Economic History Association

The Impact of Globalization in the Roman Empire, 200 BC-AD 100 Author(s): Ryan M. Geraghty Source: The Journal of Economic History, Vol. 67, No. 4 (Dec., 2007), pp. 1036-1061 Published by: Cambridge University Press on behalf of the Economic History Association Stable URL: http://www.jstor.org/stable/40056408 . Accessed: 28/08/2013 13:42
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Cambridge University Press and Economic History Association are collaborating with JSTOR to digitize, preserve and extend access to The Journal of Economic History.

http://www.jstor.org

This content downloaded from 159.178.22.27 on Wed, 28 Aug 2013 13:42:23 PM All use subject to JSTOR Terms and Conditions

TheImpact inthe Roman ofGlobalization 100 200 BC-AD Empire,

Ryan M. Geraghty
to a general The article modelto describe Italy'sresponse employs equilibrium of theRoman and factor market theexpansion commodity integration during a comprehensive of the Italian constructs Empire.This novel approach story on controthat and sheds corroborates established economy developments light versial andunanswered The success of the model supports arguments questions. was a that Romans wererational economic actors andthat theRomaneconomy market well-integrated system.

standard tenet ofeconomics is that ortheintegration globalization, ofworldcommodity and factor has profound andpredictmarkets, able effects an andinresource mix, allocation, upon economy's output come distribution. To studyglobalization and its effects, economists construct that of the domestic models tell the general equilibrium story ofmarket historian has employed Yet no economic impact integration.1 of theRoman to the general equilibrium techniques study development ofimpeoverthedegree imperial economy. Sparsedataanduncertainty rialmarket theapand Romans' rationalism make economic integration of these modelsdifficult.2 While acknowledging plication traditional thisarticle a generalequilibrium modelto tell a limitations, employs coherent of Italian Rome's expanstory during peninsular development sionfrom 200 BCto AD 100.The novelty is theuse ofa ofthisapproach of Italianecoformalized economicmodelto portray a broadpicture of nomicdevelopment, and thepurpose, questions beyondaddressing a Romanhistory, is to arguethat such models is powerful constructing method ofstudying ancient history. Historians make valid arguments thatRome's economicexpansion farmers had muted effects on theItalianeconomy, becausesubsistence did notrespond to market incentives and becauseRomansweremotiIn addivatedmoreby cultural thanby economics.3 and social factors theories contend thatit is a mistake to project modern tion,historians
The Journalof EconomicHistory, Vol. 67, No. 4 (December2007). The Economic Association. All rights reserved. ISSN 0022-0507. History 2124 Ashland SantaMonica,CA 90405. E-mail:ryan.geraghty Avenue, RyanM. Geraghty, @gmail.com. I am grateful for continuous adviceandcriticism from Foote, Williamson, Christopher Jeffrey Peter All errors, ofcourse, areminealone. andMatthew Weinzierl. Temin, Jones, Christopher andO'Rourke, "Did theGreat Irish Famine." Williamson, Inequality: 2 theDebate." Sailer, "Framing andFinley, Ancient "Famine"; Garnsey, Economy.

1036

This content downloaded from 159.178.22.27 on Wed, 28 Aug 2013 13:42:23 PM All use subject to JSTOR Terms and Conditions

Impact ofGlobalization

1037

in fundamentally different ontopast societiesthatfunctioned ways. muchof their recent have dedicated Economichistorians work,howRomeand forother theseideas forancient to challenging ever, places.5 that Romans economic rasomescholars Forexample, argue displayed in their of landedestatesand in their decisions management tionality aboutmigration.6 in debatefocuses on theimportance ofmarkets controversial Another to whichimperial markets functioned as a Romanlifeand theextent somehistorians theimpeunit."7 Whereas arguethat "singleeconomic nor reliant markets for thefulwas neither rialeconomy upon integrated that was an conclude themarket fillment ofneeds,others system imporintheempire.8 he modeofexchange ifnotthedominant, tant, Although and Peter Temin cataworked that markets slowly imperfectly, accepts and concludes use of pricesacrossthe empire loguesthewidespread "functioned as partof a singlecomprehensive thatRomanprovinces market."9 Mediterranean whatRoman thesedebatesdirectly knowing requires Approaching of their ina more or werethinking citizens comprehensive knowledge that we possesstoday. Rather than thecontradictory teractions snippets I and then formalize a conto these than debates, develop speakdirectly based on theassumptions of theItalianeconomy understanding ceptual market The consistency and economic ofperfect integration. rationality andthehistorical evidence of theresulting predictions (or lackthereof) thus model's and of the an assessment assumptions indirectly provides a themodelprovides described above.In addition, thedebates addresses thataddresses view of Italianeconomicdevelopment single,coherent and are which models on of the existing explanations questions many whatwe focused.The success of the model in predicting singularly the on more coneconomic knowofItalian history supports predictions and martheassumptions of economic troversial rationality questions, the model,and the approachitself.Of thatunderlay ket integration a modelis an iterative course, processand themodelpreconstructing is onlyto hereis basic and oversimplified. sented My aim,however, has to ancient that this readers convince history enough potenapproach andapplicable models. ofmore refined thedevelopment tialtomerit
4 Land. andWhittaker, Ancient Finley, Economy, 5 Scheideland von Traditional and Schultz,Transforming Reden,"Introduction"; Agriculture. "; andKehoe,Management. Rathbone, Morley, "Metropolis "Development"; Ancient Economy, p. 10. Finley, Land. Ancient Cities;andWhittaker, Garnsey, Economy; Finley, 9 "Market Temin, Economy," p. 169.

This content downloaded from 159.178.22.27 on Wed, 28 Aug 2013 13:42:23 PM All use subject to JSTOR Terms and Conditions

1038

Geraghty

Economichistorians models haveyetto employ general equilibrium in thestudy oftheancient anduncerbecausethelimited past, probably retaindatamaketheprecise in of the more studies predictions possible howcentpast unattainable. thanprecisepredictions, More important of theItalianeconomy and its development ever,is thebroadpicture thatour economicreasoning and the availableevidenceconvey.The lackofdocumentation from ancient Romemakestheavailableevidence a diverse of andgovernsources array archaeological findings, primary ment and other of morererecords, models,and studies demographic centagricultural with economies. the of Together application economic these a of the logic, fragments provide picture past. A GENERAL MODEL OF ROME EQUILIBRIUM The general modelpresented hereis an integrated collecequilibrium tionofcommodity andfactor markets driven and and demand bysupply theentirety oftheItalian The modelretogether representing economy. several that are of quires simplifying assumption typical general equilibrium in whatfollows.10 The model's analysisand will be described such as factor and endowments inputsare the exogenousvariables, tradable in theempire determined as a wholeandincommodity prices, of the Italian itself. thecausedependent economy By understanding and-effect in the Italian andestablishing howthe relationships economy variables evolvedovertimeas a result of imperial exogenous expantheeffects of expansion on themodel'sendogesion,one can measure nousoutputs, thepricesoftradable including goodsandresource inputs andthequantities ofresources dedicated to eachsector. mainfactors of consists of three My modelof theRomaneconomy and slaves four final and sectors productionland, labor, goods and urbangoods.Two intermediate wine,livestock, grain, goods,feed and manure, realistic links between final sectors. The provide goods modelalso includes theannona,free to the grain provided populaceof Romeby theimperial the taxesexpropriated from government. Finally, close the model and finance trade deficit. provinces Italy's and livestock reflect themainsectors Grain, wine,urban production, of theeconomy, the model but realistic. thepriGrain, making simple in thesubsistence-based andwine, mary commodity imperial economy, theprimary in constituted 85 of the luxury good, percent agriculture cultivable areas.11 Livestock was prevalent further from citiesin Italy's
10 "Did theGreat Irish Economics. andSodersten andReed,International O'Rourke, Famine"; 1' andRathbone, Jongman, Economy, "Development."

This content downloaded from 159.178.22.27 on Wed, 28 Aug 2013 13:42:23 PM All use subject to JSTOR Terms and Conditions

Impact ofGlobalization

1039

all producaltitude pastureland.Urbangoods,whichrepresent higher in and towns tion cities cloth,furniture, including pottery, glassware, and other metalproducts, urbanand rural ironware, help differentiate over wine,andurban Although grain, goodsweretraded populations.13 thelack of refrigeration andthehighcostof landtranslongdistances, in livestock between made trade Italyand therestof theempire port Italian livestock was delivered Transported "bythehoof," prohibitive.14 towns for local consumption.15 tonearby and determined within the Thus, livestock pricesare not tradable whereas and urban are modelitself, tradable wine, grain, goods prices in imperial and determined markets. This assumesthat Italywas small to theimperial as a wholeanddidnotimpact economy compared prices in imperial markets. In fact, a determined was Italy significant imperial fiveofits50 million muchofits market, housing peopleandpossessing theimpact in underwealth.16 of domestic markets ThoughI consider I leavethesmallopeneconomy andwineprices, standing grain assumpthemodel. tioninplaceto simplify that factor The modelalso assumes endowments of labor, slaves,and landwereexogenous to theItalianeconomy. The quantity of laborwas such driven as and shocks, bypolitical primarily conscription colonizacostsprevented from in tion;hightransportation provincials migrating numbers to take of Italian economic Similarge advantage prosperity.17 thequantity of slaveswas largely on political shocks, larly, dependent of in the slaves wars of for territorial mostly winning conquests fought or defense.18 The quantity of landin use was motivated both expansion economic the desire to new lands under cultiforces, by primarily bring to agriculture vationas thereturns forces, grew,and by noneconomic thepushing of peasants their old ancestral landto ejectedfrom namely moremarginal thatfactor endowments are Overall,assuming plots.19
12 An excavation at the PalaMorley, Metropolis; and Yeo, "Land and Sea Transportation." tine indicatesthe relativeimportanceof the different animals: 40 percentpigs, 30 percentsheep or goats, 20 percentfowl,and 10 percentcattle(Morley, Metropolis,p. 152). Frank,Economic Survey. Yeo, "Land and Sea Transportation." 15 Morley,Metropolis. 16 Italian Manpower. Brunt, 17 Scheidel, "Human Mobilityin Roman Italy, 1". Scheidel, "Human Mobility in Roman Italy,2"; and Harris,"Roman Slave Trade." The desire foradditionalslaves partiallymotivatedsome Roman wars, but the quest forMediterranean In fact, supremacy,political prestige,and other formsof booty were usually more important. prisonerswere oftenunforeseenconsequences of conflicts,and Roman victoriesoftenprovided thousands of slaves from single battles. Wars were also highly expensive, implyingthat the slave supply only responded to immense variationin price. See Shorter, Rome; Toynbee, HanLand. nibal 's Legacy: and Whittaker, 19 Italian Manpower; and Toynbee, Hannibal's Legacy. Morley,Metropolis; Brunt,

This content downloaded from 159.178.22.27 on Wed, 28 Aug 2013 13:42:23 PM All use subject to JSTOR Terms and Conditions

1040

Geraghty

withwhatwe knowof Italianlabor, consistent exogenousis largely and at themodel. landuse this time andusefully slaves, simplifies in latersocieties slavesand laborare often close substitutes, Though thismodelpositsa differentiation between freeand slave laborand the income that basedon shares receive from andwineproduction, they grain in Romansociety.20 harder their different characteristics Slaves worked than wars andlonger wouldtolerate.21 Peasant absences to fight peasants reduced their on farm.22 Peasants and their overall the reliability presence excluded themselves from on largeestatesalongside slavesfor working cultural intense reasons.23 whilegrain Furthermore, production requires laboronlyduring inwine constant and peakseason, production requires tensive laborthroughout thegrowing season.24 Forthisreason, slavelarather than free labor with and more lax work its absences bor, frequent was moreprofitably in wine production. The model habits, employed this free and slave labor and captures understanding byseparating byposdifferent functions for andwine. iting production grain The modelincludes theannona,orthedole offree by grain provided theimperial becauseitwas a significant sourceof income government, for urban residents andprobably an important factor mopsychological to At the its some 300,000Rotivating peasant migration capitol. peak man residents, or 5-10 percent of theItalianpopulation, received the imand its in that its the historical record annona, prominence suggest oftheRomanpopulacewas greater than that.25 pacton thecontentment In addition, theprospect to of freegrain was probably appealing peasantsfarfrom their theannona'sreRome,motivating despite migration striction to onlya slice of Romanresidents. of the For thepurposes I assume that the and annona was free its existence model, throughout distributed urban residents. evenly among With themodelbroadly I nowdelveintoanddescribe its established, incommodity andfacwhich determine howchanges parameters, prices torendowments The most affect and factor imporcommodity prices.26 tant define whichand in whatproportion ofprothefactors parameters
20 Phillips, Slavery. Arable Cultivation; thatpeasants and Rathbone, Spun*, "Development." Spurrestimates worked150 daysperyear, 250 daysperyear. whereas slaveswerefully and worked employed This shedslight on calculations of laborand slave costs.For example, Rathbone assumesthat laborand slavesworked that than thesameamount and calculates slavesweremoreexpensive free for make slave labor the shorter work would however, peasants. Accounting peasant year, than labor. cheaper peasant
22 Hopkins, Conquerors. 23 Morley,Metropolis. 24 Rosafio, "Slaves." 25 Rickman, Corn Supply.

26See Tables4-7 for a complete ofthemodel'sparameters. Appendix summary

This content downloaded from 159.178.22.27 on Wed, 28 Aug 2013 13:42:23 PM All use subject to JSTOR Terms and Conditions

Impact ofGlobalization

1041

- receiveincomefrom - labor,slaves,and land theproduction duction A of Romancontemporary and sale of each commodity. combination with later subsistence and texts, economies, comparisons agronomist commodities of the themselves us knowledge agricultural provide the revenues weresplit.Whatmatters of how commodity hereare picture buttheoverallpicture of how resources combine to notexactfigures output. generate a combination of freelaborand landwith Grainproduction requires Romanpeasantstended of slaves and manure. to prosmalladditions thanwine forthe reasonsdescribed duce grainrather previously.27 in laterperiods on grainproduction textsand research Agronomist this andadd information on howagricultural corroborate picture largely the force and landowners.28 labor is divided amongst output on theother demands slavesand Wineproduction, hand,intensively the other factors of Wine's on excludes production. dependence largely and laboris described slavesoverfree above,anditslaborintensiveness make slaves more than land as land value over important density high Romanfarmers on their texts well.Agronomist educating contemporary from which we can infer a of wine production paint picture vineyards inwineoutput.29 ofproduction ofeachfactor theshare of freelaborand Urbangoods are produced by some combination I free relative slaves.As no dataexiston their importance,assumethat of urban theoutput laborand slavesshared production evenly. Finally,
27 Again, grain's cyclical naturebettersuited theirabsences fromthe farmforwar, grain rethe growing season, and, unlike wine, quired less intensive labor than did wine throughout to theirstatus. growinggrainwas not viewed as detrimental to a varietyof subsistenceeconomies, The calibrationof neoclassical productionfunctions includingItaly and Africa, shows thatland's share of gross product is approximatelyone-half. See Clark and Haswell, Economics. Jongman'sstudyof Pompeii provides evidence thatthe agworkforceand land rentseach absorbed half of total grain yields. Including manure ricultural the to 0.4, with manure and slaves contributing and slaves requires reducingthese proportions remaining0.2 of total product. See Jongman,Economy. A more detailed analysis of Roman An average plebeian familyof grainfanningbased on the agronomisttextssupportthis finding. fourpossessed a plot of about ten iugera, which J. K. Evans estimatesproduced approximately 250 modii of grain per annum forrevenue of HS 1,000 per annum. See Evans, "Wheat Producamount to 0.56 of total production"; and Duncan-Jones,Economy. Subsistence requirements tion; as peasants satisfiedtheirsubsistence only partiallywith grain,peasant labor amounts to withour earlierestimate. around0.4-0.5 of theirproduct,which is consistent 29Wine was far more labor-intensive than grain, requiring25 to 35 days per iugerumcompared to grain's 10 to 15. See White, Roman Farming',and Spurr,Arable Cultivation.Cato's ideal vineyardof 100 iugera withyields of 1.6 cullei per iugerumearns a revenue of HS 16,000 and at a slave cost of HS 6,400 (assuming 16 slaves at a price of HS 2,000 each amortizedover ten years withan additional subsistencecost of 50 modii of grain at HS 4 per modius), or 0.4 of total revenue. See Duncan-Jones, Economy; and Rathbone, "Development." Columella's prescriptionof one part free labor foreach threeparts slave labor leads a value of 0.15 for labor, and his estimateof land rentsamountto HS 5,000 per annum fora 100 iugera plot for0.3 of total revenue.See Rathbone,"Development."

This content downloaded from 159.178.22.27 on Wed, 28 Aug 2013 13:42:23 PM All use subject to JSTOR Terms and Conditions

1042

Geraghty

livestock production intensively employsland and feed and requires little labor. As limited data existforRome,I applyKevin relatively O'Rourke's findings on livestock's factorincome shares in preindustrial Ireland totheRomaneconomy.30 Otherparameters are the initialconditions of theRomaneconomy andtheincome and slaveowners, elasticities of free labor, landowners, to whomall slave incomeaccrues.The initialconditions an describe Italian in wine,urdominated butlacking economy bygrain production and livestock in and abundant labor and land butscarce ban, production, in slaves.31 Incomeelasticities definehow each factor of production the it earns and thus incomes are redirected affects how spends money to the consumption of othergoods. Free labor primarily spendsits on subsistence and then on other small money goods spends portions once subsistence aresatisfied.32 Landlords andslaveownrequirements ersspendtheir incomes on wine,urban reflecting goods,and livestock, their wealth on covluxuries and theease with which greater spent they eredtheir subsistence needs.33 Moredetailed andexplanations areincluded parameter specifications in theAppendix. Giventheimprecise the nature of thedata,I examine model'soutput for a broadrange ofparameters.
IMPERIAL EXPANSION AND THE EXOGENOUS SHOCKS TO THE ITALIAN ECONOMY

the natureof the Italianeconomyin Imperialexpansionaffected model howthedirect effects ofexpansion, powerful ways.My explains or theexogenous shocksto theItalianeconomy, affected thedecisions of economicactorsand caused ripplesthroughout To the economy. their the we must first understand more study consequences, completely
30 O'Rourke,"Did theGreatIrishFamine," p. 15. I use a landshareof 0.5, a feedshareof 0.35 anda slaveshare of0.15 basedon O'Rourke'sfigures. I estimate a 0.10 urbanlaborsharebased on thesize of Romansewersystems, withthe remainder of thelaborforcealmostentirely Mededicated to grainproduction. See Morley, I estimate that about85 percent of Italy'slandunder cultivation tropolis. produced grain; apthat amount was required to support itspopulation of 4.9 million without proximately persons See Jongman, The remainder of thelandwas spenton wineor beef grainimports. Economy. production. 32Historians 75 percent of their on grain. incomes agreethatlaborers spentapproximately See Garnsey, andSpurr, ArableCultivation. I assumethat with nearsubtheir Famine, laborers, sistence urbangoods at home,leaving25 percent of their incomes forwine wages,produced andlivestock consumption. The wealthy a trivial formed amount ofthepopulation, that their consumpimplying grain tionwas similarly trivial. that are notknown, texts ancient Thoughincomeelasticities suggest winewas themostimportant followed consumption good forthewealthy, by urban products andthen livestock. See Morley, andPurcell, "Wine." Metropolis;

This content downloaded from 159.178.22.27 on Wed, 28 Aug 2013 13:42:23 PM All use subject to JSTOR Terms and Conditions

Impact ofGlobalization

1043

andthedegree of certainty andprecision shocks to theItalian economy I know them.3 First describe those that we with whichwe knowmore in the rise the slave the of landunpopulation, expansion confidently thedeclinein freelabor,and growth of theannona dercultivation, those that aremore uncertain theambiguous before addressing change in and the rise wine and urban in grain goodsprices. Throughout, prices cause-and-effect that drive I describe boththeunderlying relationships and the available data. these changes mostapparent in the The exogenousshockto the Italianeconomy is thetremendous of the slave after record historical growth population influx of slaves from wars than more 200 BC. The constant provincial rates andhigh rates ofmanumission andin fact slaves' highdeath offset all of Italy'spopulation forvirtually thispeaccounted growth during thatItaly's slave population estimate was roughly Historians riod.35 of slave 200 in 200 BC. Estimates 500,000 Italy's population yearslater, records slaveswonfrom victobasedon warandeconomic cataloguing 2 1.2 to million or to 30 from riesandtrades, range people, up percent I havemostcontheestimates, of Italy'stotal Examining population.36 of thedemographic in Walter Scheidel'sdetailed facfidence analysis whichleads himto a conservative theslave population, torsaffecting theprecisefigures slaves in 1 BC.37 Of course, of 1.2 million estimate whatmatters are unimportant; is the calculations based on historians' intheItalian slavepopulation oftwoormore. cleargrowth bya factor that thefree ofItalystayed wellestablished It is also fairly population A after 200 BC. balanceof natural or declinedslightly constant reproand colonization of the death duction rates, conscription, provagainst we do thepathofItalian incesdetermined population growth.38 Though or fell,it is clear remained constant thepopulation notknowwhether
34 See Appendix Table 2 fora summaryof the exogenous shocks consideredin the model. Italian Manpower, and Scheidel, "Human Mobilityin Roman Italy,2." Brunt, 36 Ibid. 37 Scheidel, "Human Mobilityin Roman Italy,2." Higher estimatesof around 2 million slaves depend on estimatesof slave populations in laterperiods, such as the antebellumSouth, Brazil, and the Caribbean, in which slave populations formedapproximately30 percent of the total costs in those economies permitteda higher degree of spepopulation. Lower transportation cialization, however,than was possible in the Roman period. This makes the 30 percentfigure improbablyhigh for Roman society. Instead, Scheidel's estimates(Human Mobility in Roman Italy, 2), based on the amount of slaves required for grain,wine, livestock,and urban production,amount to 1.2 million, a high figuregiven the rapid death and manumissionrates experienced by the slave population. 38 The main source of growth was natural increase. Based on eighteenth-century data, Scheidel estimatesthatthe ruralnet rate of increase produces a 100 percentrise in the population by ad 14. A combinationof high urban death rates,forcedcolonization, and high militarization rates reduce offsetthis source of growthand resultin a small decline by most estimates. See Scheidel, "Human Mobilityin Roman Italy, 1."

This content downloaded from 159.178.22.27 on Wed, 28 Aug 2013 13:42:23 PM All use subject to JSTOR Terms and Conditions

1044

Geraghty

that theburden ofimperial which that free peasants expansion, required in was a significant serve the armyand populateforeign provinces, drainon thefreelaborpopulation and stymied population anymaterial Most estimates to a modest decline gain. point by 1 BC as thedrainof overwhelmed natural imperial expansion reproduction.39 folThe expansion oflandunder cultivation this clearly during period little reliable lows from ourknowledge of Italiandevelopment, though and the evidenceexiststo support theclaim.Gains in totalpopulation of provincial markets raisedthereturns of agriculture, making opening land and slave ownlands and wealthier marginal profitable motivating of Italy ers to push peasantsto thosemarginal lands.40 The recovery to from theHannibalic Warin thelatethird BCalso contributed century in the size of land use.41 Based on the number of soldiers, plots, growth landand thedegree to whichthesesettlements didnotdisturb existing the of after wars soldiers settlement holders, practice granting rights by 2 itself expandedland undercultivation by 5 percent. Thoughthearchaeologicalrecordprovidesevidenceof expansionat thistime,the lack of preserved remains from theprior perioddoes notnecessarily indicatethatareas were uninhabited.43 However,the major factors theuse of land all pointto greater landuse after 200 BC,and, driving the data are lacking,an expansionin land undercultivation though occurred. probably The prominence of theannonain thehistorical record allowsus to its the sold Senate precisely quantify growth. Originally, up to 40 modii of grainto about50,000Romanresidents at a costof HS 1.5 permodius. By 54 BC,theannonaprovided 40 modiiof freegrainto nearly a number later to200,000.44 reduced 300,000Romancitizens,
39 ItalianManpower, inRoman andScheidel, "Human Brunt, Hopkins, Conquerors', Mobility 1." I use Brunt's in225 BCand4.3 million of4.4 million inad 14. Other historians Italy, figures arrive at different butroughly similar estimates. Scheideluses 4 millionin both225 BC and ad 14 in his calculation ofnetmigration rates. that estimates Hopkins Italy'spopulation ranged from 4.0 million to4.4 million within I settled on this byad 14. Havingexamined figures range, Brunt's middle-of-the-road estimate. andToynbee, Hannibal 's Legacy. Morley, Metropolis; Hannibal'sLegacy. Toynbee, ItalianManpower. Brunt determines that ofCaesarandAugusBrunt, onlythesettlements tusin 59 BCand30-25 bc succeeded in avoiding theappropriation oflandsfrom thus, peasants; soldier landsettlements contributed inthelandendowment to a gainof 5 percent each assuming landallotment consisted of5-10 iugera. 43 "Famine." Garnsey, 44 CornSupply. To include theannonain themodel,I assumethat itinvolved free Rickman, from and compute thequantity of graindistributed graindistributions inception percapita.In thefirst theannonadistributes twomillion or 40 modiitimes50,000recipients, modii, period, of grain to an urban of 400,000foran averageof 5 modiipercapita.In thesecond population theannonadistributes million or 40 modiitimes to 1.2 modii, 200,000recipients, period, eight million an average of6 2/3modii a gainofone-third. peoplefor percapita, representing

This content downloaded from 159.178.22.27 on Wed, 28 Aug 2013 13:42:23 PM All use subject to JSTOR Terms and Conditions

Impact ofGlobalization

1045

200 BC is less clearthan The evolution of commodity the pricesafter the moreestablished changesin the slave and freelaborpopulations, and the annona. The lack of consistent of land data use, price degree from drawn tentative. makesanyconclusions Howpricecomparisons of theevidence that there was a signifiever,an examination suggests the empire, to both cant expansionin tradethroughout contributing in Italyandtheprovinces between correlations commodity prices strong Italianand provincial between 200 BC, and convergence prices.After thedispersion ofthemilitary andcoinage, the ofprovinces, thecreation of the Mediterraof roads and canals,and thepacifying construction in imperial led to a substantial nean, among otherfactors, growth risein thenumber of shipdata displaya threefold trade.45 Shipwreck from the third for the number of BCto a wrecks, proxy century voyages, The growth ofthemoney which was used BC.46 thefirst supply, century a for exchanges, suggests significant expansion especially long-distance of traded in thequantity AD,commodity By thefirst century goods.47 werehighly correlated. Forexample, in Italy andintheprovinces prices a significant correlation between andItalDavid Kesslerfinds Egyptian to shocksin Egyptin a prewithItalianpricesresponding ian prices, In his analysisof supplying the city of Rome, dictablemanner.48 in Romewas . . . [that] that "thesituation concludes Rickman Geoffrey the were not so relevant the local resources by periodof earlyEmthefailure of theRomanmarket morewas first pire. . . Whataffected of in provinces overseas. . . ; secondly, harvests inadequate supplies at seas . . ." 9 ofcornshipsin storms andthedestruction shipping, thatcommodity the growth of tradesuggests In addition, pricesin were and the markets, primarily converging. Imperial provinces Italy of for 50-75 accounted and North Africa, grain imports percent Egypt forItalian wine.50 Thissugof demand to Romeandat least25 percent deof provincial led to an influx grain, expansion geststhatimperial for new markets Italian wine,raising grainprices;provided pressing Italianwineprices;and caused grainand wine pricesin Italyand its of the to Kessler,90 percent to converge. Indeed,according provinces and was due to between transportaItaly Egypt pricedifferential grain x In the of classical ecoAD. tioncostsby the first language century in and urban wine a had nomictheory, advantage comparative Italy
45 Hopkins, "Taxes"; Greene,Archaeology; and Laurence, "Land Transport." 46 AncientShipwrecks. Parker, 47 Temin, "Market Economy"; and Duncan-Jones,Economy. 48 Kessler. "Feeding Rome." 49 Rickman,Corn Supply,p. 154. Kessler, "Feeding Rome"; Temin, "Market Economy"; and Tchernia,"Italian Wine." Kessler, "Feeding Rome."

This content downloaded from 159.178.22.27 on Wed, 28 Aug 2013 13:42:23 PM All use subject to JSTOR Terms and Conditions

1046

Geraghty

whatItalyposgoods production (because theyemployed intensively sessedinabundance, slavesandan urban laborforce) anda comparative in whatItaly intensively disadvantage grain(because grainemployed in land). possessed scarcity, With I nowexamine affectthis broadpicture thefactors established, in and the data more detail. Economic available ing commodity prices forces of directions. The opening pulledItaly'sgrain pricesin opposite theprovinces Italthe of substantially expanded supply grain reaching ian markets andplaceddownward on grain Meanwhile, pressure prices. indomestic in the total factors, primarily growth Italy's population, creaseddemand, and theconversion of resources from grain producing to other reduced Italian The goods supply, pushing priceshigher. path of grain becomesan empirical on which thedataareinprices question conclusive.52 For thepurposes of the model,I followthe smallopen and hold thatgrainprices fell. However,the economyassumption model'sresults are consistent inacrossa rangeof grain pricechanges, constant and moderate cluding prices pricegrowth. Unlikethecase of grain, economic forces unambiguously pressured wineandurban After 200 both wine and urban BC, goodspriceshigher. faced dodemand from the rest of the and from goods greater empire mesticconsumers.53 Whereasthe data tentatively indicatethatwine rose over our the data on urban prices period, goods pricesare nonexistent.54 Our understanding of directional changesand approximate magnitudesis sufficient to drawlogicalconclusions of aboutthedevelopment theItalianeconomy from ourconceptual framework of howtheItalian worked. The following thepredictions ofthe sections address economy - first model that those areconsistent with evidence establish to existing themodel's credibility and thenthosethataddressmorecontroversial issues.
52DuncanJones uses Trajan'salimentary which ofHS 16 scheme, provided boysallowances andgirlsHS 10 permonth, to infer a priceofHS 4 permodius at theclose ofthefirst permonth ad. See Duncanthefoodshortages of ad 19 and ad 64, TiJones, century Economy. During berius andNero imposed of HS 2 and HS 3 permodius, low pricefigpriceceilings curiously uresthat low market HS 1 andHS 10 Pricesfluctuated between suggest correspondingly prices. andafter theHannibalic War.See Garnsey, Famine.Few dataexistforItaly permodius during inthelatesecondcentury andearly first BC,buta pricein SicilyofHS 3-3.5 permodius century thattheItalianpricewas HS 5-6 per modiusin thefirst BC. See Rickman, suggests century CornSupply. Thistentatively toa decline from HS 5 to HS 4 overourperiod. points 53 "Wine." Purcell, in thesecondcentury an averagewinepriceof HS 10 peramCato,writing BC,employs Two centuries Columellauses an averagewine later, phora.See Rathbone, "Development." an exaggeration. Pricescouldvarysignificantly around thesebenchpriceof HS 15,probably from HS 8 to HS 64 inthefirst See Duncanmarks, Jones, ranging century. Economy.

This content downloaded from 159.178.22.27 on Wed, 28 Aug 2013 13:42:23 PM All use subject to JSTOR Terms and Conditions

Impact ofGlobalization
RESULTSAND INTERPRETATION

1047

does themodelpredict? The model'sresults turn outto What, then, acrossa broadrangeof plausibleparameter be consistent values and the model's directional are Moreover, predictions exogenousinputs. with the available evidence. On consistent the other the hand, largely of magnitude overshoot thehistorical remodel'spredictions generally in an the flaws our of Italian small thus assumptions ality, exposing andeconomic markets rationality. openeconomy, perfect in success the model's known economic deNonetheless, predicting in Italyafter 200 BC lend credence to its implications for velopments its successes are two morecontroversial questions. Among comforting in the Italian economyafter developments major,well-documented in thecountryside to wineproduction from and 200 BC: theshift grain rural to urbanareas. The modelalso of freelaborfrom themigration due to makesmorecontroversial predictions: peasantsmigrate mostly slave from exnot owners factors and gain pushfactors; imperial pull lose; and thebulkof Italianwealthderives pansionwhilelandowners andpeoplefrom theprovinces. ofresources theexpropriation from thatlend credence to the modelis the Firstamongthepredictions in Italian wine the to from shift countryside, though production grain In themodel, the thedegree ofthistransition.56 themodeloverestimates to grain risein winepricesrelative pricesmakeswinemoreprofitable in slaves relative to land and freelabormakes The growth thangrain. In reless expensive. slaves cheaperand wine production relatively from to wine and redirect resources Romans grain production, sponse itsgrain andimports itswinesurplus Italy exports shortage.57
55 See Appendix Table 3 fora summaryof the model's predictions. of 36 percentand a fourfold The model's average resultis a decline in grainproduction gain in wine production, gain to a fivefoldgain. These changes resultfroma rangingfroma threefold This magnitudeis fargreater of 51 percentof Italy's land fromgrainto wine production. transfer follows fromits simplifyThe model's overestimate thanthemostlikelyhistorical directly reality. ing assumptionsand is an example of the model's known imprecision.Jongman'sanalysis estimates thatno more than 5 percentof Italian land was convertedto vineyardsat this time. He forItaly's urbanpopulationrequiredonly 1 percentof its culshows thatenoughwine production tivable area, and that20 percentof Italy's cultivablearea would have produced enough wine for where few resources are able to satisfy the entireempire. In this saturatedmarketenvironment the amountof of more wine would reduce prices,in turnlimiting market demand,the production resourcesdedicatedto wine. The model, however,does not account forthisphenomenonbecause on imperialmarket does not have an effect it assumes thatItalianproduction prices. This is given which is typicalin computablegeneralequilibriummodby the small open economyassumption, els. As I discuss, because of thisand othersimplifying assumptionsthe model cannot be used as an analytical tool to make quantitative predictions;however, these limitationsdo not call into thatthemodel conveys. See Jongman, "Slavery." questiontheconceptualstory The model predictsthatwine exportsgrow by a factorof 20 and grain importsby a factor of 17; in fact,grainimports grew by a factorof 6-12. See Garnseyand Sailer, Roman Empire.

This content downloaded from 159.178.22.27 on Wed, 28 Aug 2013 13:42:23 PM All use subject to JSTOR Terms and Conditions

1048

Geraghty

The archaeological evidenceof Italiantradewiththeprovinces and land use and contemporary Romanaccounts thisstory. Grain support to in from five million modii 200 BC to 30imports Italygrew peryear 60 million modiiper yearby 1 BC.58 Wine exports also peakedat this theexcavation ofamphorae, orwinecontainers, revealsignificant time; Italianwine exports to Gaul thatdeclinedoverthe subsequent centuries.59 data also showthatoverhalfof shipwrecks on shipShipwreck routes from to in Gaul occurred the second and first centuries ping Italy BC. Examining theItaliancountryside, archaeological digsthroughout and Etruria a villa indicate of Campania proliferation wine-producing sitesthrough thefirst 22 in with of land Etruria and BC, century percent 15 percent of land nearPompeiidedicated to wine production.61 The most elaborate villas in were inhabited the same 100 largest, roughly withthegrowth, yearperiodthat corresponds peak,and declineof the Italian wine industry. Romans were aware of these Contemporary As Pliny theElderwrites, "Thencomethefavoured of changes. country in this those vine clad hills with their Campania; valleybegins glorious wine. . . famous all theworld and... thesceneoftheseverest comover, between Father Liber and ."62 Father Ceres . . . petition [wine] [grain] A secondmajorprediction of themodelis a significant of migration free laborfrom rural tourban areas.63 In themodel, a combination ofrisurban the of the and annona, ing goodsprices, expansion falling grain thancountry rewhilehigher life, pricesmakescitylifemoreattractive turns to wineproduction motivate landowners to pushpeasants offtheir landsand convert them intovilla estates. thequestion of why Although remains the mass of controversial, peasants migrated migration peasants Based on cityand townrepredicted by themodelis well established. mains andcontemporary written historians estimate that thepenrecords, insula'surban in from 200 BC to 1.2 million two 500,000 population grew 4 centuries later. Usingestimates ofurban death Scheidel calculates rates, that urbanization absorbed 1.8 to 2.2 million between 200 rapid peasants and 1 BC. The massmigration of peasants from thehinterland to Rome andother cities was a phenomenon notseenagaininWestern unhistory tilLondon's boommore than a millennium later.
58 Gamsey and Sailer, Roman Empire. Morley,Metropolis; Tchernia,"Italian Wine." Parker,"Ancient Shipwrecks." Romans: and Jongman, Greene,Archaeology; Arthur, Economy and Society. 62 Morley,Metropolis. The model predictsthatthe proportion of free labor living in urban rises by 20 percentof the total population. The model's predictionof nearly 30 percentof peasants living in cities is consistentwith historians'estimates.See Hopkins, Conquerors, foran estimateof Italy's urban population. Italian Manpower; and Scheidel, "Human Mobilityin Roman Italy, 1." Brunt,

This content downloaded from 159.178.22.27 on Wed, 28 Aug 2013 13:42:23 PM All use subject to JSTOR Terms and Conditions

Impact ofGlobalization

1049

withthe are consistent In twomajorareas,themodel'simplications overestimate thehistorical thepredicted evidence, magnitudes although it does about the more controversial and unWhat, then, say reality. of who and who known lost, migrated, gained questions whypeasants theories aboutwhypeasbecamewealthy? andhowItaly Manyexisting thatprofit-hungry aristocrats and low grain antsmigrated emphasize landand that from their left imperial expansion peasants pricespushed 5 More the first AD. destitute recent theoRomans century by ordinary in the literature, thatpull factors, ries,less prevalent suggest notably urban in cities and most The towns, migration.66 inspired opportunities us to estimate the fraction modelcan speakto thequestion by allowing driven to urban from oftheshift bypushandpull facgrain production therise in urban to themodel,pull factors, tors.According primarily drive 80 percent ofpeasoftheannona, andtheexpansion goodsprices after 200 BC. antmigration to thisconclulendssupport Scheidel'srecent analysis demographic inthe focus onpeasant dislocation ofthepushtheory sion.Thevariations Italian as a result of Hannibal's of the second BC,either pillaging century or constant militarization of thepeasfalling grainprices, countryside, on suchas theories focus growing opportunities, By contrast, pull antry.67 urban which becamemore theriseoftheannonaandItalian production, AS Scheidel's BC and afterwards. in the first analysis century prevalent an increasing of immiItalian citiesrequired showsthat supply growing death ratesexceeded becauseurban their to replenish populations grants with the of more consistent a conclusion birth rates, timing pulltheories favors ofmigration Scheidel's than pulltheories analysis pushtheories.69 that on theories andcastsdoubt overthepushtheories imperial expansion from their lands. them Roman driving peasants by impoverished on who gainedand who lostfrom is unclear The literature imperial lostwhilearistoit is generally expansion, though agreedthat peasants
65 Italian Manpower, and de Neeve, Colonus. Hopkins, Conquerors; Brunt, Scheidel, "Human Mobility in Roman Italy, 1"; Morley, Metropolis; and Jongman, "Slaves." 67Arnold Toynbee argues thatpeasants fled theirfarmsforthe safetyof citywalls duringthe Hannibalic War around in the late thirdcenturyBC. During theirabsence, Hannibal wrecked theirland and aristocrats boughtabandoned, devastatedplots at firesale prices, leaving peasants of in cities and towns. However, P. W. de Neeve observes thatHannibal's destruction destitute is an insufficient explanationforthe widespread ruraldepopuonly a limitedarea of countryside lation apparentin the archaeological record. He argues thatfallinggrain prices motivatedpeasin urban areas. Anotherplausible push theants to abandon theirfarmsforbetteropportunities two centuries made peasants unable to militarization over burden of that the constant is ory sustaintheirlivelihoods on theirfarms,forcingthemto abandon theirhomes and migrateto urban centers.See Toynbee, Hannibal 's Legacy; de Neeve, Colonus; and Hopkins, Conquerors. Morley,Metropolis. "Slaves." Scheidel, "Human Mobilityin Roman Italy, 1"; and Jongman,

This content downloaded from 159.178.22.27 on Wed, 28 Aug 2013 13:42:23 PM All use subject to JSTOR Terms and Conditions

1050

Geraghty

cratsgained.Our model,however, It acknowltellsa different story. because edges thatfalling grainpricesreducefreelabor'sreal income This is almostenmostlaborers were employed in grainproduction. in urban areasand the offset, however, tirely by growing opportunities of the annona,fora minimal net effect on freelabor inexpansion come.70 The model'sstory thus contradicts thosewhobelievethat peasantswere made destitute and slave by falling grainprices,evictions, But it fits PeterGarnsey's that conclusion imports. peasantlivelihoods remained reaunaffected fordifferent by imperial expansion, although sons. WhereasGarnsey to farm their believesthatpeasantscontinued themodelpredicts that maintained undisturbed, smallholdings peasants their incomes conditions and to changing economic onlyby responding to urban from the areas.71 The massive of free labor moving migration to citiesandtowns around over200 twomillion countryside peasants in at the same rate as years, approximately Englishpeasantmigration thenineteenth thatthemodel's story is moreaccucentury suggests rate.72 themodeltellsa morecomplexstory aboutland and slave Overall, owners.Landowners lose while slave ownersgain, as fallinggrain less profitable labormakegrain pricesand morescarcefree production andrising winepricesandmoreabundant slavesmakewineproduction morelucrative. Wealth to own thusshifts to aristocrats wealthy enough slavesat theexpenseof middling landowners who cannot afford slaves and for whomselling their landsandmigrating to citiesis an unappealinthesepredictions ontothepast,aristocrats' ing option.73 Projecting comeswouldhave declinedin thesecondcentury as BC, grainprices as wineproduction flourished. fell, onlytoriseagaintonewheights in thesecThereis someevidence of trouble forRomanlandowners ond century BC and substantial returns evidence forhighand climbing to owning slaves.Landlords'ability to purchase largeplotsof landin thesecondcentury at that time.74 The decline suggests depressed prices inthestructure andsize ofrural ofproresidences andthedepopulation ductiveareas indicate thatformerly had fallen on aristocrats wealthy hardtimes.75 On theother thepriceof slaves,orthepresent value hand,
70The model that realwagesdeclineby 9 percent becausedownward from predicts pressure winepricesmorethan offsets from urban falling grain pricesand rising upward pressure rising andtheexpansion oftheannona. prices 71 Cities. Garnsey, Fora comparison to English inRomanItaly, 1." see Scheidel, "Human migration, Mobility 73The model that landrents decline whileslaveprices predicts bya totalof46 percent, grow of 160percent. As a result, aristocrats' realincome bya total grows by92 percent. Hopkins, Conquerors. andde Neeve,Colonus. Morley, Metropolis',

This content downloaded from 159.178.22.27 on Wed, 28 Aug 2013 13:42:23 PM All use subject to JSTOR Terms and Conditions

Impact ofGlobalization

1051

thisperiod.A catalogueof of slave returns, during grewsignificantly at Delphiindicates a 50 percent their freedom thepricesslavespaid for 200 BC.76 or the value of slave after in slave returns, present prices, gain of slave pricesin contemporary Mentions sources, including Livy and between 200 BC an eighttotenfold Tacitus, gainin slaveprices suggest andAD200. and the returns to slave owning flourished As the wine industry enormous and forowners accumulated slave unprecedented grew, tunes.RichardDuncan-Jones's catalogueof Romanprivatefortunes of RomanswithmorethanHS 200 million revealsthatthe number era.78 Sitesin Latiumand Campania bytheearlyimperial grewsixfold and of muchhigher morecommon, became larger, qualityconstrucafter100 BC, and praiseof magnificent villas in tionand decoration of fertile theseareas replaceddescriptions Indeed,the grainlands.79 and finer witnessed the construction of more elaborate BC first century villasthan everbefore. becameimmensely Romansociety becamewealthier As slaveowners at the time. Seneca theunimpresa evident compares unequal, reality circa of general and consulScipioAfricanus, sive size of theresidence AD.81 com200 BC,to thoseof his own day in thefirst century Livy doubled durforSenatemembership that thewealth ments requirements based on A calculation of Gini coefficients this Livy's figperiod.82 ing rosefrom oftheItalian that theGinicoefficient uressuggest population theaverBC to 0.23 100 yearslater. 0.11 in thesecondcentury Indeed, 200 timesmoreincomethana peasant'ssubsisgenerated age senator Romanfortunes ofHS 400 million in the tence early imperiumP wages totals and sixteenthby a factor seventeenth-century English surpassed and Gini coefficients of Romanlandof 18-72 in wheatequivalent, The from the modern to British early age.84 figures compare ownership an explanation modelalso provides whyand how thisoccurred not theimporting of vast butthrough of the through exploitation peasants,
76 that rosefrom 403 drachmae indicate prices p. 159. The inscriptions Hopkins, Conquerors, BC. inthefinal twocenturies to641 drachmae that Romanslave pricesin the twoconclusions: Jones's HereI combine Jones, "Slavery." BC than Athenian ad wereeight to tentimes secondcentury century pricesin thefourth higher totheAthenian BCwerelow compared inthesecondcentury andthat Roman prices. prices 7 DuncanJones, Economy, p. 343. 79 andMorley, Rathbone, "Development"; Metropolis. 80 Dench,andPatterson, Curti, "Archaeology." on senator was therateofreturn I assumefrom Cato that landholdings Economy. Jongman, 6 percent. 84Duncaninpopulation exceededBritain Jones, bya facEconomy, p. 5. The RomanEmpire torof 18.
Frank,Economic Survey. Shatzman,Senatorial Wealth.

This content downloaded from 159.178.22.27 on Wed, 28 Aug 2013 13:42:23 PM All use subject to JSTOR Terms and Conditions

1052

Geraghty

markets for of slaves and the openingof new provincial quantities slave-intensive production. the model addresses thequestion of the sourcesof Roman Finally, from twopowealth after 200 BC. In themodelRomanwealth derives tential sources theexpropriation of provincial wealth, including grain the literaandslaves,andthegainsfrom themarket economy. Although tureoftendescribesRome's seizureof vast quantities of provincial of how much wealthin the form of tribute and slaves,the questions or marits gainsderived from Italygainedand whether expropriation factor ketsare hardly that total addressed.85 The modelpredicts Italy's market inteincome rises25 percent this 5 from during period: percent In other and 20 percent from seizuresof wealthand people.86 gration the Italian is driven words, economy's expansion primarily by theexof provincial but also significantly wealth, propriation by gains from trade. The prediction further as onlyprovokes questions, fewdataexist to corroborate ordeny thisinterpretation. That the model's storyfit the historical evidence in two wellestablished Italian from aspectsof developmentthe ruraltransition and the migration of peasantsto citiesand grainto wine production - makesitsmorecontroversial towns Theseconpredictions plausible. troversial inthehistorical as well.The find evidence predictions support model's success also lends credenceto its assumptions of a wellmarket in theRomanempire rationand of economic integrated system the Romans. The to which the model's alityamong predictions degree of magnitude reovershoots observed historical however, phenomena, veal thatits oversimplified the more comfail to assumptions capture ofRoman times. plexreality CONCLUSION In thisarticle I construct a simplified modelof general equilibrium theRomaneconomy, in order to shedlight bothon thevalidity of my on and thebig questions of Italian economic assumptions development. I proceedto defend and explainthemodeland to describe thechanges to the model that from Rome's resulted exogenous expansion. imperial the model's predictions obof magnitude tendto overstate Although servedphenomena, its directional the conclusions drawn predictions - tell a broad from theunderlying logic on whichthemodelis based
85 Land. Whittaker, In this calculation, I count gains fromchanges in commodityprices and returns to labor, and gains fromchanges in taxes collected and land, and slaveryas gains frommarketintegration the slave endowmentas gains fromprovincialexploitation.

This content downloaded from 159.178.22.27 on Wed, 28 Aug 2013 13:42:23 PM All use subject to JSTOR Terms and Conditions

Impact ofGlobalization

1053

thatis confirmed of Romaneconomic development by theavailstory fit This between themodeland theavailable evidence. able historical themodel's other and its assumption of evidencesupports predictions Roman markets and economic motivation and imperial well-integrated rationalism. research can applysuchmodelsto study other Further aspectsof anADorthe ofItalyafter thefirst suchas thedecline cient century history, in inflation the later or use them to refine of empire, impact monetary markets suffered abouttheRomaneconomy. ourassumptions Imperial in this and costs not reflected fromimmensecommunication lags but economics also Romanswerenotsolelymotivated model.87 by by andconventions.88 socialpressures, rules-of-thumb, Theywerenotpure on maintaining their subsistence butinstead focused maximizers profit their riskof ruin.89 and reducing Italywas a majorpartof theRoman theemaffected and commodity pricesthroughout economy probably makeita abouttheserealities assumptions pire.The model'ssimplified of Italianeconomic thebroadstory tool fortelling useful development tool.Introducing thesecomplexiitsuse as a precise butlimit predictive of a morerealistic ties intothemodelwouldmakeits story portrayal and would refine, butnot substantially Italianeconomicdevelopment ourunderstanding. change, of ourunderstanding The basic powerof a modelis itsformalization underofformalized The combination ofcause andeffect relationships. make more allows us to available data the and powerful prestanding thanwe can makewiththeavailabledata alone. This makes dictions ourdataarelimwhere ofancient for thestudy modelsa useful history, modelsuch as thatused thepowerof an economic ited.Furthermore, that thecomplex considers of a coherent hereis itscrystallization story in an and delivers of manysimultaneous interactions economy changes This quantified a broad seriesof testable storyis more predictions. fohistorical models often than the more testable and comprehensive I bein these cused on singlequestions isolation. advantages, Despite to employ ecoalikehavehesitated and economists historians lievethat because the the ancient world of in their models nomic degreeof study that era. for is unknowable that thesemodelstypically require precision outthe not necessitate do models economic However, knowing precise Thebroad andaccurate totella basic,coherent, lineofan economy story.
87 andDuncanJones, Yeo, "LandandSea Transportation"; Economy. Morley, Metropolis', Economy;Temin,"MarketEconomy";Purcell,"Wine"; and Hopkins, Finley,Ancient Conquerors. Cities. Garnsey,

This content downloaded from 159.178.22.27 on Wed, 28 Aug 2013 13:42:23 PM All use subject to JSTOR Terms and Conditions

1054

Geraghty

use of economic modelsin thestudy of thedistant pastwouldmakea on ourunderstanding ofancient history. significant positive impact

A Computable General Appendix: Equilibrium ModelofRoman Italy

- grain The modelhas four sectors (#),andurban (G), wine(W), livestock produc- land(R), labor(L), andslaves(S). Two intion(U) andthree ofproduction factors is termediate as factor Manure also function (M) and feed(GB)9 inputs. goods,manure beand functions livestock feed consists of The produced by entirely grain. production low areCobb-Douglas in form with constant returns to scale.
+ QMG + 0SG= 1 G = AcLeLGReRGKfMG?fSG where 0LG+ 6RG W = AwLeLWReRW\fMWtf)sw where0LW+0RW+0w+0sw = 1 + 0GB + 0SB=\ B = ABRdRBGeGB^SB where 0RB U = AuL6W^slJ where 0LU+ 0su=l M=AmB

In equilibrium, factor markets balancesupply, anddemand, degivenexogenously, rivedfrom thesectors in whicheach factor is employed. Labordemand derives from thegrain wine(Z,^),andurban landfrom thegrain wine(Rw), (LG)9 (Ly) sectors; (RG), and livestock and slaves from thegrain(SG),wine (Sw),livestock (RB) sectors; (SB), and urban(Su) sectors. Manure'ssupply is givenby theproduction function formanure aboveandbydemand from thegrain (MG)andwine(Mw)sectors
L = LG + Lw + Lu S = SG + Sw + SB + Su R = RG+ Rw + RB M=MG + MW

markets are suppliedby imports and deand domestic Commodity production mandedby domestic consumers and exports. The grainsupplycomes from private market forfree distributions (GM),government imports (GA),andItalian imports grain aredetermined ofgrain (G). Government production imports bythepercapitaamount distributed and slavesdemand {Lv). Feed (GB),laborers, (gA)andtheurban population with laborandslavesconsuming their grain, percapitarequirements (gi andgs) multitotal factor endowments (Z,andS): pliedbytheir + GM+G = GB+ gLL+gsS GA where GA=gALu. Wineandurban goodsare supplied (W and U) anddesolelyby Italian production manded andlaborers. Landlords andlaborers (WxandUx),landlords, byexports spend fractions andrjLW on wineandr/RU andrjLU on urban total incomes (rjRW goods)oftheir (YRand YL)on wineandurban goods:
W=Wx+fiRwYR+ riLwYL U=Ux+rjRUYR + rlLUYL

This content downloaded from 159.178.22.27 on Wed, 28 Aug 2013 13:42:23 PM All use subject to JSTOR Terms and Conditions

Impact ofGlobalization

1055

is nontraded. Otherwise thelivestock Unlike wine,andurban grain, goods,livestock is similar: market

+ rjLBYL B = rjRBYR

totalincome(YR) from land rents to Landlords receivetheir (dR) and thereturn costs.At equilibrium slavereturns slaves(m) minus subsistence equateto slave costs, costs(pGgs) andprice(ps). The nominal which consist ofsubsistence wagerate(wL)is and other commodities subsistence grainrequirements (pGgi) (yL). The spentupon to the at rural has two nominal urban wage, equilibrium equivalent wage, components, sector thewage paid by theurban (wu) and theincome provided by thegrain namely dole(pGgA)' = dR + m-pGgsS YR m=PGgs+Ps
WL=PGgL+yL = Wu+PGgA WL

that whichstates thevalue of is thebalanceof tradecondition, The finalequation valueincludes wineexports urban Export (pwWx), equalsthevalueofimports. exports value includes (7). Import grain (puUx),and taxesleviedupontheprovinces exports markets (pgGa),grainimported by theprivate (pgGm), by thegovernment imported theprovinces from andslavesimported (psS): + psS PwWx+PuUx+T=pGGA+pGGM which Excel to examine itsresults, I solvedthemodelby handandused Microsoft Tables 1-7. inAppendix areshown
Table 1 Appendix VARIABLE DEFINITIONS Variables Exogenous Variables Endogenous

R: Landendowment G: Totalgrain byItalian producers produced Totalwineproduced L: Laborendowment W\ byItalian producers S: Slave endowment B: Totallivestock producers produced byItalian U: Totalurban producers production byItalian price pG:Grain ofmanure M: Totalamount Wineprice pw\ feed ofgrain usedfor goodsprice GF Totalamount Pu: Urban market amount of Total subsisPer GM: imported bytheprivate grain capita gs: from Totalamount ofwineexported for Wx: tence Italy requirement from ofurban slaves Ux:Totalamount Italy production exported theprovinces from oftaxescollected T: Totalamount gL:Percapitasubsisincome oflandlord Totalamount for YLL: tence requirement than on goodsother oflaborincome Percapitaamount labor spent grain yL\ al- ps'.Annualized priceofa slave gA:Percapitagrain rent on land free labor d: Annual lotment for inurban centers pB:Priceoflivestock Pm.Priceofmanure wL:Wagerate paidto labor sector wy.Wagerate production paidto laborin urban to slave m:Rateofreturn

This content downloaded from 159.178.22.27 on Wed, 28 Aug 2013 13:42:23 PM All use subject to JSTOR Terms and Conditions

1056

Geraghty
Appendix Table 2 EXOGENOUS SHOCK DIRECTION AND MAGNITUDES

Variable Exogenous Laborendowment (I) Landendowment (R) Slave endowment (5) Grain price(pG) Wineprice0M Urban goodsprice(pv) Annona allotment (gA) Sources:See thetext.

% Change Average -2 +6.5 +140 -20 +20 +10 +33

Low % Change -5 0 +100 -20 +10 +5 +10

High% Change +10 +10 +200 +20 +30 +15 +50

Appendix Table 3 MODEL OUTPUTS Variables Endogenous Output Av. % Low% High% Change Change Change Sector Shares Factor Av. Low High Sector Share Initial Final Final Final

Grain -36 -55 -5 0.875 0.43 0.37 0.51 (G) kLG +410 +290 +500 kLW 0.025 0.29 0.08 0.30 Wine(JF) 0.41 Livestock +16 -5 +26 0.10 0.28 0.0 (5) kw Urban +100 +40 +200 0.31 0.26 0.59 (U) kRG 0.85 +16 -5 +26 0.56 0.43 0.61 kRW 0.05 Feed(G5) Manure +54 +33 +60 0.13 0.01 0.18 kRB 0.10 (M) GrainIm.(GM) 0.58 0.38 1 +1,670 +830 +2,270 kMG 0.80 Wineexports 0.42 0 0.62 0.20 +2,080 +970 +2,560 Aw (Wx) 0.29 Urban +560 +590 0.05 0.00 0 +1,030 (Ux) kSG exports Taxes (7) 0.60 0.23 0.82 +440 +140 +480 0.40 ksw Ruralwage(wL) +12 0.15 0.10 0.03 0.24 -90 -15 kSB Urban -44 0.40 0.30 0.15 0.39 -60 +23 ksu wage(wt/) -46 -58 +21 Rent(rf) Slave return +64 -16 +81 (m) Livestock -12 -11 +16 price(pB) - 12 - 11 Manure +16 price(pM) Landlord income +92 -45 +170 (YR) Slave price(ps) +160 -42 +200 Sources:See thetext and modelas specified in theAppendix. I solvedthemodelusingMicrosoft Excel. The modelis basedon exogenous income factor income elasticities, shocks, shares, andother initial conditions stated inAppendix Tables2, 4, 5, 6, and7, respectively.

This content downloaded from 159.178.22.27 on Wed, 28 Aug 2013 13:42:23 PM All use subject to JSTOR Terms and Conditions

ofGlobalization Impact
Appendix Table 4 EXOGENOUS FACTOR INCOME SHARES Income Share Factor Value Average Low Value

1057

HighValue

0.4 0.3 0.5 6LG 0.4 0.3 0.5 6RG 0.15 0.1 0.2 Omg 0.05 0.01 0.10 0SG 0.15 0.1 0.2 6LW 0.3 0.2 0.4 0RW 0.1 0.05 0.15 Bmw 0.45 0.3 0.6 0sw 0.5 0.4 0.6 Orb 0.35 0.3 0.4 6GB 0.15 0.1 0.2 0SB 0.5 0.4 0.7 6W 0.5 0.3 0.6 0su of income derived from thesale of an output that income shareis theproportion Note:A factor in itsproduction. For example, theshareof totalrevenue from is earned grain by each factor sees as income is labor'sfactor income shareofgrain that thelaborproducing grain production Sources:See thetext.
(Ota).

Table 5 Appendix INITIAL FACTOR SECTOR SHARES Factor Sector Share Value Average Low Value HighValue

0.8 0.95 0.875 kLG 0.01 0.1 0.025 kLW 0.05 0.15 0.10 kLU 0.9 0.85 0.7 kRG 0.10 0.01 0.05 kRW 0.15 0.05 0.1 kRB 0.9 0.7 0.8 kMG 0.3 0.1 0.2 Xmw 0.10 0.05 0.01 kSG 0.3 0.5 0.4 ksw 0.1 0.2 0.15 ksB 0.5 0.3 0.4 ksu ofa sector. dedicated to theproduction ofthefactor shareis theproportion sector Note:A factor shareof is thefactor sector to grain of labordedicated theproportion For example, production tograin laborwith (kLG). respect Sources:See thetext.

This content downloaded from 159.178.22.27 on Wed, 28 Aug 2013 13:42:23 PM All use subject to JSTOR Terms and Conditions

1058

Geraghty
Appendix Table 6 EXOGENOUS INCOME ELASTICITIES

Income Elasticity

Value Average

Low Value

HighValue

0.7 0.6 0.5 nLW 0.4 0.3 0.5 rjLB 0.6 0.5 0.7 nRW 0.25 0.2 0.3 nRU 0.15 0.1 0.2 nRB Notes:An income is thefraction the oftotal income on eachgood.Forexample, elasticity spent fraction of their totalincomethatlandowners of landspendon wineis theincomeelasticity owners forwine{nRW). The wineand livestock incomeelasticities of laborare calculated after forlabor'sgrain theassumption is that laborers amount accounting consumption; spenda fixed on grain for subsistence before wineandlivestock. consuming Sources:See thetext.

Appendix Table 7 CALCULATION OF OTHER INITIAL CONDITIONS a Value Method

The annonadistributed subsistence or40 modii, to its approximate requirements, Laborreceived subsistence to 40 modiin also equated recipients. wages,which real terms.a of Thus,urbanwages and freegrainprovided equal proportions total wages. freeand slave,was 4.9 million.5 Each person aGA 0.015 In 200bc Italy'stotalpopulation, consumedan average of 35 modii of grain per annum. Italy imported aGM 0.03 fivemillion modiiof graintotal, of whichat most1.6 million aG 0.96 approximately fortheannona.0 and theannonaformed 3% and 1.5% Thus,imports provided ofthetotal anddomestic theremainder. grain supply production was slightly less thannine timesits slave population.5 aGB 0.05 Italy's freepopulation consumednine timesmore grainthanthe slave Thus, the freepopulation aG$ 0.1 I Feed was a relatively smallpartof theeconomy at thattime.0 aGL 0.85 population. assumethat laborcomprised slaves 10%,and feed 85% of totalgrain demand, 5%. landlord incomewas HS 267 million of which60% was aWR 0.56 Minimum per annum, Labor incomewas HS 200 per capita,of which15% was awi 0.24 spentuponwine.d amounted to approximately 20% of wine demand.f awx 0.20 spenton wine.eExports This providesthatwine demandwas 56% landlords, 24% labor,and 20% exports. Landlords and laborers incomes on livestock.*1 Given 60% and 10% oftheir aBR 0.6 spent thefigures demandcomes from landlords and 40% aBL 0.4 above,60% of livestock from laborers. I assume that,like wine, 20% of urbanproducts were exported. Landlords aUR 0.8 consumed theremaining 80%. aux 0.2 amounted to approximately inflows based on 10% of provincial axw 0.075 Total exports calculations of slave and grainimport and wineexport values.^ I assumethat axu 0.025 wineexports werethree times theamount ofurban olxt 0.9 exports. five aMA 0.06 In 200 bc, Italy possessed 500,000 slaves and imported approximately million modiiof grain.bc Slaves cost HS 200 per annum and grainHS 5 per aMG 0.14 modius.g consisted of 80% slaves, 14% private and Thus,totalimports aMS 0.8 grain, 6% government grain.

awG 0.5 awU 0.5

This content downloaded from 159.178.22.27 on Wed, 28 Aug 2013 13:42:23 PM All use subject to JSTOR Terms and Conditions

Impact ofGlobalization
Appendix Table 7 - continued a Value Method

1059

40 million I assume 40% was cultivated. iugeraof land,ofwhich Italycontained theaverage was HS 400, andthat thereturn that on landwas priceperiugerum costHS 400 perannum.8 Thismeans 6%.hItalypossessed500,000slavesthat for income andslaves35%. that landaccounts 65% oflandlord amortized Slave pricesof HS 2,000 provide amQ 0.5 pricesof HS 200. Slave subsistence to HS 200.gThus,slave costswere50% priceand 50% costsalso amounted amp 0.5 subsistence. 75% of theirincomeon grainand 25% on other awG 0.75 Labor spentapproximately goods.a consumption any 0.25 e Its of 4.4 million earnedHS 200 percapitaperannum. anL 0.60 Italy'sfreepopulation g of earned returns of 400 slave HS annum. ans 0.14 population 500,000 percapitaper 40 million of of which was under 40% cultivation. 0.25 land, iugera Italy possessed anR landpricewas HS 400, andtheaverage As above,theaverage return was 6%.h anT 0.01 Taxespaidfor offivemillion modii at HS 4 permodius.c grain imports on theItalianeconomy'sinitial Notes'. This tableprovides conditions around myassumptions 200 bc where notalready income factor sector or income elasticshares, shares, givenbyfactor in thistable(in order)are theproportion ities.Included of urbanlabor'stotalwages derived theproduction of urban of grain consumed in Italy from goodsand theannona;theproportion andtheannona;theproportion from Italian ofeachcommodity farms, imports, grain originating andurban and (for wine,livestock, slaves,landlords, goods) consumed by labor, output (grain, in theform of wine,urban of totalexports goods,and taxes;the grain)as feed;theproportion in theform oftotalimports of theannona,other and slaves;theproportion of grain, proportion landand slaves;theproportion earned from of slave pricesderived landlord income from subof laborincome sistence costsand slave prices;theproportion on grain and other spent goods; oftotal GDP derived from andtheproportion labor, slaves,land,andtaxes. Sources'. PeasantsandFood, p. 96. Cities, aGarnsey bBrunt ItalianManpower, p. 121. c CornSupply, Rickman, pp. 36-42. d and Society, Economy p. 195;andTable 6. Jongman, PeasantsandFood, p. 96; andTable 6. , Cities, eGarnsey f "Italian Tchernia, Wine," p. 92. gRickman, CornSupply, 151;andDuncanJones, p. Economy, p. 12. hDuncanJones, Economy, pp. 51-52. 1 Roman andSailer, Empire. Garnsey ] Scheidel, inRomanItaly, "Human 2." Mobility olrr 0.65 aRS 0.35

REFERENCES
and Land-UseAround the Paul. Romansin Northern Arthur, Campania:Settlement Basin.London:British Schoolat Rome,1991. Massicoand theGarigliano 225 B.C.-A.D. 14. Oxford: Clarendon P. A. ItalianManpower, Brunt, Press,1971. Haswell.TheEconomics Third Colin,andMargaret Clark, ofSubsistence Agriculture. 1967. Edition. London:Macmillan, EmmaDench,andJohn R. Patterson. "TheArchaeology ofCentral Emmanuele, Curti, Trends and Approaches." TheJournal RomanItaly:Recent and Southern ofRomanStudies 86 (1996): 170-89. DuncanRichard.The Economy Studies. Jones, of theRomanEmpire:Quantitative Press,1982. Cambridge: Cambridge University

This content downloaded from 159.178.22.27 on Wed, 28 Aug 2013 13:42:23 PM All use subject to JSTOR Terms and Conditions

1060

Geraghty

in theRomanWorld." and its Social Consequences Evans,J.K. "WheatProduction TheClassicalQuarterly 31 (1981): 428-42. of California Press, Finley,Moses I. The Ancient University Economy. Berkeley: 1973. V: Romeand Italyof An Economic Rome,Volume Frank, Tenney. Survey ofAncient theEmpire. NJ:Pageant Paterson, Books,1959. Peter.Cities, Ed. Walter ScheiPeasants,and Food in Classical Antiquity. Garnsey, del. Cambridge: Press,1998. Cambridge University . Famineand Food Supplyin theGraeco-Roman World: Responsesto Risk and Crisis.Cambridge: 1988. Cambridge University Press, and Culand Richard Sailer.TheRomanEmpire:Economy, Peter, Society, Garnsey, ture. of California 1987. Press, Berkeley: University Kevin. TheArchaeology London:B. T. Batsford Ltd, Greene, of theRomanEmpire. 1986. WilliamV. "Towardsa Studyof theRomanSlave Trade."In TheSeaborne Harris, Commerce Rome:Studies inArchaeology and History, edited byJ.H. ofAncient D'ArmsandE.C. Kopff. Rome:American at Rome,1990. Academy Keith.Conquerors Vol. and Slaves: SociologicalStudies in RomanHistory. Hopkins, 1. Cambridge: 1978. Press, Cambridge University . "Taxes and Tradein theRomanEmpire (200 B.C.-A.D. 400)." TheJournal Studies 70 (1980): 101-25. ofRoman A. H. M. "Slaveryin theAncient Review9, World."The EconomicHistory Jones, no. 2 (1956): 185-99. Willem.TheEconomy and Society J. C. Gieben, Jongman, ofPompeii.Amsterdam: 1988. . "Slavery andtheGrowth ofRome:The Transformation ofItalyintheSecond and Firstcenturies EdB.C.E." In Rome the Cosmopolis, editedby Catherine 2003. wardsandGregWoolf,100-22.Cambridge: Press, Cambridge University the and Investment on EstatesinRomanEgypt Kehoe,DennisP. Management during Bonn:Dr. Rudolf HabeltGmbh, 1992. EarlyEmpire. 200 BC-200 AD." HarKessler,David. "FeedingRome: The RomanGrainMarket, vardUndergraduate Economic 2004. in Archives. Available Harvard Thesis, "Land in Roman and the Laurence, Ray. Transport Italy:Costs,Practice, Economy." In Trade,Traders, and theAncient andChristopher edited City, byHelenParkins Smith. London:Routledge, 1998. Neville.Metropolis and Hinterland: TheCity Morley, ofRomeand theItalianEcon200 B.C-A.D.200. Cambridge: Press,1996. omy, Cambridge University de Neeve,P. W. Colonus:Private theRepublic in Roman Farm-Tenancy Italyduring and theearly Amsterdam: J.C. Gieben,1984. Principate. This Journal 51, no. 1 O'Rourke,Kevin. "Did the GreatIrishFamineMatter?" 1-22. (1991): A. J.Ancient and theRoman Provinces. OxParker, Shipwrecks oftheMediterranean BAR International ford: Series,1992. WilliamD. Slavery Trade. Phillips, fromRoman Timesto theEarly Transatlantic ofMinnesota Press,1985. Minneapolis: University N. "Wine and Wealthin Ancient Purcell, Italy."TheJournal of RomanStudies75 (1985): 1-19. DominicW. "The Development of Agriculture in the'AgerCosanus' durRathbone, the Roman Problems of Evidence and TheJournal ing Republic: Interpretation." Studies 1\ (1981): 10-23. ofRoman

This content downloaded from 159.178.22.27 on Wed, 28 Aug 2013 13:42:23 PM All use subject to JSTOR Terms and Conditions

Impact ofGlobalization

1061

The Corn Supplyof Ancient Rome. Oxford:Clarendon Rickman, Press, Geoffrey. 1980. In Landusein theRoman Rosafio, Pasquale."Slaves and Coloniin theVilla System." edited Peter ErikSkydsgaard, and Jens 145Carlsen, Orsted, Empire, by Jesper 58. Rome:Analecta RomanaInstituti Danici,1994. theDebate overGrowth in theAncient In The Sailer,Richard. "Framing Economy." Ancient editedby WalterScheideland Sittavon Reden.New York: Economy, 2002. Routledge, in RomanItaly,1: The Free Population." The Scheidel,Walter."HumanMobility Roman Studies 94 Journal of (2004): 1-26. in RomanItaly, 2: The Slave Population." . "HumanMobility TheJournal of Roman Studies 95 (2005). and Sittavon Reden."Introduction." In TheAncient edWalter, Scheidel, Economy, itedby Walter Scheideland Sittavon Reden.Edinburgh: Edinburgh University 2002. Press, W. Transforming Traditional New Haven,CT: Yale Theodore Schultz, Agriculture. Press,1964. University and RomanPolitics.Brussels:LatomusRevue Israel.SenatorialWealth Shatzman, 1975. D'ErudesLatines, New York:Longman, David.Romeand herEmpire. 2003. Shorter, Reed. International Economics. ThirdEdition. London: Sodersten, Bo, and Geoffrey 1994. Macmillan, inRoman M. S. ArableCultivation Italyc.200 B.C-c.A.D. 100. London:SociSpurr, thePromotion ofRomanStudies, 1986. etyfor in at "Italian Wine Gaul the End of theRepublic." In Tradein the Andre. Tchernia, edited Peter Keith and Ancient C. R. Whittaker, Economy, by Garnsey, Hopkins, ofCalifornia 87-105. Berkeley: Press,1983. University in theEarlyRomanEmpire." "A Market TheJournal Peter. Temin, Economy ofRomanStudies 91 (2001): 169-81. 's Legacy: TheHannibalicWar Arnold J.Hannibal 's Effects on Roman Life, Toynbee, II: Romeand herNeighbors Hannibal'sExit.London:Oxford UniVolume after 1995. Press, versity NY: Cornell K. D. Roman Ithaca, Press,1970. White, Farming. University and Tradein theRomanEmpire.Brookfield, VT: VaC. R. Land, City, Whittaker, 1993. rorium, of theCornLaws just Prior to Repeal."ExploraG. "The Impact Williamson, Jeffrey 27 (1990): 123-56. tions inEconomic History TheKuznets Memorial Lectures and History: . Inequality, Poverty, oftheEcoYale University. Oxford: Basil Blackwell, 1991. nomic Growth Center, in Imperial and ProYeo, Cedric."Land and Sea Transportation Italy."Transactions 11 221-44. the American Philological Society (1946): of ceedings

This content downloaded from 159.178.22.27 on Wed, 28 Aug 2013 13:42:23 PM All use subject to JSTOR Terms and Conditions