Endogenous Growth in the Ancient Roman Economy

Senior Honors Thesis

Presented to

The Faculty of the School of Arts and Sciences

Brandeis University

Undergraduate Program in the Humanities – Classical Studies

Undergraduate Program in Economics

George Hall, Geoff Clarke, Joel Christensen, Advisors

In partial fulfillment of the requirements for the degree of Bachelor of Arts

by
Draken Garfinkel

May 4th, 2023

Copyright by
Draken Garfinkel

Committee members:

Name: George Hall

Name: Geoff Clarke

Name: Joel Christensen

Name: Jeremy Swist

Name: Andrew Molinsky

 Garfinkel 2

Contents
1 Introduction 3
1.1 History, Endogenous Growth, and its Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.2 The Variables in Question . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.3 Previous Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.4 Methodology, Structure, and Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

2 First Principles 20
2.1 Foundational Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.2 The Problem of Slavery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.3 Differences between Industrialized and Nonindustrialized economies . . . . . . . . . . . . . . . 25
2.4 Governance in the Ancient Roman Economy . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

3 Population Growth and Savings Rate (n and s) 30

3.1 The Data: n . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.2 The Data: s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

4 Innovative Synergy (ϕ) 37

4.1 The Data: ϕ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
4.2 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

5 Innovative Congestion (λ) 43

5.1 The Data: λ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
5.2 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

6 Conclusion 51
6.1 Review of Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
6.2 Modern Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
6.3 Avenues of Future Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
6.4 Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
 Garfinkel 3

1 Introduction

As attested in the archaeological and historical records, the Early Roman Empire (approx.
25 BCE – 200 CE) represents a zenith of economic development and activity in the Ancient
Mediterranean.1 Built off successive generations of economic growth, this period is often
compared in scope to various 17th and 18th century European nations like Britain and Por-
tugal.2 Despite these comparisons, despite Rome’s wealth of natural resources, cheap labor,
and learned aristocracy whose works are still widely read today, the Ancient Roman Econ-
omy never advanced beyond one firmly rooted in subsistence agriculture. This analysis will
attempt to give an explanation as to why using Romer and Jones’ endogenous growth model,
as well as how Ancient Rome can serve as a case study on contemporary nonindustrialized
economies and governance more broadly.

Through a dissection of the various records available, this analysis proposes that Ancient
Rome’s lack of industrialization–or broader sustained economic development–resulted from
a matrix of institutional and cultural shortcomings which failed to properly incentivize inno-
vation, protect its profit-generating ability, or efficiently use Rome’s human resources. The
growth it did experience was based on growing the capital and labor stocks, not total factor
utility; horizontal, not vertical expansion. Perspiration, not innovation.

Successful development initiatives often rely on both immense monetary investment and
large groups of highly coordinated external specialists; The Ancient Roman Economy lacked
these. As a case study, the Ancient Roman Economy suggests that endogenous growth is
mainly reserved for industrial and post-industrial economies, with the jump from nonindus-
trial to industrial being incredibly difficult without external assistance.
1
For instance, histograms of extant shipwrecks point to a greater concentration in this time period (see
Hopkins, 1980 “Taxes and Trade in the Roman Empire (200 B.C.-A.D. 400)," The Journal of Roman Studies
vol. 70, p. 106) and analysis of Greenland ice cores point to higher residues indicative of industrial processes
at this time too (see Malmendier, 2009 “Law and Finance ‘at the Origin.’ ” Journal of Economic Literature,
vol. 47, no. 4, p. 1082.
2
Scheidel, 2007 “A Model of Real Income Growth in Roman Italy.” Historia: Zeitschrift Für Alte
Geschichte, vol. 56, no. 3, p. 343.
 Garfinkel 4

1.1 History, Endogenous Growth, and its Factors

Although economic growth models date back to Thomas Malthus,3 the first major growth
model relevant to this analysis is the Cobb-Douglas Production Function which states:

Y = F (K, L) = ALα K 1−α (1)

Where Y = Total Output, A = Total Factor Productivity, K = Capital Stock, L = Labor

Stock, and α = the proportion of income devoted to labor or capital. In effect, output
growth is caused by increasing K, L or A(their effective utilization.) Oftentimes, however,4
the model is expanded to account for different types of capital:

γ 1−α−β−γ
Y = F (L, Kν , KH , KM ) = ALα Kνβ KH KM (2)

Where Y = Total Output, A = Total Factor Productivity, L = Total Labor Stock, Kν

= Stock of Land-Based Capital, KH = Stock of Human-Based Capital, KM = Stock of
Machinery-Based Capital, α = Percentage of income devoted to L, β = Percentage of income
devoted to Kν , and γ = Percentage of income devoted to KH . Although elaborated, the
basis remains the same: Total output derives from the efficient use of various factors, each
taking a different amounts of available investment funds depending on their utility.

In 1956, Robert Solow5 would expand on this model through his concept of the “steady
K
state," that at a given technological level the ratio increases to an asymptotic end, upon
L
which economic growth is preeminently determined by n, population growth, and investment

3
See Malthus, 1798 “An Essay on the principle of population."
4
E.g. Romer, 1990 “Endogenous Technological Change," University of Chicago Press, vol. 98, no. 5, p.
S80.
5
See Solow, 1956, “A Contribution to the Theory of Economic Growth," The Quarterly Journal of Eco-
nomics, vol. 1 p. 65-94.
 Garfinkel 5

is wholly devoted to preserving the capital stock according to the following equation(s):

sy = (n + δ) ∗ k (3)

s
k/y = (4)
n+δ

Where s = savings rate, y = output per capita, n = population growth rate, δ = the collective
depreciation rate of capital, and k = the stock of collective capital per capita. Another way
of describing this would be:
sy = (n + gA + δ)k (5)

Where gA = is the growth rate of A, total factor productivity. With this nuance, the growth
of total capital in a steady state is only determined by population growth n and the growth
of total factor productivity A, whereas the capital per person k is only affected by gA

Solow’s model however doesn’t account for gA , with that variable remaining “exogenous,"
defined by parameters outside the model. Drs. Paul Romer and Charles Jones’ models6 –the
focus of this analysis–however define gA as “endogenous," derived from parameters within
the model, most notably innovation. In effect, therefore, Solow’s “exogenous" growth model
emphasizes population as the driving factor, while Romer and Jones’ model7 emphasizes an
increase in A “knowledge,"8 via technical and technological innovation9 (R&D) and economic
integration.

This analysis primarily concerns itself with Jones’ definitions of dA and gA as drivers
of economic growth and will spend the majority of its time analyzing each of the factors
6
See Romer, 1990 “Endogenous Technological Change," University of Chicago Press, vol. 98, no. 5, p.
S71–S102 and Jones, 1995 “R D-Based Models of Economic Growth.” Journal of Political Economy, vol.
103, no. 4, p. 759–84.
7
Jones, 1995 p. 760 puts forward the names of “the Romer/Grossman Helpman/Aghion-Howitt models.
8
Rivera-Batiz and Romer, 1991 p. “Economic Integration and Endogenous Growth.” The Quarterly
Journal of Economics, vol. 106, no. 2, p. 531-532.
9
For this thesis “technical innovation" refers to improvements in administration, allocation of resources,
etc., while “technological innovation" specifically refers to the development and utilization of more efficient
capital.
 Garfinkel 6

comprising gA and how they manifest in the Ancient Roman Economy. To that end, Jones
defines A’s velocity and growth as:

dA = ΘLλA Aϕ (6)

n∗λ
dA/A = (7)
1−ϕ

Where Θ = some constant, LA = labor stock devoted to increasing A (effectively “R&D,")

ϕ = “innovative synergy" with −1 < ϕ ≤ 1, λ = “innovative congestion" with 0 ≤ λ < 1,
and dA/A itself equalling gA .

In Jones’ 1995 paper, ϕ generally refers to the synergy between innovations themselves,
how one innovation helps inspire the next. For this analysis, however, ϕ also represents
how innovations further increase the productivity of existing capital. Similarly, whereas λ
in Jones’ model refers to innovative congestion in caused by redundant research, in this
analysis λ is expanded, referring to all congestive forces on innovation, including cultural
factors. Also worth mentioning are two more equations:

dK = sY − δK (8)

sy
dK/K = dk/k = −δ (9)
k

Using terminologies already established, Equation 8 states that the velocity of capital stock
equals available funds for investment (sY ) minus quantity of capital lost (δK) in any given
time period. Put another way, the change in quantity of capital is contingent on how fast
capital decays and how much investment offsets this decay. Equation 9 states much of the
same. While this analysis has already separated capital (K ) into various subsets (KH , KM ,
Kν ,) the above equations effectively describe the growth of each.

While not in the equations above, two more variables also are worth mentioning due
 Garfinkel 7

to their impact on economic practices in nonindustrialized economies, particularly that of

Ancient Rome: “Risk," represented by κ, and “information asymmetry," represented by ξ.
While both κ and ξ are internalized within (and not mutually independent of) λ, they are
significant enough to be mentioned in of themselves. Similar to λ, however, they act as
negative modifiers on A and gA by creating conditions less likely to promote innovation.

Taken together, the following comprise the main variables influencing endogenous eco-
nomic growth in a nonindustrialized society such as that of Ancient Rome:

Term Definition

n Population growth rate

s Savings rate

δ Depreciation rate of capital (and its various forms)

λ Elasticity of change in ideas / innovative congestion

ϕ Elasticity of ideas / innovative synergy

While other variables, such as KH , KM , Kν , LA , κ, and ξ have sizable impacts as well,

they will not directly be discussed due to this analysis’ focus on innovation-led/endogenous
growth’s role (or lack thereof) in increasing output in the Ancient Roman Economy. Now
that these main five variables have been identified, they will be placed within the model and
their significance explained.

1.2 The Variables in Question

Population growth (n) As described in Solow’s growth model, at a constant techno-

logical level (A) or in the steady state, n becomes the engine of growth, with the rate of
economic expansion directly corresponding to n (i.e. gL and gK .) In Jones’ model, however,
n also directly relates to gA , with this direct correlation representing n’s growth of LA (labor
devoted to innovative investment.)

In other words, irrespective of technological innovation, economic growth is highly de-

 Garfinkel 8

pendant on a positive, if not relatively high, rate of n. n also importantly comes with
various agglomeration effects. Higher population growth creates higher population density
and urban areas which themselves help develop phenomena such as financial institutions,
normalized prices, standardized currencies, and so on.10 While a nonindustrialized economy
such as Ancient Rome could increase n through innovations designed to increase agricultural
productivity, (as mentioned in subsequent sections,) other avenues also included cultivation
of peripheral lands, conquest, importation of slaves, etc.

Savings rate (s) Irrespective of technological level, savings rate s is a second major engine
of economic growth and development, as it directly impacts the amount of money saved
per time period, with the implication that saved incomes are necessarily devoted towards
investment that either grows A or maintains or grows K. For this reason higher savings rates
K
generally increase the ratio for as long as it continues to increase output per worker (that
L
is until a steady state is reached.)

In economies such as Ancient Rome’s, where rapid technological advancement wasn’t

present, higher savings rates (as will be elaborated on in subsequent sections) nevertheless
have numerous other effects. These include maintaining the capital stock, increasing the level
of liquidity, and raising the level of monetization–especially if savings and their payments to
investors take the form of hard currency.

Depreciation rate of capital (δ) While δ, the depreciation rate of capital, is more
complicated than either n or s, it is nevertheless fundamental for understanding the prospect
of growth and development in the Ancient Roman Economy. Conceptually, δ represents the
% of investment funds which must go to maintaining the current stock of K rather than
increasing it, with higher δ corresponding to higher allotment until δ = 1 and all investment
10
As such, coinciding with the 8th-7th century BCE expansion of urban centers we see the first emergence
of standardized currency in Lyida (see Gills and Frank, 1992 “World System Cycles, Crises, and Hegemonial
Shifts, 1700 BC to 1700 AD.” Review (Fernand Braudel Center), vol. 15, no. 4, p. 641.)
 Garfinkel 9

must be used to maintaining the current capital stock (if it even can.) Higher δ lowers K in
the steady state, as well as negatively impacts gA .

The catch is that δ differs for different types of capital, particularly human (δKH ) and
mechanical capital (δKM .) In nonindustrialized economies especially, mechanical capital,
human capital, and general labor are to some extent interchangeable, causing economies to
disproportionately invest into whichever have the lowest rates of δ.

As will be elaborated in Section 2.2, for the Ancient Romans this meant an over-reliance
on slavery with slavery “crowding out" KM , as δKM would have been higher without modern
materials or preservation methods (just as the Romans relied on nondurable record-keeping
media like wax and papyrus which don’t survive in the archaeological record outside extraor-
dinary conditions.)11 Another catch is that, just like n or s, δ is susceptible to “decay shocks"
where the rate of depreciation for specific inputs greatly increases over a short period of time.

Innovative synergy (ϕ) Unlike the previous variables, which all function the same ir-
respective of technological level, ϕ and λ only properly function in a world where A is not
static. As shown in Jones (1995) and Equations 6-7, ϕ acts as a modifier on gA . According
to the equations, when > 0, ϕ acts as an increasingly powerful modifier on gA . When < 0, ϕ
acts as a negative modifier, but to not the same extreme. Altogether ϕ’s ability to help gA
is more important to this analysis.

Expanding on Jones’ model, for this analysis ϕ represents how novel innovations both
help create new ones and increase the productivity of existing, unrelated technologies. In
other words, ϕ represents more than simple synergy, rather the proper implementation and
utilization of technologies once they have been created.

ϕ is likewise derived from many factors: Cultural readiness to adopt new technologies,
11
Temin, 2006 p. 134. These “extraordinary conditions" include the hermetically-sealed towns of Pompeii
and Herculaneum, as well as the dry and hot climate of Egypt.
 Garfinkel 10

information costs/asymmetries, effectiveness of innovative institutions, etc. ϕ is also nec-

essarily tied to λ, “innovative congestion," as will be elaborated in Section 5. Finally, ϕ is
derived from any given society’s reception to “la Technique."

As defined by French Philosopher Jacques Ellul in his 1954 (rev. 1965) treatise The
Technological Society,12 “la Technique" is the concept that, in competitive environments,
the adoption of new technologies by one actor require all other actors to do so in order to
remain competitive, such that environments become increasingly homogeneous and advanced
over time. A modern example would be how the industrialization of Sub-Saharan Africa was
and is often contingent on locals adopting Western business practices at the expense of their
own cultures.13

In other words, “la Technique" is directly tied to the rate of innovation, with Ellul stating
Early Modern Europe was unique in that they developed a culture wholly devoted to “la
Technique." Conversely, Ellul mentions that Classical Antiquity had a relatively poor recep-
tion to “la Technique" due to underlying cultural blocks to be discussed in Section 5.1, with
such receptions impacting ϕ and λ.14

Innovative congestion (λ) Innovative congestion λ represents the difficulties associated

with creating new innovations. Ranging [0,1) it, unlike ϕ, is an exclusively negative modifier
on gA . As λ approaches 1, it has an increasingly negligible impact on gA ; λ = 0 conversely
represents a world where innovation isn’t possible. Similar to the aforementioned factors,
high λ (low innovative congestion) has numerous spillover effects derived from a higher
quantity of LA , most notably the creation of ancillary industries such as museums and
academies. These synergies become less potent the more decentralized LA becomes due to
higher transaction and information costs.
12
See Ellul, 1965, The Technological Society, trans. Wilkinson, Jonathan Cape publishers.
13
Jarausch, 2015 “Disappointing Decolonization.” Out of Ashes: A New History of Europe in the Twentieth
Century, REV-Revised, Princeton University Press, p. 504.
14
Ellul, 1965, The Technological Society p. 28-29, 70-71.
 Garfinkel 11

Like ϕ, λ is also affected by a matrix of socioeconomic factors, one of the most prominent
being confidence levels: To occur, one must believe innovation to not only be possible, but
profitable.15 Others include whether the proposed innovation addresses current needs, its
perceived cost of implementation, cultural willingness to invest in innovation as opposed to
simple “horizontal expansion," how many parties have the ability to fund research at all,
social demand for innovation as informed by reception of “la Technique," and the strength of
intellectual property rights (IPRs) and rule of law to protect innovations’ profit-generating
ability. New Institutional Economics abounds in Section 5.1, which will discuss λ at length.16

1.3 Previous Literature

The previous subsections have already discussed the history of economic growth theories;
now, to provide further context, the history of economic analyses of Ancient Rome (or the
Ancient Mediterranean more broadly) must be too.

While economic analyses of Ancient Rome are nearly as old as Economics itself,17 models
of endogenous growth such as Jones/Romer’s are rarely applied due to their emphasis on
technological innovation–itself being mostly absent. Instead analysts have focused on the
effect of institutions or governance, factors which Romans such as Strabo were themselves
aware of such as in the following passage about Roman Egypt:18
15
If, on the other hand, innovative investment is forced without the underlying belief that it will actually
work, researchers are incentivized to fabricate results, such as what happened during the SDI tests of the
1980s (see Weiner, 1993 “Lies and Rigged ‘Star Wars’ Test Fooled the Kremlin, and Congress," The New
York Times.
16
As mentioned in Hollander, 2017 “The Roman Economy in the Early Empire: An Overview.” Paul and
Economics: A Handbook, edited by Thomas R. Blanton and Raymond Pickett, 1517 Media, p. 4, New
Institutional Economics (NIE) has already been used by ancient historians.
17
For instance, Max Weber in Die Stadt constructs his tripartite urban model (classifying cities are “con-
sumer," “producer," or “merchant,") and then specifically applies it to the Ancient Mediterranean to explain
the decline of Classical civilizations. See Erdkamp, 2001 “Beyond the Limits of the ’Consumer City’. A
Model of the Urban and Rural Economy in the Roman World." Historia: Zeitschrift für Alte Geschichte vol.
50 no. 3 p. 333.
18
Strabo, 2014 (trans. Roller) Geography, Cambridge University Press, 17.1.13. For this thesis, ancient
sources will be cited using English names and a book.chapter.section or book.sentence structure.
 Garfinkel 12

“If then a king, who administered his government in the worst possible manner,
and with the greatest negligence, obtained so large a revenue, what must we
suppose it to be at present, when affairs are administered with great care?"

Early Writers Originally published in 1864, The Ancient City 19 by Fustel de Coulanges
is a proto-ethnography which uses textual and archaeological data to discuss the formation
and perpetuation of Classical civilizations along socioeconomic and religious lines. Like
future works, Fustel de Coulanges argues that, while profit-maximizing “ancient capitalism"
existed, fundamental cultural, technological, and political differences between the ancient
and modern worlds prevent their economies from being compared wholesale.20

Fustel’s main conclusion is that centralized, wealthy city-states primarily drove economic
growth in the Ancient Mediterranean, as their resources and institutions offered the neces-
sary stability and protection to promote trade. This is very similar to the modern belief that
effectively incentivizing economic growth and development comes from governmental bod-
ies/polices and less “market forces."21 Even so, in Fustel’s model mainly ignored economic
practices, believing that social and religious ones were the true foundations of Classical
city-states.

Rostovtzeff Excluding some 18th or 19th century texts,22 the first major analysis of the
Ancient Roman Economy is Michael Rostovtzeff’s Social and Economic History of the Roman
Empire (1926.) Rostovtzeff states he wrote the Social and Economic History to “[create] a
single book or monograph treating of the social and economic life of the Roman Empire as
19
See Fustel de Coulanges, 1980 The ancient city: A study on the religion, laws, and institutions of Greece
and Rome, Johns Hopkins University Press, Baltimore.
20
Erdkamp, 2001 p. 333.
21
Pelizzo, Kinyondo, and Nwokora, 2018 “Development in Africa." World Affairs, vol. 181, no. 3, p. 268
mentions numerous preferably growth initiatives which all involve governmental oversight and legislation;
clean technology, anti-corruption, increased sustainability, etc.
22
While Finley, 1985 The Ancient Economy. 2nd ed., University of California Press, p. 17. mentions
Hutcheson’s 1742 “The Principles of Oeconomics and Politics," its own reference to Xenophon’s Oikonomikos,
as well as texts written by David Hume and others, these aren’t true economic analyses as we understand
them.
 Garfinkel 13

a whole and tracing the main lines of its evolution."23 While the text does highlight the
administrative institutions behind the Roman Empire’s Economy throughout its tenure, it
doesn’t readily discuss economic growth nor its endogenous causes.

While Rostovtzeff’s narrative is one of how ineffective governance leads to factor decay, it
is burdened by Gibbonesque emphases on mounting “degeneracy"24 and his own Soviet influ-
ences. Writing in 1920s Ukraine, Rostovtzeff anachronistically refers to the Roman Empire’s
economy as a “pure [state] capitalist economy based on slave labor," calls Roman provincial
and urban elites “bourgeoisie," and at length discusses the importation of Ukrainian (“South-
ern Russia") grain into the Eastern Mediterranean throughout Classical Antiquity to forcibly
insert one of the Soviet Union’s “precursors" into the western historical narrative.25

Rostovtzeff’s research, although heavily dated, does provide some useful insights, such
as a foundational belief in effective governance’s importance for nonindustrialized economies
and the importance of upkeeping KH . He also catalogues various administrative practices
throughout the Roman Empire. His conclusions on Scythia (Ancient Ukraine’s) role in
feeding Eastern Mediterranean city-states have been vindicated with time and builds onto
the greater narrative of an ancient parallel to the modern economy’s growth via international
connectivity and unified markets.

Finley Adapting the theories of Rostovtzeff and Weber, in 1973 Cambridge University
Professor Sir Moses Finley released Then Ancient Economy, a landmark treatise which de-
picted urban collocation, transport costs, and societal factors as the primary drivers behind
economic practices and growth in the Ancient Mediterranean.26 Finley also popularized
within Classics the idea of Weber’s “consumption city," i.e. that ancient cities, unlike their
23
Rostovtzeff, 1926 Social and Economic History of the Roman Empire, The Clarendon press p. xi.
24
For example, Rostovtzeff dates the stagnation of Ptolemaic Egypt to when “the first able kings were
replaced by epigonoi (lit. “those born besides," referring to rulers born into their station, p. 283) and
unashamedly refers to “the Oriental despotism of the fourth and fifth centuries" [i.e. Late Antiquity.] (p.
xiii.)
25
Rostovtzeff, 1926 p. xv, 2, 279, and 249 respectively for some examples.
26
Finley, 1985 p. 107.
 Garfinkel 14

industrial-era counterparts, served as centers of consumption rather than production or ex-

port.27

As elucidated in Garney, Hopkins, and Whittaker’s anthology Trade in the Ancient Econ-
omy (itself dedicated to Finley,) these conclusions point to Finley’s own paradigm. Finley
puts forward a model of “cellular self-sufficiency” in the Ancient Mediterranean, that “each
farm, each district, each region grew and made nearly all that it needed."28 In other words,
Finley believed in a minimalist, fairly static, unintegrated economy, where massive environ-
mental pressure and innovative congestion (λ) stifled growth, potential economies of scale,
and the sharing of new techniques. About the presence of long-distance trade Finely had
this to say:

“Trade over long distances is of course a necessary condition for interdependence

but...trade of itself does not warrant such jargon as ‘a large unified economic
space’ unless it can be demonstrated."29

The ensuing “Primitivist vs. Modernist" debate over the veracity of Finley and other’s
works is essential for understanding the various theories surrounding the Ancient Roman
Economy, even if “no one self-identifies as part of either camp."30 “Primitivists" generally
believe in a disunified, underdeveloped, and stagnant model of the Ancient Mediterranean’s
Economy, where production is overwhelmingly at the subsistence level, there are few if any
unified markets, and that city-states were largely autonomous and operated like frogs around
a pond irrespective of larger political units.31

“Modernists" conversely believe the Ancient Mediterranean, especially under Roman ad-
27
Garnsey et al., 2015 “An Underdeveloped Economy.” The Roman Empire: Economy, Society and Culture,
2nd ed., University of California Press, p. 74.
28
Peter Garnsey, 1983 Trade in the Ancient Economy, University of California Press p. xi.
29
Fitzpatrick, 2011 “Provincializing Rome: The Indian Ocean Trade Network and Roman Imperialism.”
Journal of World History, vol. 22, no. 1, p. 27, citing Finley, 1985 p. 177.
30
Hollander, 2017 p. 2.
31
The comparison of Mediterranean economic and political activity to frogs around a pond is borrowed
from Plato’s Phaedro section 109d.
 Garfinkel 15

ministration, was highly-integrated, institutionally complex, locally specialized, and overall

analogous to various 17th or 18th century European nation-states like England or Portugal,32
especially in reference to their maritime trade.33 Modernists also believe that, due to these
similarities, contemporary economic vocabularies can describe the Ancient Roman Economy,
whereas Primitivists believe that fundamental cultural and institutional differences undercut
the validity of such analyses, and that “different concepts were (sic.) needed to analyse a
qualitatively different object [economy.]"34

The Ancient Economy is therefore a decidedly Primitivist text, where high risk and in-
formation asymmetry created disunified markets in the Ancient Mediterranean,35 not “in-
tellectual failing" but “institutional behaviour"36 led to high innovative congestion (low λ,)
and a cultural disinterest in optimal factor utilization pervaded the economy. While Finley’s
model is compelling, it doesn’t discuss the underlying causes behind many of its claims to
the greatest degree possible.

Hopkins For this analysis, University of Cambridge Professor and noted co-editor of Trade
in the Ancient Economy Keith Hopkins’ main contribution is his 1980 article “Taxes and
Trade in the Roman Empire (200 B.C.– A.D. 400.)" In it, Hopkins outlines his “taxes and
trade" model, one which provides a theory of endogenous growth in the Ancient Roman
Economy.

According to the model, the Ancient Roman Economy existed in two concentric rings.
The core provinces possessed the majority of mechanical and human capital (KH , KM )
32
Scheidel, 2007 “A Model of Real Income Growth in Roman Italy.” Historia: Zeitschrift Für Alte
Geschichte, vol. 56, no. 3, p. 343.
33
Parker, 2002 “Ex Oriente Luxuria: Indian Commodities and Roman Experience.” Journal of the Eco-
nomic and Social History of the Orient, vol. 45, no. 1, p. 56.
34
Hobson, 2014 “A Historiography of the Study of the Roman Economy: Economic Growth, Development,
and Neoliberalism" in TRAC 2013: Proceedings of the Twenty-Third Annual Theoretical Roman Archaeology
Conference, London 2013 p. 15.
35
Temin, 2013d “Per Capita GDP in the Early Roman Empire." The Roman Market Economy, Princeton
University Press, p. 249, citing Finley, 1985 p. 22.
36
Finley, 1985 p. 23.
 Garfinkel 16

through a higher concentration of roads, education, mints, population centers, etc., while
the peripheral provinces conversely provided land-based capital (Kν ) and non-labor inputs.
Increased cooperation between these rings led to increased monetization, factor utility (A,)
and overall integration via “the commercialization of exchange, an elongation of the links
between producers and consumers, the growth of specialist intermediaries (traders, shippers,
bankers), and an unprecedented level of urbanization."37

While decidedly Modernist in principle, Hopkins’ analysis unfortunately doesn’t readily

discuss innovation-generated growth in the Ancient Roman Economy, instead focusing on
the growth of K. For instance, Hopkins mentions how the large fixed costs incentivized
investment into large vessels designed for long-distance arbitrage,38 but not how those same
factors could incentivize new technologies overall.

Temin Unlike the previous scholars, who were some combination of Classicist, Historian,
or Sociologist, Temin is an economist, former associate of Solow’s,39 and a Modernist’s
Modernist, frequently describing the Ancient Roman Economy in terms of 17th Century
Europe.40 In terms of endogenous growth, one of Temin’s main contributions is the claim
that, despite a relatively high level of urbanization41 and monetization,42 the Ancient Roman
Economy’s emphasis on slavery disallowed it from participating in our conception of a “market
economy" and limited economic growth beyond the subsistence level.43

While seemingly accurate, elsewhere Temin also cites slavery itself–alongside technical
37
Hopkins, 1980 p. 102.
38
Garnsey et al., 2015 p. 77 citing Hopkins, 1980 p. 105.
39
Temin, 2013c “Growth Theory for Ancient Economies.” The Roman Market Economy, Princeton Uni-
versity Press, p. 211.
40
Hollander, 2017 p. 5, citing Temin, 2013f The Roman Market Economy, Princeton University Press p.
259. Note that Temin, 2013d “Per Capita GDP in the Early Roman Empire," among others, are chapters in
Temin, 2013f The Roman Market Economy.
41
Temin, 2013d p. 252.
42
In many places, such as in Temin, 2013d p. 243 citing Hopkins, 1980 p. 121’s claim of a lower aggregate
tax rate during the Late Empire Period (Roman Dominate,) Temin cites Hopkins’ research and conclusions,
such as that of a relatively high monetization level.
43
Temin, 2004 “The Labor Market of the Early Roman Empire.” The Journal of Interdisciplinary History,
vol. 34, no. 4, p. 513, 525.
 Garfinkel 17

innovations like the expansion of contract law–for Roman economic expansion.44 Taken
together, however, Temin’s main claim is that Ancient Rome’s economic growth derived
mainly from increases in factor inputs, (Kν , L, etc.) rather than an increase in A.

1.4 Methodology, Structure, and Limitations

The past two subsections have outlined the model used for this analysis and previous litera-
ture on this subject. This subsection will go over this analysis’ methodology, structure, and
limitations.

Methodology and Structure The majority of this analysis will be spent articulating the
factors comprising gA in Romer/Jones’ model and how their presence (or lack thereof) in
the Ancient Roman Economy indicates a lack of innovation-led output growth in exchange
for output growth caused by accumulation of factor inputs. For each of these components
of gA , the following process will be observed:

First, the factor will be discussed in terms of how it impacts gA and what that means for
the Ancient Roman Economy or nonindustrialized economies more broadly. After that, the
section will go through various historical examples to indicate how and why the analyzed
factor contributed to, didn’t contribute to, or actively worked against innovation-led growth
of A. Finally, the section will conclude with a review of the findings and expansion of the
discussion to economic development in nonindustrialized economies and how that relates to
governance.

Since no cohesive, quantitative dataset exists on the Ancient Roman Economy (nor could
be created in any reasonable amount of time,) the individual analyses will instead follow a
more historiographic approach, utilizing the material and historical records to do so. The
44
Malmendier, 2009 p. 1079 citing Temin, 2001 “A Market Economy in the Early Roman Empire.” The
Journal of Roman Studies, vol. 91, pp. 169–81 and Temin, 2006 “The Economy of the Early Roman Empire.”
The Journal of Economic Perspectives, vol. 20, no. 1, p. 133-151.
 Garfinkel 18

material record includes things artifacts (surviving pieces of culture) and epigraphy (sur-
viving, non-literary inscriptions.) The historical record, on the other hand, includes a wide
variety of personal accounts, correspondences, and histories written at the time.

Before discussing the factors composing gA , however, Section 2 will begin by discussing
several first principles about the Ancient Roman Economy, the fundamental differences be-
tween it and our own, and some of the requisite factors underlying innovation-led growth
which the Romer/Jones model doesn’t explicitly account for. Section 2 is meant to epito-
mize many of this analysis’ arguments, with subsequent sections articulating them in greater
detail. Section 2 will also briefly discuss δ.

Sections 3–5 will then discuss the factors composing gA and innovation-led economic
growth more broadly. Section 3 will discuss population growth n and savings rate s together
as a baseline, as both influence influence the growth of capital stock and investment irre-
spective of technological innovation. Section 4 will discuss ϕ, innovative synergy, and its role
in propagating and creating innovations. Section 5 will discuss λ, innovative congestion, the
wide variety of social, cultural, and environmental factors which can impact gA , and expand
on some of Section 2’s concepts. After these Section 6 will conclude.

Limitations Compared to traditional economic analyses, this analysis has many clear
limitations, primarily its lack of a large set of quantitative data to run regressions against.
Whereas traditional analyses use census data, parish records, sales data, etc. to draw con-
clusions, for the Ancient Roman Economy such records were either never created or do not
survive. What data does survive is fragmentary, scattered, and clouded by bias.45 Due to
the often contradictory or woefully incomplete nature of our primary sources, the frustrat-
ing concession must be made that, while all sourced data is canon, not all sourced data is
45
While a large portion of data survives from Roman Egypt, administrative differences between the
province and the rest of the empire disallow most forms of extrapolation. For instance, Bowman, 1996
“Provincial administration and taxation" in The Cambridge Ancient History, Vol. 10, Cambridge University
Press, p. 346 mentions how, due to national security concerns, Roman Egypt was managed directly by
persons of lower (equestrian) rank, with this aberrant practice remaining for centuries.
 Garfinkel 19

necessarily true.

For instance, even something as simple as the price of grain cannot be readily determined
despite having a relatively stable demand per capita across contexts. While many recorded
prices orbit around 3-4 sesterces (HS) per modius (approx. 8.73L,)46 several of these instances
come from records of price ceilings or other government interventions.

While this figure might seem reasonable, Pliny the Elder–one of our chief primary sources
for economic data in the Ancient Roman Economy–complicates things. Near an oft-quoted
passage mentioning a supposed annual trade deficit of 100M HS with the Indian Subconti-
nent,47 Pliny states the price of black pepper, when first imported to ports Berenike,48 has
a price of 4 denarii (16 sesterces) per Roman pound (approx. 329g.) The notion that 8.73L
of wheat would ever cost the same as 82g of black pepper, even today, is absurd. One must
disregard Pliny’s numbers here, even if we are unable to disregard the rest of his testimony.

As shown, such problems with the sources can be addressed through a combination of
cross-referencing, critical analysis, and honesty about this analysis’ scope. A copious amount
of modern scholarship will also be used to augment the ancient sources available. In many
ways therefore, this analysis is using a historiographic approach to discuss economic concepts,
specifically how they relate to nonindustrialized economies using a specific example from two
millennia ago. Both historical and economic in nature, this analysis will use each’s tools when
applicable.

46
Levick, 1982 “Domitian and the Provinces.” Latomus, vol. 41, no. 1, p. 57 cites 4 HS/mod; Temin,
2013d p. 245 and Tacitus Annals 15.39 cite 3 in the 1st century CE; and Temin, 2006 p. 138 cites 4-6
instead.
47
Pliny the Elder, NH 12.84, 12.29. Based on calculations by Temin (see Temin, 2013d “Per Capita GDP
in the Early Roman Empire") and others, a rough estimate of GDP in the Early Roman Empire would be
10 billion HS, meaning this figure of 100M HS per year would be a trade deficit of 1%.
48
See Sidebotham, 2011 “Trade in Roman Berenike." Berenike and the Ancient Maritime Spice Route,
University of California Press, p. 221-258.
 Garfinkel 20

2 First Principles

This section will discuss a variety of first principles and other foundational ideas about the
Ancient Roman Economy, fundamental differences between it and modern economies, the
role of governance in its economy, some of the factors underlying innovation-led growth
and industrialization, and how these aforementioned factors were not wholly present in the
Ancient Roman Economy. This section is meant as a prelude and summary for the overall
analysis, with subsequent sections expanding on specific ideas presented in this one. After
these have been discussed, this section will conclude and tie in these concepts to the various
factors comprising gA , which will be discussed afterwards.

2.1 Foundational Concepts

The Ancient Roman Economy, much like all other nonindustrialized economies, was primar-
ily agrarian. The vast majority of the population lived near a subsistence level, devoting
themselves to the production of grain and other staple crops. While the exact percentage
is uncertain, a rough estimation is 85% of arable (usable) land being solely used for the
cultivation of grain and grapes49 with a similar percentage of the labor pool working the
land due to the prevailing lack of labor-saving machinery.50

This has many implications. Such a high percentage of L devoted to agriculture lowers
LA , meaning fewer innovations are made overall. Subsistence agriculture also has high op-
portunity costs, as any non-agricultural activities directly impact one’s food security. This
combined with variable weather conditions creates high and unpredictable transportation
49
Geraghty, 2007 “The Impact of Globalization in the Roman Empire, 200 BC-AD 100.” The Journal of
Economic History, vol. 67, no. 4, p. 1038.
50
H.W. Pleket (see Peter Garnsey, 1983 p. 132) cites A.H.M. Jones’ 1964 statistic of 95% agricultural em-
ployment based on available records from the Late Roman Empire, but also questions its veracity. Erdkamp
(see Erdkamp, 1999 “Agriculture, Underemployment, and the Cost of Rural Labour in the Roman World.”
The Classical Quarterly, vol. 49, no. 2, p. 556) cites between 80–90%. Lebergott, 1966 “Labor Force and
Employment, 1800–1960" in Output, Employment, and Productivity in the United States after 1800, National
Bureau of Economic Research, p. 131 cites a similar level of agricultural employment for the Southern US
in 1800.
 Garfinkel 21

costs51 that prevented unified, price-clearing markets from properly forming;52 this leads
leads to less efficient resource allocation and potential for investment. High agricultural
opportunity costs also reduce the stock of KH , which depresses total factor utility (A.)53

Subsistence agriculture, especially in technologically-poor settings, is also made inherently

risky by large, hard to control variation in crop yields. This risk lowers surplus income and
prevents the vast majority of the population from engaging in innovative investment at all.
When surplus income did exist, therefore, it would likely be invested in things with high
and quick rates of return, notably horizontal expansion and infrastructure projects. In other
words, investment would primarily go towards “tiling out" production (increasing L and K
stocks) rather than increasing A through innovation.

This is reflected in the Ancient Mediterranean’s economic-political practices. With low

KM levels, population (labor-units,) not technology, is more readily seen as the source of
power, with loss of population through disease, war, systems collapse, etc. of major con-
cern.54 Population therefore also correlates to a nation’s ability to project market demand
and source resources from beyond its borders.55

Throughout Classical Antiquity the primary methods of population (and thereby eco-
nomic) growth were the foundation of allied city-states (colonies) and cultivation of peripheral
lands, such as the Maghreb region of Northern Africa.56 The resurgence of Mediterranean
trade following the Bronze Age Collapse likewise coincided with the 8th century BCE es-

51
Transport costs were generally higher overland, but more variable by sea. See Garnsey et al., 2015 p.
72. Geraghty, 2007 p. 1045 additionally notes how in the 1st century CE, 90% of the wheat price between
Egypt and Rome was due to transportation costs.
52
Peter Garnsey, 1983 p. 4-5.
53
This notion that the Roman Empire, while extensive, had a low technological index overall has been
widely understood, especially towards agriculture and shipping (see Finley, 1985 p. 108.)
54
Gills and Frank, 1992 p. 632. The Bronze Age Collapse is a key example of this, where a systems
collapse led to a near-universal cessation in macro-economic activity.
55
E.g. Pliny the Elder Natural History Book 12 extensively discusses Rome’s trade with India and the
Arabian Gulf, as is supported by the presence of Roman artifacts in those contexts.
56
Peter Garnsey, 1983, Trade in the ancient economy p. 82-84.
 Garfinkel 22

tablishment of Greek colony-states around the basin,57 58

a practice later adopted by the
Romans to great effect.

These colonies were also established to address overpopulation, as limited farming and ad-
ministrative technologies prevented cities from growing beyond a certain limit without exten-
sive food imports. In many ways, the Ancient Roman Economy existed inside a “Malthusian
world,59 where (in line with his theory,) population growth can only be sustained through
increasing the quantity of cultivated land and keeping urban populations low to minimize
urban congestion costs.

Due to these hard limitations, to an even greater extent than modern economies, output
derived from the total number of labor units available, with KM effectively acting as a
substitute for additional labor.60 As pointed out by a passage of Columella,61 this equation
of KM and L was rather internalized within the Ancient Roman Economy, as while Columella
accounts for a farm’s startup and variable costs, he doesn’t differentiate between labor or
capital costs.

Despite all these factors, economic growth did nevertheless occur, but to discuss the
engines of Ancient Rome’s economic development (or lack thereof,) slavery must first be
discussed.

57
Peter Garnsey, 1983, Trade in the ancient economy p. 7-9. See also Saprykin, 2017 “Ancient Sea Routes
in the Black Sea.” The Sea in History - The Ancient World, edited by Philip de Souza et al., NED-New
edition, Boydell Brewer, p. 345.
58
Much of this is also attested to in Thucydides History of the Peloponnesian War 1.1-7. See Thucydides,
1996, trans. Strassler,The Landmark Thucydides: a Comprehensive Guide to the Peloponnesian War. Free
Press for a translation.
59
Temin, 2013a The Roman Market Economy, Princeton University Press p. 220-242, 243.
60
Veen, 2010 “Agricultural Innovation: Invention and Adoption or Change and Adaptation?” World Ar-
chaeology, vol. 42, no. 1, p. 3.
61
Columella, 1745 (trans.) De Re Rustica 3.3.8-10. The De Re Rustica (or DRR) is an ancient agricultural
manual, one of many.)
 Garfinkel 23

2.2 The Problem of Slavery

Besides for the overall difference in technological level, a main distinction between the An-
cient Roman and modern economies is the former’s systemic utilization of slave labor. While
this subsection and thesis more broadly discusses slaves as a form of capital, neither the hor-
ror of Roman Slavery nor the millions of lives it ruined should be discounted. While the
children of freedpersons did possess legal status to other nonelites,62 slaves were nevertheless
systemically disadvantaged and employed in low-KH fields (e.g. mining and agriculture)
more than the general population, entailing a massive under-utilization of human resources.

While slavery long predated Rome in the Ancient Mediterranean, Roman slavery gained
popularity during the Middle Republic as a substitute to the labor Roman citizens could no
longer provide due to the Republic’s wars of increasing scale and time.63 In other words,
slaves effectively functioned as KM , multiplying the productivity of existing L or substituting
it altogether.64 Even so, slavery’s use as a substitute for KM however constituted a massive
liability. Not only did slavery de-incentivize innovation overall,65 by the 1st century slaves
and freedpersons made up a sizable percentage of the workforce, even in urban and educated
environments. This coincides with Temin’s argument that, in a (mechanical) capital-scarce
economy like Ancient Rome’s, the primary divider between labor inputs is level of education,
as when KM is low KH is used to compensate.66

Building off Section 1.2, slaves as a form of “capital" are uniquely vulnerable to “decay
shocks," circumstances where a rapid increase in δ negatively impacts the capital stock. For
a form of “capital" internalized within humans, such a shock would be caused by famine,
plagues, or other mass casualty events which disproportionately affected slaves. Among the
62
Scheidel, 2007 p. 323.
63
Duncan-Jones, 2019 “The Antonine Plague Revisited." Arctos, the Classical Association of Finland vol.
52, p. 327-328.
64
As pointed out by Pliny (Pliny the Elder, NH 18.37,) its assumed family members would also be working
the farm, partially to lower costs.
65
Gills and Frank, 1992 p. 646. This effect could naturally also be self-perpetuating.
66
Temin, 2013c p. 538.
 Garfinkel 24

more dramatic causes of such disasters were volcanic eruptions.

According to Cassius Dio,67 besides for sealing Pompeii and Herculaneum, the ash ejected
by Vesuvius’ eruption in 79 CE traveled to Africa, Egypt, Syria, and Rome itself, where–
after it filled the air and darkened the sun–a pestilence followed.68 Dio here unwittingly
describes “global dimming," a phenomenon where a Plinian eruption ejects enough ash into
the atmosphere to noticeably lower global temperatures and light levels, often leading to crop
failures, famine, and disease. Another example of this occurrence was the 1815 eruption of
Mount Tambora, which led to the “Year Without a Summer" in 1816 and its famines.

Whereas the most severe plagues were liable to kill up to roughly 30% of the population,69
slaves seem to have been unique vulnerable due to their systemic abuse and poorer nutrition.
70
For example, in Pompeii71 survives a mention of pan puero, slave bread made with various
admixtures (chalk, dried beans, etc.)72 to “stretch" the wheat or overall improve presentation
at the expense of nutritional content. If a common practice, the end result would be a slave
population more susceptible to disease, as is suggested by the historical record.

When capital is excessively invested into human rather than mechanical forms, plagues
become even more dangerous, as they not only lower L and cause inflation,73 but also dam-

67
Cassius Dio Roman History 66.22.2-66.23.5 (likewise cited by Duncan-Jones, 2019 p. 60-61, 81.)
68
This might be the same plague mentioned in Orosius 7.9.11 (see Rogers, 1984 “Domitian and the Finances
of State.” Historia: Zeitschrift Für Alte Geschichte, vol. 33, no. 1, p. 63.)
69
For instance, the Antonine Plague of the 2nd-3rd centuries CE is recorded as having killed this amount
(see Ammianus Marcellinus, 1950, trans. Rolfe p. 363 / 23.6.24, Magie, 1950b “From Gold to Iron." Roman
Rule in Asia Minor, Princeton University Press, p. 663, Duncan-Jones, 2019 p. 44.
70
Dionysius of Halicarnassus, Roman Antiquities 10.53 (cited by Duncan-Jones, 2019 p. 50) states that a
plague once killed Rome’s entire slave population. While certainly an exaggeration, this likely means slaves
were disproportionately affected.
71
Insula IX.vii.24-5, recorded in the Corpus inscriptionum Latinarum, 1862, 4.5380. See Etienne, 1992
“Pompeii: The Day a City Died," H.N. Abrams, p. 184 for a full translation.
72
Spurr, 1983 p. 10 mentions some of these, as well as Pliny the Elder NH 18.117, which the passage cites.
73
Temin, 2013e “Price Behavior in the Roman Empire.” The Roman Market Economy, Princeton University
Press, p. 74 states “the mechanism by which plague causes inflation is not clear;" this occurs by plagues
damaging KH -reliant industries, lowering output asymmetrically. This passage itself cites Rathbone, 1997,
“Prices and Price Formation in Roman Egypt," Economie antique: Prix et formation des prix dans les
economies antiques (Saint-Bertrand-de-Comminges, France: Musée archéologique départmental,) p. 183-
244.
 Garfinkel 25

ages A and land-based capital by respectively destroying KH and the labor to maintain
it.74 Even outside of plagues, the use of slaves as capital still contains built-in inefficiencies,
as slaves commonly rebel and (bluntly speaking) make poor collateral on loans.75 Despite
all these liabilities, the Romans had trouble phasing out slavery compared to peripheral
peoples, as pointed out by a passage of Ammianus Marcellinus (31.2.25) where several Ro-
mans are accosted for still owning slaves in the 4th century CE. In conclusion, beyond being
unconscionable, slavery constituted a massive tax on Rome’s economic potential.

2.3 Differences between Industrialized and Nonindustrialized economies

Beyond slavery or the cost-induced slower spread of information, there are several more
fundamental differences between nonindustrialized and industrialized economies, how they
function, and how industrialized economies are uniquely oriented to promote innovation-led
growth. Before discussing those, however, energy capture must be discussed.

“Energy capture" is the process of how humans extract and use our environment’s energy,
with higher energy consumption distinctly correlating to higher indexes of economic and
technological development. Industrial and post-industrial economies are defined by their
utilization of nonfood calories (coal, nuclear, etc.,)76 while nonindustrialized economies rely
more on food-based calories such as animal traction.77 As will be elaborated later, the
limited, agricultural basis of a society’s energy incentivizes maximizing agricultural output
as soon as possible, promoting investments with the fastest rates of return.78 Failure to see
immediate returns would negatively impact the food supply, exacerbated by much of the
population still living at a near-subsistence level.

Since harnessing nonfood calories is a fundamental aspect of industrialized economies–

74
Temin, 2013e p. 85.
75
Temin, 2013c p. 213.
76
Morris, 2013 Morris, 2013 “Energy Capture.” The Measure of Civilization: How Social Development
Decides the Fate of Nations, NED-New edition, Princeton University Press, p. 58.
77
Morris, 2013 p. 55-56.
78
Veen, 2010 p. 3
 Garfinkel 26

nonetheless through unique pieces of technology–it follows that these industrial economies
are maintained through the upholding of property rights, specifically intellectual property
rights (IPRs) and patents which protect the profit-generating potential of such technologies
for their creators.79 Even so, the concept of patents and property rights was wholly absent
from the Ancient Roman Economy, as it emerged in 15th century Italy only to then spread
to England by the 17th.80 While not present in the Ancient Roman Economy, modern
scholarship indicates that, even in developing economies, strong IPRs incentivize innovation,
even if the utility gained may differ from industrialized ones.81

Even if IPRs do exist in a nonindustrialized economy, to effectively support innovation-

led growth they need to be backed by an impartial justice system able to uphold the rule of
law. This is problematic, as effective rule of law entails impartial and speedy prosecution of
criminals, clear demarcations between overlapping legal jurisdictions, and the reinforcement
that agents of the law (police, judges, lawmakers, etc.) are not above it.82

This becomes especially difficult in nonindustrialized economies such as Ancient Rome’s

where legal authority is inherently embodied in political figures who often also hold religious
positions. As will be elaborated in Section 5.1, introducing new technologies often requires
new legal or administrative frameworks which can be fought by those in power to protect
their own economic interests.83 IPRs in of themselves are also not a cure-all, as they aren’t
always effective and when improperly implemented can undermine innovation.84

79
Jones, 1995 opens his paper (p. 759) by stating his theory of endogenous growth is based off this ability
for innovators and early adopters to gain additional profits.
80
Merges, 1995 “The Economic Impact of Intellectual Property Rights: An Overview and Guide.” Journal
of Cultural Economics, vol. 19, no. 2, p. 105-106.
81
Janjua, Samad, and Ullah, 2019 “Intellectual Property Rights (IPRs) and Economic Growth in Pakistan.”
The Pakistan Development Review, vol. 58, no. 3, p. 231, 235-6.
82
Taylor, 2017 “Regulatory Rule of Law.” Regulatory Theory: Foundations and Applications, edited by
Peter Drahos, ANU Press, p. 398, 403-4.
83
Verhulst et al., 2017 “Innovations in Global Governance: TOward and Distributed Internet Governance
Ecosystem." Who Runs the Internet?: The Global Multi-Stakeholder Model of Internet Governance, Centre
for International Governance Innovation, p. 99.
84
Menell, 2011 “Governance of Intellectual Resources and Disintegration of Intellectual Property in the
Digital Age.” Berkeley Technology Law Journal, vol. 26, no. 4, p. 1538, 1555.
 Garfinkel 27

Finally, the transition from a nonindustrialized economy to an industrialized one founded

on rule of law and protection of property rights which allows for innovation at all rests on
the shoulders of a large, highly coordinated cadre of high-KH specialists.85 These specialists,
such as in the case of Colombia in recent decades, are often external to the economy, used to
root out previous corrupt systems, set up new institutions, and train the next generation of
specialists once they leave.86 The Ancient Roman Economy lacked such external specialists.

2.4 Governance in the Ancient Roman Economy

With these previous points in mind, the following question emerges: If not for (entirely)
upholding the rule of law or providing the optimal environment for innovation, what role
did governments play in the Ancient Roman Economy? Based off the available sources, four
main functions emerge: Maintaining religious functions, projecting military might, settling
legal disputes (particularly over contracts,) and providing food security for portions of the
population.

Towards these ends, governmental functions often entailed mitigating risks (such as natu-
ral disasters’) which were omnipresent. A surviving mining contract (CIL 3.2.948) carves out
specifications for flooding. Pliny the Younger’s letter 3.19 weighs the economic advantages
(lowered travel costs, management synergies, etc.) of buying an adjacent farm against the
risk of a single storm being able to destroy both. Even though cargo shipping at scale was
dramatically cheaper by water than land,87 loans were often written to excuse borrowers in
case of shipwreck, even if this functionally resulted in interest rates as high as 12.5-30%.88
Crop failures could result in workers going without wages, instead receiving a portion of the
85
Rubin, 2006 “Governance, Rule of Law, and Human Rights.” Afghanistan’s Uncertain Transition From
Turmoil to Normalcy, Council on Foreign Relations, p. 24-26. See also Taylor, 2017 p. 404 and Verhulst
et al., 2017 p. 97.
86
Marcella, 2009 Democratic Governance and the Rule of Law: Lessons from Colombia. Strategic Studies
Institute, US Army War College, p. 19-27.
87
Acknowledged in Diocletian’s Price Edicts (see Kropff, 2016 (trans. Kropff)“An English translation of
the Edict on Maximum Prices, also known as the Price Edict of Diocletian," Academia.edu.)
88
Peter Garnsey, 1983 p. 36.
 Garfinkel 28

surviving crop.89

Piracy was also common, with Plutarch and Cassius Dio writing on the Roman Republic’s
efforts to curtail their power.90 Beyond hunting pirates, the Roman state also tried to lower
maritime risk by selling shipping insurance and not taxing merchant vessels.91 Alongside
these state institutions, non-state firms such as joint-stock companies existed,92 although
their reach was limited due to the natural difficulties of loaning to people outside one’s social
network. Rebellions were also common, with the Roman policy of establishing allied colonies
a clever method of reducing the risk of rebellion through increased economic integration.93
94
While such projects echo modern international development initiatives, the Roman state
would naturally only be able to develop peripheral regions up to their own technological level
through the expansion of infrastructure and legal systems.

As for the state’s obligation to keep certain persons fed,95 this didn’t constitute an eco-
nomic so much as a welfare program meant to help the local urban poor present in Rome
(and other cities through parallel programs.)96 The Roman program, called the annona
or “dole" was a state subsidy on grain established in 139 BCE to assist landless plebeians
pushed off their land following the wide series of land enclosures over the previous century
similar to their 18th century English counterparts.98

Originally this program benefited 50,000 Romans by selling them wheat at 1.5 sesterces

89
Pliny the Younger Letters 9.37.
90
Plutarch, Life of Pompey 24-25.1 and Cassius Dio Roman History 36.20.1-36.23.4.
91
Suetonius Life of Claudius 18.2 and Tacitus Annals 13.50-51.
92
Temin, 2013b p. 160, 162.
93
Magie, 1950a p. 570 and Erdkamp, 2001 p. 341. I also recommend Ian Morris’ 2014 book War! What
is it Good For: Conflict and the Progress of Civilization from Primates to Robots, as it touches on much of
these same subjects.
94
To likewise help economic integration, the Romans likewise based their own currencies off of preexisting
Hellenistic models. See Katsari, 2003 “The Organisation of Roman Mints during the Third Century CE: The
View from the Eastern Provinces.” Classics Ireland, vol. 10, p. 27-28.
95
Examples of this obligation date back to Aristotle Constitution of the Athenians 43.4.
9697
98
Scheidel, 2007 p. 329. Section 3 will expand on this, as well as why these enclosures led to industrial-
ization in England but not Rome.
 Garfinkel 29

(HS) per modius (about half price,)99 By the Late 1st century BCE, however, the program
had expanded to 300K Romans (5-10% of Italy’s population)100 and a general policy of at-
tempting to alleviate local famines through a series of price controls and subsidies.101 Outside
of select urban areas, however, food insecurity kept the vast majority of the population at a
subsistence level due to the high risk of moving to commercial agriculture.102

2.5 Conclusion

If anything, this section has begun outlining the “developmental trap" that keeps nonin-
dustrialized economies such as Ancient Rome’s from developing into industrialized ones. A
combination of poor resource management, poor diffusion of information, lack of coordina-
tion, lack of human capital, and lack of rule of law able to uphold property rights all make
innovation-led growth practically impossible in these circumstances without substantial out-
side assistance which (at this time) would be impossible to receive.

Taken together, this section (like previous scholarship)103 begins to suggest multiple equi-
libria of economic development between purely subsistence agriculture and a purely postin-
dustrial economy, with the equilibria founded on different technological indexes indicated
by A. Now that these foundational concepts have posed why innovation-led growth would
be so difficult, Section 3 will begin to analyze Jones’ model of gA and how, even without
innovation-led growth, traditional avenues such as increasing L still remain feasible.

99
Levick, 1982 p. 52 cites a price of 4HS/modius, Tacitus (Annals 15.39) cites a price ceiling of 3HS/modius,
Temin, 2013d p. 245 likewise cites a passage of Cicero (In Verrem 2.3) and other literature to surmise a
general price of 3HS/modius.
100
Geraghty, 2007 p. 1040, 1044. Garnsey et al., 2015 p. 84 mentions the sheer amount of grain amphorae
at Rome which attest to this massive importation.
101
See Suetonius Life of Domitian 4.5, Finley, 1985 p. 33 referencing Julian II’s actions, Tacitus Annals
15.18.2, 15.39.2, and other passages cited above.
102
Bang, 2007 p. 226.
103
Bang, 2007 p. 24
 Garfinkel 30

3 Population Growth and Savings Rate (n and s)

Now that Section 2 has outlined the basis of why innovation-led output growth is extremely
difficult in nonindustrialized economies, this section will discuss the alternative: Factor-based
growth led by population growth n and savings rate s. While these two variables determine
the growth of L and K respectively in the Solow model, in Jones’ model they are also relevant
for determining innovation-led gA . This section will look at both aspects and elaborate on
how the former (growth through expanding L and K ) is more relevant to nonindustrialized
economies such as Ancient Rome’s. This section will also expand on the discussion of why,
despite similar circumstances, an industrial revolution occurred in England and not Ancient
Rome.

3.1 The Data: n

As elaborated in Section 1.2, in models either emphasizing gA or gK & gL , n directly cor-

responds to gY . n was nevertheless extremely low in the Ancient Roman Economy, leading
to low growth overall. Although we lack specific numbers, a metric often floated is 0.5%104
which is roughly equivalent to extant population growth rates from 14th-16th century Eng-
land and 18th century Italy.105 Such a low rate of n is caused by Malthusian pressures,
where at equilibrium in a subsistence agrarian economy the expectation is that all currently
available lands are being cultivated at what is approximately the highest level of efficiency
possible.106

n did however change over time, especially during the 2nd-1st centuries BCE, which led
to what Polybius terms “[how] in less than fifty-three years nearly the whole world was

104
Monson, 2007 “Rule and Revenue in Egypt and Rome: Political Stability and Fiscal Institutions.”
Historical Social Research / Historische Sozialforschung, vol. 32, no. 4 (122), p. 246 citing Bagnall, 1994
The demography of Roman Egypt, Cambridge University Press, p. 81-90.
105
Laslett, 1971 The World We Have Lost. 2nd ed., Methuen, p. 107.
106
As mentioned by Temin (see Temin, 2006 p. 135,) “subsistence level" refers to a range of different output
levels united by their lack of commercial sale.
 Garfinkel 31

overcome and fell under the single dominion of Rome."107 Following the Second Punic War,
Rome received simultaneously, as pointed out by Sallust and others,108 a massive influx of
wealth, a population depression due to large military defeats,109 and a massive increase in
the population of available slaves/slave labor.110 These factors combined to create a massive
economic reorganization resulting in a vast increase to Rome’s level of urbanization and
output, with any growth in A resulting from administrative changes rather than technological
ones.

Since many of the previous small landowners had died in the wars, afterwards various
Roman families acquired these lands and reconstituted them into latifundia, massive slave-
operated plantations capable of immense profit.111 Writing two and a half centuries later,
Pliny the Elder puts the price of these estates at 1.3M sesterces although it’s unclear whether
this refers to the average annual output, cost of the plantation’s land, or the plantation itself
including its slaves, buildings, etc.112

These latifundia had many effects, including increasing overall output through economies
of scale. They also created a subclass of fabulously wealthy aristocratic clans, a large,
unlanded urban underclass, and (in the long term) political instability, the latter of which is
itself detrimental to growth.113 The increased population density, increased food supply, and
massive influx of slaves likely all contributed to a spike in Italy’s population during this time
period, ultimately transforming into military and economic supremacy. While increasing n
107
Polybius The Histories 6.2.3.
108
Sallust Catilinian War 10.1.
109
Livy when recounting the battle of Lake Trasimene (see Livy Ab Urbe Condita (AUC ) 22.7)) reports a
total of 15,000 Roman casualties in one day. Polybius likewise in his Histories (see Scheidel, 2007 p. 324
citing Polybius) reports a Roman loss of 650 Roman ships carrying up to 420 men each between 255 and
249 BCE, resulting in easily over 100,000 Roman casualties.
110
Geraghty, 2007 p. 1043 (citing Brunt, 1971 Italian Manpower, Oxford University Press) puts the esti-
mated change from 500K in 200 BCE to 1.2-2.0M in 0 CE.
111
Rosenstein, 2008 “Aristocrats and Agriculture in the Middle and Late Republic.” The Journal of Roman
Studies, vol. 98, p. 1.
112
Pliny the Elder Natural Histories (NH ) 13.92.
113
Pelizzo, Kinyondo, and Nwokora, 2018 p. 273 states that in recent decades greater economic growth in
Africa has been tied to more stable governments, whether they were democracies or dictators supported by
external powers.
 Garfinkel 32

increases L in subsequent time periods, more nuance is required, as a large portion of these
new workers were slaves and simply increasing n doesn’t increase the quantity of LA or K
by itself, even if more K is required to effectively utilize the increases in L.114

Astute readers will however realize that these characteristics echo those in Britain on
the Industrial Revolution’s eve;115 why then did the Industrial Revolution not happen here?
As mentioned in Section 2.3, Ancient Rome lacked the proper legal framework to protect
innovators or their inventions. Pliny the Elder notes how–in his own time–the monopolistic
owners of latifundia were liable to being killed and having their property expropriated by
the state.116 Hardly optimal conditions.

In an earlier paper,117 Temin notes how, while British industrialization started in a variety
of industries, it most prominently began in labor-intensive industries like metallurgy and
textiles where Britain was attempting to gain a comparative advantage over relatively labor-
rich countries like France.118 Whereas Rome seemingly leveraged its slave populations to the
detriment of KM (even when such devices as the steam engine were invented,)119 Britain did
the inverse, leveraging increased KM against lower L.

Also unlike Britain, the Roman consolidation of the ager publicus wasn’t entirely legal;
Pliny mentions a law limiting private holdings to 500 jugera, with one jugerum being the

114
As mentioned by Pelizzo, Kinyondo, and Nwokora, 2018 p. 265, an additional problem is that, as indus-
trialization occurs, most of the population is still in agriculture while agriculture simultaneously becomes
less and less profitable compared to alternatives.
115
I recommend Heblich, Redding, and Voth, Sept. 2022 “Slavery and the British Industrial Revolution."
National Bureau of Economic Research, no. 30451, which goes into extensive detail on the various factors,
including the massive influx of slave labor, which contributed to the Industrial Revolution in Britain.
116
Pliny the Elder, NH 18.35.
117
Temin, 1997 “Two Views of the British Industrial Revolution.” The Journal of Economic History, vol.
57, no. 1, p. 63–82.
118
O’Brien, 1996 “Path Dependency, or Why Britain Became an Industrialized and Urbanized Economy
Long before France.” The Economic History Review, vol. 49, no. 2, p. 213–49 goes into great detail
on the underlying legal and cultural differences which resulted in Britain having a much faster rate of
industrialization than France. Many of the factors attributed to France also apply to the Ancient Roman
Economy.
119
See Heron, 1851, trans. Woodcroft (trans. Woodcroft) p. 72, or Heron Pneumatica section 50. The
device mentioned is an ancient version of the steam engine yet nevertheless wasn’t widely adopted.
 Garfinkel 33

amount of land one man and a yoked animal could effectively plow in a day.120 During
a political upset in the 130s BCE, and later too, this policy was reiterated.121 By Pliny’s
time however the practice was still widespread, stating that all of Africa’s122 latifundia were
owned by six landlords.123 Also unlike Britain, Ancient Rome had a strong moralizing against
wealth consolidation, with Pliny in the same passage stating the latifundia “destroyed Italy"
and citing a section of Vergil124 which states that it is socially undesirable to own large
estates (and therefore derives any economies of scale from them.)

Finally, whereas in Britain increased land productivity (and hence increased n) through
consolidation and administrative economies of scale eventually reinforced by innovation, in
Ancient Rome few innovations followed such reforms (see Section 4.1.) Instead, Rome’s
great economic “boon"–its massive influxes of slave populations–had mainly ceased by the
1st century CE, with the majority of slaves now being born into their status.125 This scarcer
supply likely decreased the rate of manumissions (impacting the growth of the regular labor
pool) as the focus shifted to maintaining current populations. To that end a law from Roman
Egypt states that female slaves would be freed if she produced at least three children.126 This
focus on slave retention also likely had a negative overall impact on aggregate KH , which
negatively impacts gA .

In conclusion, while n grew at times during the Ancient Roman Economy, these were
fundamentally due to administrative changes, not technological ones; n remained low overall
in accordance with Malthusian pressures, based off extant data for other nonindustrialized
economies. In either model this would result in low growth of K or L or gA , with a primary
growth of L resulting from slavery. Even so, while similar conditions led to high gA in

120
Pliny the Elder, NH 18.17.
121
For more information see Plutarch, 2017 Parallel Lives, “Lives of Tiberius and Gaius Gracchus."
122
Likely referring to the province roughly equivalent to modern-day Tunisia.
123
Pliny the Elder, NH 18.35.
124
Vergil’s Georgics, 2008 (trans. Volk, Katharina) 2.412.
125
Temin, 2013c p. 530.
126
Temin, 2004 p. 532.
 Garfinkel 34

Britain in the form of industrialization, this didn’t occur in Ancient Rome for a variety of
reasons. Growth primarily occurred through expansion of the capital and labor stock or
from administrative reforms, not technological innovation. Now that n has been discussed,
we will continue onto s, its impact, and how it failed to contribute to gA over simply growing
the K stock.

3.2 The Data: s

The previous subsection outlined conditions where combined population influx and economic
reorganization resulted in the creation of a fabulously wealthy aristocratic class, one who
would have the necessary savings to invest in new capital and even innovation. Even so,
the evidence suggests that excess funds were primarily used for “horizontal expansion," i.e.
tiling out production without actually increasing A except for administrative economies of
scale.

Before that, why does investment occur? At its simplest, people invest their savings
believing that it will yield returns greater than the initial cost. With that said, what in
Ancient Rome’s nonindustrialized economy is there to invest in? Construction projects,
such as infrastructure or real estate, for one. According to Suetonius,127 Augustus’ conquest
of Egypt greatly lowered interest rates in Rome due to the quantity of looted wealth, enabling
many of Augustus’ and others’128 subsequent building projects, which in turn would increase
the demand for slaves,129 as slaves commonly made up much of the construction workforce.130
In other words, during the early 1st century CE, wealthy families had the perfect opportunity
to invest into large housing complexes and the ownership of slaves.

127
Suetonius, 1914 (trans. Thomson) Life of Augustus 41
128
According to Plutarch Life of Crassus 2.4-5, 3.1, Crassus was one such investor who became rich through
investing into crediting and real-estate management.
129
Finley, 1985 p. 70.
130
Thornton and Thornton, 1990 “The Financial Crisis of A.D. 33: A Keynesian Depression?” The Journal
of Economic History, vol. 50, no. 3, p. 656. See also Livy Ab Urbe Condita (AUC ) 1.57, where he mentions
the grievance of Romans being forced to do construction work they believed was more fitting of slaves.
 Garfinkel 35

The second most obvious investment opportunity is land/agriculture. Even so, the level of
productivity for Roman agriculture remained fairly constant throughout Classical Antiquity
with a grain seed-to-harvest ratio of about 1:4.131 As pointed out by Finley, if productiv-
ity per land area remains constant, the preferred form of investment would be “horizontal
expansion:"132 More land, more slaves, even if y remains constant. Morris however points
out that agricultural investment without economies of scale can result in a Malthusian trap
where geometric population growth outstrips food supply, increasing the demand for even
more cultivation.133 In other words, while investment into agriculture would raise profits for
the landowner, it could create other problems in the long run.

Investment and s however likely remained low overall due to a lack of liquidity in the
Ancient Roman Empire. This is in stark contrast to modern development initiatives, which
are often contingent on massive amount of liquidity and direct investment.134 In a pre-
dominantly agrarian society, few have access to the security and savings to fund substantial
investment, let alone innovative research.135 While the overall level of monetization is un-
clear, it’s believed to be relatively low since the Roman state accepted taxes in kind.136
While Hopkins claims taxes in cash could facilitate higher monetization and investment,137
other scholars believe monetization was localized with few mechanisms for currency to travel
long distances.138 The monetization of rural wages is believed to be irregular, stagnant, and
based on convention rather than market prices.139 Money supplies were likewise limited by

131
Finley, 1985 p. 108, Hopkins, 1980 p. 118.
132
Finley, 1985 p. 116.
133
Morris, 2013 p. 94.
134
Pelizzo, Kinyondo, and Nwokora, 2018 p. 269-270.
135
Erdkamp, 2001 p. 340.
136
Bang, 2007 “Trade and Empire: In Search of Organizing Concepts for the Roman Economy.” Past
Present, no. 195, p. 20.
137
Hopkins, 1980 p. 102.
138
Paolilli, 2008 “Development and Crisis in Ancient Rome: The Role of Mediterranean Trade.” Historical
Social Research / Historische Sozialforschung, vol. 33, no. 4 (126), p. 277, citing Duncan-Jones, 1994 Money
and Government in the Roman Empire, Cambridge University Press, p. 172-179.
139
Howgego, 1992 “The Supply and Use of Money in the Roman World 200 B.C. to A.D. 300.” The Journal
of Roman Studies, vol. 82, p. 26. Temin, 2006 p. 140. Finley, 1985 p. 23.
 Garfinkel 36

the amount of precious metals mined140 and seized in war, with the latter seen as a notable
means of “investment" for expanding the state’s money supplies and power.141

Since their purchasing power is founded on individual demand, taxes and wages in kind
cannot stimulate trade (or investment) as easily as their currency-based kin.142 Even so,
by the metrics above, the greatest holder of liquid assets in the Ancient Roman Economy
would be the state through their ability to print hard currency and use it to fund namely
soldiers, bureaucrats, and laborers (see Section 5.1.)143 As testament to the lack of liquidity
for non-government actors, Tacitus144 states how the reinforcement of an anti-usury law and
2
a requirement for senators to hold of their land in Italy resulted in a Keynesian-esque
3
financial crisis marked by a lack of liquidity only ended by the government’s intervention.145
Low liquidity means low savings, low investment, and low economic growth. This becomes
even worse when realizing that certain types of infrastructure, namely those pertaining to
public health, require constant investment to prevent disaster, much as has become apparent
in Sub-Saharan Africa.146

Finally, in a world of fairly low K per capita, where injections of wealth and capital (in the
form of slaves) comes from military conquest, there becomes a final key form of “investment:"
Corruption. Unlike other forms of investment, corruption doesn’t increase K or grow the
economy, undermines rule of law, and is often ancillary to other crimes. As illuminated by a
a passage of Cicero,147 Caesar spent much of his Gallic spoils bribing people (to eventually
capture the Roman state,) as in a nonindustrialized society control over the levers of power
can often be more lucrative than any form of normal investment.

140
Howgego, 1992 p. 4
141
Thucydides History of the Peloponnesian War 6.24.
142
Hopkins, 1980 p. 103.
143
Thornton and Thornton, 1990 p. 656.
144
See Tacitus, 1873, trans. Alfred John Church et. al. Annals 6.16-17.
145
Thornton and Thornton, 1990 p. 660 and Frank, 1935 “The Financial Crisis of 33 A. D.” The American
Journal of Philology, vol. 56, no. 4, p. 336-337.
146
Pelizzo, Kinyondo, and Nwokora, 2018 p. 279.
147
Cicero, 2001 Letters to Friends 7.13.1.
 Garfinkel 37

3.3 Conclusion

Population growth n and savings rate s are both integral economic growth, whether that
be growth of L and K or increasing gA through increasing LA or the available funds for
investment. Even so, in the Ancient Roman Economy several factors prevent the latter from
occurring, such as a lack of the proper legal or monetary basis for supporting innovation. The
prevalence of subsistence agriculture likewise incentivizes investment being directed towards
growing K with as immediate returns on investment as possible rather than longer-term
growth by growing gA . Between that, the incentivization for corruption, and the lack of
ability to save, there simply isn’t much room left for innovative research.

Similarly, n remained low due to a lack of major advances in agricultural technique, with
the majority of increased food supply either coming from imports or administrative reforms
locally irrespective of their long-term viability. This narrative of A primarily growing through
effective governance likewise pairs with Strabo’s belief in its efficacy as detailed in Section
1.3.

Even so, this section leaves open two primary questions. First, if and when innovations did
occur, what factors stopped them from being readily adopted? Second, what were more of
these cultural or economic factors which prevented excess funds from being directed towards
innovative research? Section 4 will focus on the first question and how it relates to Jones’
notion of ϕ in his model of gA , while Section 5 will focus on the second and how it relates to
λ in the same model.

4 Innovative Synergy (ϕ)

While n and s can contribute to either growth by either “perspiration" (growing L and K ) or
“innovation," increasing A, innovative synergy ϕ in Jones’ equation of gA exclusively works
towards the latter. This section will demonstrate how, while ϕ was present in the Ancient
 Garfinkel 38

Roman Economy, it seemingly had little impact on economic growth, let alone innovation-
led gA . This section will analyze why that was likely the case and relate it to this analysis’
ongoing discussion of the difficulties preventing development from a nonindustrialized to an
industrialized economy.

4.1 The Data: ϕ

There are many ways to indicate ϕ in the historical record: Recording extant innovations,
their adoption, possible loci of innovation, etc. Temin however epitomizes this matter:
According to Temin, one way of looking at “the impact of technological change is to look at
the extent of consumption over subsistence it allowed."148 In other words, ϕ can be partially
intuited by the extent to which the development and spread of novel technologies enables
economic diversification, intensification, or development.

Temin in the aforementioned passage also mentions two examples: Ancient Rome’s mas-
sive infrastructure projects (roads, bridges, aqueducts, city planning, etc.) and the prop-
agation of water mills, wheels, and so on around the empire. While both are decidedly
labor-saving KM which creates greater A, the former is more complicated. While it is true
that Ancient Rome’s infrastructure utilized (as mentioned by Vitruvius)149 a unique form
of hydraulic cement,150 the technology was not the focus. Rather, their successes were con-
tingent on the effective administration and mobilization of Rome’s military and logistical
networks en masse,151 itself a great feat of synergy through economies of scale.

While these infrastructure projects (e.g. the roads themselves) provided economic benefit,
their primary function was always military utility. Construction costs were often passed
148
Temin, 2006 p. 147.
149
Vitruvius Pollio, 1956 (trans. Granger) On Architecture, Loeb classical library, 2.6.
150
New scholarship (see Seymour et al., 2023 “Hot mixing: Mechanistic insights into the durability of ancient
Roman concrete" vol. 9, no. 1, doi 10.1126/sciadv.add1602.) appears to show that this hydraulic cement
had self-repairing properties.
151
Roth, 1999 The Logistics of the Roman Army at War (264 B.C. - A.D.235). Leiden, The Netherlands,
Brill Publishing, p. 214-217.
 Garfinkel 39

onto local cities152 similar to how local agriculture sustained inland legions at detriment to
their own subsistence. This narrative of “synergy through scale" (lowered transaction costs
through unified weight measures, language, political administration, etc. taking precedent
over technological improvement) is likewise repeated with respect to Rome’s Red Sea and
Indian Ocean trade, citing documents such as the Muzuris Papyrus.153

This “agglomeration" narrative is also reflected in sources of the time. In a section of his
agricultural handbook Res Rustica, Varro154 advises to grow flowers and other crops close
to cities to reduce transportation costs, take advantage of volatile urban demands, and to
capitalize on the presence of other seasonally/contractually employed persons like doctors
and artisans, a clear example of firm agglomeration. Thucydides likewise states that, to
his understanding, complexities such as shipping and trade emerged under the protection of
city-states,155 while Fustel de Coulanges states cities emerged to protect and empower local
elites.156 These passages reveal an understanding that, at the time, economic development
emerged from urban centers and their agglomerative synergies, not technological innovation
outright.

That isn’t to say there weren’t innovations or autarkic technological differences which
promoted technological exchange. Cato the Elder, in his handbook on agricultural husbandry
de Agricultura, says Pompeii’s volcanic stone mills were the highest quality and encourages
his readers to buy from them.157 Pliny the Elder cites a variety of different localized pottery
industries, such as Arretium’s,158 whose high demand and prevalence in the archaeological

152
Roth, 1999 p. 215.
153
Manning and Renfrew, 2014 “At the Limits: Long-Distance Trade in the Time of Alexander the Great
and the Hellenistic Kings.” Reconfiguring the Silk Road: New Research on East-West Exchange in Antiquity,
edited by Victor H. Mair and Jane Hickman, University of Pennsylvania Press, p. 10.
154
Varro, RR 1.16.3-4
155
Thucydides History of the Peloponnesian War 1.2, 1.6.
156
Coulanges, 1980 The Ancient City p. 126-134.
157
Mayeske, 1972 “Bakeries, bakers, and bread at Pompeii: A study in social and economic history,"
University of Maryland, p. 12 and Cato, 1935 - 1967 (trans.) On Agriculture 135.2.
158
Pliny the Elder NH 35.160-161.
 Garfinkel 40

record points to their prevalence in or generation of trade.159

More interestingly, however, Pliny also mentions a variety of unique agricultural techno-
logical innovations across the empire, such as a labor-saving form of plow in Raetia (NH
18.172,) a rudimentary combine harvester (NH 18.296,) cost-saving viticulture practices de-
veloped to offset volatile labor costs (NH 17.213-5,) and a type of Gallic scythe able to more
efficiently harvest grain than Italian ones (NH 18.261.) In each of these examples, however,
the narrative is never how, by widely adopting these innovations, land could be “freed up"
to cultivate cash crops or food for the army. Instead these technological innovations are
modeled by Pliny as modifiers on the natural productivity of the region, in the mind that
“Region X would normally only produce m units of grain per year, but due to a peculiar
invention of the locals the region instead produces more, making it roughly equivalent to
Region Y instead."

While there are examples of widely-adopted technologies resulting in economic growth,

such as the lateen sail,160 they are rare and often based off immediate needs–such as the
lateen sail’s lower relative cost–rather than an actual increase in overall efficiency. While
this dips into the concept of innovative congestion (λ) (see Section 5,) another passage of
Pliny helps explore two cultural reasons as to why innovative synergy remained so low. In a
passage on agricultural grafting,161 Pliny states that the field has already reached its zenith
(pervenit ad columen,) that all discoveries had already been made (expertis cuncta,) that
seemingly all species had been discovered (nullum certe pomum novum diu iam invenitur,)
and that any further experimentation would defy the cosmic order (neque omni insitu misceri
fas est.)

In short, we have several biases directly lowering ϕ. First is the belief that all inventions,
159
For example, see Humphries, Dec. 2017 “The Internal Organization of the Arretine Terra Sigillata In-
dustry: Problems of Evidence and Interpretation.” The Journal of Roman Studies, vol. 87, pp. 111–55.
160
Roberts, 2011 “ “The Hellenistic Business Environment.” The Origins of Business, Money, and Markets,
Columbia University Press, p. 108.
161
Pliny the Elder NH 15.57.
 Garfinkel 41

to use Jones’ terminology, come from a limited stock of possible ideas; beyond a certain point
investment into R&D is pointless because new ideas cannot physically be made.162 As will
be elaborated in Section 5.1, Pliny’s belief likely derives from his own ideological context,
whose beliefs such as the Theory of Forms is conducive to a theorized limit on potential
innovation. Obligational respect towards the natural order (fas) would also negatively impact
ϕ, especially in a culture where deviation from the natural order invited divine wrath.163

Beyond these cultural factors, there were other built-in inefficiencies decreasing ϕ, namely
a lack of information symmetry or centralized repositories to track and record data or inno-
vations. While certain sites did exist at Alexandria or Rome, they were exclusive and hard
to access.164 Like today, innovative research was also primarily focused towards medicine,
military tech, and agriculture,165 as then as now these greatly and immediately benefit those
funding the research, i.e. the state.

The latter two fields, however, also point at an underlying truth about the spread and
utility of innovations. Innovations and institutional reforms only work when they prop-
erly address the population’s needs, such as in the case of many international programs in
Africa.166 In a technologically-poor society such as Ancient Rome, the same end which in-
novation creates could often also be attained through increasing L or the “good governance"
mentioned by Strabo.167 Once again, perspiration over innovation.

Rome’s economic advantage furthermore didn’t come from superior technology: There

162
Modern worldviews, on the other hand, dismiss this notion, coming from a scientific context which
axiomatically holds technological progression to be theoretically limitless.
163
Seneca the Younger Theyestes 110-121 gives several examples of how inversions of the natural order
(notably killing one’s family,) is reflected in natural disruptions such as the sun refusing to rise and rivers
refusing to flow.
164
The Library of Alexandria, for instance, Strabo mentions (Geography 17.1.8) as explicitly being part of
the royal palace, presided over by a priest, and altogether run similar to a temple complex than a public
library.
165
Roberts, 2011 p. 111.
166
Matthews, 2004 “Post-Development Theory and the Question of Alternatives: A View from Africa.”
Third World Quarterly, vol. 25, no. 2, p. 376.
167
Strabo Geography 17.1.13.
 Garfinkel 42

isn’t anything terribly impressive about the pilum, lorica segmentata or other uniquely Ro-
man innovations. Rome’s power came from its massive labor pool and its incredible admin-
istrative ability to utilize it at scale for large projects.168 The Roman aqueduct is in of itself
an example of this concept: A foreign design, built at scale and in great quantity, to achieve
a goal more efficiently than otherwise.

4.2 Conclusion

While innovations and innovative synergies did certainly exist in the Ancient Roman Econ-
omy, their actual effect of augmenting A is, as before, outshone by two other factors: Output
growth through expanding the K and L, and greater factor utility created by effective ad-
ministration and logistics. As is showcased by Rome’s vast infrastructure projects, their
economic growth emerged from the replication of extant technologies at scale and wiring
them into an increasingly complex network of synergies.

Unlike industrialized economies, it seems Ancient Rome lacked the internal mechanisms
to effectively promote innovative investment and share its fruits. While this might have
partially emerged from underlying social conditions (as will be discussed in Section 5.1) or a
perceived lack of global technological progression, it also likely emerged from Ancient Rome’s
poor IPRs and over-utilization of slave labor. In much the same way as modern developing
nations are over-reliant on one particular technology or industry, Rome’s reliance on slave
labor probably stifled the perceived need to create new technologies or spread existing ones.

Localized optimizations or innovations were undoubtedly made, such as Pliny’s mentioned

agrarian innovations or the wider controlled breeding of livestock,169 these seemingly are
exceptions and nevertheless limited in scope or application. This section has however only
partially explained the various cultural factors which limited innovation-led gA in the Ancient
168
Peter Garnsey, 1983 p. 130-131.
169
MacKinnon, 2010 “Cattle ‘breed’ Variation and Improvement in Roman Italy: Connecting the Zooar-
chaeological and Ancient Textual Evidence.” World Archaeology vol. 42, no. 1, p. 56-58, 63, citing Varro
RR 2.5.9. and Pliny the Elder NH 8.167.
 Garfinkel 43

Roman Economy. The next section will elaborate on this topic.

5 Innovative Congestion (λ)

This section will elaborate on the various factors controlling innovative congestion λ in the
Ancient Roman Economy and how, even though it limited gA , still left room for output
growth caused by an increase of the capital and labor stocks. This section furthermore
argues that, while numerous factors greatly impeded the rate of technological innovation in
the Ancient Roman Economy, these factors are by no means unique to it, rather features
that would pervade many technologically poor or otherwise nonindustrialized economies.

5.1 The Data: λ

To determine λ’s rather detrimental effect on gA , a baseline must first be established. At

face value, one of Ancient Rome’s largest boosts for potential innovation was its sociopolit-
ical unification, as unified governments, laws, languages, etc. lowers congestion,170 allowing
for the faster spread of ideas and people. On the other hand, as an agrarian society with
a relatively low technological level and high travel/transaction costs, Ancient Rome lacked
many of the modern economic institutions associated with low innovative congestion, namely
safe and reliable access to low-interest loans. While banks did exist, they were often associ-
ated with temples171 (making them in effect government-owned) or aristocratic holdings and
based on personal acquaintance.172

Beyond just banks, businesses in general were often run by freedpersons as fronts for their
patrons by the cultural belief that non-agricultural work was unseemly; according to one of

170
Hollander, 2017 p. 6.
171
Hollander, 2017 p. 17.
172
Silver, 2011 “Finding the Roman Empire’s disappeared deposit bankers." Historia: Zeitschrift Für Alte
Geschichte, vol. 60, no. 3, p. 303, citing Reden, 2002 “Money in the ancient economy: A survey of recent
research" Klio, vol. 84, no. 1, p. 148.
 Garfinkel 44

Cicero’s writings, the acceptance of wages was similarly “itself the mark of slavery."173 Worse
still, due to greater risk interest rates were both high and volatile.174 Ceteris paribus, higher
interest rates decrease the amount of loans given, which lowers quantities invested.175

As today, the actor with the greatest ability to fund innovative investment would be
the state via their access to taxes and seigniorage.176 Beyond investment, however, the
state seemingly needed to balance its desire for expansion against security threats, real
or perceived. According to a passage of Philostratus,177 Domitian at one point restricted
investment into viticulture in Asia due to a perception that the wine would prompt people to
conspire against the state. Regardless of this passage’s veracity, it points to possible internal
biases which could put economic growth at odds with security, a perceived dilemma in many
authoritarian, developing countries today.

Even so, the question therefore arises: Did the Ancient Roman state, or other Ancient
Mediterranean governments more broadly, invest into innovation? If so, what did they invest
into? If not, what did they do instead, and why? While reports of designated innovators or
engineers existed, such as in a passage of Diodorus Siculus about a specialized “think tank"
assembled by Dionysius I of Syracuse, these are the exception rather than the rule.178 In
fact, the passage of Diodorus states that Dionysius I’s think-tank was explicitly focused on
developing military technologies.

Finley gives another example (Ptolemaic Egypt,)179 stating that despite improvement to
173
Cicero, On Occupations 1.150, cited by Hollander, 2017 p. 18. While Hollander’s conclusion is correct,
he misrepresents Cicero’s quote, as in the same passage Cicero states that wage labor, like slave labor, doesn’t
respect the personhood/creative input of the worker.
174
For example, another Cicero passage (Cicero, 1979 (trans. Roy) Letters to Atticus 4.15.7) discusses the
constant fluctuation of interest rates in Late Republican Rome.
175
Maximums on interest rates were only established, much less standardized, in the 1st century BCE at a
(rather high) rate of 1% per month or 12% per year (see Plutarch, Life of Lucullus 20.)
176
The Roman state taxes many things; imports, exports, manumissions on slaves, etc. At times they also
used networks of “publicani " to collect taxes (see Magie, 1950a “Centralization and prosperity under the
Flavians." Roman Rule in Asia Minor, Princeton University Press, p. 567.)
177
Levick, 1982 p. 69 citing Philostratus Vit. Soph. 520.
178
Veen, 2010 p. 5 citing Diodorus, 1935 (trans. Oldfather)) Library of History 14.41.3-4, 42.1.
179
Finley, 1985 p. 138.
 Garfinkel 45

the existing irrigation systems, the introduction of new crops, and various administrative
reforms (all meant to increase state revenue by simply “giving Egypt the advantages of
already existing Greek technologies and Greek processes,") the actual innovations to come
out of Alexandria’s Museion were “military technology...and mechanical toys."180 For a more
Roman example, while Caesar does briefly mention military engineers in his commentaries,181
they are in an explicitly non-innovative role. What is more, the Roman army, despite its
heavy use of munitions and impromptu fortifications (see Caesar DBG 7.75-89,) “had no
separate engineering units, or sub-units."182

From these examples a clear picture emerges: Technological innovations, when they did
occur through state funding, were limited in scope, while most “engineers" in the modern
sense were instead devoted to creating new instances of existing technologies, notably public
works. While numerous projects were constructed, many–such as the theaters at Pompeii,
Puteoli, and various baths around Rome–were in essence built by elites for elites,183 184
with–
yet again–mass reception a secondary concern. Recorded examples of state investment also
include a federal management of papyrus supply and distribution during a shortage185 and
granting shipping insurance to increase grain imports to Rome during a famine.186

In these instances emerges the trend of short-term policies meant to address immediate
concerns (shortages, infrastructure demands from the wealthy, etc.) rather than a preference
for long-term innovative investment. While this strategy may result from the high discount

180
According to Athenaeus (see Athenaeus and Olson, 2019 (author and translator) Deipnosophistae 5.198f-
199a,) one such “toy" was a clockwork automaton made for Ptolemy II Philadelphus able to sit, stand, and
pour libations.
181
Caesar Commentaries on the Gallic War 5.11 (Gilliver, 2004 Caesar’s Gallic Wars 58-50 BC. Taylor
and Francis used.)
182
Roth, 1999 p. 215.
183
Koloski-Ostrow, 2015 “Understanding Roman sanitation from archaeology: Toilets, sewers, and water
system." The Archaeology of Sanitation in Roman Italy: Toilets, Sewers, and Water Systems, University of
North Carolina Press, p. 53.
184
Humphries, Dec. 2017 “Roman Senators and Absent Emperors in Late Antiquity”. Acta Ad Archaeolo-
giam Et Artium Historiam Pertinentia, vol. 17, no. 3, p. 29.
185
Pliny the Elder NH 13.89.
186
Suetonius Life of Claudius 18.2.
 Garfinkel 46

(interest) rates of the times187 incentivizing short-terms loans, they also relate to Section
2.3’s discussion of energy capture. Such limitations would incentivize large infrastructure
projects requiring little KH or KM to build, yet would grant more efficient energy usage and
immediate returns on investment. These projects would also naturally crowd out riskier,
longer-term endeavors into innovative research.

Innovations–agricultural ones in particular–also often require localization to address spe-

cific environments,188 similar to how modern development projects in Africa have needed to
take into account local cultural values rather than imposing external ones a priore.189 Due to
these concerns, it often preferable to invest in already proven infrastructure projects rather
than adapting novel ones despite evident gains from potential agricultural innovations.

The same passage cited above also hints at a λ-based “poverty trap," where innovations do
not become viable until sufficient economic complexity has already been reached,190 pointing
at traction-based rotary mills as an example. Despite being well-known since before the
4th century BCE, these demonstrably more effective mills weren’t widely adopted in Roman
Italy (replacing the far less efficient hand-quern ) until the mid 2nd century BCE.191 Even
though rotary mills could likewise help develop wine, oil, and livestock industries,192 they
nevertheless took centuries to be adopted due to only being efficient at scale and once a
certain complexity index had already been reached.

Even so, none of this answers the fundamental questions: Why were superior technologies
at large adopted despite their clear advantage? Why did the Roman state (among others)

187
Monson, 2007 p. 252-254.
188
Veen, 2010 p. 7. As pointed out by Mayeske (see Mayeske, 1972 “Bakeries, bakers, and bread at Pompeii:
A study in social and economic history" p. 33-37) and Spurr (see Spurr, 1983 “The Cultivation of Millet in
Roman Italy.” Papers of the British School at Rome, vol. 51, p. 1-2) this extended to the grain variation
across the Mediterranean Basin, so different farming practices would likewise have to be adopted.
189
Matthews, 2004 p. 380.
190
Veen, 2010 p. 7
191
Mayeske, 1972 p. 4-5.
192
On Roman livestock industries, see MacKinnon, 2010 p. 56-58, citing sections of Varro, Columella, etc.
See also Pliny the Elder NH 8.167.
 Garfinkel 47

devote so little to true innovative research? Environmental factors can’t explain it on its
own. What were the cultural biases which resulted in high levels of innovative congestion in
the Ancient Roman Economy?

Despite the unprecedented complexities of their times, the Roman literary tradition, from
Homer to Cato to Cicero to Pliny to Cassius Dio to even Symmachus in the Late 4th Century
CE, contains a deep belief in global stagnation/decline naturally evolving into a de-emphasis
or disbelief in the viability of technological innovation.193 While these passages are several
among many, three more in particular point to a prevailing Roman belief in technological
innovation’s impotence.

The first comes from Petronius’ Satyricon.194 In it, a story is related about how a craftper-
son once approached the Roman Emperor with a vial of unbreakable glass whose imperfec-
tions could be hammered out as if bronze. After admitting that only they knew the material’s
secret, the Caesar had him killed as to not devalue gold (“aurum pro luto haberemus.") While
the story is undoubtedly fictitious, it highlights several perceptions about innovation at the
time.

First, people believed that innovations, even if useful, could not be adopted lest they
destabilize previous industries (e.g. slavery.) Second, preserving the elites’ fortunes took
precedent over overall output growth. Third, this passage hints at a mistrust of Rome’s
kleptocratic government,195 elaborating on previous’ sections discussion on rule of law and
the state’s contemporary habit of expropriating numerous elites’ holdings,196 lowering overall
193
In order: Homer, 1961 - 1951 (trans. Lattimore) Iliad 12.445-449 references modern degeneracy com-
pared to ancient, heroic strength. Cato DAC 1. begins with moralizing about the superior morals of the
ancients. Cicero In Catilinam 1.2 likewise criticizes a believed moral decline. Pliny’s aforementioned pas-
sages on grafting (NH 15.57) among others reinforces a belief in limited possible knowledge, most of which
has been already found. Cassius Dio, 1914 (trans. Cary) Roman History 72.36.4 likewise frames history as
decline from gold to “iron and rust." Symmachus’ letter 1.4 (see Salzman and Roberts, 2011 The Letters
of Symmachus: Book 1, Society of Biblical Literature) likewise refers to the lack of virtue in modern times
compared to older ones.
194
See Petronius, 1913 (trans.) Satyricon 51.
195
Scheidel, 2007 p. 332.
196
See Tacitus Annals 4.19-20 and Suetonius Life of Nero 32, 34.5 for some examples.
 Garfinkel 48

confidence in potential investment.

The second passage, Suetonius Vesp. 18.1, is one where Vespasian opts for a high labor,
low capital solution for transporting building materials by the belief that he has an obligation
to provide gainful employment for his people. While this anecdote is likely also fictitious, it
nevertheless points to a belief that must have been prevalent enough for Suetonius to endorse
it: High employment and labor participation took precedent over overall growth. This is
mirrored by Cato On Agriculture 2.7 where he endorses selling old and sick slaves rather
than nursing them back to health, further putting an emphasis on keeping workers occupied,
even if not necessarily with meaningful work. The final passage comes from Vergil’s Aeneid,
a landmark of Roman literature:197

excudent alii spirantia mollius aera Let others make bronze statues that can breathe,
(credo equidem), vivos ducent de marmore vultus, Carve marble to become a human face,
orabunt causas melius, caelique meatus
Be better orators, and chart the heavens,
describent radio et surgentia sidera dicent:
Astronomers predicting when stars rise.
tu regere imperio populos, Romane, memento
(hae tibi erunt artes), pacique imponere morem, Remember, Roman, your task, ruling nations,

parcere subiectis et debellare superbos. Your artistry to set the terms for peace,

Spare those who yield, and beat resisters down.

This passage normatively states that Romans as a whole should be wholly uninterested
in art, science, and anything that isn’t related to dominating other peoples. In stating
“these [i.e. dominion] are your arts" (“hae tibi erunt artes,") Vergil implies that all the other
mentioned professions, from innovative engineering to artistry to oratory to astronomy, are
inherently un-Roman.198
197
See Vergil Aeneid 6.847-853, cited by Roberts, 2011 p. 112.
198
Shipping (see Peter Garnsey, 1983 p. 136) was also stigmatized, as it prevented the capitan from
engaging in local politics.
 Garfinkel 49

None of this is to say that the Romans were uninnovative. The opposite is quite true,
whether it be the Romans ready adoption of foreign military, administrative, and infrastruc-
ture technologies or the continued evolution of Roman civic and military structure through-
out the centuries. Time and again the Romans showed themselves as highly adaptable and
willing to adopt new ideas. This “cleverness" (ingenium) or resourcefulness is also quite
similar to the Indian concept of “jugaad " or the Russian “smekalka," even if–while each con-
cept connotes clever resourcefulness–they often refer to corner-cutting or ad hoc solutions to
unaddressed systemic problems.199

No, as put by Garnsey, in his, Hopkins’, and Whittaker’s 1983 anthology Trade in the
Ancient Economy, it seems that the Romans ultimately lacked an internal cultural push
towards innovation, that “[there was] no moral reinforcement for productive investment, nor
for profit maximization...high status involved competitive and ostentatious expenditure."200

The “competitive and ostentatious expenditure" refers to the “consumer city" theory of
the Ancient Roman Economy. In short, the theory states that (slavery notwithstanding,) a
difference in mentality among ancient elites (compared to their medieval or modern coun-
terparts,) led to cities in the Ancient Roman Economy becoming centers of conspicuous
consumption (luxuria,) not production, and that all production inside urban centers was
merely the spin-off of rural production.201

This “consumer city" theory is however not well-supported by the historical record; recent
literature has also pointed out how the theory plays into a wider narrative of homogeniz-
ing nonindustrialized economies to justify their exploitation by industrial or post-industrial
nations.202 While conspicuous consumption in the Ancient Roman Economy did exist, and
was talked about at length, much of the time it was talked about in a moral lens, not an
199
“Innovation Culture”, 2011 “Innovation Culture.” Research Technology Management, vol. 54, no. 4, p.
59–60.
200
Peter Garnsey, 1983 p. xiii-xiv.
201
Erdkamp, 2001 p. 333-335, 339.
202
Hobson, 2014 p. 14.
 Garfinkel 50

economic one, even if the two were connected.203 The “spin-off" idea also disregards urban
collocation overall: Urban centers are habitats built around markets and manufacturing sites
for the processing of agricultural goods, by definition.

5.2 Conclusion

Across these numerous examples, a singular narrative emerges: Innovative congestion in the
Ancient Roman Economy was incredibly detrimental to the emergence of innovation-led gA
due to a combination of cultural and administrative factors which disincentivized innovative
investment. Instead, there prevailed a reliance on the known: Mass infrastructure projects
and slavery, expansion of K and L. The role of slavery in stifling innovative investment can
not be overstated, especially when the recorded areas of state-funded research (primarily
medical and military technologies) are two of the major fields where additional labor inputs
cannot solve a problem.

Besides for these factors, and a lack of intellectual property rights–especially versus a
kleptokratic state– and a general growth agnosticism seemed to permeate the entire society.
To quote the historian Willem Jongman:204

[W]hy were the Romans not as prosperous as we are...Were they poor because
they or their rulers failed to maximise their incomes, because they lacked the
economically rational desire to improve their lot, or were they poor because they
did not have any alternatives?

If this section has suggested anything, it’s that innovation is a process that can fail to
manifest for many kinds of reasons. Even if there were sufficient personnel and funding for
203
While conspicuous consumption can include hosting games and festivals, in this context the more apt
term would be luxuria, extravagance. Numerous sources (e.g. Plutarch, Life of Cato the Elder 18.2-3, Life
of Lucullus 7, 14, Sallust Catilinian War 5, 11) discuss luxuria’s effects, almost always characterizing it
as damaging to Roman national character by the logic that, when the time came, those infected by their
extravagances would not be able to put them aside for the good of the nation.
204
Jongman, Jan. 2006 (revised 2007) The rise and fall of the roman economy: population, rents and
entitlement p. 237–238, cited by Hobson, 2014 p. 14.
 Garfinkel 51

innovative research, without the belief that their research would bear fruit or that they could
profit off it205 given current technological or administrative limits, no innovation would easily
take place. And these problems are by no means unique to Ancient Rome.

Solving these problems, a general disbelief in technological innovation, mistrust of a gov-

ernment accustomed to expropriating wealth, etc. requires long-term cultural shifts which
simply adding new technologies into a society won’t solve. The steam engine was invented
in Classical Antiquity,206 but failed to garner attention due to the prevalence of substitute
goods (slavery) and a lack of the proper institutions to capitalize on it. If anything can be
learned from this example it’s that inventions cannot increase A by themselves.

6 Conclusion

6.1 Review of Findings

For a nonindustrialized economy such as Ancient Rome’s, the prospect of “endogenous

growth," output growth caused by innovation-led gA , is incredibly difficult. Industrialized
societies chiefly maintain economic growth through the creation of profit-generating innova-
tions with this potential protected through IPRs, a state that upholds the rule of law, and
the diffusion of this technology throughout society coinciding with a strong reception to “la
Technique." Nonindustrialized societies such as Ancient Rome, on the other hand, are for the
most part only able to economically grow through the increasing of K and L, perspiration
instead of innovation, as they lack the proper institutions to do otherwise. While the reasons
for this are numerous and interlocking, they can for the most part be narrowed down to these
core factors which transcend Ancient Rome and apply to nonindustrialized more broadly:

1. High rates of subsistence agriculture limiting LA .

205
Jones, 1995 p. 759 explicitly mentions endogenous growth as “technological change that results from the
research and development efforts of profit-maximizing agents."
206
See Heron, 1851, trans. Woodcroft p. 72, or Heron Pneumatica section 50.
 Garfinkel 52

2. High information travel costs and information asymmetry.

3. Systematic under-utilization of the labor pool.

4. Over-reliance on KH over more durable inputs.

5. Cultural disbelief in the potential for innovation.

6. Over-reliance on specific forms of capital, labor, or industries.

7. Lack of potential for innovation due to poor monetization.

8. Perception that innovations would economically challenge existing power structures.

9. Poor protections for the profit-generating ability of innovations.

10. Governmental institutions’ unwillingness to protect innovators.

Many of these factors derive from the same two core principles: Low agricultural output and
ineffective governance. Even so, while some of these points can be addressed by the injection
of new technologies into an economy, many instead need to be addressed by systemic political,
social, and economic reorganization. Even when done correctly, such reorganizations often
occur at the detriment to the nonindustrialized economy’s original culture, as by the model
of “la Technique" economic development often entails adopting the norms and cultures of
developed nations.

Escaping the “trap" of a technologically-poor economic index is an incredibly complex,

time-consuming process which’s effectiveness is contingent on a variety of socioeconomic
factors. Even in regions like postcolonial Africa, although several countries gained their
independence concurrently, they industrialized at rapidly different paces.207 Similar to the
Roman example of their own colonial efforts, however, success stories are often contingent
on large amounts of foreign investment to bring a nonindustrialized economy up to the
207
Pelizzo, Kinyondo, and Nwokora, 2018 p. 264.
 Garfinkel 53

level of whatever outside force helps them. Although the Romans possessed an extensive
ability to bring others up to their own economic level, numerous internal factors (namely
slavery) prevented them from developing their own economy–not even mentioning how their
developmental efforts often necessitated the obliteration of local cultures.

6.2 Modern Implications

If the examples of Ancient Rome and modern developmental projects suggest anything,
it’s that their effectiveness is based on the full utilization of many talented and coordinated
specialists who can establish new socioeconomic institutions and lay the groundwork for their
perpetuation, with particular focus on a legal system able to protect the profit-generating
capabilities of novel technologies.

Even so, beyond the use of talented specialists or the reform of legal/economic institutions,
there are two main factors which must be addressed. First, the labor pool must be freed
to research innovations and conduct complex business; to that end agricultural productivity
must be increased and more sustainable policies promoted.208 Second, there must be wide
access to funds to invest, as without ready access to loans or liquid capital development often
stagnates, even if those funds are coming from external countries.209

Underneath both of these is a deeper truth. Whether with regards to innovation or growth
of K and L, the source of an economy’s power is its population; to succeed, therefore, an
economy must enable its population to be as able to work as effectively as possible, rather
than denying them government assistance on the basis of outdated beliefs or theories.210

As has been shown throughout this analysis, the Ancient Roman Economy lacked all these
208
Pelizzo, Kinyondo, and Nwokora, 2018 p. 268.
209
Pelizzo, Kinyondo, and Nwokora, 2018 p. 279.
210
Bielefeld, 2016 “Neoliberalism and the Return of the Guardian State: Micromanaging Indigenous Peoples
in a New Chapter of Colonial Governance.” Engaging Indigenous Economy: Debating Diverse Approaches,
edited by Will Sanders, vol. 35, ANU Press, p. 157 gives such an example, focusing on the ill treatment of
Australian Aboriginals from their government.
 Garfinkel 54

things. They systematically underutilized their labor pool. They didn’t manage to ensure
food security for large sections of the population. They forever remained an oligarchical
government with very little ability for everyday people, let alone non-citizens, to give input
let, redress grievances, or protect their economic interests. There was little basis for their
government’s existence at all save for the continued “idea" of Rome–and although it evolved
over the centuries, Western Rome nevertheless fell in 476 when upholding that fiction no
longer became advantageous for its elites. This idea nevertheless remained strong enough to
keep its state and economy united for over a millennium.

Unfortunately, such things like national “ideas" or other bases for economic or political
cooperation and consolidation, as well as their policies, cannot be simply transplanted to
other contexts. Like a building it must be built from local materials and made in consid-
eration of local needs. And even so, to change a nation you inevitably need to change the
minds of its people and build trust in whatever new system is being created.

6.3 Avenues of Future Study

Since this thesis has covered many subjects in moderate depth, immediate avenues for future
study include further depth on any of the subjects covered. The first, main one is slavery’s
overall detriment to long-term growth, taking a deep dive into the textual and archaeological
record to find specific examples of its impact. The second would be a further look at the
innovations created under Roman control of the Mediterranean Basin; in short, a full tally
of what innovations are mentioned, where they emerged, how they emerged, where they
survive in the archaeological record, and how widely they were adopted. The third would
be a closer look at the relationship between Ancient Rome’s worldview and Ellul’s concept
of “la Technique," specifically as it relates to competition and innovation.

While this analysis has taken a decidedly history-focused stance, I also would want to
preform a hard data analysis relating to the Ancient Roman Economy. While many ques-
 Garfinkel 55

tions are difficult for want of a dataset, one feasible analysis would be looking at how the
presence of Roman infrastructure (roads, rivers, etc.) affects the distribution of coins in the
archaeological record. Another option would be to look at assemblages from specific contexts
(Caesarea, Rome, Athens, etc.) and chart where their material culture is sourced across the
centuries, using that to reconstruct the volume of trade across certain routes.

While not as data driven, another analysis could be an examination of Rome’s legal codes
and how they helped (or hindered) the growth of business and industry, then comparing that
to the legal codes of other nonindustrialized economies. Rome’s economic situation could be
further compared to contemporary or historically developing nations as well. Finally, this
analysis hardly touched on the role of private businesses in promoting trade and economic
activity; their stories should be told too.

In many ways this analysis can serve as a jumping-off point for future explorations into
economic growth and development inside the Ancient Mediterranean, analyses that will
hopefully bridge the gap which Classics and Economics together have been trying to close
since even before Adam Smith. Too often have these sorts of analyses been preformed by
Classicists who know nothing of Economics, or Economists who know nothing of Classics.
To fully understand the economic picture of Ancient Rome, and how that relate to the
developing economies of today and yesteryear, an understanding of both must be applied.

6.4 Acknowledgements

This thesis underwent many revisions, starting as a model built from first principles and
ending here. I thank my advisors, Dr. George Hall, Dr. Geoff Clarke, and Dr. Joel
Christensen. I also thank the numerous researchers on whose works my own stands: Drs.
Peter Temin, Charles Jones, Paul Romer, Robert Solow, and Moses Finley. I also thank
Brandeis University’s faculty, students, and all who helped me this far. Gratias tibi ago.
 Garfinkel 56

Works Cited

Articles
Bang, Peter Fibiger (2007). “Trade and Empire: In Search of Organizing Concepts for the
Roman Economy”. In: Past Present 195, 3–54. issn: 00312746, 1477464X. http://www.
jstor.org/stable/25096668 visited on 05/19/2022.
Duncan-Jones, Richard (2019). “The Antonine Plague Revisited”. In: Arctos 52, 41–72.
https://journal.fi/arctos/article/view/84955 visited on 05/18/2022.
Erdkamp, Paul (1999). “Agriculture, Underemployment, and the Cost of Rural Labour in
the Roman World”. In: The Classical Quarterly 49.2, 556–572. issn: 00098388, 14716844.
http://www.jstor.org/stable/639879 visited on 02/19/2023.
Erdkamp, Paul P. M. (2001). “Beyond the Limits of the ’Consumer City’. A Model of the Ur-
ban and Rural Economy in the Roman World”. In: Historia: Zeitschrift für Alte Geschichte
50.3, 332–356. issn: 00182311. http : / / www . jstor . org / stable / 4436621 visited on
05/18/2022.
Fitzpatrick, Matthew P. (2011). “Provincializing Rome: The Indian Ocean Trade Network
and Roman Imperialism”. In: Journal of World History 22.1, 27–54. issn: 10456007,
15278050. http://www.jstor.org/stable/23011677 visited on 05/19/2022.
Frank, Tenney (1935). “The Financial Crisis of 33 A. D.” In: The American Journal of
Philology 56.4, 336–341. issn: 00029475, 10863168. http://www.jstor.org/stable/
289972 visited on 06/07/2022.
Geraghty, Ryan M. (2007). “The Impact of Globalization in the Roman Empire, 200 BC-AD
100”. In: The Journal of Economic History 67.4, 1036–1061. issn: 00220507, 14716372.
http://www.jstor.org/stable/40056408 visited on 05/18/2022.
Gills, Barry K. and Andre Gunder Frank (1992). “World System Cycles, Crises, and Hege-
monial Shifts, 1700 BC to 1700 AD”. In: Review (Fernand Braudel Center) 15.4, 621–
 Garfinkel 57

687. issn: 01479032, 2327445X. http://www.jstor.org/stable/40241240 visited on

05/30/2022.
Heblich, Stephan, Stephen J Redding, and Hans-Joachim Voth (Sept. 2022). “Slavery and
the British Industrial Revolution”. In: Working Paper Series 30451. https://doi.org/
10.3386/w30451. http://www.nber.org/papers/w30451.
Hopkins, Keith (1980). “Taxes and Trade in the Roman Empire (200 B.C.-A.D. 400)”. In:
The Journal of Roman Studies 70, 101–125. issn: 00754358. http://www.jstor.org/
stable/299558 visited on 05/18/2022.
Howgego, Christopher (1992). “The Supply and Use of Money in the Roman World 200
B.C. to A.D. 300”. In: The Journal of Roman Studies 82, 1–31. issn: 00754358. http:
//www.jstor.org/stable/301282 visited on 09/12/2022.
Humphries, Mark (Dec. 2017). “Roman Senators and Absent Emperors in Late Antiquity”.
In: Acta ad archaeologiam et artium historiam pertinentia 17.3 N.S. 27–46. https://
doi.org/10.5617/acta.5688. https://journals.uio.no/acta/article/view/5688.
“Innovation Culture” (2011). In: Research Technology Management 54.4, 59–60. issn: 08956308,
19300166. https://www.jstor.org/stable/26504921 visited on 03/18/2023.
Janjua, Pervez Zamurrad, Ghulam Samad, and Nazakat Ullah (2019). “Intellectual Property
Rights (IPRs) and Economic Growth in Pakistan”. In: The Pakistan Development Review
58.3, pp. 225–237. issn: 00309729. https://www.jstor.org/stable/27125024 visited
on 03/18/2023.
Jones, Charles I. (1995). “R D-Based Models of Economic Growth”. In: Journal of Political
Economy 103.4, 759–784. issn: 00223808, 1537534X. http://www.jstor.org/stable/
2138581 visited on 02/13/2023.
Katsari, Constantina (2003). “The Organisation of Roman Mints during the Third Century
CE: The View from the Eastern Provinces”. In: Classics Ireland 10, 27–53. issn: 07919417.
http://www.jstor.org/stable/25624494 visited on 06/08/2022.
 Garfinkel 58

Levick, Barbara (1982). “Domitian and the Provinces”. In: Latomus 41.1, 50–73. issn: 00238856,
22944427. http://www.jstor.org/stable/41532326 visited on 10/10/2022.
MacKinnon, Michael (2010). “Cattle ’breed’ variation and improvement in Roman Italy:
connecting the zooarchaeological and ancient textual evidence”. In: World Archaeology
42.1, 55–73. issn: 00438243, 14701375. http : / / www . jstor . org / stable / 25679727
visited on 06/10/2022.
Malmendier, Ulrike (2009). “Law and Finance "at the Origin"”. In: Journal of Economic
Literature 47.4, 1076–1108. issn: 00220515. http://www.jstor.org/stable/40651533
visited on 05/18/2022.
Marcella, Gabriel (2009). “Democratic Governance and the Rule of Law: Lessons from Colom-
bia”. In: http://www.jstor.org/stable/resrep11338 visited on 03/18/2023.
Matthews, Sally (2004). “Post-Development Theory and the Question of Alternatives: A
View from Africa”. In: Third World Quarterly 25.2, 373–384. issn: 01436597, 13602241.
http://www.jstor.org/stable/3993687 visited on 04/01/2023.
Menell, Peter S. (2011). “Governance of Intellectual Resources and Disintegration of Intellec-
tual Property in the Digital Age”. In: Berkeley Technology Law Journal 26.4, 1523–1559.
issn: 10863818. http://www.jstor.org/stable/24118665 visited on 03/18/2023.
Merges, Robert P. (1995). “The Economic Impact of Intellectual Property Rights: An Overview
and Guide”. In: Journal of Cultural Economics 19.2, 103–117. issn: 08852545, 15736997.
http://www.jstor.org/stable/41810541 visited on 03/18/2023.
Monson, Andrew (2007). “Rule and Revenue in Egypt and Rome: Political Stability and
Fiscal Institutions”. In: Historical Social Research / Historische Sozialforschung 32.4
(122), 252–274. issn: 01726404. http : / / www . jstor . org / stable / 20762249 visited
on 06/02/2022.
O’Brien, Patrick Karl (1996). “Path Dependency, or Why Britain Became an Industrialized
and Urbanized Economy Long before France”. In: The Economic History Review 49.2,
 Garfinkel 59

213–249. issn: 00130117, 14680289. http://www.jstor.org/stable/2597914 visited on

03/18/2023.
Paolilli, Antonio Luigi (2008). “Development and Crisis in Ancient Rome: the Role of Mediter-
ranean Trade”. In: Historical Social Research / Historische Sozialforschung 33.4 (126),
274–288. issn: 01726404. http://www.jstor.org/stable/20762323 visited on 05/20/2022.
Parker, Grant (2002). “Ex Oriente Luxuria: Indian Commodities and Roman Experience”.
In: Journal of the Economic and Social History of the Orient 45.1, 40–95. issn: 00224995.
http://www.jstor.org/stable/3632707 visited on 05/18/2022.
Pelizzo, Riccardo, Abel Kinyondo, and Zim Nwokora (2018). “Development in Africa”. In:
World Affairs 181.3, 256–285. issn: 00438200, 19401582. https : / / www . jstor . org /
stable/26572790 visited on 04/01/2023.
Rathbone, Dominic (1997). “Prices and Price Formation in Roman Egypt”. In: Economie
antique: Prix et formation des prix dans les economies antiques, 183–244.
Reden, Sitta von (2002). In: Klio 84.1, 141–174. https://doi.org/doi:10.1524/klio.
2002.84.1.141. https://doi.org/10.1524/klio.2002.84.1.141.
Rivera-Batiz, Luis A. and Paul M. Romer (1991). “Economic Integration and Endoge-
nous Growth”. In: The Quarterly Journal of Economics 106.2, 531–555. issn: 00335533,
15314650. http://www.jstor.org/stable/2937946 visited on 02/13/2023.
Rogers, Perry M. (1984). “Domitian and the Finances of State”. In: Historia: Zeitschrift für
Alte Geschichte 33.1, 60–78. issn: 00182311. http://www.jstor.org/stable/4435871
visited on 05/18/2022.
Romer, Paul M. (1990). “Endogenous Technological Change”. In: Journal of Political Econ-
omy 98.5, S71–S102. issn: 00223808, 1537534X. http : / / www . jstor . org / stable /
2937632 visited on 12/05/2022.
Rosenstein, Nathan (2008). “Aristocrats and Agriculture in the Middle and Late Republic”.
In: The Journal of Roman Studies 98, 1–26. issn: 00754358. http://www.jstor.org/
stable/20430663 visited on 05/30/2022.
 Garfinkel 60

Rubin, Barnett R. (2006). “Afghanistan’s Uncertain Transition From Turmoil to Normalcy”.

In: 24–30. http://www.jstor.org/stable/resrep05730.9 visited on 03/18/2023.
Scheidel, Walter (2007). “A Model of Real Income Growth in Roman Italy”. In: Historia:
Zeitschrift für Alte Geschichte 56.3, 322–346. issn: 00182311. http://www.jstor.org/
stable/25598399 visited on 05/19/2022.
Seymour, Linda M. et al. (2023). “Hot mixing: Mechanistic insights into the durability of
ancient Roman concrete”. In: Science Advances 9.1, eadd1602. https://doi.org/10.
1126/sciadv.add1602. eprint: https://www.science.org/doi/pdf/10.1126/sciadv.
add1602. https://www.science.org/doi/abs/10.1126/sciadv.add1602.
Silver, Morris (2011). “Finding the Roman Empire’s Disappeared Deposit Bankers”. In: His-
toria: Zeitschrift für Alte Geschichte 60.3, 301–327. issn: 00182311. http://www.jstor.
org/stable/41342853 visited on 10/10/2022.
Solow, Robert M. (1956). “A Contribution to the Theory of Economic Growth”. In: The
Quarterly Journal of Economics 70.1, 65–94. issn: 00335533, 15314650. http://www.
jstor.org/stable/1884513 visited on 04/05/2022.
Spurr, M. S. (1983). “The Cultivation of Millet in Roman Italy”. In: Papers of the British
School at Rome 51, 1–15. issn: 00682462. http://www.jstor.org/stable/40310790
visited on 06/10/2022.
Temin, Peter (1997). “Two Views of the British Industrial Revolution”. In: The Journal of
Economic History 57.1, 63–82. issn: 00220507, 14716372. http : / / www . jstor . org /
stable/2951107 visited on 03/18/2023.
— (2001). “A Market Economy in the Early Roman Empire”. [inlangeng]. In: The Journal
of Roman studies 91, 169–181. issn: 0075-4358.
— (2004). “The Labor Market of the Early Roman Empire”. In: The Journal of Interdis-
ciplinary History 34.4, 513–538. issn: 00221953, 15309169. http : / / www . jstor . org /
stable/3656762 visited on 05/18/2022.
 Garfinkel 61

— (2006). “The Economy of the Early Roman Empire”. In: The Journal of Economic Per-
spectives 20.1, 133–151. issn: 08953309. http : / / www . jstor . org / stable / 30033637
visited on 05/18/2022.
Thornton, M. K. and R. L. Thornton (1990). “The Financial Crisis of A.D. 33: A Keyne-
sian Depression?” In: The Journal of Economic History 50.3, 655–662. issn: 00220507,
14716372. http://www.jstor.org/stable/2122822 visited on 09/12/2022.
Veen, Marijke van der (2010). “Agricultural innovation: invention and adoption or change
and adaptation?” In: World Archaeology 42.1, 1–12. issn: 00438243, 14701375. http :
//www.jstor.org/stable/25679724 visited on 05/30/2022.
Verhulst, Stefaan G. et al. (2017). “Who Runs the Internet?: The Global Multi-stakeholder
Model of Internet Governance”. In: 95–117. http://www.jstor.org/stable/resrep05243.
11 visited on 03/18/2023.
Weiner, Tim (1993). “Lies and Rigged ‘Star Wars’ Test Fooled the Kremlin, and Congress”.
In: 1–. https://www.nytimes.com/1993/08/18/us/lies-and-rigged-star-wars-
test-fooled-the-kremlin-and-congress.html.

Books
Bagnall, Roger S. (1994). The demography of Roman Egypt. [inlangeng]. Cambridge studies
in population, economy, and society in past time ; 23. Cambridge [England] ; Cambridge
University Press, 1994. isbn: 0521461235.
Brunt, P. A. (1971). Italian manpower, 225 B.C.-A.D. 14. [inlangeng]. London: Oxford
University Press, 1971. isbn: 0198142838.
Corpus inscriptionum Latinarum (1862). [inlanglat]. Berolini: apud G. Reimerum, 1862.
Coulanges, Fustel de (1980). The ancient city : a study on the religion, laws, and institutions
of Greece and Rome. [inlangfre;eng]. Baltimore: Johns Hopkins University Press, 1980.
isbn: 0801823048.
 Garfinkel 62

Duncan-Jones, Richard (1994). Money and Government in the Roman Empire. Cambridge
University Press, 1994. https://doi.org/10.1017/CBO9780511552632.
Ellul, Jacques (1965). The technological society. [inlangeng]. London: Jonathan Cape, 1965.
Etienne, Robert. (1992). Pompeii: The day a city died. [inlangeng]. Discoveries. New York:
H.N. Abrams, 1992. isbn: 0810928558.
Finley, M. I. (Moses I.) (1985). The Ancient Economy. [inlangeng]. 2nd ed. Sather classical
lectures ; v. 43. Berkeley, Calif: University of California Press, 1985. isbn: 0520054520.
Laslett, Peter (1971). The world we have lost: England before the Industrial Age. [inlangeng].
2nd ed. London: Methuen, 1971. isbn: 0416083900.
Peter Garnsey Keith Hopkins, C.R. WHittaker (1983). Trade in the ancient economy. [inlangeng].
Berkeley: University of California Press, 1983. isbn: 0520048032.
Rostovtzeff, Michael (1926). Social and Economic History of the Roman Empire. [inlangeng].
Accessed: 2022-04-251[]Accessed: 2022-04-25. Biblo-Moser, 1926. isbn: 978-0819621641.
Roth, Jonathan P (1999). The logistics of the Roman army at war (264 B.C.-A.D. 235).
[inlangeng]. Vol. 23. Columbia studies in the classical tradition. Leiden: Brill, 1999. isbn:
9781423729730.
Temin, Peter (2013f). The Roman market economy. [inlangeng]. The Princeton economic
history of the Western world. Princeton, N.J: Princeton University Press, 2013. isbn:
1400845424.

Book chapters
Bielefeld, Shelley (2016). “Neoliberalism and the return of the guardian state: Micromanag-
ing Indigenous peoples in a new chapter of colonial governance”. In: Engaging Indige-
nous Economy: Debating diverse approaches. Vol. 35. ANU Press, 2016, 155–170. isbn:
9781760460037. http://www.jstor.org/stable/j.ctt1d10hpt.18 visited on 04/01/2023.
 Garfinkel 63

Bowman, Alan K (1996). “Provincial administration and taxation”. [inlangeng]. In: The
Cambridge Ancient History. Vol. 10. Cambridge: Cambridge University Press, 1996, 344–
370. isbn: 0521264308.
Garnsey, Peter et al. (2015). “An underdeveloped economy”. In: The Roman Empire: Econ-
omy, Society and Culture. 2nd ed. University of California Press, 2015, 71–90. isbn:
9780520285989. http://www.jstor.org/stable/10.1525/j.ctt9qh25h.12 visited
on 05/18/2022.
Hobson, Matthew S. (2014). “A Historiography of the Study of the Roman Economy: Eco-
nomic Growth, Development, and Neoliberalism”. In: TRAC 2013: Proceedings of the
Twenty-Third Annual Theoretical Roman Archaeology Conference, London 2013. Oxbow
Books, 2014, 11–26. isbn: 9781782976905. http://www.jstor.org/stable/j.ctvh1dwm3.
4 visited on 05/18/2022.
Hollander, David B. (2017). “The Roman Economy in the Early Empire: An Overview”.
In: Paul and Economics: A Handbook. 1517 Media, 2017, 1–22. isbn: 9781506406039.
http://www.jstor.org/stable/j.ctt1kgqtgr.6 visited on 05/18/2022.
Jarausch, Konrad H. (2015). “Disappointing Decolonization”. In: Out of Ashes: A New His-
tory of Europe in the Twentieth Century. REV - Revised. Princeton University Press,
2015, 480–505. isbn: 9780691173078. http://www.jstor.org/stable/j.ctvc77dk5.23
visited on 04/01/2023.
Jongman, Willem (Jan. 2006). “The rise and fall of the roman economy: population, rents
and entitlement”. In: Jan. 2006, 237–254. isbn: 8872284880.
Koloski-Ostrow, Ann Olga (2015). “Understanding Roman sanitation from archaeology: Toi-
lets, sewers, and water systems”. In: The Archaeology of Sanitation in Roman Italy: Toi-
lets, Sewers, and Water Systems. University of North Carolina Press, 2015, 52–83. isbn:
9781469621289. http://www.jstor.org/stable/10.5149/9781469621296_koloski-
ostrow.7 visited on 06/07/2022.
 Garfinkel 64

Lebergott, Stanley (1966). “Labor Force and Employment, 1800–1960”. In: Output, Employ-
ment, and Productivity in the United States after 1800. NBER, 1966, 117–204. http :
//www.nber.org/chapters/c1567.
Magie, David (1950a). “Centralization and proesperity under the Flavians”. In: Roman Rule
in Asia Minor. Princeton University Press, 1950, 566–592. isbn: 9780691051505. http:
//www.jstor.org/stable/j.ctt17t75f4.28 visited on 05/30/2022.
— (1950b). “From Gold to Iron”. In: Roman Rule in Asia Minor. Princeton University Press,
1950, 659–687. isbn: 9780691051505. http://www.jstor.org/stable/j.ctt17t75f4.32
visited on 06/03/2022.
Manning, J.G. and Colin Renfrew (2014). “At The Limits: Long-Distance Trade in the Time
of Alexander the Great and the Hellenistic Kings”. In: Reconfiguring the Silk Road: New
Research on East-West Exchange in Antiquity. University of Pennsylvania Press, 2014,
5–14. isbn: 9781934536681. http://www.jstor.org/stable/j.ctt7zw7mj.7 visited on
01/18/2023.
Morris, Ian (2013). “Energy Capture”. In: The Measure of Civilization: How Social Develop-
ment Decides the Fate of Nations. NED - New edition. Princeton University Press, 2013,
53–143. http://www.jstor.org/stable/j.ctt1r2fj0.8 visited on 05/30/2022.
Roberts, Keith (2011). “The Hellenistic Business Environment”. In: The Origins of Business,
Money, and Markets. Columbia University Press, 2011, 101–115. http://www.jstor.
org/stable/10.7312/robe15326.14 visited on 06/02/2022.
Saprykin, Sergey (2017). “Ancient sea routes in the Black Sea”. In: The Sea in History - The
Ancient World. NED - New edition. Boydell Brewer, 2017, 345–361. isbn: 9781783271573.
http://www.jstor.org/stable/10.7722/j.ctt1kgqszh.34 visited on 01/10/2023.
Sidebotham, Steven E. (2011). “Trade in Roman Berenike”. In: Berenike and the Ancient
Maritime Spice Route. 1st ed. University of California Press, 2011, 221–258. isbn: 9780520244306.
http://www.jstor.org/stable/10.1525/j.ctt1ppv29.17 visited on 10/10/2022.
 Garfinkel 65

Taylor, Veronica L (2017). “Regulatory rule of law”. In: Regulatory Theory: Foundations and
applications. ANU Press, 2017, 393–414. isbn: 9781760461010. http://www.jstor.org/
stable/j.ctt1q1crtm.33 visited on 03/18/2023.
Temin, Peter (2013a). “Economic Growth in a Malthusian Empire”. In: The Roman Market
Economy. Princeton University Press, 2013, 220–242. isbn: 9780691147680. http://www.
jstor.org/stable/j.ctt1r2g35.16 visited on 01/10/2023.
— (2013b). “Financial Intermediation”. In: The Roman Market Economy. Princeton Univer-
sity Press, 2013, 157–190. isbn: 9780691147680. http://www.jstor.org/stable/j.
ctt1r2g35.13 visited on 01/09/2023.
— (2013c). “Growth Theory for Ancient Economies”. In: The Roman Market Economy.
Princeton University Press, 2013, 195–219. isbn: 9780691147680. http://www.jstor.
org/stable/j.ctt1r2g35.15 visited on 05/18/2022.
— (2013d). “Per Capita GDP in the Early Roman Empire”. In: The Roman Market Economy.
Princeton University Press, 2013, 243–262. isbn: 9780691147680. http://www.jstor.
org/stable/j.ctt1r2g35.17 visited on 05/18/2022.
— (2013e). “Price Behavior in the Roman Empire”. In: The Roman Market Economy. Prince-
ton University Press, 2013, 70–92. isbn: 9780691147680. http : / / www . jstor . org /
stable/j.ctt1r2g35.8 visited on 05/27/2022.

Miscellaneous and Ancient Texts

Aristotle (n.d.). The Athenian Constitution. [inlangeng, trans. Sir Frederic G. Kenyon].
Accessed: 2022-05-241[]Accessed: 2022-05-24. http://classics.mit.edu/Aristotle/
athenian_const.2.2.html.
Athenaeus and S. Douglas Olson (2019). Athenaeus Naucratites: Deipnosophistae. [inlangeng].
Berlin, 2019.
Cato, Marcus Porcius (1935 - 1967). Marcus Porcius Cato, On agriculture; Marcus Terentius
Varro. On agriculture. [inlangeng]. Cambridge, Mass, 1935 - 1967.
 Garfinkel 66

Cicero (n.d.). On Occupations. [inlangeng, trans. Walter Miller (1913)]. Accessed: 2022-
05-241[]Accessed: 2022-05-24. http://www. perseus. tufts .edu/hopper/text?doc=
Perseus%5C%3Atext%5C%3A2007.01.0048%5C%3Abook%5C%3D1%5C%3Asection%5C%
3D150.
Cicero, Marcus Tullius (2001). Cicero–letters to friends. [inlanglat;eng]. Cambridge, Mass,
2001.
— (1979). Cicero’s Letters to Atticus. [inlanglat;eng]. Harmondsworth, Eng. ; 1979.
Columella, Lucius Junius Moderatus. (1745). L. Junius Moderatus Columella of husbandry.
In twelve books: and his book concerning trees. Translated into English, with several il-
lustrations from Pliny, Cato, Varro, Palladius, and other ancient and modern authors.
[inlangeng]. London, 1745.
Dio, Cassius (1914). Dio Cassius: Roman history. [inlangeng]. Massachusetts, 1914.
Diodorus (1935). The Library of History. [inlanggre;eng]. London, 1935.
Elder, Pliny the (n.d.). Natural Histories. [inlangeng, trans. Bostock et al. (1855)]. Accessed:
2022-05-241[]Accessed: 2022-05-24. https://www.perseus.tufts.edu/hopper/text?
doc=Plin.+Nat.+toc.
Gilliver, Kate (2004). Caesar’s Gallic Wars 58-50 BC. [inlangeng]. 2004.
Halicarnassus, Dionysius of (n.d.). Roman Antiquities. [inlangAncient Greek]. Accessed:
2022-05-241[]Accessed: 2022-05-24. http://www.perseus.tufts.edu/hopper/text?
doc=Perseus%5C%3Atext%5C%3A2008.01.0572.
Heron (1851, trans. Woodcroft). The Pneumatics of hero of Alexandria / Bennet Woodcroft.
[inlangeng]. London, 1851, trans. Woodcroft.
Homer (1961 - 1951). The Iliad of Homer. [inlangeng]. Chicago, 1961 - 1951.
Kropff, Antony (2016). An English translation of the Edict on Maximum Prices, also known
as the Price Edict of Diocletian. Accessed: 2022-10-091[]Accessed: 2022-10-09. 2016. http:
//kark.uib.no/antikk/dias/priceedict.pdf.
 Garfinkel 67

Livy (n.d.). From the Fouding of the City. [inlangeng, trans. William Heinemann (1919)]. Ac-
cessed: 2022-05-241[]Accessed: 2022-05-24. https://www.perseus.tufts.edu/hopper/
text?doc=Liv.+1&fromdoc=Perseus%5C%3Atext%5C%3A1999.02.0152.
Malthus, T. R. (Thomas Robert) (1798). An Essay on the principle of population. [inlangeng].
London, 1798.
Marcellinus, Ammianus (1950, trans. Rolfe). Ammianus Marcellinus. [inlangeng]. Cam-
bridge, Mass, 1950, trans. Rolfe.
Mayeske, Betty Jo (1972). Bakeries, bakers, and bread at Pompeii: A study in social and
economic history. 1972.
Petronius (1913). The Satyricon of Petronius. [inlangeng]. 1913.
Plato (n.d.). Phaedro. [inlangeng, trans. Harold North Fowler et al. (1966)]. Accessed: 2022-
05-191[]Accessed: 2022-05-19. http://www. perseus. tufts .edu/hopper/text?doc=
Plat.+Phaedo&fromdoc=Perseus%5C%3Atext%5C%3A1999.01.0170.
Pliny (n.d.). Epistulae. [inlangeng, trans. Firth et al. (1900)]. Accessed: 2022-05-251[]Accessed:
2022-05-25. http://www.attalus.org/info/pliny.html.
Plutarch (2017). Complete collection of Plutarch’s Parallel lives. [inlangeng]. New York,
2017.
Pollio, Vitruvius (1956). Vitruvius, On architecture. [inlangeng ; lat]. London, 1956.
Polybius (n.d.). The Histories. [inlangeng;grc]. Cambridge, Mass. ;
Sallust (n.d.). Catilinian War. [inlanglat]. England.
Salzman, Michele and Michael Roberts (2011). The Letters of Symmachus: Book 1. [inlangeng].
Atlanta, 2011.
Strabo (2014). The Geography of Strabo. [inlangeng]. Cambridge, United Kingdom, 2014.
Suetonius (1914). Suetonius: The Lives of the Twelve Caesars. [inlangeng]. England, 1914.
Tacitus (1873, trans. Alfred John Church et. al.). Annals. Accessed: 2010-11-041[]Accessed:
2010-11-04. 1873, trans. Alfred John Church et. al. http://www.perseus.tufts.edu/
 Garfinkel 68

hopper/text?doc=Perseus%5C%3Atext%5C%3A1999.02.0078%5C%3Abook%5C%3D15%5C%
3Achapter%5C%3D72.
Thucydides (1996, trans. Strassler). The Landmark Thucydides: A Comprehensive Guide to
The Peloponnesian War. [inlangeng]. 1230 Avenue of the Americas, New York, NY 10020,
1996, trans. Strassler.
Vergil’s Georgics (2008). [inlangeng]. New York, 2008.
Younger, Seneca the (n.d.). Thyestes. [inlangLatin]. Accessed: 2023-02-231[]Accessed: 2023-
02-23. http://www.perseus.tufts.edu/hopper/text?doc=Perseus%5C%3Atext%5C%
3A2007.01.0010.