenth century. Still, there was a much larger increase
in the ownership of coffee cups in the upper wealth group. In 1820 almost half of the estate
owners from this group possessed cups. Strangely, this figure drops by half in the next period,
although the change in the composition of the sample in terms of the individual characteristics
of the estate owners might be an explanation.
Copperware
As Faroqhi (2002a: 301) states, until the nineteenth century, there was very little
furniture in Ottoman houses that could be defined as “furniture in today’s norms.” These were
some “chests and boxes, a yer sofrası (a traditional dining place which is used on the ground),
made up of a wooden or leather stool for putting pots, pans and trays on, carved shelves on
the wall for putting lamps and books.” In contrast to the simplicity of the home furnishings, a
large variety and quantity of copperware was present in an average Ottoman house even at
the beginning of the eighteenth century. Almost all estates possessed a few pots, cauldrons,
and pans, in different sizes, with or without lids. As wealth grew, the quantity and variety of
the copperware increased, leading to differentiations in function. The variety of items we come
across in these sources proliferated: hamam leğeni (bathtub), çamaşır leğeni (washtub), hamam
tası (bath bowl), abdest leğeni (ablution tub), maşraba (dipper), el tabesi/tabe dest (hand pan),
pekmez tabesi (molasses pan), yol tabesi (journey pan), börek tebsisi (pastry tray), baklava
tebsisi (baklava tray), kadayıf tebsisi (kadayıf tray), kevgir/süzgü/süzgeç (colander), ayaklı sahan
(free standing shallow pan), el ibriği (hand ewer), şerbet ibriği (syrup ewer), kahve ibriği (coffee
ewer), abdest ibriği (ablution ewer), çorba tası (soup bowl), hoşab tası (compote bowl), kahve
tebsisi (coffee tray), ocak güğümü (oven billycan), hamam güğümü (bath jug), sofra tası (table
vessel), yumurta tabesi (egg pan), and lokma tabesi (plum pan).
Although, as mentioned previously, a systematic review of the ownership of different

types of copperware is impossible due to the inconsistency in the recording of the inventories,
the improvement in the possession of kitchenware and other household goods made of
copper can be traced by looking at the lump-sum amounts recorded (see Table 17 and Figure
5). An average estate owner in lower wealth group possessed 8.2 kilograms of copperware in
around 1700; within a century, this figure had tripled, reaching 24.4 kilograms by around 1820.
In the upper wealth group, the amount of copperware per estate rose from 26.7 kilograms in
1700 to 50.3 kilograms in the early nineteenth century. From 1820 to 1850, copperware
declined for all estate owners – albeit to levels still substantially higher than those in the initial
period.
Table 20- Average amount of copperware possessed (in kg)
Lower wealth Upper wealth

group group
1700 8.2 26.7
1730 3.2 23.2
1760 12.5 28.6
1790 12.8 41.6
1820 24.4 50.3
1850 18.3 38.8
Figure 5- Average amount of copperware possessed (in kg)

Lower and upper wealth groups
Overall, it can be suggested that in the early the eighteenth century, an average
Ottoman house consisted of matresses, duvets, rugs or felts, chests, a number of cushions,
towels, and some copperware. The richer households contained, alongside these essential
goods, carpets, curtains, and sheets, and occasionally candlesticks and mirrors. Around this
date, differences in wealth and status of the estate owners manifested themselves not in the
composition but in the quantity, quality, the degree of wornness, the ornamentation and the
raw material of the household goods.
From the mid-eighteenth to mid-nineteenth centuries, the durables present in the

Ottoman houses from different wealth groups diversified and multiplied, with the households
acquiring many new types of furniture and consumer durable. Not all commodities followed
the same trajectory, however. Patterns of ownership varied by the type of good.
4. Ownership of household goods at constant wealth levels
4.1. Regression model
In this section, I seek to better identify the degree to which ownership of household
goods grew independently of the wealth of individuals. To do so, I conducted a pooled, cross-
sectional regression analysis that evaluates how the index scores of consumer goods and the
amount of copperware owned changed over time, independently of changes in the
composition of the sample in terms of wealth, gender, religious status, and social status. As
was mentioned earlier, the pooled cross-sections are obtained by collecting samples
independently of each other at different points in time. The fact that the random samples are
collected independently of each other implies that they need not be of equal size and will
usually contain different statistical units at different points in time. The data can be analysed
like ordinary cross-sectional data, except that we must use dummies in order to account for
shifts in the distribution between different points in time. Regressing index scores and the
amount of copperware upon time dummies and controls for wealth yields information about
improvements in the ownership of domestic goods unexplained by an increase in wealth.
Separate OLS regressions are run for the index scores A and B and the amount of
copperware owned, using the form:
(1)
𝐿𝑁𝐼𝑁𝐷𝐸𝑋𝑆𝐶𝑂𝑅𝐸𝑖
= 𝛽0 + 𝛽1 𝐿𝑁𝑊𝐸𝐴𝐿𝑇𝐻𝑖 + 𝛽2 𝑊𝑂𝑀𝐸𝑁_𝑀𝑈𝑆𝐿𝐼𝑀𝑖 + 𝛽3 𝑀𝐸𝑁_𝑁𝑂𝑁𝑀𝑈𝑆𝐿𝐼𝑀𝑖
+ 𝛽4 𝑊𝑂𝑀𝐸𝑁_𝑁𝑂𝑁𝑀𝑈𝑆𝐿𝐼𝑀𝑖 + 𝛽5 𝑇𝐼𝑇𝐿𝐸_𝑀𝐸𝑁𝑖 + 𝛽6 𝑇𝐼𝑇𝐿𝐸_𝑊𝑂𝑀𝐸𝑁𝑖
+ 𝛽7 𝐸𝐿𝐻𝐴𝐶𝑖 + Σ𝛽𝐾 𝑃𝐸𝑅𝐼𝑂𝐷𝐾 + 𝑢
Logged index scores (LNINDEXA, LNINDEXB) and logged copperware in kilograms

(LNCOPPERKG) are regressed against variables for real wealth, dummies combining the gender
and religious status of the estate owners, as well as the titles that were acquired. LNWEALTH
is the logged net estate value in constant akçe 61 . WOMEN_MUSLIM, MEN_NONMUSLIM,
WOMEN_NONMUSLIM are dummy variables indicating gender and religious status. These
variables assume the value 1 if the estate owner was, respectively, a Muslim woman, a non-
Muslim man, and a non-Muslim woman. Muslim man is the reference category, meaning it is
omitted from the regression. As has been discussed above, if there was a meaningful
relationship between the religious status or the gender of the estate owners and the quantity
and variety of the domestic goods owned, the evolution of the index scores would also reflect
61 It must be noted that in computing the CPI, Pamuk (2000a) uses a consumer basket which consists
predominantly (80 percent) of agricultural products. Thus, the deflation process does not remove the
potential impact of changes over time in the prices of manufactured goods relative to agricultural goods
on the ownership of consumer durables. It is assumed that this impact will be captured by the time
dummies.
changes in the composition of the sample. Incorporating these variables controls for across-
period variations in the number of female and non-Muslim inventories in the sample.
TITLE_MEN, TITLE_WOMEN and ELHAC are dummy variables that are included in the model as
an indicator of social status. TITLE_MEN and TITLE_WOMEN indicate men and women with
official and religious titles, and ELHAC indicates pilgrims. The reference category is men and
women without titles62.
The effect of time-dependent factors other than wealth is captured by the period
dummies in the regression for each sample period (1730, 1760, 1790, 1820, and 1850). The
reference category is our initial period, 1700. Including time dummies allows the intercept to
have a different value in each period. In so doing, we can observe changes in the possession
of household goods over different periods. If the number and variety of household effects
owned by the estate owners increased independently of wealth and the changing character of
the estate owners, we would expect the coefficient of the dummies for later periods to be
positive and significant.63
One important point needs to be made at the outset about employing this functional
form. In the regression WEALTH, our control variable is not independent from the index scores,
since the total value of the domestic durable stock is a share of the total estate value. A rise in
index scores and the amount of copperware signifies that the estates contained a larger
amount of domestic chattels, which would be expected to have increased the value of the
domestic goods stock and thus the total wealth. However, the aim of the exercise is to identify
62 Some of the individual characteristics of the estate owners that one might expect to be closely
associated with the level of domestic comfort and conveniences enjoyed are not included in the model
due to absence of systematic information in the inventories. Age, marital status, and the occupation of
the deceased are three such characteristics. This should be noted as a limitation of the present study.
63 As has been shown above, the wealth measure used here closely reflects the real wages, the only
proxy available for income in the Ottoman realm. However, we do not know whether and how the
relation between income and wealth changed over time. It is entirely possible that during a period of
significant social and economic transformation, this relationship was markedly altered. Thus, in
controlling for wealth, we are not completely isolating the impact of changing income levels. For
instance, if saving rates were declining over time, a constant wealth level would indicate rising lifetime
earnings. This changing relationship between wealth and income can also be captured by the “time
effect.”
any the change over time in ownership of goods at constant wealth levels, rather than explain
the relationship of wealth to consumption per se. In other words, I am not trying to establish
a causal relationship, but to estimate the level of domestic comfort enjoyed by Ottomans with
comparable wealth levels in different time periods.
That said, the relationship between total estate value and the stock of household
durables deserves discussion. First, it should be noted that the link between wealth and the
value of the stock of domestic goods was weaker in Ottoman inheritance inventories than in
European and American probate inventories. Overton et al. (2004) state that consumption
goods made up over half of the total value of an average English inventory. In our sample, the
share of the value of household durable stock held within the overall estate was less than one
third (29 percent). The lower shares constituted by domestic durables in the Ottoman
inventories can be explained by the inclusion of real estate, which was usually the most
valuable item in an inventory.
Second, it should be underlined that the positive association between the total estate
value and the stock of household durables is not merely a result of the endogeneity between
these two variables. Even when the value of the household durables is excluded, the remaining
wealth comprised of real estate, net debts, cash, capital goods, and personal goods (such as
clothing) still correlates positively with the index scores and the amount of copperware owned,
although the strength of the relation is slightly weaker in this second case, as would be
expected (see Table 21 and Figures 6 to 8). This is in accordance with the general wisdom that
wealthier people lived in a richer domestic environment than their poorer compatriots.
Table 21- Correlations between the (pooled) stock of household durables and (pooled)
wealth (in constant akçe)
Total estate value less the value of

Total estate value household durable stock
Index A 0.34 0.29
Index B 0.36 0.31
Copperware (kg) 0.49 0.43
Figure 6- Index A, Total estate value, and total estate value less the value of household
durables stock (in constant akçe)
Figure 7- Index B, Total estate value, and total estate value less the value of household
durables stock (in constant akçe)
Figure 8- Copperware (kg), Total estate value, and total estate value less the value of
household durables stock (in constant akçe)
4.2 Results
Table 22 reports the regression results. As expected, wealth is strongly and positively
associated with the number and variety of household durables owned. In all three regressions
separately run for index scores A and B, and copperware, LNWEALTH1 is positive and
significant at the 1 percent level. When wealth doubles, index scores A and B, and copperware
owned rise by 20 percent, 24 percent and 44 percent respectively.
Another important finding was the difference between Muslim men and Muslim
women. While no statistically significant difference exists between Muslim men and non-
Muslim estate owners (both men and women) is detected in the ownership of household
goods, the results show that Muslim women owned more domestic goods compared to
Muslim men who possessed the same level of wealth. The index scores A and B assigned to
estates belonging to Muslim females are 44 and 49 percent higher than the scores assigned
to the estates belonging to Muslim males. In 1700, an estate owned by a woman that was
worth 1,000 constant akçe had index scores 8 (A) and 11 (B), whereas these figures were 5 and
7 for a Muslim man’s estate of comparable total value. Unlike other types of household goods,
Muslim women did not display superiority in terms of ownership of copperware over Muslim
men. Yet, there was a significant difference between these latter and non-Muslim Ottoman
women. Women in the non-Muslim community owned about 43 percent less copperware than
Muslim men of comparable wealth.
Table 22- OLS Regression results
Dependent variable LNINDEX A LNINDEXB LNCOPPERKG

CONSTANT -0.127 -0.157 -1.360***
(0.174) (0.194) (0.352)
LNWEALTH 0.262*** 0.308*** 0.505***
(0.022) (0.024) (0.043)
WOMEN_MUSLIM 0.364*** 0.458*** -0.086
(0.1) (0.112) (0.202)
MEN_NONMUSLIM 0.039 0.106 -0.01
(0.106) (0.118) (0.216)
WOMEN_NONMUSLIM 0.049 0.096 -0.567***
(0.116) (0.13) (0.242)
TITLE_MEN 0.183* 0.222** 0.039
(0.103) (0.115) (0.216)
TITLE_WOMEN -0.096 -0.122 0.163
(0.123) (0.137) (0.235)
ELHAC -0.248** -0.290** -0.051
(0.116) (0.129) (0.23)
1730 -0.220* -0.323** -0.069
(0.122) (0.136) (0.237)
1760 0.061 0.025 0.364**
(0.095) (0.106) (0.187)
1790 0.306*** 0.304*** 0.752***
(0.104) (0.116) (0.204)
1820 0.388*** 0.410*** 0.851***
(0.11) (0.123) (0.215)
1850 0.411*** 0.443*** 0.411*
(0.11) (0.123) (0.218)
N 380 380 319
Adjusted-R2 0.38 0.4 0.42
Notes: *, **, *** indicates significance at the 10 percent, 5 percent and 1 percent levels,
respectively. 1700 period is the omitted category.
Acquiring religious and official titles, which were indicative of social status, was only
significantly associated with the ownership of domestic chattels in the case of men. TITLE_MEN
is significant at the 10 percent level in the first two regressions. The estates of men with titles
had 20 percent (A) and 25 percent (B) higher scores than estates belonging to the deceased
without titles. TITLE_WOMEN is not significant at the 10 percent level, suggesting that
acquiring titles had no statistically meaningful effect on the ownership of domestic goods by
women. Interestingly, the pilgrim status was inversely related to the quantity and variety of
household goods contained in an inventory. At the same level of wealth, pilgrims’ estates had
lower scores for A and B. Titles and pilgrim status had no significant effect on the ownership
of copperware.
However, the main purpose of the exercise is to detect improvements over time in the
ownership of domestic goods that cannot be explained by variations in real estate values or
by differences in the composition of the sample in terms of gender, religious status, or titles.
This is the time effect, which is assumed to be captured by the period dummies. When
controlled for changes in wealth, and differences in the sample across periods, the coefficients
of the period dummies 1790, 1820, and 1850 are positive and highly significant for all three
series. This shows that at constant wealth values, the index scores are higher in the last three
periods compared to the initial period. Estates of the same total value contained more and
more various domestic goods around the mid-nineteenth century. Like several empirical
studies on European consumerism, these findings point to the second half of the eighteenth
century as the period when consumer goods proliferated in the Ottoman Empire.
Another interesting finding concerns the 1730 period, which is often associated in the
literature with conspicuous consumption and extravagant lifestyles of the rich, and more
recently, with the rise of the consumerist desire among ordinary Ottomans. The results,
however, contradicted these arguments. The regression detected a decline in the ownership
of household goods at constant wealth levels during the Tulip Era. This, of course, does not
rule out the possibility of progress in the possession of household durables that occurred in
parallel to the rise in wealth levels. Having said that, it should be also recalled that the limited
sample size for this period might be responsible of this situation.
We can probe the change over time further. The estimated coefficient on a time
dummy corresponding to a particular period is an estimate of the difference between the
intercept in that period and the intercept in the omitted time dummy. Hence, the coefficient
of the period dummies reflects changes in the level of index scores compared to the initial
period (1700), and does not give us an idea about changes between consecutive periods. For
instance, the fact that the dummies for 1820 and 1850 are significant and positive only
indicates that in these periods, the index scores were higher than the level in 1700 when
controlled for the wealth, religious status and gender of the estate owner, and does not
provide evidence of a statistically significant increase from 1790 to 1820 or from 1820 to 1850.
However, by altering the time dummy omitted, we can observe changes between consecutive
periods. When the 1730 dummy is omitted instead of the one from 1700, the coefficient of the
1760 dummy will now reflect the variation in the scores between 1730 and 1760, provided
other variables remain constant. The coefficients of all other variables will remain the same,
while the constant and the coefficients of the time dummies vary.
Table 23 examines the exponential coefficients of the time dummies when the dummy
for the previous period is omitted.64 This allows us to identify statistically significant changes
between successive periods. At the 1 percent level, the intercepts of the 1790 period are
significantly different from the intercepts of the 1760 period. The results show that while
wealth remained constant, index scores A and B rose by 28 percent and 32 percent from 1760
to 1790. The intercepts for 1760 are significantly higher than those for 1730 at the 1 percent
level for index B, and at 5 percent level for index A, although the scores for this period are no
different than the initial period (see Table 10). The results also suggest a decline in scores from
1700 to 1730 at constant wealth levels, as demonstrated by negative and significant
coefficients at the 10 percent level. The third regression employing the amount of copperware
owned as the dependent variable yields similar results. Like the ownership of other household
goods, the ownership of copperware improved at constant levels of wealth between 1730 and
1790. However, from 1820 to 1850, the copperware appearing in the estates of comparable
real wealth declined by 36 percent.
64 Only CONSTANT and the coefficients of the time dummies vary. Coefficients of LNREALWEALTH,
MEN_NM, WOMEN_M, and WOMEN_NM remain the same. CONSTANT indicates the intercept for the
omitted period.
Table 23- Exponential coefficients of time dummies (exp(n)-1)
Omitted variable Time dummy Dependent variable

LNINDEXA LNINDEXB LNCOPPERKG
1700 1730 -0.197* -0.276** -0.07
1730 1760 0.325** 0.417*** 0.542*
1760 1790 0.277*** 0.321*** 0.474**
1820 1820 0.085 0.112 0.104
1850 1850 0.023 0.034 -0.356***
respectively.
The findings of this study demonstrate that Ottoman individuals who were equally
wealthy owned a greater variety and quantity of domestic durables from the second half of
the eighteenth century onwards. The ownership of goods at constant wealth levels declined
from 1700 to 1730 (for household goods other than copperware), rose between 1730 and
1790, and then remained roughly the same until around the mid-nineteenth century.
As a check on the robustness of these results, the index scores and the amount of
copperware owned are regressed against LNWEALTH2, an alternative measure of wealth.
(2)
𝐿𝑁𝐼𝑁𝐷𝐸𝑋𝑆𝐶𝑂𝑅𝐸𝑖
= 𝛽0 + 𝛽1 𝐿𝑁𝑊𝐸𝐴𝐿𝑇𝐻2𝑖 + 𝛽2 𝑊𝑂𝑀𝐸𝑁_𝑀𝑈𝑆𝐿𝐼𝑀𝑖 + 𝛽3 𝑀𝐸𝑁_𝑁𝑂𝑁𝑀𝑈𝑆𝐿𝐼𝑀𝑖
LNWEALTH2 is constructed by excluding the value of the domestic durables stock. As

mentioned above, LNWEALTH, our control variable in regression (1) is not independent from
the index scores, since the total value of the domestic durable stock is a share of the total
estate value. Using LNWEALTH2 as a control variable, we will now look at whether Ottomans
with the same amount of assets other than household goods, owned a greater quantity and
variety of domestic durables.
The results presented in Table 24 are consistent with the results from the previous
regressions employing the total estate value as the independent variable. The index scores
indicating the quantity and the variety of domestic goods, as well as the amount of copperware
possessed, are strongly and positively associated with the value of the assets other than
household durables. When WEALTH2 doubles, the index scores A, B, and C rise by 13 percent,
16 percent, and 17 percent, respectively. This suggests that only a small share of the change
in the value of other types of assets was reflected in the ownership of domestic durables. As
in the previous model, the coefficients of WOMEN_MUSLIM, TITLE_MEN, and the period
dummies 1790, 1820, and 1850 are positive and significant. The coefficients of the
independent variables other than LNWEALTH2 are quite close, and are slightly higher than the
coefficients in the previous regression. This alternative model produces lower R 2 compared to
the first model since the value of the household effects and WEALTH2 are independent.
Dependent variable LNINDEXA LNINDEXB LNCOPPERKG
CONSTANT 0.546*** 0.642*** -1.491***

(0.159) (0.179) (0.364)
LNWEALTH2 0.178*** 0.208*** 0.524***
(0.02) (0.022) (0.044)
WOMEN_MUSLIM 0.362*** 0.452*** -0.075
(0.109) (0.122) (0.208)
MEN_NONMUSLIM 0.063 0.132 -0.053
(0.114) (0.128) (0.223)
WOMEN_NONMUSLI
0.036 0.077 -0.584**
M
(0.126) (0.142) (0.251)
TITLE_MEN 0.230** 0.275** 0.043
(0.111) (0.125) (0.224)
TITLE_WOMEN -0.066 -0.082 0.158
(0.133) (0.149) (0.242)
ELHAC -0.193 -0.226 -0.076
(0.125) (0.14) (0.234)
1730 -0.207 -0.310** -0.071
(0.133) (0.15) (0.242)
1760 0.097 0.066 0.360*
(0.102) (0.115) (0.191)
1790 0.292*** 0.285** 0.679***
(0.112) (0.126) (0.211)
1820 0.420*** 0.445*** 0.838***
(0.119) (0.133) (0.22)
1850 0.440*** 0.477*** 0.403*
(0.119) (0.134) (0.223)
N 376 376 316
Adjusted-R2 0.29 0.3 0.42
respectively. 1700 period is the omitted category.
In order to understand better what these results mean, we can estimate the changes
in the level of domestic comfort enjoyed by Ottoman men without any titles and with a wealth
equal to the median wealth (949 constant akçe) in 1700. The estimated index scores for each
period comes from the initial equation (1).
𝐿𝑁𝐼𝑁𝐷𝐸𝑋𝑆𝐶𝑂𝑅𝐸𝐼
The estimated intercept term (0) is the intercept in the period indicated by the
omitted dummy, namely the 1700 period. The estimated coefficient on an included time
dummy corresponding to a particular period is an estimate of the difference between the
intercept in that period and the intercept in 1700 (e.g. for 1730 this equals 0+ k1).
The estimates are presented in Table 25, Figures 9 and 10; and use the coefficients
presented in Table 22.
Table 25- Estimated index scores for 949 constant akçe (Muslim male)
Copperware
Index A Index B
(Kg)
1700 5.32 7.06 8.1
1730 4.27 5.11 7.6
1760 5.66 7.24 11.8
1790 7.23 9.57 17.4
1820 7.84 10.63 19.1
1850 8.03 10.99 12.2
Increase as %
of the initial 0.51 0.56 0.51
score
In 1700, a Muslim Ottoman man with an estate worth 949 constant akçe and without
any titles possessed roughly five different types of domestic durables, excluding copperware.
About a century later, a Muslim man with comparable wealth owned eight different types of
domestic durables. During the same period, index score B, which proxies for both the variety
and the quantity of the goods owned, rose from 7 to 11. The amount of copperware owned
at this wealth level was 8 kilograms in 1700 and 19 kilograms in 1820. This figure declined to
12 kilograms in the mid-nineteenth century. Put differently, the level of domestic comfort, as
measured in terms of index B, enjoyed by a Muslim male inhabitant of Üsküdar with a wealth
of 949 constant akçe in the mid-nineteenth century could only be attained by his ancestor in
the early century if he was four times wealthier. The amount of copperware owned by this
individual would be equal to the amount of copperware owned by his counterpart who was
twice as wealthy in 1700. From the early eighteenth to the mid-nineteenth century, the
estimated scores of A and B, and the amount of copperware possessed rise by 51 percent, 56
percent, and 38 percent, respectively. When the fact that the early modern house contained
very few objects is taken into account, the extent of the change becomes evident. As such,
every additional item should have significantly contributed towards a more comfortable
domestic environment.
Figure 9-Estimated index scores for an estate of 949 constant akçe (Muslim male)
Figure 10 – Estimated amount of copperware (kg) owned for an estate of 949 constant
akçe (Muslim male)
5. Conclusion
Using evidence from inheritance inventories, this study uncovered substantial changes
in the domestic environment in the Ottoman town of Üsküdar from 1700 to 1850. The interiors
of houses grew much richer whether assessed by the more conventional method of counting
the frequency of goods, or by using an index of amenities that provides a more comprehensive
picture of the domestic environment. By looking at the change over time in the ownership of
consumer durables after controlling for variations in wealth, gender, religious status, and titles,
I identified a marked increase in the acquisition of goods that did not follow from rising wealth
in households and from the change in the sample in terms of the characteristics of estate
owners.
Three caveats need to be recognized in considering these findings. First, Ottoman

inventories capture individual not household possessions. Male and female estates were
complementary, and would ideally be taken together to evaluate changes in the material
environment. Second, I have considered household durables, but as Karababa (2006) has
suggested it is possible that the consumption of personal items might have followed a
different trajectory. Third, it is uncertain how well these findings can be generalized beyond
Üsküdar to the less well-connected regions that made up the majority of the Ottoman Empire.
The rise over time in consumer goods as observed in inventories does not simply
reflect an upward bias in the sample of inventories that became more pronounced in later
periods. The improvement in the material environment is observable even at the same wealth
level. In fact, rising wealth was not the major determinant of the growth of household goods
over time.
From the second half of the eighteenth century onwards, Ottomans who were no
richer than their counterparts in 1700 owned a greater quantity and variety of domestic goods,
and thus, enjoyed higher levels of domestic comfort. The amount of household effects owned
by a Muslim male with a wealth of about 1,000 constant akçe in the mid-nineteenth century
could only be attained by ancestors living in around 1700 who were roughly four times
wealthier. In accordance with this picture, the frequency counts of selected household goods
demonstrated that several of these goods spread throughout society during the same period.
By around the 1850s, an ordinary Ottoman household contained most of the basic necessities,
and in several cases, some of the luxuries that were reserved for the rich in the early
seventeenth century, such as coffee utensils and clocks.
The results suggest that the growth of consumer goods in the early-modern era was
not a phenomenon restricted to Europe and America, and can be observed in certain regions
outside the Western world. Another striking finding concerns the timing of the improvements.
The most rapid growth in the household effects owned at constant wealth levels occurred in
the second half of the eighteenth century. This is also when, according to several historians,
the consumer boom became revolutionary in the West. A comparison of the extent of the
improvements in the material environment between Western Europe and the Ottoman Empire
is much more difficult. The picture depicted here gives the impression that throughout the
period, Ottoman domestic interiors remained quite simple when compared to their European
and American counterparts, both in terms of the number and variety of pieces present, and
the level of elaboration of the material surroundings.
Several time-dependent factors may have undergirded the introduction of new

consumption habits. Possible candidates include price reductions, product innovations,
enhanced distribution and retail networks, and increased interest in and information about
material goods that led households to reallocate their resources in favour of these goods. An
increase in the lifetime earnings that is not reflected in the total estate values due to several
factors might also be a factor that helps explain the phenomena.
The improvement in the domestic environment from the second third of the
eighteenth century onwards overlaps with the upward trend in urban construction workers’
real wages that lasted until around the 1850s. Interestingly, both the inheritance inventories
and wage series depict an optimistic picture for this part of the century, characterized by a
general economic decay. This might be indicative of a declining demographic trend that
created labour shortages on the one hand and resulted in industrial goods becoming relatively
more abundant on the other.
Finally, if increasing acquisition of consumer goods by all segments of the society

independently of wealth is considered as the trademark of the early modern consumer
revolution, then the Ottoman Empire experienced this revolution in the second half of the
eighteenth century.
219
CHAPTER 5
RELATIVE PRICES AND THE RISE OF HOUSEHOLD GOODS
From the second half of the eighteenth century onwards, inhabitants of Üsküdar
owned a greater variety and quantity of domestic goods compared to their counterparts
possessing the same level of real wealth in the early century. In this paper, I analyse the
relationship between the possession of household durables and their overall value in the
estates and trace the evolution of the price of consumer goods as a possible explanation for
the improvements in the domestic environment.
As Shammas (1989) highlighted, an important lacuna in the historical literature on

consumption is the lack of attempts to distinguish changes in consumption due to alterations
in price or income from those produced by a shift in tastes. If higher income levels are set
aside, the most straightforward explanation of the growth of consumer goods would perhaps
be a shift in the consumer strategies driven by changing attitudes towards consumption and
the domestic environment. In the face of higher standards imposed by a new understanding
of a comfortable and desirable domestic life, we might expect that in the eighteenth century,
individuals redefined their priorities and thus reallocated their resources in favour of domestic
durables. A second (or complementary) explanation can be sought in prices. A decline over
time in the prices of consumer durables would mean that households might have acquired a
greater quantity of household goods than they had earlier, despite the constant – and
sometimes even shrinking – overall value of their total stock of such goods.
These two alternative explanations are of a different nature and have different
implications for the study of Ottoman consumption and living standards. If the increase in the
ownership of domestic goods occurred due to greater resources devoted to these goods,
changes in preferences finding their source in transformations in the sociocultural sphere
would appear to be the most important factor underlying eighteenth- and nineteenth-century
Ottoman consumerism. This would provide support for the relatively autonomous cultural
approach to the consumer change (McKendrick 1982; McCraken 1988). In this case, the
ownership of domestic goods would not point to an increase in living standards but merely to
Chapter 5-Relative prices and ownership of household goods 220
a change in priorities, since the individuals had to abandon a share of their spending on other
items in order to acquire more consumer goods.
However, if the ownership of greater quantities of domestic goods was made possible
by the declining prices of such goods, then our attention needs to focus on the economic
dynamics and structural changes in production and trade that might have provided the driving
force for Ottoman consumerism. In this case, individuals would have achieved an increase in
their living standards by acquiring more goods and achieving higher levels of domestic
comfort without a trade-off between different expenditure categories.
Changes in consumer preferences are conventionally traced in expenditure patterns.

In an ideal world, we would look at household budgets to investigate whether the rise of
consumer goods in the early-modern period can be explained on the basis of a demand shift
that resulted from sociocultural transformations in early-modern societies. At constant income
and price levels, an increased preference for household durables would be shown by an
increase in the share of household budgets spent on these goods at the expense of other
items of consumption. Unfortunately, historical household budgets are rarely available in most
cases, and are certainly not extant for Ottoman households.
In the absence of budgets, we are forced to explore what inventories can tell us about
this question. As De Vries and other historians have emphasised, probate/inheritance
inventories are not ideal sources to determine whether more goods were consumed because
prices declined, income levels rose, or consumers changed their budget allocations and spent
more on that category of goods. Consumption is a flow concept linked to income rather than
wealth. Probate inventories are snapshots of the possessions of the deceased at the moment
of death, and as such, they capture only a part of all the consumer goods acquired during a
person’s lifetime. Since early-modern consumption changes were characterized by the
acceleration of fashion cycles, the diminishing durability of goods, and a lower intrinsic or
resale value, the evolution of the value of the stock of consumer goods in the estates cannot
be understood as a reflection of the evolution of expenditures (De Vries 1994, 2008). In other
words, the relative importance of consumer goods within overall wealth is not an indicator of
consumer preferences.
Notwithstanding the pitfalls and dangers of conducting an analysis based on the value
of the consumer goods stock recorded in the inventories, their analysis can yield important
insights regarding the relationship between the possession of household goods and their
value. Numerous studies on early-modern consumerism revealed that in seventeenth- and
eighteenth-century Europe and America, the growing prominence of consumer goods was
tied to the reduced (or constant) values of the consumer goods stock. Here, I explore whether
this characteristic feature of Western consumerism was also present in Ottoman inheritance
inventories.
In this paper, I examine both the changes over time that occurred in the total value of
the stock of household durables and the evolution of prices and inventory valuations. I
investigate whether Ottomans could increase their consumption without a rise in expenditures
on domestic goods. This question can be addressed independently of whether Ottomans
raised their overall spending on consumer durables. A decline in the prices of household
durables would mean that individuals were able to enjoy more domestic chattels, even when
they did not experience a rise in their incomes, or when they did not increase the share of
income allocated to these goods.
This investigation ultimately aims to understand whether early-modern Ottoman

consumerism can be understood only as a sociocultural phenomenon. Despite the growing
interest in material culture and consumption in the Ottoman realm in recent decades, the
Ottoman world of goods has not been explored from an economic perspective. Rather,
Ottomanists are inclined to portray consumerism as a social and cultural phenomenon linked
to rising social mobility and changing modes of sociability. These studies adapt the theoretical
framework offered by McKendrick (1982) to the Ottoman context, and demonstrate how
consumption gained a social role and began to be used for social differentiation (Artan 1998,
2000; Baram 1999; Ellis 2011; Exertzoglou 2003; Karababa 2006, 2012; Karababa and Ger 2011;
Micklewright 2000; Salzmann 2000; Sajdi 2014; Zilfi 2000). Little consideration has been given
to the alternative possibility: that falling prices might have encouraged greater consumption.
In the following sections of this chapter, I first look at the share of wealth tied up in
durables, and investigate whether it grew parallel to the increase in the variety and quantity of
household goods possessed. At a second step, I focus on prices of luxuries, non-food items
and domestic goods, and explore whether a declining trend, that might be explicative of the
rising ownership of household goods, was observable.
1. Evolution of value of household durable stock over time
This first question we address here is whether the growth in consumer durables that
we observed in Ottoman households is simply explicable by a shift towards buying more
goods. To do so, I look at the overall value of domestic goods reported in the estates and
explore the evolution of the share of household durables within the whole estate.
Numerous studies on early-modern consumerism have revealed that in seventeenth-

and eighteenth-century Europe and America, the growing prominence of consumer goods
was paired with their reduced (or constant) significance to the total wealth of those that
bequeathed estates. Shammas (1982) was the first historian to discover this vital fact of early-
modern consumerism. She calculated that the percentage of total wealth in her inventories
formed by consumer goods remained fairly steady from the sixteenth century until the
eighteenth century at around 25 percent. Similarly, Main and Main (1988) determined that
consumption goods declined not only as a percentage of total probated wealth, but also in
absolute values, in southern New England between 1640 and 1774. The decline of the share
of consumer goods, they supposed, could be explained by the process of investing in farm
improvements, which would raise the value of farm assets as frontier settlements aged. They
also thought that the decline in absolute values was surprising and warranted explanation.
Carr and Walsh (1980) showed that in colonial Chesapeake, the value of consumer durables as
a percentage of total estate value rose from the 1650s to the 1670s before falling in every
county examined from the 1670s to the late 1770s. Here, too, the absolute values often – but
not always – fell. In England, Earle (1989) found that “there was little change over time [in the
value of domestic goods], a rather surprising result since…there was a considerable qualitative
change in domestic goods.” Here, I ask whether this characteristic of the consumer revolution
was also present in the Ottoman Empire, or whether they highlight an economic shift that was
characteristic of European and North America but not observable elsewhere in the early-
modern world.
In the first exercise, I examine whether the share of domestic goods in the total estate
value increased over time. This refers to the relationship between the value of domestic
durables and other components of wealth (such as real estate, production goods, jewellery,
books, and the like). However, one needs to be cautious when interpreting this relationship.
The proportion of material wealth represented by domestic goods might reveal more about
the other components of wealth than it does about domestic durables (Overton 2002). For
instance, a rise in the value of real estate recorded from one period to another might mask the
increase in the value of domestic durables in absolute terms, by keeping the share constant or
reducing it. Looking at the change in the share of domestic goods at constant wealth levels
solves the problem only partially.
The nominal values of domestic durables are a better guide to understanding whether
the increased quantity and variety of household goods in the estates can be coupled with the
rising values of such goods, as one might expect. In the second exercise, I explore the evolution
of the value of the domestic durable stock expressed in terms of Pamuk’s consumer basket,
which consists of basic consumer goods 65.
1.1. Methodology
In order to understand whether the increased quantity and variety of household

goods in Ottoman estates occurred in tandem with the rising values of such goods, the share
of household durables within the whole estate and the value of household durables in terms
of CPI are regressed against variables indicating total wealth, gender, religious status, titles,
and the period dummies.
65 Although the exercise is the same as deflation, I prefer not to call the resulting figure “real” or “
absolute” value, unlike most other studies. The method of deflation, dividing a nominal quantity by
the price level, is applied to express the quantity in real terms. However, when the nominal value of
consumer durable stock is divided by an index of agricultural prices, as seen in many of the studies
above, this does not give us real quantities, unless the relative prices of agricultural and consumer goods
are constant (meaning agricultural prices can act as a proxy for manufactured goods prices). The greater
the decline or increase in the prices of manufactured goods relative to agricultural goods is, the greater
the gap between real terms and the resulting figure. Therefore, here, the quotient is interpreted to
represent the number of staple goods (constituting Pamuk ’s consumer basket) that the stock of
consumer durables possessed by estate owners corresponds to at different periods. Since staples
constituted by far the greatest share of the total expenditures of a household, and since the majority of
the inhabitants of Üsküdar were engaged in agricultural activities – meaning they had recourse to
agricultural incomes during the period – this can provide a cogent indicator of the phenomenon at play.
(1)
𝐿𝑁𝑆𝑇𝑂𝐶𝐾𝑉𝐴𝐿𝑈𝐸𝑖
(2)
𝑆𝐻𝐴𝑅𝐸𝑖 = 𝛽0 + 𝛽1 𝐿𝑁𝑊𝐸𝐴𝐿𝑇𝐻𝑖 + 𝛽2 𝑊𝑂𝑀𝐸𝑁_𝑀𝑈𝑆𝐿𝐼𝑀𝑖 + 𝛽3 𝑀𝐸𝑁_𝑁𝑂𝑁𝑀𝑈𝑆𝐿𝐼𝑀𝑖

As the stock of household durables is correlated with wealth, the incorporation of a

variable indicating the total estate value allows us to observe whether the value stock of
household durables possessed by individuals at the same level of wealth, increased over time.
This also helps us to avoid any distortions that might occur due to possible upward or
downward biases in the samples across different periods. If the share of wealth allocated to
domestic goods and the value of household durable stock in terms of CPI increased over the
course of time, we would expect the coefficients of the dummies for later periods to be
significant and positive when controlled for wealth. Interaction dummies bringing together the
gender and religious statuses of the estate owners (MEN_M, WOMEN_M and NON-MUSLIM)
are included as control variables.
The descriptive statistics of the regression variables are reported in Table 1.

N Mean S.D. Min Max

LNSTOCKVALUE 380 5.916 1.360 -2.263 9.064
LNWEALTH 380 7.549 1.387 4.158 11.369
WOMEN_MUSLIM 380 0.300 0.459 0 1
MEN_NONMUSLIM 380 0.174 0.379 0 1
WOMEN_NONMUSLIM 380 0.126 0.333 0 1
TITLE_MEN 380 0.261 0.440 0 1
TITLE_WOMEN 380 0.126 0.333 0 1
ELHAC 380 0.074 0.262 0 1
SHARE 380 29.689 23.341 0.003 100
1.2. Results
Regression results are presented in Table 2. In both regressions, the coefficient of LNWEALTH
is significant at the 1 percent level. As expected, wealth is positively associated with the overall
value of the household good stock and negatively associated with the share this stock held
within the total estate. When other variables are held constant, the share of household
durables declined by 5.3 points while the value of domestic durables rose by 60 percent when
wealth doubled.
Dependent variable SHARE LNSTOCKVALUE
CONSTANT 87.073*** 0.608*

(6.413) (0.321)
LNWEALTH -7.651*** 0.676***
(0.8) (0.04)
WOMENMUSLIM 7.206** 0.526***
(3.703) (0.186)
MENNONMUSLIM -0.586 0.105
(3.896) (0.195)
WOMENNONMUSLIM -5.495 -0.223
(4.294) (0.215)
TITLE_MEN 6.274* 0.173
(3.805) (0.191)
TITLE_WOMEN -0.212 -0.171
(4.531) (0.227)
ELHAC -9.526** -0.3
(4.267) (0.214)
1730 -7.891* -0.207
(4.506) (0.226)
1760 -3.978 -0.069
(3.5) (0.175)
1790 -6.012 -0.029
(3.828) (0.192)
1820 1.326 0.273
(4.055) 0.203
1850 2.147 0.245
4.075 0.204
N 380 380
Adjusted-R2 0.29 0.47
respectively. 1700 is the reference category.
The interaction dummies show that Muslim women were distinctive in terms of the
composition of their estates as well as the value of the household goods they owned. At
constant levels of wealth, household durables occupied a greater place (about 8 percent more)
in the estates of Muslim women, while the value of domestic durables possessed by this group
was 52 percent higher compared to Muslim men. These results are in accordance with the
findings on index scores of consumer goods in chapter 4, which revealed that Muslim women
possessed a greater quantity and variety of household chattels. The overall value of the stock
of household effects possessed by a Muslim male estate owner with a wealth of 949 constant
akçe (the median wealth in 1700) is estimated at 209 constant akçe, while the same figure was
319 constant akçe for a Muslim female. At this wealth level, the value of the domestic durable
stock constituted 34 percent of the total value of Muslim male estates, and 42 percent of
Muslim female estates. No significant difference is detected between male and female non-
Muslims and Muslim men.
When wealth remained constant, household durables held a greater share in the
estates of men with official and religious titles (about 6 percent more), and a smaller share
(about 10 percent less) in the estates of pilgrims compared to those individuals who held no
titles.
At constant levels of wealth, the value of the domestic durable stock or its relative
share of overall wealth did not increase over time. Neither of the regressions detected a
statistically significant difference between 1700 and the later periods. These results are similar
to the findings of previous research for large parts of early-modern England, North America,
and the Netherlands, which have shown that both the absolute value of probated consumer
goods and their share within the overall estate remained stable or declined, while people
owned increasingly more goods.
As pointed out by De Vries, the constant stock values of consumer goods do not
necessarily imply that the flow of household expenditures was also constant. The size of the
stock can be a misleading proxy for expenditures when the rate of change of this stock is not
taken into consideration. As snapshots of the possessions of the deceased at the moment of
death, probate inventories capture only a part of all the consumer goods acquired during a
lifetime. The turnover associated with the accelerating changes in fashion cycles throughout
the long eighteenth century resulted in an increasingly smaller number of goods purchased
over the course of a lifetime being included in the inventories. This implies that long-term
growth in the volume and diversity of possessions in inventories is an underestimated measure

of the real increase in consumption, both in real and monetary terms.
This notwithstanding, it should be recalled that the purpose of the exercise is not to
observe changes in household expenditures on domestic goods from the stock values; rather,
it is to explain the contradiction between material and monetary expressions of the household
durable stock in inventories. As we saw in chapter 4, the quantity and variety of the household
effects contained in Ottoman inheritance inventories increased over time independently of
any growth in wealth. The results of this exercise suggest that this growth was not driven by a
rising shares of the stock of these goods within estates. We can thus reject one obvious
explanation for how consumption might increase, and again see a close parallel between
Ottoman and Western patterns of consumption in this period.
2. Prices of luxuries, non-food items and domestic goods
The other potential explanation for this paradox of rising consumption despite
stagnant wealth is a decline in the prices of such goods. We can consider the stock of
household durables in each inventory as a bundle. The overall value of this bundle would be
equal to the sum of the per-unit-price of each good multiplied by its quantity.
𝑉𝐴𝐿𝑈𝐸 = 𝑄1 ∗ 𝑃1 + 𝑄2 ∗ 𝑃2 + ⋯ 𝑄𝑛 ∗ 𝑃𝑛
In this formula, the quantity (and the variety) of domestic goods owned represents
the relationship between the total value of the bundle and the prices. If everything else
remained constant, an increase in the amount of goods owned implies either an increase in
the overall value of the bundle or a decline in prices (or a combination of these two). At a
constant price level, people who owned more goods would see the total value of their
possessions increase. On the other hand, if prices were declining, the same total value would
indicate ownership of a larger quantity of goods.
Falling prices has a long been identified as one driver of rising consumption in Europe
and America. Shammas (1994) and Overton (2000), showed that the prices of manufactured
goods, particularly of textiles, declined substantially in early-modern England and America.
More recent findings on the changes in relative prices of agricultural and non-agricultural
prices in Europe (Clark 2004; Hoffman et al. 2002; Malanima and Pinchera 2012) have
supported the hypothesis that declining prices played a crucial role in explaining the rising
consumerism in the seventeenth and eighteenth centuries.
Does the evolution of prices also explain consumption in the Ottoman Empire? Were
the Ottomans able to increase their consumption without incurring rising expenditures on
domestic goods? This section of the paper explores the evolution of the inventory valuations
of household effects and the prices of some other manufactured and traded goods in the
eighteenth- and nineteenth-century Ottoman Empire, in an effort to understand how the
changes in relative prices were linked to the growth of domestic goods.
At this juncture, it should be recalled that most estate owners were engaged in
agricultural activity to some extent, as suggested by the frequent appearance of rural assets
and means of production in the inventories, and one would expect the semi-rural character of
Üsküdar to enhance the effect of relative price trends. A rise in agricultural prices relative to
manufactured goods would not only expand the ability to purchase manufactured goods but
would also lead to an increase in agricultural incomes.
Comparing the levels of relative prices (price of silver and manufactured goods relative
to wheat) across nations, Allen et al. (2004), suggest that in addition to being a region of high
wages, rents and cheap silver, Northwest Europe was also a region with the world’s most
expensive food grains vis-à-vis the world’s cheapest non-food industrial goods. Allen et al.
(2004) conclude that the discrepancy in the relative prices is a source of misleading findings
regarding global inequality based on grain wages that depict a picture of non-divergence prior
to the Industrial Revolution. They argue that wages deflated by a consumer basket involving
non-food goods, as well as staple goods, would provide a more accurate picture of inequality
between nations, allowing us to observe divergence well before the nineteenth century 66.
However, in their analysis, Allen et al. (2004) focus on levels rather than trends. That
manufactured goods were cheaper relative to grains in Northwest Europe compared to other
parts of the continent, including Ottoman Istanbul – and presumably other parts of the world
66 In contrast to the stylized facts this study highlights, the tables presented demonstrate that the
region that was home to the cheapest manufactured products in almost every case was Spain (rather
than England or the Netherlands). Spain was followed by England or the Netherlands depending on the
item. For the tables of the paper, see http://www.iisg.nl/hpw/globalhistory.php.
– does not necessarily mean that the latter did not experience the same trend in prices during
the same period. By investigating the evolution in the prices of manufactured and traded
goods and the valuations of domestic goods in the Ottoman realm between 1700 and 1850, I
aim to understand whether the downward trend in the relative prices of industrial goods was
a pattern particular to Northwest Europe or can be expanded to other regions.
3.1. Manufactured and traded goods
We can obtain an initial sense of the movement of the prices of manufactured against
agricultural products if we look at a snapshot of the start and points of our period. Table 3
compares the 10-year average of the real prices of a number of manufactured and traded
goods (sugar, coffee, soap, nails, charcoal, wood, woollen cloth, imported cloth, and velvet)
from the early seventeenth (1600-1610) and mid-nineteenth (1850-1860) centuries. The price
series are taken from Pamuk (2000a), while the real prices are obtained by dividing the nominal
prices by Pamuk’s consumer price index, his indicator of the general price levels in Ottoman
Istanbul.
Apart from wood, itself an agricultural product, the real prices of all the goods
dropped during the 1600-1860 period (see Table 3). The largest drop occurred in imported
cloth (Londrina) and velvet. Between these dates, the price of imported cloth declined by 97
percent while velvet diminished by 82 percent. The decline (42 percent) in woollen cloth prices
was less emphatic, yet still significant. Compared to the price levels in the early seventeenth
century, 10-year average real prices of coffee and sugar were about 75 percent lower in the
mid-nineteenth century. Similarly, the real price of soap, nails and charcoal decreased by 43
percent, 52 percent, and 33 percent, respectively.
Table 3- Comparison of prices (in constant akçe)

Early seventeenth to mid-nineteenth centuries
Change as % of
initial price
Early 17th c. Mid-19th c. (1600-1860)
Sugar/CPI (kg) 11.2 3 -0.73
Coffee/CPI (kg) 13.8 4 -0.77
Soap/CPI (kg) 20.9 11.9 -0.43
Nails/CPI (kg) 12.5 6 -0.52
Charcoal/CPI (100kg) 19.1 12.8 -0.33
Wood/CPI (250 kg) 3.8 9.1 1.37
Woollen cloth/CPI (meter) 91 48.6 -0.47
Imported cloth/CPI (meter) 49.1 1.5 -0.97
Velvet/CPI (meter) 143.3 45 -0.82
Source: Prices are taken from Pamuk’s (2000a) database.
Figure 1, looking at the 10-year average real prices of manufactured and traded goods
contained in Pamuk’s consumer basket clearly reveals the decline in the real prices of most of
these goods between 1600 and 1860. Furthermore, the figure demonstrates that the average
price of sugar, coffee, soap, charcoal, and nails followed an inverse U-shape pattern, rising
from the early sixteenth to mid-eighteenth century, before subsequently falling until the end
of the period. The same pattern, albeit less markedly and with a peak in the early eighteenth
century, is also visible for woollen cloth and Londrina67.
67 It is quite striking that a similar inverse U-shape pattern in real prices of industrial and traded goods
is identified in England and Italy by Clark (2004), Hoffman et al. (2002), and Malanima and Pinchera
(2012) for the same period. The “bundle of modern goods,” as Clark puts it, became more expensive
relative to staples in England between 1650 and 1730 before becoming more inexpensive until the last
decade of the eighteenth century. Similarly, in Central-Northern Italy, the increasing trend of real textile
prices was reversed around 1730, and the price of textiles declined relative to the general price level
between 1730 and 1790.
Figure 1- Manufactured and traded goods’ prices (in constant akçe)
Sources: Prices are taken from Pamuk’s (2000a) database.
The inspection of Figure 1 also makes the considerable volatility of the manufactured
and traded goods’ real prices, manifest. Therefore, at a further step, the time trend of the yearly
real prices is investigated by a regression. To understand the trajectory of price movements
better, we can break down the period into shorter durations. For each of the goods, time
trends in yearly prices in constant akçe are estimated using the regression below:
(3)
𝑃
𝐿𝑁 (𝐶𝑃𝐼𝑡 ) = 𝛽0 + 𝛽1 𝑇𝐼𝑀𝐸 + 𝑢
𝑡
Regressions are conducted separately for the seventeenth, eighteenth and nineteenth
centuries, as well as for the whole period (1600-1860). The estimated time trend coefficients
indicating the yearly rate of the change in prices are given in Table 4. For the whole period,
the downward trend in real prices is confirmed by the results. Except for nails and wood, the
time coefficient is negative and significant at the 1 percent level.
Table 4- Estimated time trend coefficients (in constant akçe)
1600-1700 1700-1800 1800-1860 1700-1860

Sugar -0.0018 -0.0017** -0.0254*** -0.0035***
Coffee -0.0022 -0.0060*** -0.0324*** -0.0055***
Nails 0.0033*** -0.0025* -0.0162** -0.0008
Charcoal -0.0034 0.0103 -0.0096***
Soap 0.0001 -0.0040*** -0.0061*** -0.0030***
Wood 0.0003 -0.0012 -0.0427 0.0032***
Woollen cloth 0.0017 -0.0089*** 0.0013 -0.0063***
Imported cloth -0.0031 -0.0149*** -0.0267* -0.0125***
Velvet -0.0033 -0.0288*** 0.0464* -0.0092***
Source: Prices are taken from Pamuk’s (2000a) database.
The results also show that despite skyrocketing grain prices due to widespread
rebellions in the countryside in the seventeenth century, drops in the relative prices of
manufactured and traded goods were not a general phenomenon during this century. The
coefficients of several goods are positive and none of the negative coefficients are significant
at the 10 percent level for the seventeenth century. This notwithstanding, there is clear
evidence that prices dropped throughout the eighteenth century prior to industrialization that
occurred, although in general, the annual rate of decline between 1800 and 1860 was higher
than the yearly rate between 1700 and 1800.
Overall, the prices of manufactured and traded goods appear to support the argument
that the growth in quantity of consumer goods possessed by the deceased may in part be
explained by falling prices. However, none of the prices in those series are for the types of
consumer goods we find in inventories, although some are for commodities used to
manufacture consumer durables. Hence, it is important to examine whether similar price trends
can be identified from the evidence in the inventories themselves. Now we turn to valuation
of household durables appearing in the inventories to test whether the decline observed in
manufactured and traded goods’ real prices throughout out this period was also reflected in
the inventory valuations of consumer durables.
4.2. Household durables
In this section, based on the valuations in the inventories, I trace how the prices of
selected household effects evolved over time. Data on industrial prices are harder to collect
and analyse than data on agricultural goods. While the unit valuations of consumer goods in
inventories were not the prices of first-hand goods in the market, they still permit us to observe
general price trends 68. The most important example of this approach is Overton’s analysis of
English prices. Employing inventory valuations, Overton (2000) constructed composite price
indices for wood (bedsteads, chairs, chests, coffers, cupboards, tables), textiles (blankets,
coverlets, cushions, napkins, pillowcases, sheets, tablecloths, towels), metals (brass pots, brass
candlesticks, dripping pans, frying pans, irons, chaffing dishes, pewter plates), agricultural
products (wheat, rye, barley, oats, cattle, horses, sheep, pigs), and capital goods (furnaces,
harrows, ladders, malt mills, ploughs and gear, spinning wheels). His findings reveal that
consumer goods became increasingly cheaper from the mid-seventeenth to the mid-
eighteenth centuries, while the price of agricultural products and capital goods rose during
the same period.
Do we see a similar change in the prices of household goods in the Ottoman Empire
from 1700 to 1850? I investigate this using my sample of inventories from the town of Üsküdar.
The unit valuation of sheets, mattresses, duvets, cushions, pillows, chests, and felts 69 are drawn
from inventories in the sample. It has been previously demonstrated that the quality of the
goods and – thus the prices – were positively associated with wealth. To limit the study to
goods of “average quality,” the valuations are taken from the interquartile range of the sample
in each time interval, while observations that are described as old, torn, ornamented, and the
like are omitted, as well as outlier values. Unit valuations are divided by Pamuk’s CPI (2000a),
and average figures are computed for each period (see Table 10).
Table 5 and Figures 2 to 6 look at unit valuations of the selected goods. The first
striking finding is that mean valuations for mattresses, cushions, pillows, duvets, and sheets
68 The reliability of inventory valuations in reflecting the trends in conventional prices is discussed in
the first chapter of the present study.
69 Throughout the period and at all wealth levels, these goods were the most common domestic goods
present in the inventories.
moved in concert. The overlap between the trends in the prices of these goods, which were
made of the same materials, supports the reliability of the findings. The first four of these items
were usually stuffed with cotton and covered with cotton textiles or occasionally silk-cotton or
wool-cotton mixtures. Sheets were usually manufactured from cotton or cotton-silk mixtures.
In all five cases, the average values declined modestly from 1700 to 1850, with two upward
shifts around 1730 and 1820.
Table 5- Unit valuations of selected household goods from Üsküdar inventories

(in constant akçe)
SHEET MATTRESS CUSHION PILLOW
Mean S.D. N Mean S.D. N Mean S.D. N Mean S.D. N
1700 9.6 5.88 40 32.77 12.76 41 22.43 13.17 68 10.22 7.09 63
1730 16.1 8.35 18 36.91 12.94 14 26.96 11.56 28 17.97 8.85 42

1760 11.9 6.27 42 34.03 11.07 32 22.41 11.66 43 14.34 9.27 53
1790 8.8 5.69 28 26.44 9.47 41 18.44 12.26 48 7.93 6.09 73
1820 12.7 7.82 23 29.66 10.26 33 21.28 12.28 17 9.56 8 44
1850 6.3 3.21 10 25.16 8.64 5 19.96 9.01 14 5.26 4.49 22
CHEST FELT DUVET
Mean S.D. N Mean S.D. N Mean S.D. N

1700 7.96 5.23 20 19.67 15.35 28 22.91 14.98 88
1730 16.16 6.57 12 21.09 17.42 14 27.36 9.6 29
1760 9.3 6.39 22 12.77 12.15 22 23.46 17.79 33
1790 14.89 9.13 30 29.39 24.24 16 19.01 11.94 47

1820 19.17 11.79 28 21.46 19.98 9 22.29 13.63 62
1850 15.36 8.56 22 48.96 24.89 7 15.34 11.11 20

Figure 2- Unit valuations of sheets, mattresses, duvets, cushions and pillows

(in constant akçe)
As the quality of the textiles contained in the estates might be shifting over time, the
same exercise is repeated for mattresses covered with beledi (a local cotton fabric) and duvets
covered with yemeni (Indians), to achieve a greater homogeneity of observations. These two
textiles were the most commonly used fabrics in Ottoman houses all through the eighteenth
century, and they appear in the inventories even as late as the 1820s. The unit valuations of
both beledi mattresses and yemeni duvets relative to basic consumption goods followed the
same pattern with all mattresses and duvets (see Figure 3).
Figure 3- Price of beledi mattresses and yemeni duvets (in constant akçe)
To check the reliability of the trends, I compare with the price of beledi mattresses in
Üsküdar with those from inventories belonging to another Ottoman town, Manisa (see Figure
4). The evidence provided by Manisa inventories 70 also confirms the trends. As beledi was
produced in Manisa, the price of an item made of this fabric was lower in this city compared
to Üsküdar. At the turn of the eighteenth century, the fabric began to be produced in Istanbul,
something that was also reflected in the series through converging prices.
Figure 4 –Unit valuations of beledi mattresses from Üsküdar and Manisa

(in constant akçe)
Source: Manisa valuations are collected by the author. See Primary Sources section for details.
Unit valuations of felts and chests, on the other hand, exhibited a rising trend (see
Figure 5). Measured in terms of staples, both of these items were substantially more expensive
at the end of the period compared to the initial period. The increase in mean valuations of
chests was consistent with the rising wood prices as demonstrated by Pamuk’s price series for
this period.
70 We need a note indicating the source: where are these held? And the number of observations.
Figure 5- Unit valuations of felts and chests (in constant akçe)
Figure 6- Valuation of copperware (in constant akçe/kg)
Copperware was indispensable in an Ottoman house and possessed an important

share within the overall value of the household durable stock. Ottoman inheritance inventories
recorded copperware owned by the deceased in weight units and with valuations. I
constructed yearly price series for copperware using two observations per year. Figure 21
shows the 10-year average for copperware prices. The price of copperware relative to staples
follows a similar pattern to that of household linen, declining from 1730 to 1790 by about 60
percent, and rising from this date onwards (see Figure 6).
4.3. Household durable basket
The price of individual goods from inventories is not easy to interpret. Most fell. Some
rose. But what did this mean for the affordability of the kind of consumer goods that Ottoman
people owned? For a more comprehensive understanding of the changes in the prices of
household durables, and how this might have affected the ability of individuals to own these
goods, we can construct a consumer basket and look at how the value of this basket evolved
between 1700 and 1850. The basket includes the essential domestic goods that appear most
frequently in the estates, according to the frequency counts in chapter 4. These goods are
presented in Table 6. Even as early as 1700, they appeared in the inventories of all Üsküdar
inhabitants independent from wealth.
Table 6- Household durables contained in basket
UNIT
Copperware
2 shallow pans with handles 1.94 kg
1 cooking pot 2.65 kg
1 frying pan 0.86 kg
1 ewer 1.2 kg
1 washtub 2.45 kg
TOTAL 9.1 kg
Household goods
Mattresses 1 piece
Sheets 1 piece
Duvets 1 piece
Pillows 1 piece
Cushions 4 piece
Felts 1 piece
Chests 1 piece
Table 7- Value of household durable basket from 1700 to 1850
VALUE OF THE
BASKET/CPI
1700 297.25
1730 346.94
1760 266.63
1790 217.48
1820 257.69
1850 251.16
Figure 7- Value of household durable basket (in constant akçe)
Table 7 and Figure 7 depict the evolution of the value of the consumer durable basket.
In the first half of the nineteenth century, the overall value of the household durable basket in
constant akçe was significantly lower than it was a century ago. The basket was worth about
300 constant akçe around 1700. This figure rose to 350 constant akçe in 1730 before falling to
220 constant akçe in 1790. Following an upward shift in the early nineteenth century, the
overall value remained at the level of 250 constant akçe in 1850. The prices dropped sharply
between 1730 and 1790. However, it was in the 1760-1790 period that the prices dropped
below their initial level.
These trends in household durable prices closely mirror the evolution of estimated
index scores and the estimated amounts of copperware owned at constant wealth levels. As
prices declined in the 1730-1790 period, index scores A and B, as well as the amount of
copperware owned for constant wealth levels, steadily rose (see Figures 8 and 9). These
findings strongly support the role of the decline in consumer durable prices in the growth of
household effects.
Figure 8- Value of household durable basket (in constant akçe), and the estimated
index A and B for an estate of 949 constant akçe (Muslim male)
Figure 9- Price of copperware (in constant akçe), and the estimated amount of
copperware (in kg) owned for an estate of 949 constant akçe (Muslim male)
5. Depreciation and decline in resale values
Alongside the decline in the original purchase price of the goods, another factor
affecting inventory valuations was lower resale values (De Vries 1994, 2008). Goods
depreciated and thus, devaluated faster in the course of the long eighteenth century because
they wore out faster and went out of fashion sooner (De Vries 2008). In the case of Üsküdar,
the traces of accelerated depreciation are visible in the 1850 sample. It is very likely that the
decline in the number of household linen items reported individually during this period (score
for this category of goods declined from 1.3 in 1820 to 0.8 in 1850 in the lower wealth group,
and from 1.5 to 1.2 in the upper wealth group), was linked to their registration as lump-sum
quantities under “some used linen.” It appears as though only newer and ornamented pieces
of linen with some resale value were mentioned item-by-item in the mid-nineteenth century
inventories, while in the previous periods, almost all items were separately recorded in detail 71.
This can be perceived as an indication that most of the used linen ceased having any
substantial worth by around the 1850s.
Accelerated physical and stylistic depreciation implies that the downward trends in
inventory valuations at least partially reflected reduced resale values. Is it possible that the
declining trend we observed in valuations is merely an indicator of higher depreciation rates
rather than a sign of a drop in consumer good prices? First, inventory valuations are not the
only evidence of the drop. As has been demonstrated above, the market prices for a number
of manufactured and traded goods also decreased during this period.
In addition, higher depreciation rates or the widening difference between sale and
resale values might not have been the case with all goods. Overton’s (2000) comparison of
second-hand pewter valuations from Lincolnshire and Worcestershire inventories and Rogers’
wholesale pewter prices show no divergence between sale and resale prices from 1550 to
1750. Like pewter in European inventories, copperware in Ottoman inventories was so durable
that it functioned as a store of value. Second-hand copperware could easily be sold at the
market (Bozkurt 2011).
Therefore, even though we accept that the general declining trend in inventory
valuations was to a certain extent resulted from speeded devaluation, we can still conclude
71 This also suggests that selected household goods’ mean valuations, as presented above, might be
understating the actual fall in prices, as quality of the composition of the household goods reported in
the estates was rising over time.
that they also reflected the decreasing trend in the purchase price of the consumer durables
relative to agricultural prices.
6. Conclusion
The growth of domestic goods in the Ottoman realm from the second half of the
eighteenth century onwards was not associated with a rise in the value of domestic durable
stock owned by people with similar levels of wealth. Instead, falling prices for consumer goods
offer us a better explanation for the growth of consumption. There was a downward trend in
both the valuations of selected household goods, as well as the prices of several manufactured
and traded goods throughout the 1700-1850 period.
Consumption in the Ottoman Empire had similar characteristics to that in early-

modern Europe and America, where a range of studies have shown that both the absolute
value of probated consumer goods and their share within the overall estate remained stable
or declined, while people owned increasingly more goods. Several studies on early-modern
European prices, which have revealed that the prices of industrial goods declined relative to
those of agricultural products, have complemented these findings while offering an
explanation as to how consumer durables spread throughout society in the face of
stagnant/declining real wages. Based on these findings, it can be argued that eighteenth-
century European consumerism was not distinctive in these respects. If the rising consumption
of durables triggered by social transformations and by declining relative prices of consumer
goods was a driving force behind the Industrial Revolution, why the same mechanism did not
operate in the Ottoman Empire, where similar trends were observed, remains an unanswered
question.
This analysis of Ottoman prices and inventory valuations refutes Allen et al. (2004)’s
argument that the decline in prices of non-food items was a phenomenon particular to pre-
industrial Northwestern Europe. 72 At the present state of research, it is too early to determine
72 It could be claimed that the decreasing trend in inventory valuations in the Ottoman realm reflected
the price drops in goods imported from Europe. Nevertheless, observations on inheritance inventories
show that even as late as the 1820s, local textiles were predominant by far among the textiles used for
household linen. Furthermore, copperware, which constituted an important share of household durables
in inventories, was made of copper that was mined and processed in the Ottoman realm.
whether the price drops prior to industrialization occurred due to productivity gains in the
manufacturing and transportation sectors or were a result of demographic trends.
Finally, these findings highlight the economic dimension of eighteenth-century

Ottoman consumerism. Depicting the decline in the prices of manufactured goods allows us
to sidestep the broad question of whether early-modern individuals increased their overall
spending on consumer durables and focus on a narrower one that concerns the ability of
individual to own goods. Based on the decline in prices of consumer goods, we can draw the
conclusion that the inhabitants of Üsküdar could own more goods without increasing their
spending on domestic durables. These findings point to an increase in living standards that
cannot be captured by consumer price indexes mainly based on food.
Although changing tastes and the disappearance of social boundaries doubtlessly

played an important role in the introduction of new consumerist desires and habits in the
eighteenth- and nineteenth-century Ottoman realm, Ottoman consumerism cannot be
conceived merely as a socio-cultural phenomenon that can be explained by a greater
preference for consumer goods resulting from higher social mobility and new modes of
sociability. These results should provide further motivation to focus on transformations in the
economic sphere, alongside sociocultural processes, in explaining the spread of consumer
goods throughout society.
245
CHAPTER 6
GENDER, RELIGIOUS STATUS AND OWNERSHIP OF HOUSEHOLD
GOODS
How did gender and religious status affect the ownership of household durables in
eighteenth and nineteenth century Üsküdar? In this chapter, we compare male and female, as
well as Muslim and non-Muslim, estate owners’ possessions in an effort to understand how
material culture varied between them.
A significant portion of the scholarly literature on the topic suggests that by the
nineteenth century, women in the West were central to consumption as makers of the middle-
class home and consumers of fashion (McGowan 2006; Leach 1984; Burman 1999; Auslander
2001; Walton 1986; Breward 1994; Roberts 1998). There is, however, no consensus as to
whether this was also true in earlier periods. While some historians argue that women were
the drivers of early-modern consumerism and possessed a striking interest in new material
goods, others are more reluctant to accept that seventeenth and eighteenth century women
played the role of independent consumers, let alone pioneers of modern consumer behaviour.
According to De Vries (1994), in an age of rising consumerism, clothing and household

goods were particular novelties, and it was women who both sought such products and who
abandoned their former domestic production of basic household goods to work in the
marketplace in order to buy the novel goods, luxuries and household goods they desired.
Comparing men’s and women’s wills in Birmingham and Sheffield, Berg (1996) argued that as
early as the eighteenth century, women bequeathed significantly more items per person than
men, including clothing, glassware, jewellery, linen, plates, and silver. Berg (1996) also claims
that women devoted greater attention to their personal effects, as more detailed and
scrupulous descriptions of clothing and other personal belongings in women's wills would
suggest.
Nevertheless, the evidence does not always support an explicit differentiation

between men and women in terms of material culture and consumer habits. Weatherill’s (1986)
findings based on 3,000 British probate inventories showed little distinctiveness in the
Chapter 6-Gender, religious status and ownership of household goods 246
possessions of female estate owners, although slightly higher proportions of British women
had new and decorative goods than did men from similar classes. She concluded that the
influence of status, wealth, and occupation on consumption was more important than gender.
Similarly, Shammas (1980) objected to the argument that early-modern women were
independent consumers who expressed themselves through consumption, suggesting that
women were unable to exercise much influence on consumption decisions as they possessed
only limited control over resources.
Recent work on eighteenth century Ottoman women’s consumerism has typically

focused on elite women. Artan (2010) and Faroqhi (2002b) point to a radical transformation in
the way Ottoman princesses related to material culture and domesticity in parallel to the
enhancement of their political role. They suggested that, from the eighteenth century onwards,
women of the dynasty were less engaged with large-scale mosque complexes or other types
of charity construction. Instead, they owned lavishly decorated seashore villas along the
Bosporus (sahilsarayı), which were designed to display political and social power. Artan (2010:
124) states that “the Ottoman princesses were becoming more and more independent not just
of ‘traditions’ but also of their husbands, and this was reflected in the way in which their own
waterfront palaces came to dwarf those of their spouses along the Bosporus and the Golden
Horn.” In doing so, they not only disassociated themselves from things past, but also from
other new elites in Istanbul. The quest for differentiation through extravagance was also
expressed in interior decoration. For instance, in decorating their palaces, these women
preferred European porcelain, unlike the new elites and the commoners, for whom Oriental
porcelain remained most desirable (Artan 2010).
Little, however, is known about ordinary Ottoman women’s attitudes towards

consumption and domesticity in the eighteenth century. This is partly due to the paucity of
primary sources which could shed light on how domestic interiors were arranged and
decorated, or on the objects and textiles used for home decoration. A veil of privacy, which
the Ottomans were keen at protecting, shadows our knowledge of Ottoman family life and,
hence, of the inner spaces of the Ottoman house. Was the increased importance attributed to
the domestic environment particular to wealthy and powerful women, or did it affect ordinary
Ottoman women in the eighteenth century? Did women from middling and lower ranks also
seek social differentiation and dissociation from established roles assigned to them through
consumption? Did women’s particular roles within the household result in different material
values? Were they more prone to changing tastes and fashions than men? Unfortunately, it is
difficult to answer these questions in the absence of comprehensive studies. In this regard,
Ottoman inheritance inventories represent an even more unique opportunity to “lift up the
roofs” and “peek into the most intimate corners of a household,” than they do in Europe
(Braudel 1967: 217).
In this regard, Establet and Pascual’s (2002) study based on a sample of Damascene
inventories from around 1700 is illuminating. They demonstrate that Damascene women 73
received a patrimony that was clearly inferior to that of men. It consisted of certain items: some
real estate, but primarily jewellery and domestic goods, which made up more than three-
quarters of their belongings. Women particularly dominated the textile interior of the home.
The authors conclude that “women ruled, with more or less variety and fantasy, the furniture
of the house, which was used not only for rest and sleep, but also as a venue for receiving
guests,” (Establet and Pascual 2002: 301).
It is not only gender-based differences in consumption and material culture, but

differences across religious communities that have attracted little attention in the literature.
The only study dealing with changing consumption patterns in Ottoman non-Muslim
communities is by Exertzoglou (2003). He identifies and explains the novel consumption
practices by the Christian Orthodox in the late nineteenth century – mainly Greek-speaking
middle class groups of the major Ottoman urban centres – as an aspect of national identity
construction.
However, it is suggested that non-Muslim communities were pioneers in adopting

modern furniture and novelties symbolizing a Western lifestyle (Exertzoglou 2003; Göçek
1996). The use of Western products was closely linked to rising Western political, economic,
and cultural influence, and as such, Christian and Jewish merchants and their communities,
being in close contact with Europe, were in an advantageous position. Accordingly, Ottoman
73 Establet and Pascual (2002) do not mention the religious status of the women estate owners in their
sample, and thus do not look at whether the estates of women from Muslim and non-Muslim
communities displayed distinctive features in terms of ownership.
Muslims are usually thought to have acquired Western products later than their non-Muslim
counterparts (Göçek 1996). The European presence in the capital during the Crimean War
(1853-1856) was a turning point in terms of the Westernization of the domestic interiors in
Muslim houses (Hornby 1863). While these interpretations are plausible, the literature contains
no solid evidence to support the argument that non-Muslims were the leaders in the
possession of Western goods.
In this chapter, I grapple with a set of questions that addresses these points: What
conclusions can we arrive at regarding the varying levels of material culture encompassing
men and women and Muslims and non-Muslims in eighteenth and nineteenth century
Üsküdar? Can we identify characteristic features in the ownership of household goods for
these groups? Do the household items included in an inventory vary by gender and religious
status? Was there a complementarity between the durables owned by men and women? Did
women and non-Muslims own more new goods associated with consumerism and
Westernization, implying that they were more inclined to adapt themselves to changing
consumer habits?
To address these questions, at a first instance, I compare Muslim and non-Muslims

and male and female estate owners in Üsküdar, in terms of wealth and ownership of real estate.
Subsequently, I look at how the overall value of household durables stock responded to
variations in total wealth. Finally, I analyse the ownership of selected household goods
according to gender and religious status.
1. Wealth and ownership of real estate
Before delving into the questions stated above, we need to examine the wealth and
ownership of real estate and ask whether there were significant differences between Muslim
men, Muslim women, non-Muslim men, and non-Muslim women.
Studies on inheritance inventories from different regions of the empire unanimously

reveal that Ottoman women were in general poorer than their male counterparts (Establet and
Pascual 2002; Grehan 2007; Karababa 2012; Ergene and Berker 2008). These studies also reveal
a significant difference in terms of ownership of real estate and production goods across
genders. However, they do not provide information about wealth disparities between Muslim
and non-Muslim communities.
As shown in chapter 4, wealth was positively and strongly associated with the quantity
and variety of the household goods that the deceased possessed. Thus, any differences in
distribution and level of wealth, and in the composition of estates across genders and religious
groups, needs to be identified and controlled for when we are exploring the effect of gender
and religious status on the ownership of household goods. We can only meaningfully speak
of an “effect of gender and religious status” when examining variations in the level of material
culture between individuals of comparable wealth.
In Table 1, the average total estate values for inventory in each category are given.
Except for the initial period, the average wealth of Muslim male inventories in constant akçe is
the highest of all four categories. Non-Muslim men and non-Muslim women came below
Muslim men. In 1700, Muslim women with a mean wealth of 1,227 constant akçe were three
times poorer than Muslim men, who left inventories worth 3,711 constant akçe on average. At
the end of the period, the gap remained the same. When all periods are taken together, Muslim
women appear as the poorest group in our sample (with a mean total estate value of 2,664
constant akçe), with Muslim men the richest. Putting it differently, Muslim women held a mere
17 percent of recorded assets, even though they represented 30 percent of the total group.
The mean value of the estates left by non-Muslim men was lower than that of those left by the
Muslim men, yet were higher than those belonging to non-Muslim women.
Out of the 38 estates in the upper wealth decile, only eight belong to women, and
out of the 162 female estates, only 21 finished in the upper quartile. This suggests that most
of the female estates included in this study belonged to women who controlled very limited
resources. In interpreting the results of this study, this fact should be taken into consideration.
Table 1- Mean total estate value by gender and religious status groups
(in constant akçe)
Muslim Non-Muslim Muslim Non-Muslim

men men women women
1700 3711 7901 1227 5390
1730 6922 5910 2023 2907
1760 4511 2618 2632 4306
1790 5192 5166 960 3089
1820 6919 5204 5837 2289
1850 9838 3365 2302 3253
ALL 6630 4962 2664 3476
N 152 66 114 48
Although evaluating the mean values can give us an idea, this can be deceptive due
to the presence of outliers, particularly when working with small samples. An OLS regression
employing dummies can provide us with a more reliable picture as to whether there is a
statistically significant difference between the total wealth of members of these four
categories. The regression model employed is as follows:
(1)
𝐿𝑁𝑊𝐸𝐴𝐿𝑇𝐻𝑖 = 𝛽0 + 𝛽1 𝑀𝐸𝑁_𝑁𝑂𝑁𝑀𝑈𝑆𝐿𝐼𝑀𝑖 + 𝛽2 𝑊𝑂𝑀𝐸𝑁_𝑀𝑈𝑆𝐿𝐼𝑀𝑖

+ 𝛽3 𝑊𝑂𝑀𝐸𝑁_𝑁𝑂𝑁𝑀𝑈𝑆𝐿𝐼𝑀𝑖 + Σ𝛽𝑘 𝑃𝐸𝑅𝐼𝑂𝐷𝑘 + 𝑢
LNWEALTH is the logged total estate value in constant akçe. MEN_NONMUSLIM,

WOMEN_MUSLIM, WOMEN_NONMUSLIM are dummy variables indicating gender and
religious status. These variables assume 1, if the estate owner is respectively a non-Muslim
man, a Muslim woman, and a non-Muslim woman, and 0 otherwise. The reference category is
Muslim males. The period dummies are incorporated into the model to control for variations
in wealth that occurred across periods.
Descriptive statistics of the regression variables are reported in Table 2.

MEN_MUSLIM 380 0.400 0.491 0 1

MEN_NONMUSLIM 380 0.174 0.380 0 1
WOMEN_MUSLIM 380 0.300 0.458 0 1
LNWEALTH 380 7.549 1.387 4.158 11.369
The results are presented in Table 3. Among the three dummy variables, only the
coefficient of WOMEN_MUSLIM is significant at the 1 percent level. According to these results,
compared to Muslim men in the sample, Muslim women were poorer. In terms of wealth levels,
there was no significant difference between Muslim men on the one side, and non-Muslim
men and women on the other.
Table 3- OLS regression results
Dependent variable LNWEALTH
Coef. Std.Err.
CONSTANT 7.2242*** 0.1815
MEN_NONMUSLIM -0.1072 0.2001
WOMEN_MUSLIM -0.5847*** 0.166
WOMEN_NONMUSLIM -0.245 0.2278
1730 0.9453*** 0.2942
1760 0.5271** 0.2271
1790 0.3229 0.2336
1820 0.8245*** 0.2256
1850 0.7736*** 0.2273
N 380
R2 0.07
respectively. 1700 is the omitted category.
The wealth inequality between men and women in the Muslim community of the town
can be partially attributed to the inheritance rules in sharia. “By law, as well as in fact, the single
or married woman inherits from her parents half the share of her brother,” (Daghestani 1932:
144). However, gender-based inequalities in inheritance practices were not particular to Islamic
law. Regulations applied by the Ottoman non-Muslim communities also reinforced economic
disparities between men and women. For instance, the Torah awards women no rights of
inheritance as long as there are male heirs in the same class. Realising that the sharia was often
more beneficial to women than the Jewish halakha, many Ottoman Jewish women appealed
to Islamic courts rather than their own communal courts to settle inheritance issues (Lamdan
2005).
In order to determine whether wealth inequality across genders was also observable
in the non-Muslim community in Üsküdar, the same regression is run this time with the non-
Muslim men as the reference category. The variable MEN_NONMUSLIM is omitted instead of
MEN_MUSLIM. The results are reported in Table 4. The coefficient of WOMEN_NONMUSLIM
is not significantly different from that of MEN_NONMUSLIM at the 10 percent level. Unlike the
Muslim estate owners observed, gender does not appear to be a significant determinant of
variations in wealth among the non-Muslim testators included in this study.
Dependent variable LNWEALTH
Coef. Std. Err.

MEN_MUSLIM 0.1072 0.2001
WOMEN_MUSLIM -0.4774** 0.2132
WOMEN_NONMUSLIM -0.1378 0.2553
1730 0.9453*** 0.2942
1760 0.5271** 0.2271
1790 0.3229 0.2336
1820 0.8245*** 0.2256
1850 0.7736*** 0.2273
CONSTANT 7.1170*** 0.2351
N 380
R2 0.07
These findings alone are not sufficient to assert that different cultural norms
regulating matters about marriage, family, and ownership in the non-Muslim communities
resulted in a more equal distribution of economic resources within the household. Besides
possible biases in the sample of inventoried estates belonging to non-Muslims, whether these
results are generalizable to the rest of Ottoman society is dubious. In Galata, located just across
the Bosporus, Baer and Göçek (1997) detected no significant difference in ownership of
immovable property and composition of estates between Muslim and non-Muslim women in
eighteenth-century inventories, concluding that the material life of women did not differ
according to religion. It would not be surprising if gender-based economic inequalities differed
from one place to another depending on the economic and social structure of the local society.
One of the fundamental reasons for Muslim women's lower wealth was their much
lower levels of ownership of real estate. The inequality in land ownership between genders is
striking. Table 5 shows the share of real estate owners as percentage of the number of
observations for each category. Whereas an important proportion of Muslim (26.3 percent)
and non-Muslim men (36.4 percent) owned houses, shops, vineyards, and gardens in Üsküdar,
immovable property appears in the estates of only 14.6 percent of non-Muslim women, and
of 7 percent of Muslim women. Putting it differently, of 79 owners of real estate in our sample,
only 15 (19 percent) were women.
Table 5- Ownership of real estate
N (Estates Share of real

recording estate owners
real estate) as %
Muslim men 40 26.3

Non-Muslim men 24 36.4
Muslim women 8 7.0
Non-Muslim women 7 14.6
Total 79
2. Wealth and value of stock of household durables
We can compare and draw out the contrasts in ownership patterns across gender and
religious status groups by analysing how the differences in the stock of domestic durables
responded to variations in total wealth. Here, I break the sample down into subsamples
according to gender and religious status in order to discover whether these categories are
distinctive in terms of the relationship between wealth and domestic durables. I regress then,
the total value of consumer durables (STOCKVALUE) against wealth in linear, semi-log, and
double-log forms separately for each of these subsamples, and compare the coefficients of
determination (R2) in order to understand which of the regression models best captures the
relationship between wealth and the value of consumer durables stock.
No time trend variable is introduced, and observations from different periods are
handled together. Putting it differently, any time effect that may alter the relation between the
value of the domestic durable stock and total wealth is ignored, and the value of the stock is
assumed to respond in the same fashion to the variations in wealth both within a period and
across periods.
Table 6 reports the descriptive statistics of WEALTH, LNWEALTH, STOCKVALUE, and

LNSTOCKVALUE, according to the four categories. When not controlled for wealth, Muslim
men had the highest mean value of the consumer durables stock with 579 constant akçe.
Muslim women and non-Muslim men followed this group with, respectively, 516 and 460
constant akçe. The average non-Muslim female estate included the smallest value of the
consumer durables stock (410 constant akçe).

according to gender and religious status
WEALTH N Mean S.D. Min. Max.

Muslim males 152 6830 12046.09 64 86341
Non-Muslim males 66 4962 5567.98 149 26117
Muslim females 114 2770 4594.1 66 24725
Non-Muslim females 48 3476 4813.92 183 23375
LNWEALTH N Mean S.D. Min. Max.

Muslim males 152 7.80 1.51 4.16 11.37
Non-Muslim males 66 7.71 1.33 5.00 10.17
Muslim females 114 7.15 1.24 4.19 10.12
Non-Muslim females 48 7.49 1.17 5.21 10.06
STOCKVALUE N Mean S.D. Min. Max.

Muslim males 152 579 862.11 2 5659
Non-Muslim males 66 460 554.49 10 2982
Muslim females 114 516 649.89 6 3574
Non-Muslim females 48 410 594.83 8 3794
LNSTOCKVALUE N Mean S.D. Min. Max.

Muslim males 152 5.46 1.54 0.66 8.64
Non-Muslim males 66 5.48 1.27 2.31 8.00
Muslim females 114 5.59 1.23 1.84 8.18
Non-Muslim females 48 5.39 1.17 2.11 8.24
Table 7 reports the regression results. When the total value of the domestic durables
stock is regressed against the total estate value, the linear form produces the highest R 2 (0.69
and 0.56) and thus, presents the best fit for both Muslim and non-Muslim women, as shown
in Table 6. For men, the best fit is the double-log model, with coefficients of determination
0.46 and 0.53, respectively. This suggests that at all levels of wealth, the value of domestic
durables possessed by women continued to rise at the same pace as increasing total estate
value, while for men as wealth rose, the stock of domestic durables expanded at a lower rate
than other assets. The propensity of Ottoman women to own domestic goods did not decline
at higher echelons of wealth, whereas increasingly a smaller portion of the increment in the
total estate value was reflected in the stock of domestic goods at male-owned estates. These
results also imply that as wealth increased, the share of domestic goods within the overall
estate declined for Muslim and non-Muslim men, whilst for Muslim and non-Muslim women
it, remained constant.
Table 7- Coefficients of determination (Adjusted R2) from OLS regressions

(Sub-samples)
LINEAR SEMI-LOG DOUBLE-LOG N
Dependent var. STOCKVALUE STOCKVALUE LNSTOCKVALUE
Independent var. WEALTH LNWEALTH LNWEALTH

Muslim men 0.30 0.31 0.46 152
Non-Mus. men 0.32 0.39 0.53 66
Muslim women 0.69 0.56 0.63 114
Non-Mus. women 0.56 0.31 0.19 48
Men 0.31 0.33 0.48 218

Women 0.63 0.46 0.39 162
Figure 1 scatters the value of household durables stock against wealth in linear, semi-
log, and double-log forms with fitted lines, separately for men and women.
Figure 1-Value of household durables against wealth (Men and Women)

We can take this question further by estimating how patterns of accumulation varied
across gender and faith. In the following step, I examine the effect of gender and religious
status on the responsiveness of the domestic durables value through the following regression
model:
(2)
𝐿𝑁𝑆𝑇𝑂𝐶𝐾𝑉𝐴𝐿𝑈𝐸𝑖 = ∅0 + ∅1 𝐿𝑁𝑊𝐸𝐴𝐿𝑇𝐻𝑖 + ∅2 𝐿𝑁𝑊𝐸𝐴𝐿𝑇𝐻_𝑁𝑂𝑁𝑀𝑈𝑆𝐿𝐼𝑀𝑀𝐸𝑁𝑖 +

∅3 𝐿𝑁𝑊𝐸𝐴𝐿𝑇𝐻_𝑀𝑈𝑆𝐿𝐼𝑀𝑊𝑂𝑀𝐸𝑁𝑖 + ∅4 𝐿𝑁𝑊𝐸𝐴𝐿𝑇𝐻_𝑁𝑂𝑁𝑀𝑈𝑆𝐿𝐼𝑀𝑊𝑂𝑀𝐸𝑁𝑖 + Σ∅𝑘 𝑃𝐸𝑅𝐼𝑂𝐷𝑘 +
𝑢
LNSTOCKVALUE is the logged domestic durables stock. Interaction terms for non-
Muslim men, Muslim women, and non-Muslim women (LNWEALTH_NONMUSLIMMEN,
LNWEALTH_MUSLIMWOMEN, LNWEALTH_NONMUSLIMWOMEN) are computed by
multiplying the LNWEALTH by gender and religious status dummies. The reference category
is LNWEALTH_MUSLIMMEN.
Table 8 looks at the descriptive statistics of the regression variables.
N Mean S.D. Min. Max.

LNSTOCKVALUE 380 5 1 0.66 8.64
LNWEALTH 380 8 1 4.16 11.37
LNWEALTH_MENNONMUSLIM 380 1 3 0 10.23
LNWEALTH_WOMENMUSLIM 380 2 3 0 10.08
LNWEALTH_WOMENNONMUSLIM 380 1 3 0 10.15
TIME 380 3.64 1.72 1 6
Table 9 reports the regression results. Only one of the interaction terms,
LNWEALTHWOMEN_MUSLIM, is statistically significant at the 1 percent level. Compared to
Muslim Ottoman men, the slope is steeper for Muslim women. In other words, a greater
proportion of the increase in wealth in the estates of Muslim women is reflected in the value
of domestic durables stock. When wealth is increased by 10 percent, the value of domestic
goods rises by 6.4 percent in the estates of Muslim men, while it rises by about 7.1 percent in
the estates of Muslim women. While great shares held by producer goods and real estate,
particularly agricultural assets, characterised wealthy Ottoman men’s estates, the composition
of the wealthy Muslim women’s inventories did not differ much from those of poor and
middle-class women.
Dependent variable LNSTOCKVALUE

Coef. Std.Err.
CONSTANT 0.652** 0.274
LNWEALTH 0.644*** 0.035
LNWEALTH_MENNONMUSLIM 0.016 0.017
LNWEALTH_WOMENMUSLIM 0.07*** 0.016
LNWEALTH_WOMENNONMUSLIM 0.003 0.02
TIME 0.043 0.03
N 380
R2 0.5
respectively. LNWEALTH_MENMUSLIM is the omitted category.
3. Ownership of selected categories of goods according to gender and religious status
While the value of household goods gives some indication of the variation in the
approach to material culture between members of each groups, it does not treat the types of
goods in question with any subtlety. In the discussion so far we have bundled all household
goods up into a single category that is measured by value. However, the type of household
good, not its price, is important as well, particularly if we are seeking to consider the speed
and impact of cultural transmission. To address this issue, in this section we focus on the
presence or absence of selected categories of goods. Which group(s) were pioneers in
adopting new material goods? Did women’s, and especially Muslim women’s, predominance
in terms of household durables result from the possession of goods associated with rising
consumerism, or from their ownership of greater amounts of traditional chattels?
I address these questions through logistic regression equations that attempt to

predict the presence of a particular item. Logistic regression is a technique designed for use
when the dependent variable is categorical and the independent variables are categorical or
continuous. Towels, household linen, chests, clocks, modern furniture, mirrors, and cutlery and
serviettes are used as the dependent variable, with real wealth (in log natural form), period
dummies, gender and the religious status of the estate owner (MEN_NONMUSLIM,
WOMEN_MUSLIM, WOMEN_NONMUSLIM), constituting the independent variables.
MEN_MUSLIM, the reference category, and is omitted.
The descriptive statistics of the logistic regression are given in Table 10.
N Mean S.D. Min Max

TOWELS 380 0.518 0.500 0 1
HOUSEHOLD LINEN 380 0.687 0.464 0 1
CHEST 380 0.542 0.499 0 1
CLOCK 380 0.218 0.414 0 1
MODERN FURNITURE 380 0.339 0.474 0 1
CUTLERY AND SERVIETTES 380 0.308 0.462 0 1
MIRROR 380 0.166 0.372 0 1
MEN_NONMUSLIM 380 0.174 0.380 0 1
WOMEN_MUSLIM 380 0.300 0.458 0 1
LNWEALTH 380 7.549 1.387 4.158 11.369
The results are presented in Table 11. In addition to the coefficient and the standard
error of each independent variable, the odds ratios (exponentiated coefficients) are reported.
Exponentiated coefficients give the change in odds resulting from a unit change in the
independent variable. A value greater than 1 indicates that as the value of the predictor
variable increases, the odds of the outcome occurring (for example the ownership of a mirror)
also increases. A value of less than 1 indicates that as the predictor increases, the odds of the
outcome occurring decrease. If non-Muslims and women were more inclined to own novel
goods, we would expect the coefficients for these groups to be positive and statistically
significant, and the odds to be greater than 1.
Table 11- Logistic regression results predicting ownership of selected goods

from total wealth, gender, religious status and period
TOWELS HOUSEHOLD LINEN

Coef. Std.Err. Odds Coef. Std.Err. Odds
CONSTANT -3.514*** 0.73 -3.203*** 0.738
MEN_NONMUSLIM -0.019 0.325 0.981 0.256 0.344 1.291
WOMEN_MUSLIM 1.385*** 0.287 3.996 0.877*** 0.296 2.404
WOMEN_NONMUSLIM 1.034*** 0.378 2.814 0.552 0.407 1.736
LNWEALTH 0.391*** 0.089 1.479 0.500*** 0.096 1.648
1730 -0.556 0.481 0.573 -0.812* 0.495 0.444
1760 0.167 0.37 1.182 -0.418 0.385 0.658
1790 0.667* 0.38 1.949 0.637 0.419 1.891
1820 0.572 0.381 1.771 0.156 0.414 1.169
1850 -0.661* 0.378 0.516 -0.366 0.391 0.694
N 380 380
PSEUDO R2 0.11 0.09
CHEST CLOCK

CONSTANT -2.284*** 0.692 -7.215*** 1.146
MEN_NONMUSLIM -0.858*** 0.331 0.424 -1.249*** 0.441 0.287
WOMEN_MUSLIM 0.802*** 0.276 2.231 -0.821** 0.356 0.44
WOMEN_NONMUSLIM 0.864** 0.387 2.372 -2.223*** 0.659 0.108
LNWEALTH 0.240*** 0.086 1.271 0.608*** 0.118 1.838
1730 -0.978* 0.526 0.376 -0.756 1.226 0.47
1760 0.379 0.362 1.461 1.518** 0.7 4.564
1790 0.913** 0.376 2.491 2.277*** 0.699 9.745
1820 0.676* 0.37 1.966 2.457*** 0.668 11.664
1850 0.858** 0.373 2.359 2.086*** 0.681 8.053
N 380 380
PSEUDO R2 0.11 0.26
MODERN FURNITURE CUTLERY AND SERVIETTES

CONSTANT -5.757*** 0.964 -4.393*** 0.824
MEN_NONMUSLIM -0.655* 0.38 0.52 -0.208 0.338 0.812
WOMEN_MUSLIM 0.075 0.321 1.077 -0.524* 0.298 0.592
WOMEN_NONMUSLIM -1.473*** 0.468 0.229 -1.696*** 0.458 0.183
LNWEALTH 0.447*** 0.105 1.563 0.400*** 0.095 1.492
1730 (omitted) -0.863 0.731 0.422
1760 1.229** 0.56 3.419 0.920** 0.445 2.51
1790 2.302*** 0.557 9.989 1.758*** 0.453 5.799
1820 2.715*** 0.539 15.109 2.084*** 0.442 8.034
1850 3.158*** 0.549 23.52 1.316*** 0.442 3.727
N 349 380
PSEUDO R2 0.24 0.19
MIRROR
Coef. Std.Err. Odds
CONSTANT -9.183 1.369
MEN_NONMUSLIM 1.539*** 0.474 4.658
WOMEN_MUSLIM 1.581*** 0.429 4.862
WOMEN_NONMUSLIM 1.216** 0.549 3.375
LNWEALTH 0.655*** 0.134 1.926
1730 -0.906 1.205 0.404
1760 1.113 0.712 3.045
1790 0.922 0.743 2.514
1820 1.441** 0.692 4.227
1850 2.840*** 0.67 17.119
N 380
PSEUDO R2 0.23
As we would expect, all the exponential values for wealth are greater than 1 and are
significant at the 1 percent level, indicating that an increase in wealth increases the likelihood
that people owned selected goods, and the relationship is stronger for new goods. Among all
the household goods included, a unit change in wealth has the greatest impact on possession
of mirrors and clocks, both of which were new goods. These have the highest change in odds
for LNWEALTH (1.926 for mirrors, and 1.838 for clocks). Wealth is measured in log natural form,
so throughout the period, an increase in the material wealth of the inventory by 100 percent
will change the odds of a mirror being present by 1.926. On the contrary, wealth had little
effect on the ownership of chests, one of the most common items in Ottoman houses, as the
odds value (1.271) that is quite close to 1 indicates.
In general, estates that belonged to Muslim men were much more likely to include
new material goods than those of Muslim women and non-Muslims. Modern furniture, cutlery
and serviettes, and clocks, which symbolized the adoption of a Western lifestyle, were more
commonly found in Muslim men’s estates. The most striking example is the clock. In the
logistic regression for clocks, the coefficients of the WOMEN_MUSLIM,
WOMEN_NONMUSLIM, and MEN_NONMUSLIM variables, which show the difference in odds
between other groups and Muslim men, are negative and statistically significant. In each case,
the change in odds for non-Muslim men, Muslim women, and non-Muslim women is
substantial (respectively 0.287, 0.440, and 0.108), meaning that Muslim men had a visible
predominance in terms of ownership of clocks, the most popular Western good by far,
together with watches (Göçek 1996). For modern furniture and cutlery and serviettes, the
results are similar. An important exception seems to be mirrors. Compared to Muslim men, in
the estates belonging to the other three groups, the likelihood of the appearance of mirrors
is substantially higher. All three coefficients are positive and significant at the level of 5 percent.
A Muslim women’s inventory was 4.862 more likely to include mirrors than a Muslim men’s
inventory.
Women, and particularly Muslim women, were the leaders in more traditional goods,
such as chests, household linen, and towels, which traditionally made part of the trousseau the
bride brought to the household. In comparison to men, the likelihood of the presence of all
three categories of goods was higher in Muslim women’s estates. The coefficients of
WOMEN_MUSLIM are positive and significant at the 1 percent level. Although not as high as
in the case of mirrors, the difference in odds for these goods are still considerable (3.996, 2.404,
2.231 respectively for towels, household linen, and chests). Non-Muslim women were more
likely to own towels and chests than Muslim men, while no statistically significant difference is
detected in the likelihood of ownership of household linen. The incidence of ownership of
chests is lower with non-Muslim men than Muslim men, whereas the ownership of household
linen and towels did not vary between these two groups.
If we look at the goodness of fit of these models, the low pseudo R 2 values for
traditional goods (0.11, 0.09, 0.11 for towels, household linen, and chests) suggest that the
equations account for a relatively small proportion of the variation in the ownership.
Considering that such goods were essential items that were present in the houses of the poor
as well as the rich throughout the period, the notion that wealth, gender, religious status, and
time had a limited impact on their possession is reasonable. The R 2 values for the novel goods
are consistently higher than those for the traditional chattels. The predictor variables in the
equations explain 26 percent, 24 percent, 19 percent, and 23 percent of the variations in the
ownership of clocks, modern furniture, cutlery and serviettes, and mirrors.
These results suggest several conclusions. First, the non-Muslim community in

Üsküdar was not a pioneer in adopting novel goods associated with Westernization and a
Western lifestyle during the 1700-1850 period. Muslim males were more inclined to obtain
such goods. Whether these findings are particular to the town or can be generalized to
elsewhere in the Ottoman realm is a question that requires further attention. Second, the
higher index scores in women’s, and particularly Muslim women’s estates resulted from the
possession of greater quantities of traditional goods, such as household linen, chests and
towels, rather than a greater engagement with new consumer goods. This suggests both that
the supposition that Ottoman women became active members of a consumer society and that
non-Muslim Ottomans’ were a vanguard for Western lifestyles were phenomena characteristic
of the late nineteenth century – if at all. Third, for a variety of goods, gender and religious
status had a statistically significant effect on the ownership of selected goods. Unlike the British
and the American household, the chattels contained in the Ottoman households were a
combination of male and female goods. This was a direct implication of the Islamic law, which
recognised married women’s right to ownership, and kept her property separate from her
husband’s. Particularly in Muslim households, the differentiation in the possessions between
men and women is striking. The gender-based division of labour within the household was
also reflected in the ownership of goods. There seems to be a complementarity between what
women and what men possessed. In general, copperware, modern furniture, and other novel
goods belonged to men, whereas women owned great quantities of household linen, towels,
and chests.
4. Explaining the difference across genders
The findings of this study show striking differences in the possession of household
durables between men and women, particularly in the Muslim community. Domestic goods
made up a significantly higher proportion in Muslim women’s estates than in other categories,
and their share did not decline as wealth rises. The stock of domestic goods possessed by
Muslim women was not only more important relative to their overall assets, but was also worth
more compared to all other three categories. This is consistent with the findings on the index
scores, which demonstrated that when wealth is kept constant, Muslim women left a greater
quantity and variety of household durables, if not more copperware 74. How can the differences
between men and women be interpreted? Were Muslim women more eager consumers than
men (and non-Muslim women), and were they the drivers of early-modern consumerism in
the Ottoman realm? Or did the legal and social norms that restricted women’s role in the
economic sphere account for the greater place occupied by the domestic durables in the
Muslim women’s estates?
First, compared to the estates of men and non-Muslim women with the same level of
wealth, the stock of household durables reported in the estates of Muslim women supplied a
greater portion of the total wealth. Does this automatically mean that Muslim women had a
higher propensity to consume household chattels? The relationship between total inventory
values and the value of consumption goods might be misleading. One needs to be cautious
when drawing conclusions about the consumer behaviour of different social groups based
merely on the share of the consumer durables held within the total estate value. The
proportion of material wealth accounted for by consumption goods might reveal more about
the stock of capital, production goods, and real estate than it does about consumption; for
example, Overton et al. (2004) found for England that geographical differences in the
proportions were caused by variations in investment in capital goods.
74 As demonstrated in Chapter 4, women in the non-Muslim community owned about 43 percent less
copperware than Muslim men of comparable wealth. No significant difference is detected between
Muslim men and women in terms of the amount of copperware owned.
Similarly, when trying to explain higher proportions in Muslim women’s inventories in

Üsküdar, we need to take into account differences in the composition of estates across
genders and religious groups. It should be noted that the main sources of Muslim women’s
property were inheritance from the family or the husband, the dowry paid by the husband to
wife, and the trousseau donated by the bride’s family (Establet and Pascual 2002; Maydaer
2006). According to Islamic law, when their husbands died, women were entitled to one-eighth
of the estate if they had any children or grandchildren, and one-fourth if they did not have
offspring (Maydaer 2006). When there were no other heirs, the state seized the remaining
property after the women’s share was awarded.
The married women also had absolute entitlement to mahr, the payment the husband
accepted to make as part of the marriage contract (Imber 1997). The amount of the dowry to
be paid depended on the socio-economic status of the bride and groom’s families, and was
fixed by an agreement among the parties. The minimum amount of mahr, as established by
the Ottoman jurists, was equal to 32 grams of gold. Based on sixteenth-century Bursa court
registers, Maydaer (2006) reports that 60 percent of the women mentioned in the registers
were awarded a dowry up to 3,200 grams of gold. Higher amounts were recorded in marriages
of the wealthy. The mahr was paid in two instalments. The first instalment was paid when the
contract was concluded, and the second one was due in a later stage in life, or upon a divorce.
When the husband died without paying this second portion of the dowry, the amount was
deducted from his estate, to be paid to the wife, in addition to her own share of the estate.
When the wife died before receiving this payment, it was included in her estate as part of the
debts owed by others to the deceased. The high frequency of the appearance of the dowry in
men’s and women’s estates imply that the settlement of the second instalment did not occur
before one of the parties passed away. At least legally, women exercised full control over their
dowry, and were free to dispose of it as they wished.
Wage incomes and returns on investments and credits were marginal as sources of
wealth for Ottoman women. Although Ottoman women were not totally isolated from
economic life – with numerous studies demonstrating that several women were involved in
property transactions (Marcus 1983; Jennings 1975), operated as credit lenders, and engaged
in commercial and manufacturing activities (Gerber 1980; Jennings 1975) in towns such as
Kayseri, Bursa, Damascus, and Istanbul – unequally distributed economic power within the
household was a reality, particularly for ordinary Ottoman subjects from lower and middling
groups, and waged work was largely inaccessible for these women .
A thorough investigation into Muslim women’s involvement in the economic life of

Üsküdar is beyond the scope of this study. To what degree Muslim women were engaged in
business investments and had control over capital, or what their role was in the network of
debt relations in the town, are important questions that cannot be answered here. However,
as demonstrated above, Muslim women in Üsküdar were significantly poorer than men, and
owned considerably less real estate compared to the latter. Women owned immovable
property usually through inheritance, and they took part in the transactions as sellers rather
than buyers (Marcus 1983; Jennings 1975). It is evident that in the absence of immovable
property – and possibly of producer goods – household durables became more emphasised
among the components of wealth. Put differently, most Muslim women in our case were
personally excluded from agricultural and commercial investments and were thus squeezed
into domestic consumption. As the linear relationship between wealth and household durables
stock shows, this did not change in the upper wealth segments.
Unlike Muslim women, the non-Muslim women in our sample were no poorer than
non-Muslim men even though men owned more real estate. There was no significant
difference between men and women in this community in terms of index scores or the value
of the domestic durables stock possessed at constant levels of wealth. Similarly, the share of
household goods held within the overall estate was no higher in female inventories. As
women’s control over resources increased, and as women’s assets diversified, the dominance
of household durables within the overall patrimony disappeared. Despite the relatively more
equitable distribution of resources between non-Muslim men and women in our sample, the
shape of the relationship between wealth and value of domestic durables possessed is not the
same for men and women. For both Muslim and non-Muslim women, the linear form best
explains the relation between the value of the domestic durables possessed and total estate
value. At higher ladders of wealth, the relative importance of household durables did not
stabilize but remained constant. This linear relationship that characterised the composition of
female estates from both communities makes the value of the domestic durables stock a more
accurate indicator and a measure of wealth for women.
However, this is only part of the story, and the higher value of household durables
stock in Muslim women’s estates and higher index scores leave no doubt that within the
household, Muslim women were predominant in terms of chattel ownership. Gender roles, as
well as the gendered division of labour within the household, played an important role. Çeyiz,
the property which a wife brings to the conjugal home and which rightfully belongs to her
(Daghestani 1932), constituted the main component of women’s household property. This
trousseau especially included textile furnishings (mattress, cushions, slipcovers, sheets, and
curtains) and handicrafts. Just as in Turkey in the early twentieth century, the wife came to the
marriage with her “bedroom” that she had made herself (Sauner-Nebioglu 1995). When the
simplicity of the average Ottoman house is taken into consideration, the occupation of
bedding and household linen of an important place among the domestic chattels would be
more understandable. The disappearance of Muslim women’s dominance in the shift to
copperware and new goods supports this view.
It seems that limited control over resources and fewer investment opportunities
combined with gender roles, which assigned women to the home and made the creation of a
comfortable domestic environment a prime female responsibility, led ordinary Muslim women
to own predominantly domestic goods. An important implication is that the stock of
household durables in the estates of Muslim women cannot be taken as a reflection of
women’s economic choices in the market. In most cases, these household goods were not
acquired through the purchase of finalised goods, but were tailored in the house, and/or were
a gift from the bride’s family to the new couple. Rather than individual consumer choices,
cultural norms and customs that regulated the role and position of Muslim women in the
economic sphere and in the household explain why the value of the domestic goods stock was
greater – both in absolute terms and relative to all assets – for this group, and why traditional
goods dominated their estates.
Considering that Muslim women were the poorest group in the society, and that in
most cases, they were deprived of immovable property, including the agricultural land that
appeared frequently in the estates of men in the town, it would be misleading to see them as
independent economic actors who had the ability to express themselves through
consumption. This is particularly true for the ordinary women from the lower and middling
groups. Overall, these findings do not support the view that women were drivers of early
modern consumerism in the context of the Ottoman Empire. These women’s association with
traditional household durables, and especially household textiles, rather than novel goods,
also support this argument. Having said that, although exceptional, Ottoman women with
considerable wealth and prestige had the material means and independence to make their
own decisions in the market place within a legal system that recognized women’s property
rights.
Finally, all of these calculations still do not enable us to arrive at definite conclusions
about how Ottoman women related to the material world in the eighteenth and early
nineteenth centuries. Ultimately, understanding whether Ottoman women possessed the
resources or the independence to acquire what they wanted or satisfy their tastes, and whether
women had different attitudes towards and desires for certain material goods, requires a
different investigation that goes beyond the capacity of this study. In addition to the presence
and frequency of goods, the quality of the domestic durables, as well as the type of the textiles,
ornaments, and colours, also reveals important clues regarding differing tastes between men
and women. Changing meanings ascribed to domestic goods and to domesticity need to be
explored. Furthermore, differences among women from different economic and social statuses
should also be taken into consideration. Most of the female estates included in this study
belonged to women who controlled very limited resources. The role of the small group of
economically active and even independent women in the rising consumer culture of the
eighteenth century, and how their ownership patterns diverged from those of men from the
same class, is worthy of further investigation.
5. Conclusion
This chapter has investigated how gender and religious status was related to the
ownership of household durables in eighteenth- and nineteenth-century Üsküdar using
evidence from inheritance inventories. The evidence suggests that throughout the period,
Muslim women were significantly poorer than Muslim men in the town. Muslim females and
their non-Muslim counterparts were also in a disadvantageous position in terms of the
ownership of immovable property. This notwithstanding, Muslim females at the same wealth
level not only owned a greater quantity and variety of household effects than Muslim males,
but the overall value of the consumer durable stock in their estates was higher. Furthermore,
unlike men’s estates, the share of this value within Muslim women’s total estates did not
decline as wealth rose. These findings reveal that Muslim women were distinctive in terms of
the composition of their estates.
When we consider the likelihood of the appearance of the new, Western goods in the
estates from different gender and religious groups, our results contradict the arguments that
non-Muslim Ottomans and/or Ottoman women were leaders in the consumption of novel
goods associated with westernization. Instead, it was Muslim males in the town who were more
inclined to acquire such goods. Muslim women’s ostensible supremacy in terms of the
ownership of household effects resulted from the possession of greater quantities of more
“traditional” goods, such as household linen, chests and towels. These findings also are in
accordance with Establet and Pascual (2002), who demonstrated that Damascene women
dominated the textile interior of the home, while men owned on average more pieces of
copperware.
One can conclude that the higher shares of household goods in the estates of Muslim
women compared to Muslim men at the same wealth level reflects the limited control over
resources and the fewer investment opportunities enjoyed by women in Ottoman society, as
well as the gender roles that consigned women to the home and made the creation of a
comfortable domestic environment a prime female responsibility, rather than a sign of greater
interest in the new consumer regime.
270
CHAPTER 7
CONCLUSION
This thesis has examined three phenomena in the context of the Ottoman Empire that
are associated with the growth path of pre-industrial Europe: integration in commodity
markets, the introduction of new consumption patterns, and the decline in the relative prices
of consumer goods. The major motivation of this thesis was to understand whether these
phenomena were a peculiarity, an indicator and a potential cause of pre-modern economic
growth in the West, or whether they were also observable in a non-Western context.
These three phenomena are central to three important accounts of pre-industrial

economic growth that have far-reaching implications not only for economic history but also
for economic development in a more general context: (1) the trade-led growth theory focusing
on the division of labour and specialisation facilitated by the rise of efficient and integrated
markets (De Vries and van der Woude 1997; Epstein 2000; Persson 1988, 1999; (2) the demand-
led growth theory emphasising the role of demand shifts triggered by sociocultural
transformations (Gilboy 1932; Sombart 1967; De Vries 1994, 2001, 2003, 2008; Horrell 1996);
and (3) the innovation-led growth theory stressing the impact of productivity gains in non-
agricultural sectors that occurred due to technical and institutional innovation even prior to
industrialisation (Clark 2004; Broadberry and Gupta 2006; Allen et al. 2004).
Questioning European “distinctiveness” in the early-modern era in terms of market

development, consumption, and the price-product structure is crucial for understanding
whether Europe was already more advanced than the rest of the world well before the
Industrial Revolution and, if so, what the sources of such early divergence in economic
performances were. In an effort to understand the extent to which these European experiences
were unique, I address the following questions in separate papers: (1) Did Ottoman commodity
markets experience long-term and sustainable integration so as to produce regional patterns
of specialisation? (2) Did the Ottomans enjoy a greater variety and quantity of goods despite
stagnant real wages, as was the case in eighteenth-century Europe? (3) Did manufactured
Chapter 7-Conclusion 271
goods become cheaper relative to agricultural goods between 1700 and 1840? As the first
long-term quantitative study on Ottoman commodity markets, consumerism, and relative
prices, this research has provided an empirical basis for discussing these topics on a solid
foundation.
This research relies on evidence from Ottoman inheritance inventories. In examining

these questions, it is necessary to first establish the potential of these inventories to expand
Ottoman price history, a field severely limited by the availability and quality of primary sources
from which historical prices can be compiled. Inventories have been widely employed in the
relevant scholarship to study questions related to wealth (for across-space and across-time
comparisons, inequality, wealth accumulation, composition of personal wealth, etc.), credit
relations, and material culture. However, the information about market prices contained in
these sources, which are available for several Ottoman towns and continued for long,
continuous periods, has attracted scant attention. In chapter 2, I demonstrated that inventory
valuations were generally consistent and were closely related to the conventional prices of the
time and, thus, can be reliably employed to study historical prices. In so doing, I have
highlighted the new opportunities in the employment of these invaluable primary sources.
Establishing credible prices for goods is essential if we are to understand trade. Prior
to this research, foreign trade in the Ottoman Empire before the mid-nineteenth century had
been studied with reference to trade volumes alone, while in the absence of such data,
domestic trade was almost entirely neglected. Although market integration and the scale of
trade are clearly related, they are not always perfectly correlated, and for a more
comprehensive picture of trade conditions, we must consider trade costs and trade volumes
together. As Bateman (2010) emphasises, an increase in trade volumes between two markets
does not necessarily entail greater integration. Trade volumes can change as a result of supply
and demand shifts that can occur for any given level of market integration. I have provided
empirical evidence to assess the development of wheat markets in the Ottoman Empire from
the mid-sixteenth to the mid-nineteenth centuries. In doing so, my research has
complemented the existing corpus of knowledge on the wheat trade both between the
empire’s provinces and with the neighbouring region of the Adriatic. Perhaps more
importantly, it has shed light on the conditions and trajectory of domestic trade – which far
outstripped international trade during the period under study – but which has remained an
almost untouched field for Ottomanists due to the extreme paucity of data.
How has this study on integration in domestic and international wheat markets in the
Eastern Mediterranean contributed to our understanding of Ottoman trade? Perhaps the most
striking findings concern the late eighteenth century and early nineteenth century, which are
widely accepted as the era during which the Ottoman Empire was incorporated into the world
economy. While this period witnessed a sharp rise in the value and volume of agricultural
exports – including raw cotton and tobacco, as well as grains – trade costs rose instead of
falling. The 1768-1774 Russo-Ottoman war seems to have particularly interrupted the sea
trade in the Ottoman Eastern Mediterranean. Even in 1780-1840, which according to Quataert
(1994) was an era of recovery and then growth in terms of the volume of international trade,
shipping costs remained at substantially higher levels than they were during the 1750s, except
in the first two decades of the nineteenth century, when there was a visible declining trend.
This contrast between trade volumes and the costs of foreign trade during this period is a
theme that requires additional research.
In the domestic arena, episodes of integration and disintegration generally coincided

with ups and downs in the overall economy. From the 1720s to the 1760s, both the
Mediterranean region and the long-distance Ottoman wheat markets became more
integrated. Yet, under the pressure of war; drought, and disease; the third quarter of the
eighteenth century was marked by fragmentation, from which Ottoman markets only modestly
recovered in the subsequent 25 years. Of greater interest, however, are the developments in
regional and interregional markets in the early eighteenth centuries. The upward shift in trade
costs during this period suggests a serious degree of disintegration, particularly between the
Balkans and the other regions of the empire.
Given that the Balkans witnessed a subsistence crisis during these years, as indicated
by skyrocketing wheat prices, the question of why the grain surplus from other provinces did
not flow to this region remains another important question requiring an answer The findings
of this thesis suggest, in line with Epstein (2000), that coordination failures resulting from the
involvement of numerous local actors in the regulation of interregional and international
wheat trade, as well as the conflicting interests of local power holders, led to extremely high
transaction costs. This impeded wheat from flowing from the Mediterranean coasts, where a
marketable surplus was available, to the Balkans, which was struggling with shortages that
occurred due to climatic factors and to the increased requirements imposed by the provisionist
policies of the Ottoman state in the early eighteenth century.
Furthermore, the grain-trading network centred on Istanbul was no more integrated

in 1774-1815, which Ağır (2013) argued to be an era of liberalisation in Ottoman grain policy,
than it was in 1748-1774, which witnessed the implementation of a comparative quota system
and the requirement of a licence to buy, transport, and sell grain. In this respect, the results
provide support to Yıldırım’s (2003) view that the established Ottoman provisionist policies
remained in practice without any radical revision and that the state continued to exercise close
control over grain markets until the adoption of a liberal economic view after 1838 under
European influence.
The third section of the thesis provided an economic perspective on Ottoman

consumerism, which has been examined within the rubric of social and cultural history in the
existing literature. I explore the existence of a “consumer revolution” and trace the key features
of the eighteenth-century European consumerism in a non-Western context, thereby
contributing to the debates on the significance of market consumption for long-term
economic development. In the absence of Asian counterparts to European probate inventories,
the divergence in terms of consumption of durables across different parts of Eurasia prior to
the nineteenth century has been primarily discussed on the basis of qualitative and anecdotal
evidence. In this regard, Ottoman inheritance inventories provide us with a unique opportunity
to bring quantitative insights into pre-industrial consumerism in a non-Western context.
Employing these sources, I have shown that from the second half of the eighteenth
century onwards, Ottomans who were not richer and who were not better-positioned in the
social hierarchy than their counterparts in 1700 owned a greater quantity and variety of
domestic goods. The rising wealth or changing characteristics of the estate owners was not
the major determinant of the growth of household goods in the course of the long eighteenth
century. Instead, the structural changes that occurred on the production, distribution, and
demand sides regarding price reductions, product innovations, enhanced distribution and
retail networks, increased incomes that were not reflected in wealth, increased attention to
and information about material goods and the like, led and allowed the inhabitants of Üsküdar
to possess more consumer goods. This picture is similar to those depicted by most studies on
European probate inventories, which depict an increase in the ownership of household
durables in the “long eighteenth century.” As such, it supports Pomeranz’s argument that
improvements in the domestic environment were not a privilege reserved for members of
European societies during the early-modern era.
Throughout the same period, the valuations of selected household goods, as well as
the prices of several manufactured and traded goods, declined relative to agricultural
products’ prices. This decline was particularly visible in the 1730-1790 period, when the
ownership of goods rose most rapidly. Although solely based on these results, it is not possible
to conclude that the multiplication of household effects was a result of the cheapening of
manufactured and traded goods rather than a reallocation of household resources as Clark
(2004) has suggested; instead, it appears that the fall in relative prices was an important factor
determining the ownership of consumer goods in the Ottoman realm. Putting it differently,
from the second half of the eighteenth century onwards, individuals could own more goods
without increasing their spending on domestic durables (via rising incomes, or by a
reallocation of household resources). This indicates that Ottoman society also experienced a
modest rise in living standards as a consequence of a greater consumption of cheaper
manufactured and traded goods. These findings accord with the literature on European and
American consumption that highlights the decline in the prices of luxuries and non-food items
in the early-modern era as a crucial factor that permitted the spread of consumer goods (Clark
2004; De Vries 1994; Malanima and Pinchera 2012; Shammas 1994).
Moreover, Muslim men in the town, rather than the non-Muslim community or Muslim
women, appeared to be inclined to acquire novel consumer goods associated with
westernization, and western lifestyle, such as clocks, mirrors, and modern furniture. The
ostensible supremacy of women – and particularly Muslim women – in terms of the ownership
of household effects resulted from the possession of greater quantities of more “traditional”
goods that were always part of the Ottoman inner house, such as household linen, chests, and
towels. Rather than a sign of greater interest in the new consumer regime, the higher shares
of household goods in the estates of Muslim women compared to Muslim men at the same
wealth level reflects both the limited control over resources and the fewer investment
opportunities available to women in Ottoman society and the gender roles that relegated
women to the home and made the creation of a comfortable domestic environment a prime
female responsibility. The lack of evidence in favour of a distinctive female consumerism that
can be associated with novel consumer patterns and observations on the composition of
female estates lead us towards Weatherill (1986) and Shammas (1980), who rejected the idea
that early-modern women were eager consumers and pioneers of the rising consumer culture.
Of course, as I have emphasised throughout, these findings should be interpreted with

caution. The main source of inventories I have used, Üsküdar, was not an ordinary Ottoman
town, and compared to other Anatolian towns, possessed several advantages in terms of the
adoption of material goods. Its position at the crossroads of trade routes and its proximity to
the capital would have facilitated access to several domestic and foreign products, as well as
the acquisition of information about them. Another important characteristic of the town was
the high level of market participation. Supplying the capital with fresh fruit and vegetables was
an important economic activity in the town. Therefore, until similar research is conducted on
other – and more remote – regions of the empire, we must be careful about assuming the
improvements in terms of possession of household effects seen there occurred elsewhere in
the Ottoman realm. It would not be unreasonable, however, to assume that improvements in
the domestic environment observed in Üsküdar were shared at least by the regions that
became increasingly more engaged in commercial agriculture throughout the eighteenth
century, such as the coastal regions of Western Anatolia and parts of the Balkans. Further
research is required to acquire a greater understanding of the regional diversity in the spread
of consumer culture during the eighteenth and nineteenth centuries.
Having said that, it should also be recalled that as Overton et al.’s (2004) study on
Cornwall and Kent has shown, regional differentiation in terms of the adoption of goods was
a significant feature of European consumerism as well. The pace of the adaptation of consumer
goods was strongly associated with the different paths of development towards capitalism in
different regions within Europe. Hence, although the extent to which the results of this study
can be applied to other parts of the Ottoman Empire remains an important question; none of
the possible answers would weaken the significance of these findings.
Overall, this research has produced important implications for the “Ottoman decline
debate.” The premise of an absolute decline that began in the late-sixteenth century that
culminated in the dissolution of the empire and which has been associated with decadence in
all spheres of life – political, military, economic, social and cultural – has been severely
challenged in recent decades. This study has contributed to the recent body of literature that
inquire as to the actual nature of the changes occurring within the Ottoman world and which
offer a more nuanced view of the Ottoman experience during the seventeenth and eighteenth
centuries.
In the field of Ottoman economic history, the argument the economy went into a
permanent downward slide that following the downturn during the late-sixteenth century from
which it never recovered is no longer convincing. A more sophisticated account of Ottoman
agriculture, manufacturing, and trade that attempts to capture the trajectory of the economy
with its ups and downs, changes and stagnation has replaced the old paradigm. In line with
this new stance, I have provided further evidence to reject the simplistic notion of a linear
decay in all areas of the economy. By demonstrating that the Ottoman inner house significantly
improved and that several traded and manufactured goods became more affordable, this
study supports Kafadar’s (1999: 68) argument that “in the eighteenth century, Ottomans lived
better than their ancestors of the Suleimanic era, in terms of material culture and means,
[although] their lot was not improving relative to the lot of those living in other parts of the
world.” Furthermore, this research has shown that Ottoman wheat markets, domestic and
international, did not deteriorate in any sustained way between the mid-seventeenth and mid-
nineteenth centuries as one would expect on the basis of the institutional dissolution depicted
by the declinist literature. In fact, the episode of integration in the first half of the eighteenth
century demonstrated that under favourable conditions, the dynamics of market development
and growth could be and were in play in the Ottoman realm.
If we return to the three theories of divergence listed above, the findings of this
research point to long-term market development (and its absence), rather than a change in
attitudes towards consumption, and productivity gains in the non-agricultural sectors as a
major source of divergence prior to the Industrial Revolution between parts of Europe and the
Ottoman Empire.
Using quantitative evidence from these inventories, I have shown that Europe and the
Ottoman Empire shared several characteristics of early-modern consumerism. The growth of
consumer goods in the long eighteenth century, which De Vries identifies as a feature of
Northwest Europe, also occurred in the Ottoman realm. The interiors of Ottoman houses grew
richer and more varied throughout this period. In both regions, a decline in the real prices of
consumer goods was a major factor, if not the only one, that triggered this change. The timing
of the improvements in consumption in the Ottoman realm overlapped with the era in which
the consumer boom became revolutionary in the West, namely the second half of the
eighteenth century. Moreover, the analysis on prices and inventory valuations refutes Allen et
al.’s (2004) argument that the decline in prices of non-food items was a distinctive pattern in
Northwestern Europe in the pre-industrial era; instead, this was mirrored in the Ottoman
Empire. Indicating that throughout the eighteenth century, the material environment in which
the Ottomans lived improved while manufactured and traded goods became more affordable,
this research has made a significant contribution to the revisionist literature, arguing that
European experiences in the pre-modern era were not as unique as once thought but were, in
some respects, shared by inhabitants of India, China, Southeast Asia, and the Ottoman Empire.
If the expansion in the consumption of durable goods was a common development

that cut across Western and non-Western regions, then the argument that it fuelled the
agricultural revolution and proto-industrialisation, and hence, set the stage for modern
economic growth and the Industrial Revolution in Northwest Europe, should be revised. Since
the improvement in consumption prior to industrialisation is not associated, in every historical
case, with these long-term developments, the rising consumerism in the eighteenth century
can at best be a “necessary but not sufficient” condition, rather than a direct cause. To shed
further light on the link between consumerism and the emergence of the Industrial Revolution,
we need to ask under what particular conditions and in which particular economic settings,
consumption can trigger production-side changes.
Assessing the scale of progress in consumption in a comparative perspective is not a

major focus of this study. However, observational evidence from inheritance inventories
suggest that despite the multiplication of objects, as well as the refinement of the materials
used to produce them, the Ottoman inner house remained quite modest compared to the
average British household throughout the period. On the other hand, prior research by Allen
et al. (2004) has shown that during the same period, non-food industrial goods were much
cheaper in certain regions of Europe than in the Ottoman Empire. Therefore, it would be more
accurate to conclude that both regions witnessed the similar trends, albeit at different rates.
When it comes to market development, European distinctiveness is much more

obvious. Around the mid-nineteenth century, domestic wheat markets in the Ottoman Empire
were no better integrated than they had been in the second half of the seventeenth century.
Neither the extended sample that incorporates coastal and landlocked regions and short-
distance and long-distance markets, nor the one restricted to the littorals of the Ottoman
Mediterranean, produced evidence to support a continuous and sustainable declining trend
in trading costs prior to the mid-nineteenth century. In contrast, studies on Northern and
Northwestern Europe unambiguously depict a growing integration of wheat markets and an
increasing synchronisation of prices in the centuries prior to the Industrial Revolution (Jacks
2004; van Tielhof, 2002; van Bochove 2008; Gonzales, Garcia-Hiernaux and Guerrero 2012). If
Europe as a whole did not become engrossed in a complete and overarching system of
markets, some regions within the continent were becoming increasingly more integrated
within themselves and among one another during the early-modern era. There is little sign
that such a process of integration occurred even in the core areas of the Ottoman Empire.
Another important point of distinction concerns integration during industrialisation.

Some studies (Chilosi et al. 2011; Shiue and Keller 2007; Studer 2008; Uebele 2009) suggest
that European markets became quickly and significantly more efficient and better integrated
between 1780 and 1820, even before the telegraph, steamship and railroads could reach their
full cost-saving potential. Uebele (2009) insists that integration in world and national markets
was stronger in the first half of the nineteenth century than in the second. Similarly, our analysis
has shown that Adriatic wheat markets began integrating in the first half of the nineteenth
century. During the same period, long-distance Ottoman markets, as well as the Ottoman
Mediterranean coasts, reflected a decline in trade costs, although this decline was quite
modest compared to Europe, and price differentials remained at levels substantially higher
around the mid-nineteenth century than a century previously. More interestingly, this period
was one of disintegration in Ottoman-Adriatic wheat markets, as discussed above. In short,
the results of this study demonstrate that the trajectory followed by Ottoman wheat markets
between 1660 and 1840 was very different than that followed by European markets
throughout this period.
The elimination/lowering of policy-promoted barriers to trade, and improvements in

the institutional environment allowed by jurisdictional centralisation have been identified as
the main sources of long-term market integration prior to the nineteenth-century
developments in communication and transportation technology. In this respect, the lack of
long-term integration in Ottoman markets directs our attention to policy-related factors and
the institutional environment to account for the absence of sustainable growth in the long run.
In contrast to a number of recent studies suggesting that there were signs of

liberalisation in the eighteenth-century Ottoman economic mind, the findings of this study
support the view that pragmatism, rather than a comprehensive economic orientation,
continued to guide policy choices until around the mid-nineteenth century. Although the
Ottomans were not unaware of mercantilist thought and practice, trade continued to draw
their interest primarily from the perspective of fiscal outcome and provisioning for the army
and urban areas. Therefore, during times of war, fiscal bottlenecks, and declining agricultural
and industrial output, as occurred in the late eighteenth century, the centre did not refrain
from intervening in commodity markets, expanding internal tariff zones, or imposing export
bans, thereby further exacerbating the overall economic conditions. This feedback mechanism
between economic growth and state intervention in the context of the pre-industrial Ottoman
Empire gives us further motivation to inquire into the real nature of the relationship between
growth and market development and to consider supply levels as a variable in explaining the
changes in the extent of integration in commodity markets.
On the other hand, in the case of the Ottoman Empire, the jurisdictional centralisation
that Epstein argued enabled the rise of efficient and integrated markets in Western Europe
was a phenomenon of the nineteenth century. It was the imperial reform edicts of 1839 and
1856 that eventually ended the political fragmentation and rivalry of the last two centuries and
ushered in a centralised bureaucracy into power. Around the mid-nineteenth century, efforts
at modernisation and centralisation were also accompanied by measures to promote domestic
and international trade, such as attempts to standardise local measurement units, and the
establishment of gendarmerie units to provide safety on trade routes.
While considering seventeenth and eighteenth-century Ottoman decentralisation as

an “alternative” and politically effective strategy towards the modern state, the revisionist
literature has ignored the consequences of this alternative path in terms of its fiscal and
economic outcomes. In fact, as Pamuk and Karaman’s (2010, 2013) study has shown, in terms
of the state’s extractive capacity, the decentralised Ottoman tax collection system was much
less efficient than the centralised tax regimes of the European states. In a similar vein, I have
demonstrated here that under the decentralised political structure of the seventeenth and
eighteenth centuries, transaction costs in the Ottoman region did not show a long-term and
steady decline – something that effectively signifies improvements in the institutional
environment.
The developments that we observe within Ottoman markets are in line with Studer’s
(2008) analysis of pre-industrial India and Europe, which revealed that Europe experienced a
gradual expansion of markets in the early-modern era and that this process accelerated with
the onset of the Industrial Revolution, while Indian wheat markets remained largely isolated
until the second half of the nineteenth century. As such, they offer additional evidence in
support of the argument that unequal market development might have been a significant
source of divergence in economic performances across different parts of the world prior to
the Industrial Revolution. Having said that, it should be also noted that unequal market
development during – as well as before – industrialisation is an important factor that might
aid us in explaining the divergence of growth paths in the long run.
281
PRIMARY SOURCES
Inheritance inventories from four Ottoman towns – Istanbul, Manisa, Ayntab, and Trabzon –
constitute the main data source of this study. These sources are included in the şer’iye registers
(judicial court registers), which are available in digital format at the Turkish Religious
Foundation Center for Islamic Studies (ISAM) Library. The details of the court registers
consulted are given below.
CHAPTER 2
The data underlying the tables and figures presented in chapter 2 are collected from
inheritance inventories included in the following court registers:
Figure 2.3- Istanbul wheat prices from inventories
Istanbul Üsküdar Court Registers No. 231, 327, 357, 365, 410, 420, 442, 457, 485, 515, 520, 529,
564, 579, 582, 585.
Figure 2.4- Trabzon copperware prices from inventories
Trabzon Court Registers No. 50, 51, 52, 54, 55, 56, 57, 58, 59, 60, 61, 62, 64, 64, 65, 66, 67, 69,
70, 71, 72, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 96, 97,
98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 117, 118,
119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133.
Figure 2.5- Beledi mattress and beledi duvet valuations from Manisa inventories
Manisa Court Registers No. 167, 168, 169, 171, 172, 173, 174, 177, 178, 179, 180, 181, 182, 183,
184, 185, 186, 188, 190, 191, 192, 193, 195, 197, 198, 199, 200, 201, 202, 204, 205, 206, 207,
208, 209, 211, 213, 214, 215, 218, 219, 220, 221, 222, 224, 225, 226, 227, 229, 233, 234, 236,
238, 240, 242, 243, 244, 245, 247, 249, 251, 252, 253, 254.
Table 2.4- Valuations of silk and cotton sheets from Üsküdar inventories
331, 332, 333, 522, 524, 526, 528, 529, 532, 535, 538, 539.
Table 2.5- Valuations of selected household goods, and total estate values from Üsküdar
inventories
331, 332, 333, 522, 524, 526, 528, 529, 532, 535, 538, 539.
Primary Sources 282
Table 2.6- Valuations of copperware and total estate values from Manisa, Trabzon, and
Istanbul inventories
184, 185, 186, 188, 190, 191, 192, 193, 195, 197, 198, 199, 200, 201, 202, 204, 205, 206, 207,
208, 209, 211, 213, 214, 215, 218, 219, 220, 221, 222.
Trabzon Court Registers No. 50, 51, 52, 54, 55, 56, 57, 58, 59, 60, 61, 62, 64, 64, 65, 66, 67, 69,
70, 71, 72, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 96, 97,
98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 117, 118,
119, 120.
342, 343, 345, 347, 349, 351, 352, 354, 355, 356, 357, 358, 361, 362, 363, 365, 366, 368, 369,
371, 374, 375, 376, 377, 378, 380, 381, 384, 385, 386, 388, 390, 392, 394, 396, 397, 400, 401,
402, , 403, 404, 405, 407, 409, 410, 413, 414, 415, 416, 420, 421, 422, 423, 424, 425, 428, 431,
432, 433, 435, 437, 438, 440, 441, 442, 444, 451, 454, 465, 467, 474, 475, 476, 477, 478, 482,
485, 487, 489, 490, 491, 492, 494, 497, 498, 499, 501, 502.
Table 2.7- Wheat prices and total estate values from Manisa and Ayntab inventories
184, 185, 186, 188, 190, 191, 192, 193,.
Ayntab Court Registers No. 35, 36, 37, 38, 39, 43, 44, 45, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56,
57, 58, 59, 60, 61, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72/A, 72/B, 73, 74, 76, 77, 78, 79, 80, 82, 83,
84, 85, 86, 87, 89, 90, 91, 92, 93, 94.
CHAPTER 3
The Manisa and Ayntab wheat price series employed for market integration analyses in chapter
3 are constructed based on inventories included in the following court registers:
Ayntab Court Registers No. 26, 27, 28, 29, 30, 35, 36, 37, 38, 39, 43, 44, 45, 47, 48, 49, 50, 51,
52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72/A, 72/B, 73, 74, 76, 77,
78, 79, 80, 82, 83, 84, 85, 86, 87, 89, 90, 91, 92, 93, 94, 97, 98, 99, 101, 102, 103, 104, 104, 105,
106, 107, 108, 109, 110, 111, 112, 113, 114, 116, 117, 118, 119, 121, 122, 123, 124/B, 125, 126,
127, 128, 136, 142, 143.
147, 148, 149, 150, 151, 152, 154, 157, 158, 159, 160, 161, 162, 163, 165, 166, 167, 168, 169,
171, 172, 173, 174, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 188, 190, 191, 192, 193,
195, 197, 198, 199, 200, 201, 202, 204, 205, 206, 207, 208, 209, 211, 213, 214, 215, 218, 219,
220, 221, 222, 224, 225, 226, 227, 228, 229, 230, 232, 233, 235, 237, 239, 240, 241, 242, 243,
244, 245, 246, 248, 250, 258, 260, 261, 262, 263, 264, 265, 266, 267, 268, 269, 270, 273, 274,
275.
Primary Sources 283
CHAPTERS 4, 5, 6
The analyses in chapters 4, 5, and 6 rely on a sample of 380 inventories from Üsküdar. These
inventories are taken from the following court registers:
1695-1705
331, 332, 333.
1725-1735
Istanbul Üsküdar Court Registers No. 378, 380, 381, 384, 385, 386, 388, 390, 392, 394.
1755-1765
Istanbul Üsküdar Court Registers No. 440, 441, 442, 451, 454, 465, 467.
1785-1795
Istanbul Üsküdar Court Registers No. 522, 524, 526, 528, 529, 532, 535, 538, 539.
1815-1825
Istanbul Üsküdar Court Registers No. 588, 589, 590, 591, 592, 594.
1845-1855
Istanbul Üsküdar Court Registers No. 606, 608, 613, 616, 618, 621, 623, 626, 629, 630.
CHAPTER 5
Data underlying Figure 5.4 comes from inheritance inventories in the following court registers:
Figure 5.4- Valuations of beledi mattresses from Manisa
227, 229, 233, 247, 249, 251, 252, 253, 254.
284
BIBLIOGRAPHY
Abel, W. 1980. Agricultural fluctuations in Europe from the thirteenth to the twentieth centuries.
London and New York: Methuen and St. Martin’s Press.
Achilles, W. 1957. Getreidepreise und getreidehandels-beziehungen europäischer räume im 16.

und 17. Jahrhundert. Zeitschrift für Agrargeschichte und Agrarsoziologie 7(1) : 32–55.
Agelán, E. L., and M. J. Méndez. 2001. El mercado de trigo en Castilla y León, 1691-1788: Arbitraje
espacial e intervención. Historia agraria: Revista de Agricultura e Historia Rural 25: 13-68.
Ağır, S. 2011. The Evolution of grain policy beyond Europe: Ottoman grain administration in the
late eighteenth century. Yale University Economic Growth Center discussion paper. Available on-
line at: http://aida.wss.yale.edu/growth_pdf/cdp999.pdf.
_______ 2013. The Evolution of grain policy: the Ottoman experience. Journal of Interdisciplinary
History 43(4): 571-598.
Akkemik, M. K., et al. 2007. May–June precipitation reconstruction of southwestern Anatolia,

Turkey during the last 900 years from tree rings. Quaternary Research 68(2): 196-202.
Aksan, V. H. 2013. Feeding the Ottoman troops on the Danube, 1768–1774. War & Society, 13:
1-14.
Aktepe, M. M. 1958. İstanbul'un nüfûs mes' elesine dair bazı vesîkalar. Tarih Dergisi 9(13): 1-30.
Alexandrescu-Dersca Bulgaru, M. 1958. Contribution au problème de l’approvisionnement en

blé de Constantinople dans la seconde moitié du XVIIIe siècle. Studia et acta orientalia.
Bucharest 1: 13-37.
___________________________________ 1992. L'approvisionnement d'Istanbul par les Principautés

roumaines au XVIIIe siècle: Commerce ou requisition?. Revue du monde musulman et de la
Méditerranée 66(1): 73-78.
Allen, R. C. 1988. Inferring yields from probate inventories. The Journal of Economic History 48(1):
117-125.
Allen, R. C., and R. Unger. 1990. The depth and breadth of the market for polish grain, 1500-
1800. In Baltic affairs: relations between the Netherlands and North-Eastern Europe, 1500-1800,
edited by J. P. S. Lemmink and J. S. A. M. Van Koningsbrugge, 1-18. Nijmegen: Institute of
Northern and Eastern European Studies.
Allen, R. C., et al. 2004. Preliminary price comparisons, 1500-1870. Paper presented at the Utrecht
Conference on Global Price and Income History. Available on-line at:
http://www.iisg.nl/hpw/papers/lindert.pdf.
Bibliography 285
Arbel, B. 2013. Venice’s maritime empire in the early modern period. In A Companion to Venetian
history, 1400-1797, edited by E.R. Dursteler, 125-253. Leiden and Boston: Brill.
Artan, T. 1998. Terekeler işığında 18. yuzyıl ortasında Eyup’te yaşam tarzı ve standartlarına bir
bakış: Orta halliliğin aynası. In 18. yuzyıl kadı sicilleri işığında Eyup’te sosyal yaşam, edited by T.
Artan, 49-64. Istanbul: Turkiye Ekonomik ve Sosyal Tarih Vakfi Yayinlari.
________ 2000. Aspects of the Ottoman elite’s food consumption: Looking for ‘staples, luxuries,
and ‘delicacies’ in a changing century. In Consumption studies and the history of the Ottoman
Empire, 1550-1922: An introduction, edited by D. Quataert, 107-181. Albany: SUNY Press.
________ 2010. Eighteenth-century Ottoman princesses as collectors: Chinese and European

porcelains in the Topkapı Palace Museum. Ars Orientalis 39: 113-147.
Asdrachas, S. I. 1972. Marchés et prix du blé en Grèce au XVIIIe siècle. Südost

Forschungen 31:178-209.
Auslander, L. 2001. The gendering of consumer practices in nineteenth-century France. In

Consumption: The history and regional development of consumption , edited by D. Miller, 157-
189. London: Taylor and Francis.
Ayalon, Y. 2008. Famines, earthquakes, plagues: Natural disasters in Ottoman Syria in the
writings of visitors. Osmanlı Araştırmaları 32: 203-227.
Aydıner, M. 2006. Küresel isınma tartışmalarına tarihten bir katkı: Arşiv belgeleri ışığında XVIII.
yüzyılın ikinci yarısında Diyarbekir ve çevresinde meydana gelen büyük kıtlık ve alınan
tedbirler. Osmanlı Tarihi Araştırma ve Uygulama Merkezi Dergisi OTAM 19(19): 123-138.
Ayduz, S. The role of translations in the 18 th century: Transfer of modern science and
technology to the Ottoman Empire. Available on-line at:
http://muslimheritage.com/topics/default.cfm?ArticleID=626.
Aynural, S. 2001. İstanbul deǧirmenleri ve fırınları: Zahire ticareti (1740-1840). İstanbul: Türkiye
Ekonomik ve Toplumsal Tarih Vakfı.
Baer, G. 1977. Popular Revolt in Ottoman Cairo. Der Islam 54(2): 213-242.
Baer, M. D. and F. Göçek. 1997. Social boundaries of Ottoman women’s experience in

eighteenth-century Galata court records. In Women in the Ottoman Empire Middle Eastern
Women in the Early Modern Era, edited by MC. Zilﬁ, 48-65. Leiden, New York, and Koln: Brill.
Balta E. 1992. Bread in Greek lands during the Ottoman rule. Ankara Üniversitesi Tarih
Araştırmaları Dergisi 27: 199-226.
Baram, U. 1999. Clay tobacco pipes and coffee cup sherds in the archaeology of the Middle
Bibliography 286
East: Artifacts of social tensions from the Ottoman past. International Journal of Historical
Archaeology 3(3): 137-151.
Barkan, Ö. L. 1964. Edirne ve civarındaki bazı imaret tesislerinin yıllık muhasebe bilançoları. Türk
Tarih Belgeleri Dergisi I: 235-377.
_____________ 1966. Edirne askeri kassamına ait tereke defterleri 1545-1659. Belgeler 5-6: 1-479.
_____________ 1975. The price revolution of the sixteenth Century: A turning point in the
economic history of the Near East. International Journal of Middle East Studies 6: 3-28.
Barkey, K. 1994. Bandits and bureaucrats: The Ottoman route to state centralization. Ithaca and
London: Cornell University Press.
Barkey, K., and R. van Rossem Ronan. 1997. Networks of contention: Villages and regional
structure in the seventeenth century Ottoman Empire. American Journal of Sociology 102(5):
1345-1382.
Bassino, J. P. 2007. Market integration and famines in early modern Japan 1717–1857.
Available on-line at: http://federation.ens.fr/ydepot/semin/texte0708/BAS2007MAR.pdf.
Bateman, V. N. 2010. Market integration in the Mediterranean: 1500–1900. Available on-line

at: www.ehs.org.uk/dotAsset/54b8f15a-76a5-4adc-8de4-f93294b18a4b.doc
_________________ 2011. The evolution of markets in early modern Europe, 1350–1800: A study of
wheat prices. The Economic History Review 64(2) : 447–471.
Bedell, J. 2000. Archaeology and probate inventories in the study of eighteenth-century

life. Journal of Interdisciplinary History 31(2): 223-245.
Belk, R. W., G. Ger and S. Askegaard. 2003. The fire of desire: A multisited inquiry into consumer
passion. Journal of consumer research 30(3): 326-351.
Berg, M. 1996. Women's consumption and the industrial classes of eighteenth-century

England. Journal of Social History 415-434.
Berov, L. 1974. Changes in price conditions in trade between Turkey and Europe in the 16th-
19th century. Etudes Balkaniques 3: 168-178.
_________ 1976. Prices in the Balkans during the 16th-19th centuries and the European
revolution of prices (in Bulgarian). Sofia: Publishing House of the Bulgarian Academy of
Sciences.
Beşirli, M. 2004. Tokat bakır kalhanesinin yönetimi (1793-1853). Fırat Üniversitesi Sosyal Bilimler
Dergisi 14: 229-258.
Bibliography 287
Boerner, L., and B. Severgnini. 2011. Epidemic trade. LSE Economic History Working Paper, No.
212. Available on-line at:
http://www.lse.ac.uk/economicHistory/workingPapers/2014/WP212.pdf
Bostan, M. H. 2012. Üsküdar. Diyanet Vakfı İslam ansiklopedisi vol. 2., 364-368. Istanbul: ISAM.
Bozkurt, F. 2011. Tereke defterleri ve Osmanlı maddi kültüründe değişim (1785-1875), İstanbul
örneği. Unpublished PhD thesis, Sakarya University.
Bozkurt, R. 1966. Osmanlɩ Imparatorluğunda kollar, ulak ve iase Menzilleri. Ankara.
Braudel, F. 1985. Civilization and capitalism 15th–18th Century. New York: Harper Collins.
Braunstein, P. 1971. A propos de l'Adriatique entre le XVIe et le XVIIIe siècle. Annales, Histoire,
Sciences Sociales 26(6): 1270-1278.
Brautaset, C., and R. Grafe. 2005. The quiet transport revolution: Norway’s sailing fleet and
19th century transport costs. University of Oxford, Discussion Paper in Economic and Social
History, No.62. Available on-line at:
http://www.economics.ox.ac.uk/materials/papers/2236/62grafe.pdf
Breward, C. 1994. Femininity and consumption: The problem of the late nineteenth-century
fashion journal. Journal of Design History 7(2): 71-89.
Broadberry, S., and B. Gupta. 2006. The early modern great divergence: Wages, prices and
economic development in Europe and Asia, 1500–18001. The Economic History Review 59(1): 2-
31.
Burke, P. 1993. Res et verba: Conspicuous consumption in the early modern world. In
Consumption and the World of Goods, edited by J. Brewer, and R. Porter, 148-61. London:
Routledge.
Burman, B. (Ed.). 1999. The culture of sewing: Gender, consumption and home dressmaking.
Oxford: Berg.
Buyst, E., S. Dercon, and B. Van Campenhout. 2000. Road expansion and market integration in
the Austrian low countries during the second half of the 18th century. Available on-line at:
https://lirias.kuleuven.be/bitstream/123456789/121684/1/dps0010.pdf.
Campbell, B. 1983. Agricultural progress in medieval England: Some evidence from eastern
Norfolk. The Economic History Review 36(1): 26-46.
Canbakal, H. 2007. Society and politics in an Ottoman town. Leiden: Brill.
_____________ 2009. Reflection on the distribution of wealth in Ayntab. Oriens 37: 237-252.
Bibliography 288
_____________ 2008-2012. Distribution of wealth in the Ottoman Empire, 1500-1840. Database

Project, Scientific and Technological Council of Turkey.
_____________ 2012. Wealth and inequality in Ottoman Bursa, 1500-1840. Available at

http://www.econ.yale.edu/~egcenter/Wealth%20and%20Inequality%20in%20Ottoman%20B
ursa-Canbakal.pdf
Carr, L. G., and L. S. Walsh. 1980. Inventories and the analysis of wealth and consumption patterns
in St. Mary's County, Maryland, 1658–1777. Historical Methods: A Journal of Quantitative and
Interdisciplinary History 13: 81-104.
___________________________ 1988. The standard of living in the colonial Chesapeake. The William
and Mary Quarterly: A Magazine of Early American History 45(1): 135-159.
Ceylan, P. 2010. In search of a‘middle class’: An inquiry into the Ottoman town of Kayseri, 17 th
and 18th centuries. Unpublished MA thesis, Sabanci University.
Cezar, Y. 1977. Bir ayanın muhallefatı: Havza ve Köprü kazaları ayanı Kör İsmailoğlu Hüseyin.
Belleten 161: 41-78.
Chartres, J.A. 1995. Market integration and agricultural output in seventeenth-, eighteenth-, and
early nineteenth-century England. Agricultural History Review 43: 117–138.
Chevet, J.M., and P. Saint-Amour. 1991. L’integration des marches du ble en France au XIX e
siecle. Histoire and Mesure 6(1/2): 93–119.
____________________________________ 1992. L’integration des marches du ble en France aux XVIIIe

et XIXe siecles. Cahiers de l’Economie et Sociologie Rurales 22: 152–173.
Chilosi, D., et al. 2011. Europe's many integrations: Geography and grain markets, 1620–
1913. Explorations in Economic History 50(1): 46-68.
Cipolla, C. M. 1974. Storia economica dell'Europa pre-industriale. Rome: Il mulino.
Clark, G. 2004. Lifestyles of the rich and famous: Living costs of the rich versus the poor in
England, 1209-1869”. Paper presented at the “Towards a global history of prices and wages”
conference. Available on-line at: http://www. iisg. nl/hpw/papers/clark. pdf.
_________ 2008. A farewell to alms: A brief economic history of the world . Princeton: Princeton
University Press.
________ 2010. The consumer revolution: Turning point in human history, or statistical
artifact?. Available on-line at:
https://mpra.ub.uni-muenchen.de/25467/1/MPRA_paper_25467.pdf
Bibliography 289
________ 2015. Markets before economic growth: The grain market of medieval
England. Cliometrica 9(3): 265-287.
Clark, G., and Y. Van Der Werf. 1998. Work in progress? The industrious revolution. The Journal
of Economic History 58(3): 830-843.
Coelho, P. R. P., and J. F. Shepherd. 1974. Differences in regional prices: The United States, 1851–
1880. Journal of Economic History 34 (3): 551–591.
Coşgel, M. M. 2006. Agricultural productivity in the early Ottoman Empire. Economics working
paper 200435. Available on-line at: http://digitalcommons.uconn.edu.
______________ 2011. The Political Economy of Law and Economic Development in Islamic
History. In Law and Long-Term Economic Change, edited by J. L. van Zanden, and D. Ma, 158-
177. Stanford: Stanford University Press.
______________ 2015. The Fiscal Regime of an Expanding State: Political Economy of Ottoman
Taxation. In Fiscal Regimes and the Political Economy of Premodern States, edited
by A. Monson and W. Scheidel, 404-428. Cambridge: Cambridge University Press.
Coşgel, M. M., and B. Ergene. 2011. Intergenerational wealth accumulation and dispersion in
the Ottoman Empire: Observations from eighteenth-century Kastamonu. European Review of
Economic History 15: 255-276.
________________________________ 2014. The selection bias in court records: settlement and trial in
eighteenth‐century Ottoman Kastamonu. The Economic History Review 67(2): 517-534.
Cox, J., and N. Cox. 1985. Valuations in probate inventories. Local Historian 16: 467-78.
Çetin, C. Ulak, yol, durak: Anadolu yollarinda padisah postalari (menzilhaneler), 1690-1750.
Istanbul: Hikmet Yayinlari.
Çiftçi, C. 2004. Osmanli döneminde İstanbul'un iaşesinde Bursa'nın rolü. Osmanlı Tarihi
Araştırma ve Uygulama Merkezi Dergisi OTAM 16(16): 153-171.
Çizakça, M. 1978. Sixteenth-seventeenth century inflation and the Bursa silk industry: A Pattern
for Ottoman industrial decline?. Unpublished PhD thesis, University of Pennyslvania.
___________ 2004. Incorporated cash waqfs and mudaraba, Islamic non-bank financial
instruments from the past to the future. Available on-line at: https://mpra.ub.uni-
muenchen.de/25336/
Daghestani, K. 1932. Etude sociologique sur la famille musulmane contemporaine en Syrie .

Paris: Librairie Ernest Leroux.
Bibliography 290
Dejongh, G., B. Van Campenhout, and M. Ceusters, M. 2000. In de greep van de markt De
integratie van de regionale landbouwmarkten in Belgiean, 1700–1850. Revue Belgie d’Histoire
Contemporaine 30: 5–47.
De Vries, J. 1975. Peasant demand patterns and economic development: Friesland, 1550-1750.
In European peasants and their markets: essays in agrarian economic history, edited by W. N.
Parker and E. L. Jones, 205-265. Princeton: Princeton University Press.
__________ 1993. Between purchasing power and the world of goods: Understanding the
household economy in early modern Europe. In Consumption and the world of goods, edited
by J. Brewer and R. Porter, 85-132. New York: Routledge.
__________ 1994. The industrial revolution and the industrious revolution. The Journal of Economic
History 54(2): 249-270.
__________ 2008. The industrious revolution: consumer behavior and the household economy,
1650 to the present. New York: Cambridge University Press.
De Vries, J., and A. Van der Woude. 1997. The first modern economy:
Success, failure, and perseverance of the Dutch economy, 1500–1815. Cambridge: Cambridge
University Press.
Demirel, O. 1990. 1700-1730 tarihlerinde Ankara’da ailenin niceliksel yapısı. Belleten 54(211):
945-961.
Der Woude, A. V., and A. Schuurman (Eds.). 1980. Probate inventories: A new source for the
historical study of wealth, material culture and agricultural development. Papers presented at
leeuwenborch Conference (Wageningen, 5-7 May 1980). Utrecht: HES Publishers.
Earle, P. 1989. The making of the English middle class: Business, society, and family life in London,
1660-1730. Berkeley and Los Angeles: University of California Press.
Ejrnas, M., and K. G. Persson. 2000. Market integration and transport costs in France 1825–1903.
Explorations in Economic History 37: 149–173.
Eldem, E., D. Goffman, and B. Masters. 1999. The Ottoman city between East and West: Aleppo,
Izmir, and Istanbul. Cambridge: Cambridge University Press.
Eldem, V. 1970. Osmanlı İmparatorluğunun iktisadi ṣartları hakkında bir tetkik. İstanbul: Türkiye İş
Bankası yayınları.
Ellis, M. 2011. The Coffee-House: A cultural history. London: Hachette UK.
Emecen, F. M. 1989. XVI. asırda Manisa kazası. Ankara: Türk Tarih Kurumu Basımevi.
Bibliography 291
______________ 2003. Manisa. In Diyanet Vakfi Islam ansiklopedisi, vol. 27, 582-591. Istanbul:
ISAM
Epstein, S. R. 2000. Freedom and growth: The rise of states and markets in Europe, 1300-1750.
London: Routledge.
Erder, L. T., and S. Faroqhi. 1980. The development of the Anatolian urban network during the
sixteenth century. Journal of the Economic and Social History of the Orient 23(3): 265-303.
Ergene, B. A. 2002. Costs of court usage in seventeenth and eighteenth-century Ottoman

Anatolia: Court fees as recorded in estate inventories. Journal of the Economic and Social
History of the Orient 45(1): 20-39.
______________ 2010. Why did Ümmü Gülsüm go to court? Ottoman legal practice between
history and anthropology. Islamic Law and Society 17(2): 215-244.
Ergene, B., and A. Berker. 2008. Wealth and inequality in 18th century Kastamonu: Estimations
for the Muslim majority. International Journal of Middle East Studies 40: 23-46.
Esenbel, S. 1994. The anguish of civilized behavior: The use of Western cultural forms in the
everyday lives of the Meiji Japanese and the Ottoman Turks during the nineteenth century.
Nichibunken Japan Review 5: 145-185.
Establet, C., and J. P., Pascual. 1992. Damascene probate inventories of the 17th and 18th
centuries: Some preliminary approaches and results. International Journal of Middle East
Studies 24(03): 373-393.
__________________________________ 1994. Familles et fortunes a Damas: 450 foyers damascains en

1700. Damascus: Institut Français de Damas.
_________________________________ 2000. Les inventaires après décès, sources froides d’un monde
vivant. Turcica 32: 113-143.
_________________________________ 2002. Women in Damascene families around 1700. Journal of
Economic and Social History of the Orient 45: 301-319.
_________________________________ 2003. Cups, plates and kitchenware in late seventeenth-and

early eighteenth-century Damascus. In The illuminated table, the prosperous house: Food and
shelter in Ottoman material culture, edited by S. Faroqhi, and C. K. Neumann, 185-197.
Würzburg : Ergon Verlag.
Establet, C., J. P. Pascual, and A. Raymond. 1994. La Mesure de l'inégalite dans la societé
ottomane: Utilisation de l'indice de Gini pour le Caire et Damas vers 1700. Journal of Economic
and Social History of the Orient 37: 171-182.
Bibliography 292
Exertzoglou, H. 2003. The cultural uses of consumption: Negotiating class, gender, and nation
in the Ottoman urban centers during the 19th century. International Journal of Middle East
Studies 35(1): 77-101.
Faroqhi, S. 1979. Sixteenth century periodic markets in various Anatolian sancaks: İçel, Hamid,
Karahisar-i Sahib, Kütahya, Aydin, and Menteşe. Journal of the Economic and Social History of
the Orient 22(1): 32-80.
__________ 1982. Camels, wagons, and the Ottoman State in the Sixteenth and Seventeenth
Centuries. International Journal of Middle East Studies 14(4): 523-539.
__________ 1984. Towns and townsmen of Ottoman Anatolia: Trade, crafts and food production
in an urban setting, 1520-1650. Cambridge: Cambridge University Press.
__________ 1994. Crisis and change. In An economic and social history of the Ottoman Empire,
edited by H. İnalcik, and D. Quataert, 411-635. Cambridge: Cambridge University Press.
__________ 2001. Trade and revenue collection in later sixteenth-century Salonica. Oriente
moderno 20(1): 97-108.
__________ 2002a. Men of modest substance: House owners and house property in seventeenth-
century Ankara and Kayseri. Cambridge: Cambridge University Press.
_________ 2002b. Consumption and elite status in the eighteenth and nineteenth centuries:
Exploring the Ottoman case. In Stories of Ottoman men and women: Establishing Status,
Establishing Control. Istanbul: Eren.
___________ 2011. Ottoman cotton textiles: The story of a success that did not last, 1500–
1800. In The Spinning world: A Global History of Cotton Textiles 1200-1850, edited by G. Riello,
and P. Parthasarathi, 89-103. Oxford : Oxford University Press.
Federico, G. 2007. Market integration and market efficiency: The case of 19 th century Italy,
Explorations in Economic History 44: 293-316.
_____________ 2008. The first European grain invasion: A study in the integration of the European
market 1750-1870. European University Institute Working Paper. Available on-line at:
http://cadmus.eui.eu/handle/1814/9247
_____________ 2011. When did European markets integrate? European Review of Economic History
15 (1): 93–126.
_____________ 2012. How much do we know about market integration in Europe?. The Economic
History Review 65(2): 470-497.
Bibliography 293
Federico, G., and K. Persson. 2007. Market integration and convergence in the world wheat
market, 1800-2000. In The New comparative economic history: Essays in honour of Jeffrey G.
Williamson, T.J. Hatton, K. H. O'Rourke, A, M. Taylor, A.M., 87- 113. Cambridge MA: The MIT Press.
Findlay, R., and K. H. O’Rourke. 2003. Commodity market integration, 1500-2000.

In Globalization in historical perspective, edited by M. D. Bordo, A. M. Taylor, and J. G.
Williamson, 13-64. Chicago: University of Chicago Press.
Frank, A. G. 1998. ReOrient: Global economy in the Asian age. Berkeley and California: University
of California Press.
Fremdling, R., and G. Hohorst. 1979. Marktintegration der preuen wirtschaft im 19 jahrhundert.
In Industrialisierung und raum, edited by R. Tilly, and R. Fremdling. Stuttgart: HSF.
Fusaro, M. 2015. Political economies of empire in the early modern Mediterranean: The decline
of Venice and the rise of England 1450–1700. Cambridge : Cambridge University Press.
Genç, M. 1984. 18. yüzyılda Osmanlı ekonomisi ve savaş. Toplumsal Araştırmalar Dergisi 49(4):
52-61.
_______ 1987. A study of the feasibility of using eighteenth-century Ottoman financial records as
an indicator of economic activity. In The Ottoman Empire and the world economy, edited by H.
İslamoğlu-İnan, 345-73. Cambridge: Cambridge University Press.
______ 1994. Ottoman industry in the eighteenth century: General framework, characteristics and
main trends. In Manufacturing in the Ottoman Empire and Turkey, 1500–1950, edited by D.
Quataert, 59-86. New York: SUNY Press.
______ 2000. Osmanlı İmparatorluğunda devlet ve ekonomi. Istanbul: Ötüken.
Gerber, H. 1980. Social and economic position of women in an Ottoman city, Bursa, 1600-
1700. International Journal of Middle East Studies 12(3): 231-244.
____________ 1988. Economy and society in an Ottoman city: Bursa, 1600–1700. Jerusalem: Hebrew
University.
Gerhard, H. J., and A. Engel. 2000. Preise als indikatoren von marktverflechtungen des
nordwestdeutschen raumes 1800 bis 1850. In Der handel im kurfurstentum, edited by K. H.
Denzel. Stuttgart: Franz Steiner Verlag.
Gilboy, E. W. 1932. Demand as a factor in the industrial tevolution. In Facts and factors in
economic history, edited by A. H. Cole. London: Metheun.
Goffman, D. 2002. The Ottoman empire and early modern Europe. Cambridge: Cambridge
University Press.
Bibliography 294
Goldstone, J. A. 2000. The Rise of the west—or Not? A Revision to socio‐economic

history. Sociological theory 18(2): 175-194.
González, R., A. Garcia-Hiernaux, and D. E. Guerrero. 2012. The integration of grain markets in
the eighteenth century: Early rise of globalization in the west. The Journal of Economic
History 72(03): 671-707.
Good, D.F. 1981. Economic integration and regional development in Austria–Hungary. In

Disparities in Economic Development since the Industrial Revolution, edited by P. Bairoch, and
M.Levy-Leboyer, 137-150. London: MacMillan.
Goodwin, B.K., and T. J. Grennes, T.J. 1998. Tsarist Russia and the world wheat market.
Explorations in Economic History 35: 405–430.
Gounaris, B. C. 2009. Reassessing wheat crises in eighteenth-century Thessaloniki. The Historical

Review 5: 41-65.
Gourvish, T.R. 1970. A note on bread prices in London and Glasgow, 1788–1815. Journal of
Economic History 30 (4): 854–860.
Göçek, F. M. 1996. Rise of the bourgeoisie, demise of empire: Ottoman westernization and
social change. New York: Oxford University Press.
Gökbilgin, T. 1994. Edirne. In Diyanet Vakfi İslam ansiklopedisi, vol. 10, 425-431. Istanbul:
ISAM.
Gradeva, R. 2005. Towards the portrait of “the rich” in Ottoman provincial society: Sofia in the
1670s”. In Provincial elite in the Ottoman Empire , edited by A. Anastasopoulos, 149-199.
Rethymno: Crete University Press.
Granger, C. W., and C. M. Elliott. 1967. A fresh look at wheat prices and markets in the eighteenth
century. The Economic History Review 20(2): 257-265.
Grehan, J. 2007. Everyday life and consumer culture in eighteenth-century Damascus.

Washington: University of Washington Press.
Griffiths, R.T. 1982. The creation of a national Dutch economy: 1795–1909. Tijdschrift voor
geschiedenis 95: 513–537.
Güçer, L. 1949. XVIII. yüzyıl ortalarında İstanbul’un iaşesi için lüzumlu hububatın temini
meselesi. İstanbul Universitesi Iktisat Fakultesi Mecmuasi 1-4: 397-416.
__________ 1952. XVI. yüzyıl sonlarında Osmanlı İmparatorluğu dahilinde hububat ticaretinin tabi
olduğu kayıtlar. İstanbul Universitesi Iktisat Fakültesi Mecmuası 13(1-4): 79.
Güran, T. 1986. İstanbul'un İaşesinde devletinrRolü (1793-1839). İktisat Fakültesi Mecmuası

44(4): 245-275.
Bibliography 295
__________ 2006. Ekonomik ve mali yönleriyle vakıflar. Istanbul: Kitabevi Yayınları.
Halaçoğlu, Y. 2002. Osmanlılarda ulaşım ve haberleşme (menziller). Ankara: PTT Genel

Müdürlüğü.
Hamadeh, S. 2004. Ottoman expressions of early modernity and the "inevitable" question of
westernization. Journal of the Society of Architectural Historians 63(1): 32-51
Hanley, S. B. 1997. Everyday things in premodern Japan: The hidden legacy of material culture.
Berkeley: University of California Press.
Hanley, S. B., and K. Yamamura. 1977. Economic and demographic change in preindustrial Japan.
Princeton: Princeton University Press.
Harris, P. M. G. 1996. Inflation and deflation in Early America, 1634-1860: Patterns of change
in the British American economy. Social Science History 20: 469-505.
Hathaway, J. 2002. The politics of households in Ottoman Egypt: The rise of the Qazdaglis.
Cambridge University Press.
Hatton, T.J., K. H. O’Rourke, and J. G. Williamson. 1994. Mass migration, commodity market
integration and real wage convergence: The late nineteenth century Atlantic economy. In
Migration and the international labor market, 1850–1939, edited by T. J. Hatton, and J. G.
Williamson, 198-215. London: Routledge.
Hinz, W. 1955. Islamische Masse und Gewichte. Leiden: E. J. Brill.
Hodne, F., and O. Gjølberg. 1981. Market integration during the period of industrialisation in
Norway. In Disparities in economic development since the Industrial Revolution, edited by P.
Bairoch and M. Levy-Leboyerd, 216-225. London: Macmillan.
Hoffman, P. T., et al. 2002. Real inequality in Europe since 1500. The Journal of Economic History
62(2): 322-355.
Horn, J. P. 1988. “The Bare Necessities”: Standards of living in England and the Chesapeake, 1650-
1700. Historical Archaeology 22(2): 74-91.
Hornby, M. E. 1863. Constantinople during the Crimean war. London: Richard Bentley.
Horrell, S. 1996. Home demand and British industrialization. The Journal of Economic History
6(3): 561-604.
Howes, D. 1996. Cross-cultural consumption: Global markets, local realities. London: Routledge.
Bibliography 296
İnalcık, H. 1953. 15. asır Türkiye iktisadi ve içtimai tarihi kaynakları. Istanbul University İktisat
Fakültesi Mecmuası 15(1-4): 51-75.
__________ 1954. Ottoman methods of conquest. Studia Islamica 2: 103-129.
__________ 1977. Centralization and decentralization in Ottoman administration. In Studies in

eighteenth-century Islamic history, edited by T. Naff, and R. Owen, 27-52. Illinois: Southern Illinois
Press.
__________ 1983. Introduction to Ottoman metrology. In Studies in Ottoman Social and

Economic History, H. İnalcık, 311-48. London: Variorum Reprints.
__________ 1994. The Ottoman State: Economy and society, 1300-1600. In An economic and social
history of the Ottoman Empire, 1300-1914, edited by H. İnalcık and D. Quataert, 9-410.
Cambridge: Cambridge University Press.
Jacks, D. 2004. Market integration in the North and Baltic Seas, 1500-1800. Journal of European
Economic History 33(3): 285-329.
________ 2005. Intra- and international commodity market integration in the Atlantic economy,
1800-1913. Explorations in Economic History 42(3): 381-413.
________ 2006. What drove nineteenth century commodity market integration?. Explorations in
________ 2009. On the death of distance and borders: Evidence from the nineteenth
century. Economics Letters 105(3): 230-233.
Jennings, R. C. 1975. Women in early 17th century Ottoman judicial records: The Sharia court
of Anatolian Kayseri. Journal of the Economic and Social History of the Orient 18(1): 53-114.
_______________ 1976. Urban population in Anatolia in the sixteenth century: a study of Kayseri,
Karaman, Amasya, Trabzon and Erzurum. International Journal of Middle East Studies 7(1): 21-
57.
Jue, J.G. 1999. Interregional trade and market integration: California and the trans-continental
Railroad, 1860–1900. Unpublished Ph.D. thesis, University of California-Berkeley.
Kafadar, C. 1999. The question of Ottoman decline. Harvard Middle East and Islamic Review 4:
30-75.
Karababa, E. 2006. Origins of a consumer culture in an early modern context: Ottoman Bursa.
Unpublished Ph.D. thesis, Bilkent University.
_____________ 2012. Investigating early modern Ottoman consumer culture in the light of Bursa
probate inventories. The Economic history review 65(1): 194-219.
Bibliography 297
Karababa, E. and G. Ger. 2011. Early modern Ottoman coffeehouse culture and the formation of
the consumer subject. Journal of Consumer Research 37(5): 737-760.
Karaman, K. K., and Ş., Pamuk. 2010. Ottoman state finances in European perspective, 1500–
1914. The Journal of Economic History 70(3): 593-629.
_________________________________ 2013. Different paths to the modern state in Europe: The

interaction between warfare, economic structure, and political regime. American Political Science
Review 107(3): 603-626.
Karpat, K. H. 1972. The transformation of the Ottoman State, 1789-1908. International Journal of
Middle East Studies 3(3): 243-281.
Keyder, Ç. 1991. Introduction: Large-scale commercial agriculture in the Ottoman Empire?.

In Landholding and commercial agriculture in the Middle East: Globalization, revolution, and
popular culture, edited by Ç. Keyder, and F. Tabak, 1-13. Albany: SUNY Press.
Khoury, D. R. 1990. The urban notables paradigm revisited. Revue du monde musulman et de la
Méditerranée 55(1): 215-230.
_____________ 2002. State and provincial society in the Ottoman Empire: Mosul, 1540-1834.
_____________ 2006. The Ottoman centre versus provincial power-holders: an analysis of the
historiography.
Kılıç, O. 2002. Osmanlı Devleti’nde Meydana Gelen Kıtlıklar. In Türkler vol. 10, 718-730. Ankara:
Genelkurmay Basimevi.
Kopsidis, M. 2002. The creation of a Westphalian rye market 1820–1870: Leading and following
regions, a co-integration analysis. Jahrbuch fur Wirtschaftsgeschichte 42 (1): 199–224.
Kostis, K. 1990. Structures sociales et retard économique. Salonique et l'économie de la laine,

XVIе—XVIIIe ss. Etudes balkaniques 1: 100-114.
_________ 1993. Sterility, High Prices and Famine (in Greek). Athens and Alexandria: Neoteri ke
Synchroni Istoria.
_________ 1995. The time of plague (in Greek). Athens: Iraklio.
Kuczynski, J. 1960. Zwei studien uber handels- und Marktprobleme. Jahrbuch fur
Wirtschaftsgeschichte 2: 113–141.
Kunt, M. 2000. Osmanlı siyasal düzeni ve değişim. In: Türkiye tarihi vol.2, edited by S. Akşin, 130-
131. Istanbul: Cem Yayınevi.
Bibliography 298
Kuran, T. 1997. Islam and underdevelopment: an old puzzle revisited. Journal of Institutional and
Theoretical Economics 153: 41-71.
________ 2000. Islamic Influences on the Ottoman Guilds. The Great Ottoman-Turkish
Civilization 2: 43-59.
________ 2003. The Islamic commercial crisis: Institutional roots of economic underdevelopment
in the Middle East. The Journal of Economic History 63(02): 414-446.
________ 2004. Why the Middle East is economically underdeveloped: Historical mechanisms of
institutional stagnation. The Journal of Economic Perspectives 18(3): 71-90.
________ 2005. The absence of the corporation in Islamic law: Origins and persistence. The
American Journal of Comparative Law 53(4): 785-834.
________ 2012. The long divergence: How Islamic law held back the Middle East. Princeton:
Princeton University Press.
Kürkman, G. 1991. Osmanlılarda ölçü ve tartılar. Istanbul: Türk ve İslam Eserleri Müzesi.
Kürkman, G., and P. M. Işın. 2003. Anatolian weights and measures. Istanbul: Suna and İnan Kiraç
Research Institute on Mediterranean Civilizations.
Kütükoğlu, M. S. 1978. 1009 (1600) tarihli narh defterine göre İstanbul’da çeşidli eşya ve hizmet
fiyatları. Tarih Enstitüsü Dergisi 9: 1-86.
_________________ 1983. Osmanlılarda narh müessesesi ve 1640 tarihli narh defteri. Istanbul:
Enderun Kitabevi.
_________________ 1996. Gumruk. Diyanet Vakfi Islam ansiklopedisi, vol. 14, 260-263. Istanbul:
ISAM.
Lamdan, R. 2005. Communal regulations as a source for Jewish women's lives in the Ottoman
empire. The Muslim World 95(2): 249-263.
Latham, A. J. H., and L. Neal. 1983. The international market in rice and wheat, 1868-
1914. Economic History Review 36(2): 260-280.
Leach, W. R. 1984. Transformations in a culture of consumption: Women and department

stores, 1890-1925. The Journal of American History 71(2): 319-342.
Lee, J., and W. Feng. 1999. One quarter of humanity: Malthusian mythology and Chinese realities,
1700-2000. Cambridge, MA: Harvard University Press.
Bibliography 299
Lemon, J. T. 1967. Household consumption in eighteenth-century America and its relationship

to production and trade: The situation among farmers in southeastern Pennsylvania. Agricultural
History 41(1): 59-70.
Main, G. L. 1983a. The Standard of living in colonial Massachusetts. Journal of Economic

History 43(1): 101-108.
_____________ 1983b. Standards of living and the life cycle in colonial Connecticut. Journal of
_____________ 1988. The standard of living in southern New England, 1640-1773. The William and
Mary Quarterly 45(1): 124-134.
Main, G. L., and J. T. Main. 1988. Economic growth and the standard of living in southern New
England, 1640–1774. The Journal of Economic History 48(1): 27-46.
Malanima, P., and V. Pinchera. 2012. A puzzling relationship. Consumptions and incomes in early
modern Europe. Histoire & mesure 27(2): 197-222.
Mantran, R. 1962. Istanbul dans la seconde moitié du XVIIe siècle: essai d'histoire institutionnelle,
économique et sociale. Paris: A. Maisonneuve.
Marcus, A. 1983. Men, women and property: Dealers in real estate in 18th century Aleppo. Journal
of the Economic and Social History of the Orient 26(2): 137-163.
Matthews, J. H. 2001. The Ottoman inheritance inventory as an exercise in conceptual

reclamation (ca. 1600-1675). Unpublished PhD thesis, Binghamton University.
Maydaer, S. 2006. Osmanlı klâsik döneminde kadınların servet edinme yolları (Bursa
örneği). Uludağ Üniversitesi İlahiyat Fakültesi Dergisi 15(2): 365-381.
Mazak, M. 2005. 1802 tarihli Üsküdar iskelesi ve Üsküdar kayikçilarinin demografik yapisi. In
Üsküdar Sempozyumu II: 12-13 Mart 2004: Bildiriler, edited by Z. Kursun, 63-68. Istanbul:
Uskudar Belediyesi.
McCabe, I. B. 2014. A history of global consumption: 1500-1800. London: Routledge.
McCants, A. 2006. After-death inventories as a source for the study of material culture, economic
well-being, and household formation among the poor of eighteenth-century
Amsterdam. Historical Methods: A Journal of Quantitative and Interdisciplinary History 39(1): 10-
23.
McCraken, G. 1988. Culture and consumption. Indiana: Indiana University Press.
McGowan, A. 2006. An all-consuming subject? Women and consumption in late-nineteenth

and early-twentieth-century western India. Journal of Women's History 18(4): 31-54.
Bibliography 300
McGowan, B. 1994. The age of the ayans, 1699-1812. In An economic and social history of the
Ottoman Empire, edited by H. İnalcik, and D. Quataert, 637-758. Cambridge: Cambridge
University Press.
McKendrick, N. 1982. The consumer revolution of eighteenth-century England. In The birth of a

consumer society. The commercialization of eighteenth-century England, edited by N.
McKendrick, J. Brewer, and J. H. Plumb, 9-33. London: Europa Publications Limited.
McKendrick, N., J. Brewer, and J. H. Plumb. 1982. The birth of a consumer society. The
commercialization of eighteenth-century England. London: Europa Publications Limited.
Menard, R. 1991. Transport costs and long-range trade, 1300-1800: Was there a European
‘transport revolution’ in the early modern era?. In Political economy of merchant empires, edited
by J. D. Tracey, 228-275. Cambridge: Cambridge University Press.
Metzer, J. 1974. Railroad development and market integration: The case of Tsarist Russia. The
Journal of Economic History 34(3): 529-550.
Micklewright, N. 2000. Public and private for Ottoman women of the nineteenth century. In
Women, Patronage, and Self-Representation in Islamic Societies, edited by D. Fairchild-
Ruggles, 155-174. Albany: SUNY Press.
Milov, L. D. 1995. Applying Quantitative Methods: the Formation of the National Agrarian Market
in Russia, 1700-1900. Research in Economic History 15: 37-70.
Mukerji, C. 1983. From graven images: Patterns of modern materialism . New York: Columbia
University Press.
Murphey, R. 1987. Provisioning Istanbul: The state and subsistence in the early modern Middle
East. Food and Foodways 2(1): 217-263.
North, D. 1958. Ocean freight rates and economic development 1730-1913. The Journal of
_________ 1968. Sources of productivity change in ocean shipping, 1600–1850. Journal of Political
Economy 76(5): 953–70.
Orbay, K. 2001.The financial administration of an imperial waqf in an age of crisis: a case study
of Bâyezîd II's waqf in Amasya (1594-1657). Unpublished MA thesis. Bilkent University.
________ 2007a. Economic development of the imperial waqfs: A study in the institutional and
local economic history in the transformation period. Unpublished PhD thesis, University of
Vienna.
________ 2007b. 16. ve 17. yüzyıllarda Bursa ekonomisi: Sultan Çelebi Mehmed Yeşil imaret’inin
Bibliography 301
mali tarihi (1553-1650). Osmanlı Tarihi Araştırma ve Uygulama Merkezi (OTAM) Dergisi 22:
125-58.
________ 2008. Tarihsel fiyatların elde edilmesi ve endekslemeler konusunda bazı notlar. Dil
Tarih Coğrafya Fakültesi Dergisi 48: 84-96.
_________ 2012. Edirne II. Beyazıd Vakfı'nın mali tarihi. GAMER 1:113-41.
O’Rourke, K.H., and J.G. Williamson. 1999. Globalization and history: the evolution of a nineteenth
century Atlantic economy. Cambridge, MA: The MIT Press.
___________________________________ 2002. When did globalisation begin?.European Review of

___________________________________ 2004. Once more: When did globalisation begin?. European

Review of Economic History 8(1): 109-117.
___________________________________ 2009. Did Vasco da Gama matter for European markets?.The

Economic History Review 62(3): 655-684.
Overton, M. 1979. Estimating crop yields from probate inventories: An example from EastAnglia,
1585–1735. The Journal of Economic History 39(2): 363-378.
____________ 1990. Re-estimating crop yields from probate inventories: A comment. The Journal
____________ 2000. Prices from probate inventories. In When death do us part: understanding
and interpreting the probate records of early modern England , edited by T. Arkell, N. Evans,
and N. Goose, 120-141. Oxford: Leopold’s Head.
____________ 2012. Household wealth, indebtedness, and economic growth in early modern
England. Available on-line at: https://ore.exeter.ac.uk/repository/handle/10036/4073
Overton, M., et al. 2004. Production and consumption in English households 1600–1750. New
York: Routledge.
Owen, R. 1992. The 1838 Anglo-Turkish convention: An overview. New Perspectives on Turkey
7: 7-14.
Özdeğer, H. 1988. 1463-1640 yılları Bursa şehri tereke defterleri. Istanbul: Istanbul Üniversitesi
Edebiyat Fakültesi Yayınları.
Özlü, Z. 2006. XVIII. ve XIX. yüzyılda Göynük’te fiyatlar. Bilig Türk Dünyası Sosyal Bilimler Dergisi
39: 127-62.
Özmucur, S., and Ş. Pamuk. 2002. Real wages and standards of living in the Ottoman empire,
1489–1914. The Journal of Economic History 62(2): 293-321.
Bibliography 302
______________________________ 2007. Did European commodity prices converge during 1500-1800?

In New comparative economic history: Essays in honor of Jeffrey G. Williamson, edited by T.J.
Hatton, K. O’Rourke, and A. M., Taylor, 59-85. Cambridge, MA: The MIT Press.
Öztürk, M. 1987. Tokat'ta Fiyatlar, 1772-1832. In Türk kültüründe ve tarihinde Tokat
semyozyumu, edited by S. H. Bolay, 237-242. Ankara : Tokat Valiliği Şeyhülislâm İbn Kemal
Araştırma merkezi.
____________ 1988. Trabzon'da fiyatlar (1771-1791). In İkinci tarih boyunca Karadeniz kongresi
bildirileri, edited by M. Sağlam, 131-139. Samsun: T.C. Ondokuz Mayıs Üniversitesi Eğitim
Fakültesi.
____________ 1989. XIX. yy’da Harput’ta fiyatlar. Belleten 52(207-208): 797-828.
____________ 1992. Bursa’da hububat fiyatları, 1775-1840. Türk Tarih Kurumu XI. Türk tarih
kongresi bildirileri c.4. Ankara: Türk Tarih Kurumu Yayınları.
Öztürk, S. 1995. Askeri kassama ait onyedinci asır İstanbul tereke defterleri, sosyo-ekonomik
tahlil. Istanbul: Osmanli Arastirmalari Vakfi.
___________ 1997. Tereke defterlerine gore XVII. asırda İstanbul’da aile nufusu, servet yapısı ve
dağılımı. Istanbul Arastirmalari 3: 21-57.
Özveren, E. 2001. Ottoman economic thought and economic policy in transition. In Economic
thought and policy in less developed Europe: The nineteenth century, edited by M. E. Mata, and
M. Psalidopoulos, 53-144. London: Routledge.
Pamuk, Ş. 1984. The Ottoman Empire in the “great depression” of 1873–1896. The Journal of
_________ 1986. The decline and resistance of ottoman cotton textiles 1820–1913. Explorations in
_________ 1994. A monetary history of the Ottoman Empire. Cambridge: Cambridge University
Press.
_________ 2000a. İstanbul ve diğer kentlerde 500 yıllık fiyatlar ve ücretler, 1469-1998. Ankara:
Devlet İstatistik Enstütüsü.
_________ 2000b. The Ottoman Empire in the eighteenth century. Itinerario 24(3-4): 104-116.
_________ 2001. The price revolution in the Ottoman empire reconsidered. International Journal
of Middle East Studies 33(1): 69-89.
_________ 2004. Prices in the Ottoman Empire, 1469–1914. International Journal of Middle East
Studies 36(3): 451-468.
Bibliography 303
_________ 2005. Urban Real Wages around the Eastern Mediterranean in Comparative Perspective,
1100-2000. Research in Economic History 23: 209-228.
_________ 2007. The Black Death and the origins of the ‘great divergence’ across Europe, 1300–
1600. European Review of Economic History 11(3): 289-317.
_________ 2009. Changes in factor markets in the Ottoman Empire, 1500–1800. Continuity and
Change 24(1): 107-136.
Pamuk, Ş., and M. Shatzmiller. 2014. Plagues, Wages, and Economic Change in the Islamic Middle
East, 700–1500. The Journal of Economic History 74(1):, 196-229.
Panzac, D. 1985. La peste dans l'Empire Ottoman, 1700-1850. Leuven: Peeters, Leuven.
__________ 1992. International and domestic maritime trade in the Ottoman Empire during the
18th century. International Journal of Middle East Studies 24(2): 189-206.
Pardailhé-Galabrun, A. 1991. The Birth of intimacy: Privacy and domestic life in early modern
Paris. Philadelphia: University of Pennsylvania Press.
Parthasarathi, P. 1998. Rethinking wages and competitiveness in the eighteenth century: Britain
and South India. Past & Present 158: 79-109.
Paterson, D.G., and R. A., Shearer. 2001. Another tale of another two cities: Liverpool and
London market in the mid-19th century. Department of Economics, University of British
Columbia discussion paper. Available on-line at:
http://papers.economics.ubc.ca/legacypapers/dp0119.pdf
Peirce, L. 2003. Morality Tales Law and Gender in the Ottoman Court of Aintab. California:
University of California Press.
Peña, D., and N., Sánchez-Albornoz. 1984. Wheat prices in Spain, 1857-1890:
An application of the Box-Jenkins methodology. Journal of European Economic History 13(2):
353.
Pennell, S. Consumption and consumerism in early modern England. The Historical Journal 42(2):
549-564.
Perkins, E. A. 1991. The consumer frontier: Household consumption in early Kentucky. The
Journal of American History 78(2): 486-510.
Persson, K. G. 1988. Pre-industrial economic growth: Social organization and technological

progress in Europe. Oxford: Blackwell.
________________ 1999. Grain markets in Europe, 1500–1900, integration and deregulation.

Bibliography 304
________________ 2004. Mind the gap! Transport costs and price convergence in the Atlantic
economy, 1850–1900. European Review of Economic History 8 (2): 125–147.
Pezzolo, L. 2013. The Venetian economy. In A companion to Venetian history, 1400-1797, edited
by E.R. Dursteler, 125-253. Leiden and Boston: Brill.
Piterberg, G. 1990. The formation of an Ottoman Egyptian elite in the 18th century. International
Journal of Middle East Studies 22(3): 275-89.
Pomeranz, K. 2000. The great divergence: China, Europe, and the making of the modern world
economy. Princeton: Princeton University Press.
Popa, M. N. 1978. La Circulation monétaire et l'Évolution des prix en Valachie. Bucarest:

Association Internationale d'études du Sud-Est Europeén.
Quataert, D. 1994. The age of reforms, 1812-1914. In An economic and social history of the
Ottoman Empire, 1300-1914, edited by H. Inalcik and D. Quataert, 759-946. Cambridge:
Cambridge University Press.
____________ 2000. Introduction. In Consumption studies and the history of the Ottoman empire,
1550-1922: An introduction, edited by D. Quataert, 1-13. Albany: SUNY Press.
____________ 2005. The Ottoman Empire, 1700-1922. New York: Cambridge University Press.
Rafeq, A. K. 2012. Women in the Sharia court records of Ottoman Damascus. Turkish Historical
Review 3: 119-142.
____________ 1994. Registers of succession (Mukhallafat) and their importance for socio-economic
history: Two samples from Damascus and Aleppo, 1277/1861. In CIEPO Osmanlı oncesi ve
Osmanlı araştırmaları uluslararası komitesi, VII. sempozyumu bildirileri, 479-491. Ankara: Turk
Tarih Kurumu.
Raymond, A. 1974. Artisans et commercants au Caire au XVIIIe siècle. Damascus: Institut francais
du Proche-Orient.
Ringrose, D. R. 1998. Capital cities, urbanization, and modernization in early modern Europe.
Journal of Urban History 24: 155-183.
Roberts, M. L. 1998. Gender, consumption, and commodity culture. The American Historical
Review 103(3): 817-844.
Roche, D. 1987. The people of Paris: An essay in popular culture in the 18th century. Berkeley
and Los Angeles: University of California Press.
Bibliography 305
_________ 2000. A history of everyday things: The birth of consumption in France, 1600-1800.
Roehner, B. 1994. Les mecanismes d’interpendences spatiale entre marches du ble au XIXe siecle.
Histoire, Economie et Societe 2: 343–394.
Rosenberg, N., L. E. Birdzell, and G. W. Mitchell. 1986. How the west grew rich: The economic
transformation of the industrial world. New York: Basic Books.
Rothenberg, W.B. 1979. A Price index for rural Massachusetts, 1750–1855. The Journal of
Rönnbäck, K. 2009. Integration of global commodity markets in the early modern era. European
Review of Economic History 13(1): 95-120.
Sahillioğlu, H. 1967. Osmanlılarda narh müessesesi ve 1525 yılı sonunda İstanbul'da fiyatlar.
Belgelerle Türk Tarih Dergisi 1: 36-40.
______________ 1999. Studies on Ottoman economic and social history. Istanbul: Organisation of
Islamic Conference Research Center for Islamic History.
Sajdi, D. (Ed.). 2014. Ottoman tulips, Ottoman coffee: leisure and lifestyle in the eighteenth
century. London and New York: IB Tauris.
Salakides, G. 2002. The provisioning of Istanbul with wheat in turbulent times: Yenisehir in
Thessaly at the end of the 18th century. Balkan Studies 43(2): 129-149.
Salzmann, A. 1993. An ancien regime revisited: “Privatization" and political economy in the
eighteenth-century Ottoman Empire. Politics and Society 21: 393-393.
_____________ 1999. İmparatorluğu ozelleştirmek: Osmanlı XVIII. Yüzyılında paşalar ve
ayânlar. Osmanlı ansiklopedisi (İktisat), C. 3, 227-235. Ankara: Yeni Türkiye Yayınları.
_____________ 2000. The age of tulips: Confluence and conflict in early modern consumer culture,
1550–1730. In Consumption studies and the history of the Ottoman Empire, 1550-1922: An
introduction, edited by D. Quataert, 83-106. Albany: SUNY Press.
____________ 2004. Tocqueville in the Ottoman Empire: rival paths to the modern state. Leiden and
Boston: Brill.
Sauner-Nebioglu, M. H. (1995). Evolution des pratiques alimentaires en Turquie: Analyse

comparative. Berlin: Klaus Schwarz Verlag.
Seng, Y. J. 1991. The Üsküdar estates (tereke) as records of everyday life in an Ottoman town,
1521-1524. Unpublished PhD thesis, University of Chicago.
Sereni, A. 1947. Il capitalismo nelle campagne (1860–1900). Turing: Einaudi.

Bibliography 306
_________ 1966. Capitalismo e mercato nazionale. Rome: Editori Riuniti.

Shammas, C. 1980. The domestic environment in early modern England and America. Journal of
Social History 14(1): 3-24.
_____________ 1989. Explaining past changes in consumption and consumer behavior. Historical
Methods: A Journal of Quantitative and Interdisciplinary History 22(2): 61-67.
______________ 1982. Consumer behavior in colonial America. Social Science History 6(1): 67-86.
______________ 1989. Explaining past changes in consumption and consumer behavior. Historical
Methods: A Journal of Quantitative and Interdisciplinary History 22(2): 61-67.
______________ 1990. The pre-industrial consumer in England and America. Oxford: Clarendon
Press.
______________ 1994. The decline of textile prices in England and British America prior to
industrialization. The Economic History Review 47(3): 483-507.
______________ 1997. Changes in English and Anglo-American consumption from 1550 to 1800.
Frontier Issues in Economic Thought (2): 122-125.
Sharp, P. R. 2008. Pushing wheat: Why supply mattered for the American grain invasion of Britain
in the nineteenth century. University. of Copenhagen Department of Economics Discussion
Paper, No. 8. Available on-line at: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1135250
Sharp, P., and J. Weisdorf. 2013. Globalization revisited: Market integration and the wheat trade
between North America and Britain from the eighteenth century. Explorations in Economic
History 50(1): 88-98.
Shiue, C. H., and W. Keller. 2007a. Markets in China and Europe on the eve of the Industrial
Revolution. The American Economic Review 97(4): 1189-1216.
_____________________________ 2007b. Market integration and economic development: A long-run

comparison. Review of Development Economics 11(1): 107-123.
Slaughter, M.J. 2001. Does trade liberalization converge factor prices? Evidence from the
Antebellum transportation revolution. Journal of International Trade and Economic
Development 10(3): 339–362.
Smiley, W. A. 2008. The janissaries and their rivals: Interest, identity, and decentralization in
eighteenth-century Ottoman Syria. Unpublished PhD thesis, University of Utah.
Solmaz, M. 1979. Tarih boyuca Asya'nin kapisi her yönüyle Üsküdar. Istanbul: Yaylacik
Matbaasi.
Sombart, W. 1967. Luxury and capitalism. Michigan: University of Michigan Press.

Bibliography 307
Stavrinidis, N. 1975. Translations of Turkish documents concerning the history of Crete (In Greek),
Tome 1. Herakleion of Crete.
____________ 1976. Translations of Turkish documents concerning the history of Crete (In Greek),
Studer, R. 2008. India and the great divergence: Assessing the efficiency of grain markets in
eighteenth-and nineteenth-century India. Journal of Economic History 68(2): 393-437.
Svorōnos, N. G. 1956. Le commerce de Salonique au XVIIIe siècle. Paris: Presses universitaires de

France.
Sweeney, K. M. 1984. Furniture and the domestic environment in Wethersfield, Connecticut,

1639-1800. Connecticut Antiquarian 36: 10-39.
Szabad, G. 1961. Das anwachsen der ausgleichstendenz der produktenpreise im habsburgerreich

um die mitte des 19. jahrhunderts. In Studien zur geschichte der Österreichisch-Ungarischen
monarchie, edited by V. Sndor, and P. Hanak, 213–235. Budapest: Akademiai Kiado.
Tabak, F. 2010. The waning of the Mediterranean, 1550–1870: A geohistorical approach.
Baltimore: Johns Hopkins University Press.
Tanyeli, U. 2003. Norms of domestic comfort luxury in Ottoman metropolises: Sixteenth to

eighteenth centuries. In The illuminated table, the prosperous house: Food and shelter in
Ottoman material culture, edited by S. Faroqhi, and C. K. Neumann, 351-374. Würzburg: Ergon
Verlag.
Teoman, Ö., and M. Kaymak. 2008. Commercial agriculture and economic change in the Ottoman
Empire during the nineteenth century: A comparison of raw cotton production in Western
Anatolia and Egypt. The Journal of Peasant Studies 35(2): 314-334.
Thirsk, J. 1978. Economic policy and projects: The development of a consumer society in early
modern England. Oxford: Clarendon Press.
Todorov, N. 1983. The Balkan city, 1400-1900. Seattle and London: University of Washington
Press.
Toprak, Z. 1992. Modernization and commercialization in the Tanzimat period: 1838-1875. New
Perspectives on Turkey 7: 57-70.
Bibliography 308
Uebele, M. 2011. National and international market integration in the 19th century: Evidence
from comovement. Explorations in Economic History 48(2): 226-242.
Unger, R. W. 1983. Integration of Baltic and Low Countries grain markets, 1400–1800. In The
Interactions of Amsterdam and Antwerp with the Baltic Region, 1400–1800, edited by J. M. van
Winter, 1–10. Springer: Leiden.
Üstündağ, E. 2003. Furniture and household goods in Istanbul at the turn of the twentieth
century. Unpublished MA thesis, Bopshorus University.
Xoplaki, E., P. Maheras, and J. Luterbacher. 2001. Variability of climate in meridional Balkans
during the periods 1675–1715 and 1780–1830 and its impact on human life. Climatic
Change 48(4): 581-615.
Valenčič, V. 1977. Žitna trgovina na kranjskem in ljubljanske zitne cene od srede 17. stoletja do
prve svetovne vojne. Ljubljana: Slovenska Akademija Znanosti in Umetnosti.
Van Bochove, C. 2008. The economic consequences of the Dutch economic integration around
the North Sea, 1500-1800. Amsterdam: Amsterdam University Press.
Van Tielhof, M. 2002. The ‘mother of all trades': the Baltic grain trade in Amsterdam from the late
16th to the early 19th century. Leiden: Brill.
Van Zanden, J. L. 1999. Wages and the standard of living in Europe, 1500–1800. European Review
_________________ 2002. The ‘revolt of the early modernists’ and the ‘first modern economy’: An
assessment. The Economic History Review 55(4): 619-641.
_________________ 2009. The long road to the industrial revolution: The European economy in a
global perspective, 1000-1800. Leiden: Brill.
Van Zanden, J. L., and M. Van Tielhof. 2009. Roots of growth and productivity change in Dutch
shipping industry, 1500–1800. Explorations in Economic History 46: 389-403.
Varlik, N. 2011. Taun. In Diyanet Vakfi Islam ansiklopedisi, vol. 40, 175-177. Istanbul: ISAM.
Vasdeavellis, I. 1952. Historical archives of Macedonia, 1695-1912. Salonika: Archives of Salonica.
Walsh, L. S. 1983. Urban amenities and rural sufficiency: Living standards and consumer behavior
in the colonial Chesapeake, 1643–1777. Journal of Economic History 43(1): 109-117.
Walton, W. 1986. “To Triumph before feminine taste”: Bourgeois women's consumption and
hand methods of production in mid-nineteenth-century Paris. Business History Review 60(4):
541-563.
Bibliography 309
Weatherill, L. 1986. A possession of one's own: Women and consumer behavior in England,
1660–1740. The Journal of British Studies 25(2): 131-156.
_______________ 1988. Consumer behaviour and material culture in England, 1660–1760. London:
Routledge.
White, C. 1846. Three years in Constantinople: Or, domestic manners of the Turks in 1844.
London: Colburn.
Wilkins, C. L. 2010. Forging urban solidarities: Ottoman Aleppo 1640-1700. Leiden: Brill.
Wong, R. B. 1997. China transformed: Historical change and the limits of European experience.
Ithaca and London: Cornell University Press.
Yıldırım, O. 2003. Bread and empire: The workings of grain provisioning in Istanbul during the
eighteenth century. In Nourrir les cités de Méditerranée: Antiquité-temps modernes, edited by
B. Marin, and C. Virlouvet, 251-272. Paris: Maisonneuve & Larose.
Zamagni, V. 1983. Ferrovie e integrazione del mercato nazionale nell’Italia post-unitaria. In Studi
in Onore di Gino Barbieri, 1635–1649. Pisa: IPEM.
Zhou, N., and R. W., Belk. 2004. Chinese consumer readings of global and local advertising
appeals. Journal of Advertising 33(3): 63-76.
Zilfi, M. C. 2000. Goods in the mahalle: Distributional encounters in eighteenth-century

Istanbul. In Consumption studies and the history of the Ottoman Empire, 1550–1922: An
introduction, edited by D. Quataert, 289-312. Albany: SUNY Press.
Zorlu, T. 2008. Innovation and empire in Turkey. London, and New York: I.B. Tauris.
ONLINE DATABASES
Allen-Unger global commodity prices database, prepared by R. C. Allen, and R. W. Unger.
Available on-line at: http://www.gcpdb.info/.
Prices and wages in Istanbul, 1469-1914, prepared by S. Pamuk. Available on-line at:
http://www.iisg.nl/hpw/data.php
Freight rates between Amsterdam and various port cities 1500-1800, and factors costs of
shipping industry 1450-1800, prepared by M. van Tielhof, and J. L. Van Zanden. Available on-
line at: http://www.iisg.nl/hpw/data.php.
Jacks commodity prices database, prepared by D. S. Jacks. Available on-line at:

http://www.sfu.ca/~djacks/data/index.html.
310
APPENDIX
Sample inheritance inventory
Inheritance inventory of Munevvere Hatun, daughter of Abdullah, and inhabitant of
Debbagzade neighbourhood in Istanbul. The inventory dates 1787.
Source: Bozkurt (2011: 392)

Appendix 311
1. Introductory protocol : identification of the deceased by given name and father’s

name and place of residence (by neighbourhood or village and the city); the names
and degree of affinity of the legatees, and the date of portioning.
2. Inventory of movable and immovable property of the deceased : Buildings (houses,
shops, watermills), vineyards, trees and crops, livestock, personal and household
goods, stores, commercial goods, as well as outstanding loans and the name of the
borrower. The valuations are recorded below each item.
3. Personal liabilities : debts incurred by the deceased (düyun), such as an outstanding
bride price to the wife, claims on the estate, bequests, and sundry expenses (medical
expenses, funeral costs, the cost of the inheritance registry process, and taxes).
4. Portioning of the net amount of the assets among the heirs.

Ceylan Essays On Prices

Uploaded by

Copyright:

Available Formats

Ceylan Essays On Prices

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ceylan Essays On Prices

Uploaded by

Copyright:

Available Formats

Essays on

markets, prices, and consumption

CHAPTER 2- Ottoman inheritance inventory as a source for price history…………………………..…..60

CHAPTER 3- Measuring market integration in the Eastern Mediterranean……………………………..86

CHAPTER 4- Was there a consumer revolution in the Ottoman Empire?........................................164

CHAPTER 5- Relative prices and the rise of household goods………………………………………..……..220

CHAPTER 6- Gender, religious status, and ownership of household goods…………………………..246

2.1. Goods reported in units in Ottoman inheritance inventories

4.2. Wealth, Descriptive statistics

6.2. Descriptive statistics of the regression variables

1.2. Ottoman towns of Üsküdar, Manisa, Ayntab, and Trabzon

3.1. Ottoman maritime network

3.2. The Eastern Mediterranean

In an effort to understand where pre-modern Northwestern Europe’s ‘distinctive

Despite several controversies in the growing body of empirical research on these

This periodization also corresponds to De Vries’ reinterpretation of the “long-

1. The Ottoman Empire: Political and Economic Environment

Map 1- The Ottoman Empire c.1683-1800

Source: Quataert (2000: 39)

1.1. The Ottoman Empire before the seventeenth century

1.2. Political environment and decentralisation

Although decentralisation, centralisation, and the changing relationship between the

For a long time, Ottomanists interpreted decentralisation as a sign of the dissolution

In this line of research, Salzmann re-conceptualised the tax-farming system, the

1.3. Economic policy

1.4. General trends in the Ottoman economy

Figure 1- Revenues of the Ottoman Central Administration (in tons of silver)

Source: Karaman and Pamuk (2010: 604)

The early nineteenth century was marked by treasury shortfalls, frequent

Figure 2- Real Daily Wages of Construction Workers: Istanbul, 1489-1914

Notes: 1489/90 Unskilled=1

1.5. The Ottoman Empire and the divergence debate

Figure 3- Real Wages of Unskilled Construction Workers in European Cities, 1450–1913

Source: Pamuk (2005: 221)

structural changes that occurred in demographic behaviour, agriculture, manufacturing, trade,

2. Literature review: pre-industrial growth, market integration, consumption, and

2.1. Market integration and trade-led growth

The “trade-led growth” theory is perhaps the best-established and most-widely

Integration in commodity markets occurs when trading costs (inclusive of

In recent decades, a wide-ranging corpus of literature on market efficiency and market

The traditional account of nineteenth-century globalisation as established by

Contrary to accounts emphasising the technological limits to integration, scholars

Within this literature emphasizing the political conditions of pre-industrial growth,

Williamson (1999, 2002, 2004) discovered no price convergence in intercontinental markets

Empirical findings on early-modern European markets are no less ambiguous with

Despite inconsistencies, existing research gives us important clues, provided a

Northern and Northwestern Europe unambiguously depict a growing integration of

Studer’s (2008) comparative study on Indian and European commodity markets, is

Finally, previous research has highlighted the importance of analysing short-term

2.2. Early-modern consumerism and demand-led growth

An alternative explanation for pre-industrial growth concentrates on demand-side

When McKendrick first introduced the idea of an early modern revolution in

According to Pomeranz’ estimates, Chinese consumption of sugar, was higher in 1750

Beyond this advantage provided by inheritance inventories, studying consumption in

Was early-modern consumerism primarily a demand-side phenomenon that was

De Vries’ argument of an ‘industrious revolution’ has been severely criticized by Clark

2.3. Changes in price-product structures and innovation-led growth

product-price pattern refutes the arguments in favour of a Northwestern European superiority

An important implication of these findings on prices concerned the previous