The Labor Market of the Early Roman...

Peter Temin

The Labor Market of the Early Roman EmpireAncient Rome was a slave society Hopkins was the rst to assertthat Rome was one of only ve slave societies in recorded historya view adopted quickly by Finley This characterization is impor-tant because slavery is used as a sign of a non-market economywhich in turn is a classi cation within the Gemeinschaft undGesellschaft conception of history as the account of progress fromone to the other The former is the informal economy of familiesand villages where social rules and obligations are dominantin uences on behavior The latter is the economy characteristic ofmodern urban societies In the classic view of Toumlnnies ldquoa periodof Gesellschaft follows a period of Gemeinschaftrdquo Toumlnnies citingMarx described the mechanism behind this progress ldquoas a processof increasing urbanizationrdquo Marx and Toumlnnies were trying tomake sense of changes in nineteenth-century society in which thedramatic rise of urbanization loomed large1

Polanyi in The Great Transformation located the center of thistransition in the labor market He argued that labor markets in themodern sense did not exist before the Industrial Revolution andthe Poor Laws that accompanied it in England This view is con-sonant with Weberrsquos judgment that a critical component of capi-talism was free labor One way to identify a period of Gesellschaft isthe prevalence of urban life but an even more important key is

copy 2004 by the Massachusetts Institute of Technology and The Journal of InterdisciplinaryHistory Inc

Keith Hopkins Conquerors and Slaves (Cambridge 1978) Moses Finley Ancient Slaveryand Modern Ideology (London 1980) Ferdinand Toumlnnies (trans Charles P Loomis) Commu-nity and Society (East Lansing 1957) 231ndash233 (orig pub as Gemeinschaft und Gesellschaft [Leip-zig 1887]) Karl Marx (trans Samuel Moore and Edwar Aveling) Capital (London 1970orig pub Hamburg 1867) I 352

THE ROMAN LABOR MARKET

Journal of Interdisciplinary History xxxiv4 (Spring 2004) 513ndash538

Peter Temin is Elisha Gray II Professor of Economics Massachusetts Institute of TechnologyHe is the author of ldquoEvolutionary Historyrdquo Journal of Interdisciplinary History XXVIII (1998)405ndash415 editor of An Economic History of New England (Cambridge Mass 2000)

The author would like to thank Roger Bagnall Alan Bowman Richard Duncan-Jones Peter Garnsey Keith Hopkins Herbert Klein Sheilagh Ogilvie Dominic Rathboneand Joshua Sosin for help and advice He also thanks the Warden and Fellows of Nuf eldCollege University of Oxford and the Simon R Guggenheim Foundation for research sup-port

the presence of a labor market in which the services of labor canbe bought and sold which Polanyi argued emerged only twocenturies ago2

The widespread use of slavery in Rome is taken as a sign thatGemeinschaft dominated the life of the Roman republic and theearly Roman empire Finley and others following his lead arguedthat ancient economies were not market economies but an alter-nate even primitive form of organization Finley stated ldquoIn earlysocieties free hired labor (though widely documented) was spas-modic casual marginalrdquo According to Hopkins the early Ro-man empire was ldquoa society which had no labor marketrdquo Hopkinsspeculated as a result ldquoIn a society without a market in free laborrecruitment by force (ie slavery) was probably the only methodof securing large numbers of full-time dependents with particularskillsrdquo3

This view is mistaken A variety of evidence indicates thatRome had a functioning labor market and a uni ed labor forceWage dispersion in the early Roman empire to the extent that weknow it is indistinguishable from that in pre-industrial EuropeRoman labor contracts have a distinctly modern allocation of risksand rewards In addition Roman slavery was so different frommodern slavery that it did not indicate the presence of non-market traditional actions Instead ancient Roman slavery was anintegral part of a labor force that shares many characteristics withlabor forces in other advanced agricultural societies Contrary toFinley who asserted ldquo[A]ncient slavery co-existed with other

514 | PETER TEMIN

2 Karl Polanyi The Great Transformation (New York 1944) Max Weber (trans Talcott Par-sons) The Protestant Ethic and the Spirit of Capitalism (New York 1930 orig pub in German1905) More recent views reject the opposition of Gemeinschaft and Gesellschaft in favor of amore complex view that sees both ideal types as present in a wide variety of societies Thisview which also derives from Toumlnnies questions which form of society is dominant notubiquitous interpreting history as the shifting balance between Gemeinschaft and Gesellschaftrather than a total transition from one to the other A society might be regarded as dominatedby one mode or the other but the sense that progress has eliminated Gemeinschaft is gone SeeThomas Bender Community and Social Change in America (New Brunswick NJ 1978)3 Ian Morris ldquoForewordrdquo in Moses I Finley The Ancient Economy (Berkeley 1999 origpub 1973) 1ndash37 68 Finley Ancient Slavery 68 In ldquoFurther Thoughtsrdquo Finley reaf rmed hispositions with ldquonuancingrdquo that ldquofree hired labor was casual and seasonalrdquo and that ldquothere wasno genuine competition between slave and free laborersrdquo (185ndash186) Following Brunt heacknowledged abundant free laborers in the largest cities but he insisted that their employ-ment was ldquostrictly speaking casualrdquo (Finley ldquoFurther Thoughtsrdquo in idem Ancient Economy177ndash207) P A Brunt ldquoFree Labor and Public Works at Romerdquo Journal of Roman StudiesLXX (1980) 81ndash100 Hopkins Conquerors and Slaves 14 111

forms of dependent labor not with free wage-laborrdquo andSchiavone who added recently that ldquoslavery led to the even-tual stagnation of the [Roman economic] system blocking offother pathsrdquo the analysis herein nds that free hired laborwas widespread and that ancient slavery was part of a uni edlabor force in the early Roman empire not a barrier to economicprogress4

the roman labor market A functioning labor market couples alabor demand with a labor supply Two conditions must be lledat least partially Workers must be free to change their economicactivity andor their location and they must be paid somethingcommensurate with their labor productivity to indicate to themwhich kind of work to choose Contemporary studies maintainthat labor needs to be mobile enough to bring wages for work ofequal skill near equality Although this stipulation does not meanthat everyone has to change jobs with great frequency enoughpeople must be able and willing to do so to keep payments to laborfrom being excessively higher or lower than the wages of compa-rable work in other locations or activities Even in the UnitedStates today which contains the most exible labor market in his-tory wages for comparable jobs are not completely equalized5

When these conditions are not ful lled there is no labor mar-ket or perhaps only local isolated labor markets People mightnot be able to change their economic activities due to hereditaryor guild restrictions They might be restricted in what they canearn or be entitled to income for reasons unrelated to their workWages in the sense of a return for labor services might be ldquospas-modic casual marginalrdquo The choice between these two alterna-tives is important because the nature of the labor market is animportant component to the nature of the economy as a wholeWith a functioning labor market an economy can respond to ex-ternal in uences like market economies do today Labor can move

THE ROMAN LABOR MARKET | 515

4 Finley Ancient Slavery 68 127 Aldo Schiavone The End of the Past Ancient Rome and theModern World (Cambridge Mass 2000) 1565 Labor productivity herein means the output of goods or services that results from a workerrsquosemployment not the average labor productivity of all workers In economics jargon it is themarginal product of labor George J Borhas ldquoDoes Immigration Grease the Wheels of theLabor Marketrdquo Brookings Papers on Economic Activity I (2001) 71 ldquoThere exist sizable wagedifferences across regions or states in the United States even for workers with particular skillslooking for similar jobs

to take an advantage of a technical change that makes an activitymore pro table or a discovery that provides an economic oppor-tunity in a new place In a local nonndashlabor market labor wouldnot be able to respond to changes in the external environmentThe economy instead would continue to act in traditional waysperhaps with a small gesture toward the new opportunities Theeconomy would be dominated by Gemeinschaft not Gesellschaft

The task of distinguishing these two conditions in the earlyRoman empire is rendered dif cult as always by the absence ofcomprehensive evidence The chief evidence for the absence of alabor market in the early Roman empire has been the mere pres-ence of slaves The question is not how many slaves were presenthowever but rather how slavery operated Slaves in the AmericanSouth before the Civil War were not part of a uni ed Americanlabor market because their activities and incomes were so re-stricted that they had no incentive to seek better working condi-tions Slaves in the early Roman empire did not suffer under thesame restrictions but despite Romersquos use of slavery free hired la-bor was the rule not the exception in the rest of the early Romanempire

The abstract conditions that de ne a labor market typicallyare related to labor markets in industrial economies they needmodi cation to apply to labor markets in agricultural economiesMost of the workers in such an economy are rural working eitherin agriculture or in associated crafts and services they rarelychange occupations or residences without strong pressure A rurallabor market exists when enough of them are free to move in re-sponse to economic stimuli thereby keeping rural wages at amoderately uniform level but also allowing for substantial geo-graphical variation in both the level and the rate of change of ruralwages For example migration and wages interacted in early mod-ern Britain to keep wages similar but by no means equal

One possible move for a substantial fraction of rural workersin advanced agricultural economies is to a city It is rare both inpast and current agricultural economies for rural and urban wagesto be equalized by migration Economists do not regard this dis-crepancy as negating the existence of a uni ed labor market theyexplain the difference by noting that new urban workers often areunemployed and that only the expected wage (that is wage 3probability of earning it) should be equalized by migration Living

516 | PETER TEMIN

costs are also typically higher in cities urban wages can exceed ru-ral wages for this reason alone Urban wages that are double ruralwages do not strain the ability of these factors to account for thediscrepancy6

Wages vary in a labor market by skill as well as by locationAlmost all workers have skills basic skills of agriculture and oftenmore advanced skills as well Economists call these skills humancapital Most ancient workers had few skills including the abilityto read that is little human capital Craftsmen and some agricul-tural workers had competencies that did not depend on literacyand would receive a higher wage in a rural labor market for themBut these skills would not earn much if anything in urban areasAlthough we tend to know more about literate workersmdashdespitethe relative paucity of themmdashthan about less-skilled workers be-cause of the literary bias of our sources the great mass of workersin the early Roman empire were illiterate andmdashby modern stan-dardsmdashunskilled

Recent scholarship has revealed the existence of many marketprices and wages in ancient Rome suggesting that the Romaneconomy was not substantially different from more recent agrarianeconomies The abstract conditions that de ne a labor market inmodern analyses however need modi cation to apply to labormarkets in agricultural economies Steinfeld has shown that work-ers were not free to change jobs at will until near the end of thenineteenth century Even in the United States and Britain two ofthe most market-oriented countries that the world has everknown the rights of workers were sharply restricted Both urbanand rural workers were subject to prosecution if they left a jobwithout their employersrsquo permission Steinfeld argued that workin these advanced economies was directed by a mixture of mone-tary and other incentives This context permits no sharp distinc-tion between free and unfree labor only a continuum alongwhich various economies or even activities within an economycan be placed In his words ldquoPractically all labor is elicited by con-fronting workers with a choice between work and a set of more orless disagreeable alternatives to workrdquo7


6 John R Harris and Michael P Todaro ldquoMigration Unemployment and DevelopmentA Two-Sector Analysisrdquo American Economic Review LX (1970) 126ndash1427 Hans-Joachim Drexhage Preise MietenPachten Kosten und Loumlhne im Roumlmischen Aumlgyptenbis zum Regierungsantritt Diokletians (St Katharinen 1991) Dominic Rathbone ldquoPrices and

Steinfeldrsquos scenario is the standard against which to evaluateRoman labor markets Wages were an important tool for the allo-cation of labor in eighteenth-century England but their use waslimited by the restrictions on labor mobility Wages in such a sys-tem would not reach equality for similar skills and most workerswould not feel free to look around for a more lucrative activity

Free urban workers in the early Roman empire were paid fortheir work and were able to change their economic activities He-reditary barriers were nonexistent and Roman guilds do not ap-pear to have been restrictive Workers in large enterprises likemines and galleys were paid wages as in more modern labor mar-kets Workers engaged in more skilled and complex tasks receivedmore elaborate compensation probably for longer units of timethan those doing wage labor again as in more modern labor mar-kets even though explicit long-term contracts were not yet estab-lished The force of competition under those circumstancesprobably brought wages and labor productivity into the same ball-park8

Some of the work in the early Roman empire was done forwages and some under the duress of slavery The early Romanempire even had salaried long-term free workers in Egypt Crafts-men sold their wares in cities and also supplied them to rural andurban patrons in return for long-term economic and social sup-port Similarly people who worked for or supplied senators andequestrians often worked for long-term rewards and advancementThe episodic nature of monumental building in Rome accom-plished largely by free laborers gives evidence of a mobile laborforce that could be diverted from one activity to another Freeworkers freedmen and slaves worked in all kinds of activitiescontemporaries saw the ranges of jobs and of freedom as sepa-

518 | PETER TEMIN

Price Formation in Roman Egyptrdquo Economie antique Prix et formatin des prix dans les economiesantiques (Saint-Bertrand-de-Comminges 1997) 183ndash244 Temin ldquoA Market Economy in theEarly Roman Empirerdquo Journal of Roman Studies XCI (2001) 169ndash181 Robert J SteinfeldThe Invention of Free Labor The Employment Relation in English and American Law and Culture1350ndash1879 (Chapel Hill 1991) 26 See also idem Coercion Contract and Free Labor in the Nine-teenth Century (Cambridge 2001) Labor mobility was reduced further in continental Europeby guilds and other restrictions See Sheilagh Ogilvie State Corporatism and Proto-Industry TheWuumlrttemberg Black Forest 1580ndash1797 (Cambridge 1997) for a description of labor conditions inearly Germany8 Tenney Frank Economic Survey of Ancient Rome (1940) V 248ndash252 Russell Meiggs Ro-man Ostia (Oxford 1973) 314

ratemdasheven orthogonal In particular rural slaves hardly comprisedan undifferentiated gang of laborers certain lists of rural slave jobsare as varied as the known range of urban or household ldquoslaverdquojobs Some rural laborers received piece rates and others dailywages Cicero anticipating Marx con ated legal and economicrelations by equating wages with servitude9

A labor market in the early Roman empire would havetended to equalize real wages in different parts of the empire Sug-gestively Cuvigny found equal wages of miners in Egypt andDacia in Eastern Europe Nominal wages of unskilled workerswere unequal in Rome and Egypt but the price of wheat andother goods differed as well Real wagesmdashthe buying power ofwages in wheatmdashwere close however the hypothesis of equalitycannot be rejected These scraps of data provide evidence of awell-functioning labor market Only the ability and willingness ofworkers to change jobs in response to wage differentials wouldproduce such uniformity10

Moreover in a functioning labor market wages increase asthe number of laborers decreases because of the competition tohire them workers are more productive when fewer of them areavailable to work It is hard to know of small changes in Romanlabor supplies but plagues led to rapid large falls in the pool ofavailable labor Egyptian wages doubled after the major Antonineplague of 165ndash175 ce This clearly is the standard labor-marketresponse to a sharp decrease in the supply of labor It demonstratesthat wages in the early Roman empire moved to clear markets inthis case to allocate newly scarce labor11

Employment contracts also give evidence of labor-market ac-tivity in which workers could choose their jobs The modern divi-sion between wages and salaries nds its analog in Roman EgyptldquoAs a general rule permanent employees of the Appianus and re-


9 Rathbone Economic Rationalism and Rural Society in Third Century AD Egypt (Cam-bridge 1991) 91ndash147 166 Brunt ldquoFree Laborrdquo Keith R Bradley Slavery and Society at Rome(Cambridge 1994) 59ndash65 Marcus Tullius Cicero (trans Walter Miller) de Of ciis (London1913 orig pub 45ndash44 bc) XXI 1150ndash151 Heacutelegravene Cuvigny ldquoThe Amount of Wages Paidto the Quarry-Workers at Mons Claudianusrdquo Journal of Roman Studies LXXXVI (1996) 139ndash14510 Temin ldquoEstimating Roman GDPrdquo paper prepared for a conference on ancient historyldquoInnovazioacutene tecnica e progresso economico nel mondo romanordquo Capri April 14ndash16 200311 R P Duncan-Jones ldquoThe Impact of the Antonine Plaguerdquo Journal of Roman ArcheologyIX (1996) 108ndash136 Walter Scheidel ldquoA Model of Demographic and Economic Change inRoman Egypt after the Antonine Plaguerdquo ibid XV (2002) 97ndash114

lated estates can be distinguished by their receipt of opsonion (sal-ary) a xed monthly allowance of cash and wheat and sometimesvegetable oil whereas occasional employees received misthos thatis lsquowagesrsquordquo some of these ldquofreerdquo workers were tied to the estatefor life like those subject to the more modern worker contractsstudied by Steinfeld but others were free to leave when their jobswere done12

Miners and apprentices had employment contracts One dat-ing from 164 ce shows that workers were paid only for workdone and that they had more right to quit than the nineteenth-century workers described by Steinfeld

In the consulship of Macrinus and Celsus May 20 I FlaviusSecundinus at the request of Memmius son of Asceplius havehere recorded the fact that he declared that he had let and he did infact let his labor in the gold mine to Aurelius Adjutor from this dayto November 13 next for seventy denarii and board He shall beentitled to receive his wages in installments He shall be required torender healthy and vigorous labor to the above-mentioned em-ployer If he wants to quit or stop working against the employerrsquoswishes he shall have to pay ve sesterces for each day deductedfrom his total wages If a ood hinders operations he shall be re-quired to prorate accordingly If the employer delays payment ofthe wage when the time is up he shall be subject to the same pen-alty after three days of grace13

Most free workers were farmers many of them tenant farm-ers although employment categories in the countryside were uid Roman tenancy contracts allocated risks between landown-ers and tenants in much the same way as analogous contracts did ineighteenth- and nineteenth-century Britain Major risks wereborne by the landowners as events beyond the tenantsrsquo controlwhereas minor risks were borne by tenants in return for the op-portunity to earn more and keep their earnings ldquoForce majeureought not cause loss to the tenant if the crops have been damagedbeyond what is sustainable But the tenant ought to bear losswhich is moderate with equanimity just as he does not have togive up pro ts which are immoderate It will be obvious that we

520 | PETER TEMIN

12 Rathbone Economic Rationalism 91ndash9213 CIL III translated in Naftali Lewis and Meyer Reinhold (eds) Roman Civilization Se-lected Readings (New York 1990 orig pub 1951) II 948

are speaking here of the tenant who pays rent in money for ashare-cropper (partiarus colonus) shares loss and pro t with thelandlord as it were by law of partnershiprdquo14

We know a lot more about wages in England before industri-alization than in the Roman empire Wages for comparable workwere similar throughout England but they were not uniform Ag-riculture was more prosperous in the South than in the North andwages were higher in the eighteenth century (This pattern was re-versed in the nineteenth century when the North industrialized)Substantial variation was evident within regions due to the im-mobility of the population A recent summary of the English datashows daily winter wages in the North to be only half of whatthey were in the South in 1700 They approached each othergradually during the next century and a half15

England is much smaller than the Roman empire was If weuse Roman data from Egypt and Dacia a more suitable compari-son is pre-industrial Europe Clearly labor had even less mobilitybetween countries than within England and wages varied morethough they did remain at the same general level Allen demon-strated that wages within Europe began to diverge in the sixteenthand seventeenth centuries By 1700 the real wages of masonsin London and Antwerp were more than double those in otherEuropean cities16

Based on this more modern evidence we do not expect to nd wages that are equal in distant places except by coincidencebut we expect wages to be similar If the early Roman empire hada labor market that functioned about as well as the labor market inpre-industrial Europe then wages in the early Roman empirewould have been approximately equal Real wages for similar tasksmight have varied by a factor of two or three as real wages did ineighteenth-century Europe but they were not different orders of


14 Peter Garnsey Cities Peasants and Food in Classical Antiquity (Cambridge 1998) 139Dennis P Kehoe Investment Pro t and Tenancy The Jurists and the Roman Agrarian Economy(Ann Arbor 1997) Gaius as represented in The Digest of Justinian D 192256 quoted inDavid Johnston Roman Law in Context (Cambridge 1999) 6415 Donald Woodward Men at Work Laborers and Building Craftsmen in the Towns of NorthernEngland 1450ndash1750 (Cambridge 1995) Gregory Clark ldquoFarm Wages and Living Standardsin the Industrial Revolution England 1670ndash1850rdquo Economic History Review LIV (2001)485xxxxx16 Robert Allen ldquoThe Great Divergence Wages and Prices in Europe from the MiddleAges to the First World Warrdquo Explorations in Economic History XXXVIII (2001) 411ndash447

magnitude As just described this presumption is consistent withthe fragmentary evidence about wages in the Principate

The army must be distinguished from the private sphere as inmodern economies Peacetime armies are often voluntary re-cruited via the standard organizational luresmdashfavorable wages andworking conditions Wartime armies by contrast often rely onconscription which is a non-market process Actions within ar-mies are directed by commands not by market transactionsArmies therefore represent at best a partial approximation to a freelabor market and typically an exception to it Since armies unhap-pily are present in almost all societies we place this exception tothe general rule to one side

The wages of the Roman army which was staffed by a mix-ture of attraction and conscription stayed constant for many dec-ades at a time When the army was not ghting which was mostof the time soldiers had to be set tasks to keep them t and out oftrouble like building roads and public monuments This construc-tion work did not interfere with the labor market in Rome orelsewhere in the center of the empire since the army was stationedat the frontiers17

Slaves appear to be like soldiers in that they are subjectto command but such was not necessarily the case in the earlyRoman empire especially in cities Unlike American slavesRoman slaves were able to participate in the labor market in al-most the same way as free laborers Although they often started atan extremely low point particularly those who were uneducatedmany were able to advance by merit Freedmen started from abetter position and their ability to progress was almost limitlessdespite some prominent restrictions These conditions createdpowerful positive work incentives for slaves in the early Romanempire

roman slavery The prevalence of slavery in ancient Rome hasstood in the way of comparisons with more recent labor marketssince it seemed to indicate that a large segment of the Romanlabor force was outside the market Classicists have used evidenceof modern American slavery to illuminate conditions in ancientRome Bradley for example opened his book on slave rebellions

522 | PETER TEMIN

17 Brunt ldquoConscription and Volunteering in the Roman Imperial Armyrdquo Scripta ClassicaIsraelica I (1974) 90ndash115 George R Watson The Roman Soldier (Ithaca 1969) 45

with a chapter on slavery in the New World Although Bradleyand Hopkins emphasized the complexity of Roman slavery theiruse of modern evidence implicitly assumed that slave economiesseparated by two millennia were essentially the same Slaveryhowever is not always and everywhere the same Roman slaverywas at the opposite extreme from slavery in the southern UnitedStates many Roman slavesmdashlike free workersmdashresponded tomarket incentives18

Slavery has two dimensions In the rst dimension whichderives from anthropology slavery varies between open andclosed Open slavery describes a system whereby slaves can wintheir freedom and enter into general society In anthropologicalterms freedmen and women are accepted into kinship groups andintermarry freely with other free persons Closed slavery deniesslaves acceptance into general society and forbids them to marryamong the general population even when freed Roman slaveryconformed to the open model Freedmen were Roman citizensand marriages of widows with freedmen were common By con-trast ldquoAmerican slavery [was] perhaps the most closed and caste-like of any [slave] system knownrdquo Relative to other workersRoman slaves were not in the same predicament as modernAmerican slaves19

The second dimension along which slavery can vary is thefrequency of manumission Frequent manumission was a dis-tinguishing feature of Roman slavery slaves in the early Romanempire could anticipate freedom if they worked hard and demon-strated skill or accumulated a peculium with which to purchase itOnce freed they were accepted into Roman society far morecompletely than the freedmen in closed systems of slavery Thepromise of manumission was most apparent for urban skilled lit-erate slaves but it pervaded Roman society20


18 Bradley Slavery and Rebellion in the Roman World 140 BCndash70 BC Bloomington1989)xxxxx19 James L Watson ldquoSlavery as an Institution Open and Closed Systemsrdquo in idem (ed)Asian and African Systems of Slavery (Oxford 1980) 7 This classi cation differs from that inWilliam V Harris ldquoDemography Geography and the Sources of Roman Slavesrdquo Journal ofRoman Studies LXXXIX (1999) 62ndash75 which classi ed a slave system as open if slaves wereimported and closed if not20 Compare Edward Gibbonrsquos magisterial pronouncement early in The Decline and Fall ofthe Roman Empire (New York 1961) 36 ldquoHope the best comfort of our imperfect conditionwas not denied to the Roman slave and if he had any opportunity of rendering himself either

The ve slave societies noted by Hopkins can be classi edaccording to these two dimensions The classi cation can be seeneasily in a two-by-two matrix as shown in Table 1 where the vesocieties are placed in the appropriate boxes Rome as an opensystem of slavery with frequent manumission is in the upper left-hand box The southern United States and the Caribbean as aclosed system with almost no manumission appear in the lowerright-hand box The slavery in ancient Greece and in modernBrazil were intermediate cases Manumission was common butfreed slaves were segregated from the larger population The re-maining possibility an open system of slavery with no regularmanumission nds no instance in these ve societies

Few people would choose to be a slave almost all Romanslaves were forced into slavery as captives children of slaves aban-doned children or debt bondage A Roman slave was subject tomore cruelty in the early Roman empire than free people wereBut even if many slaves were at or near the bottom of societyand the economy it makes sense to ask how hopeless was theirposition

All people even slaves need to have incentives to work Freepeople may work to increase their income slaves may requireother incentivesmdashpositive incentives or ldquocarrotsrdquo (rewards forhard or good work) and negative incentives or ldquosticksrdquo (punish-ment for slacking off or not cooperating) There is a large litera-ture on the incentive structures of modern American slaverypossibly because its emotional content makes consensus elusiveBut while disagreements remain on many points most scholarsagree that negative incentives dominated the lives of modernslaves in the Americas

Positive incentives were more important than negative inmotivating Roman slaves Sticks can get people to work but notto perform skilled tasks that require independent action It is hardto distinguish poor performance from back luck when work iscomplex and carrots are far more effective than sticks in motivat-ing hard work Consider a managerial job like a vilicus (super-visor) A slave in such a position motivated by negative incentivescould claim that adverse outcomes were not his fault Beating him

524 | PETER TEMIN

useful or agreeable he might very naturally expect that the diligence and delity of a fewyears would be rewarded with the inestimable gift of freedomrdquo

or exacting worse punishment would lead to resentment ratherthan cooperation andmdashone might con dently expectmdashmore ldquobadluckrdquo A vilicus motivated by positive incentives would anticipatesharing success he would work to make it happen The effort ofan ordinary eld hand however could be observed directly andeasily slackers could be punished straight away Since eld handstypically work in a group positive incentives that motivate indi-viduals to better efforts are hard to design21

The cruelty in ancient slavery as in early modern indenturehas been described often because it contrasts sharply with ourmodern sense of individual autonomy Cruelty was a hallmark ofthe early Roman empire as it has been of most nonindustrial soci-eties Imperial Rome appeared particularly to celebrate its crueltyprobably because of its military orientation ancient cruelty was byno means reserved for slaves The vivid examples of violence to-ward slaves however do not offset the many competing stories ofmore benevolent slave conditions The miserable condition ofslaves working in the bakery overseen by Apuleiusrsquo golden ass donot so much illustrate the harsh conditions of Roman slavery asthe dismal conditions of ordinary labor in pre-industrial econo-mies In these Malthusian economies greater productivity resultedin larger populations rather than gains in working conditions orreal wages Almost all workers before the industrial revolution andthe demographic transition lived near what economists call subsis-tence which was not necessarily the edge of starvation but thelimit of endurance Slaves were not the only ones to do the dan-gerous work in Roman mines convicts and wage earners sufferedthere too22


Table 1 Varieties of Slavery in the Five Slave Societies

frequent manumissiononly exceptionalmanumission

Open systems early Roman empire

Closed systems classical Greece nineteenth-century Brazil

southern United Statesthe Caribbean

21 Stefano Fenoaltea ldquoSlavery and Supervision in Comparative Perspective A ModelrdquoJournal of Economic History XLIV (1984) 635ndash66822 Slave revolts do not give evidence of predominantly negative incentives The attestedslave revolts were concentrated in a short span during the late republic a time of great social

Some poor people regarded the life of a slave as better thanthat of a free man Ambitious poor people sold themselves intolong-term slavery that promised however uncertainly more ad-vancement than the life of the free poor Such a choice howeverrare in the early Roman empire would have been inconceivablein a closed system of slavery built on negative incentives It wasmore like the process of apprenticeship in early modern Europerevealing the integration of Roman slavery with the overall labormarket23

These observations relate mostly to urban slaves We do notknow how large a share of Roman slaves was urban but it wassubstantial Conventional estimates place the population of Italy inthe Principate at around 6 million 1 million of whom lived inRome itself Slaves are estimated to have constituted as much asone-third of Italyrsquos population and one-half of Romersquos These es-timates imply that one-quarter or less of the Italian slaves lived inRome the rest lived in smaller cities and the countrysidemdashwherethey were less than one-third of the rural labor force If so slaveswere not the dominant labor force either in the city or the coun-tryside of the early Roman empire Slaves in Egypt appear fromsurviving census returns to have composed about 10 percent ofthe population spread thinly among the households Two-thirdsof the listed slaves were women most likely working in house-holds rather than elds All of these educated guesses are highlyuncertain24

Manumission into Roman citizenship played an importantpart in urban Roman slavesrsquo and some rural slavesrsquo incentivesManumission was common but not universal The state set rulesfor manumission but left the decision of which slaves to free in thehands of individual slave owners who could use it to encouragethe most co-operative and productive slaves Slaves were often

526 | PETER TEMIN

upheaval See Bradley Slavery and Rebellion Lucius Apuleius (trans J Arthur Hanson) Meta-morphoses (ldquoThe Golden Assrdquo) (Cambridge Mass 1989 orig pub 2d century) II 92Garnsey and Richard Saller The Roman Empire Economy Society and Culture (Berkeley 1987)119 use this example to show the conditions of Roman slaves Oliver Davies Roman Mines inEurope (Oxford 1935) 14ndash1623 Jacques Ramin and Paul Veyne ldquoDroit Romain et Socieacuteteacute les Hommes Libres quiPassent pour Eacutesclaves et lrsquoEacutesclavage Volontairerdquo Historia XXX (1981) 49624 Hopkins Conquerors and Slaves 101 Roger S Bagnall and Bruce W Frier The Demogra-phy of Roman Egypt (Cambridge 1994) 48ndash49 71 Scheidel ldquoProgress and Problems in Ro-man Demographyrdquo in idem (ed) Debating Roman Demography (Leiden 2001) 49ndash61

able to purchase freedom if they could earn the necessary funds ina peculium which served as a tangible measure of slave productiv-ity The right of slaves to accumulate and retain assets was an im-portant part of the incentive structure Slaves who were sold orfreed kept their peculium even though they technically could notown property Slaves even owned slaves25

Hopkins asked ldquoWhy did Roman masters free so manyslavesrdquo His answer was complex On one hand he noted that thepromise of freedom was a powerful incentive ldquoThe slaversquos desireto buy his freedom was the masterrsquos protection against laziness andshoddy workrdquo Thus did he distinguish Roman slavery from thatin the southern United States On the other hand however heemphasized the similarity of these two types of slavery emphasiz-ing the role of cruelty and negative incentives He devoted morespace to slavesrsquo resistance and rebellions than to their achievementand cooperation Although he perceived the apparent sharp linebetween slavery and freedom as only part of a continuum of laborconditions he failed to break away from the prevalent view ofAmerican slavery at the time of his writing This imperfect anal-ogy still haunts the eld when Bradley wanted to explain whyRoman slaves rebelled he quoted the accounts of modern Ameri-can slaves26

Garnseyrsquos argument that ancient slavery was less harsh thanslavery in the southern United States appeared near the end of anintellectual history of ancient Greece and Rome that did not em-phasize Roman slavery as a distinct labor system Garnsey notedldquoThe prospect of manumission gave [Roman] slaves an incentiveto work and behave wellrdquo He drew out the implications of thisproposition for the idea of slavery particularly for Christians Thisarticle draws implications for the economic role of Roman slaveryin the Roman labor force27

Bradley devoted a chapter to manumission in his study ofRoman slavery but he minimized its role as an incentive Stressingthe uncertainty in the lives of slaves he concluded ldquoAbove all


25 John A Crook Law and Life of Rome (London 1967) 187ndash191 If a slave used hispeculium to purchase his freedom his former owner acquired possession of the slaversquos earnings26 Hopkins Conquerors and Slaves 115ndash132 Bradley Slavery and Rebellion27 Garnsey Ideas of Slavery from Aristotle to Augustine (Cambridge 1996) 87 97 RonaldFindlay ldquoSlavery Incentives and Manumission A Theoretical Modelrdquo Journal of PoliticalEconomy LXXXIII (1975) 923ndash934 determined the optimal timing of manumission for apro t-maximizing owner

therefore it is this absence of psychological and emotional securityin slave life which offers the key to the continuing ability of Ro-man slave-owners to control and keep in subjection their slavesrdquoHe confused the random part of Roman life to which all Romanswere subject to a large degree with the systematic part that moti-vates behavior After all even though getting an education carriesa lot of risk even today we urge our children to get an educationto improve their chances of living a good life28

The Roman incentive mechanism operated with consider-able uncertainty Some valuable Roman slaves received theirfreedom without having to pay for it as a reward for specialachievement or for noneconomic reasons Manumission in theearly Roman empire was not unlike starting a new company to-day Success is a product of both skill and luck the latter can bethe more important factor Yet success only comes to those whotry that is those who are willing to take risks Manumission rep-resented the same kind of opportunity for Roman slaves If a slavetried both skill and luck would play a part in his eventual successor failure The risks inherent in the process would not necessarilyhave discouraged many slaves29

Freedmen were granted Roman citizenship on an almostequal basis The well-known association of freedmen with formermasters worked to their mutual bene t When people engaged intrade or made arrangements for production they needed to knowwith whom they were dealing Information about buyers and sell-ers was scarce It was mitigated partially by identifying possiblecustomers or vendors as members of known families Slaves com-ing to freedom without a family naturally associated with theirformer ownersrsquo families Because they retained the names of andconnections with their former owners they could be associatedwith their ownersrsquo family Thus were former slaves able tointegrate into the economy Moreover a productive freeman re-turned the favor by increasing the reputation and income of his

528 | PETER TEMIN

28 Bradley Slaves and Masters in the Roman Empire (Brussels 1984) 14329 This argument can be viewed as an expansion of remarks in John R Hicks A Theory ofEconomic History (Oxford 1969) ldquoThere are two ways in which labor may be an article oftrade Either the laborer may be sold outright which is slavery or his services only may behired which is wage-paymentrdquo (123) Hicks acknowledged that slavery typically is a cruelbrutal institution but he softened this indictment when slaves have personal relations withtheir owners and can take economic initiative as in the early Roman empire ldquoPerhaps itshould be said when this point is reached the slave is only a semi-slaverdquo (126n)

former owner Freedmen of course could marry other Romancitizens and their free children were accepted fully into Romansociety Many widows married freed slaves to carry on a businessactivity30

Freedmen probably identi ed themselves as such on theirtombstones for two reasons First they were still part of theirformer ownersrsquo business family and second they were proud ofhaving shown the ambition and ability to win their freedom Liketodayrsquos self-made man the freed slave worked for what heachieved he was not the recipient of inherited wealth This op-portunity is the hallmark of open slavery and a functioning labormarket31

The continuum of incentives ranges from all negative as ina Nazi concentration camp or the Soviet gulag to all positive asin a progressive school where no child is criticized and all childrenare winners Most working conditions fall somewhere betweenthese two extremes Modern jobs clearly are near but not at thepositive end non-performance carries its penalties American slav-ery was near the opposite end the threat of punishment was ubiq-uitous whereas rewards for good service were rare Romanslavery by contrast was far closer to the positive end althoughhardly as close as modern jobs The lives of rural illiterate un-skilled slaves in the early Roman empire were like those in Ameri-can slavery The working conditions of educated urban slavesmay well have approached those of free men Hence most Ro-man slaves particularly urban slaves participated in a uni ed Ro-man labor market

comparisons with other systems of slavery Table 1 com-pares manumission rates and the conditions of freedmen in differ-ent societies The unique position of Roman slaves is clear fromthe availability of education for them which resulted from the in-centive structure of the Roman system and led to the integrationof Roman slaves into the overall labor market A few examples of


30 See Garnsey Cities Peasants and Food 30ndash3731 Lilly Ross Taylor ldquoFreedmen and Freeborn in the Epitaphs of Imperial Romerdquo Ameri-can Journal of Philology LXXXII (1961) 113ndash132 ldquoMany freedmen are known from theirnames alone but we know of so many of them because they wanted to memorialise their lifeand achievements in the same way as more august senators and knights erecting tombstonesthat have survived until todayrdquo

ancient working conditions for both free and slave workers illus-trate the operation of this uni ed labor market

Modern American slavery was a closed system The NewWorld slaves did not enter Eurocentric American society on easyterms their opportunities were severely limited Their descen-dants in the United States are still awaiting complete integrationinto society The descendants of former African slaves have faredmuch better in Brazil where manumission was more frequentEven in Brazil however slaves only began to be freed with anyregularity in the nineteenth century when pressure for the aboli-tion of slavery rose Yet since freed slaves were still excluded byformer Europeans few positive incentives were available tothem32

The frequency with which Greek slaves were set free is un-known but freed slaves in Athens did not become members ofGreek society They inhabited ldquoa limbo world in which full politi-cal and economic membership of the community was deniedthemrdquo Unlike Athenian citizenship Roman citizenship was in-clusive This fundamental difference between the two may havedetermined how each society interpreted slavery In any case theprevalence and visibility of manumission among Roman slavesmade Roman slavery far different than slavery in Athens33

By the time of the Principate most slaves were probablyslaves from infancy either as the children of slaves or unwantedchildren of free parents since captives were few by then A debateabout whether slaves were replenished through reproduction ormaintained through foundlings and the slave trade persists butmost scholars agree that the supply of captives had dwindled

530 | PETER TEMIN

32 Roman slavery shared attributes with another modern institution indentured servicePoor Englishmen who wanted to emigrate to North America in the eighteenth century butcould not afford it often mortgaged their future labor to pay for their passage Indentureslasted a xed number of years often fewer than ve and immigrants were able to resume lifewithout stigma after their indenture was nished While indentured however immigrantscould not move choose occupations or even determine the certain particulars of their cir-cumscribed lives They were in a descriptive oxymoron short-term slaves David WGalenson White Servitude in Colonial America (Cambridge 1981)33 Garnsey Ideas of Slavery 7 A few older ancient historians noted the comparatively be-nign quality of ancient slavery although without referring to manumission and without dis-tinguishing between Greek and Roman slavery If Greek slavery was more similar to Romanthan to modern slavery featuring reduced positive incentives of an open slave system the rea-son is by no means obvious See Alfred Zimmern ldquoWas Greek Civilization Based on SlaveLaborrdquo Sociological Review II (1909) 1ndash19 159ndash176 (repr in idem Solon and Croesus [London1928]) A H M Jones ldquoThe Economics of Slavery in the Ancient Worldrdquo Economic HistoryReview IX (1956) 185ndash204

Rules for manumission became explicit Augustus enacted a law(lex Fu a Caninia) restricting the proportion of slaves that a slaveowner could manumit at his death but also preserving the struc-ture of incentives by forcing owners to decide which of theirslaves to set free Rights of freedmen were expanded The incen-tive for slaves to act well became clear Freedmen moved intoskilled and well-rewarded trades and other activities and theirchildren born after manumission entered society with all of theirrights34

Manumission was common and well known in the earlyRoman empire Livy recounted a legend about a slave who wasfreed in 509 bce the rst year of the republic as a reward forfaithful service albeit of a political rather than an economic na-ture Although Livy could not have known whether the story wastrue he thereby revealed attitudes in his own time A legal princi-ple of the era dealt with the status of a child born to a womanwho conceived while a slave was freed and then enslaved againbefore giving birth For this to have been an interesting questionthe boundary between slavery and freedom must have beenpermeable35

No counts of Roman manumission exist but the myriadreferences to manumission and freedmen in the surviving recordsattest to its frequency Scheidel assumed that 10 percent of slaves inthe early Roman empire were freed every ve years starting atage twenty- ve in a demographic exercise Some of Scheidelrsquos as-sumptions have attracted vigorous rebuttal but not this oneThese estimates and opinions apply to the totality of urban and ru-ral Roman slaves In the judgment of a modern observer ldquoMosturban slaves of average intelligence and application had a reason-able expectation of early manumission and often of continued as-sociation with their patronrdquo In the judgment of another ldquoRomanslavery viewed as a legal institution makes sense on the assump-tion that slaves could reasonably aspire to being freed and henceto becoming Roman citizensrdquo36


34 Scheidel ldquoQuantifying the Source of Slaves in the Early Roman Empirerdquo Journal of Ro-man Studies LXXXVII (1997) 157ndash169 Harris ldquoDemographyrdquo Bradley Slavery and Society10 asserted that the intent of Augustusrsquo law was to restrict manumission only to those slaveswho had proved that they deserved freedom35 Titus Livius (trans B O Foster) History (London 1919 orig pub 27 bc) I 23ndash5Ernst Levy (ed) Pauli sententiae (Ithaca 1945) II 24336 Scheidelrsquos estimate of 10 was an intermediate one the actual level could have beenhigher or lower (ldquoQuantifyingrdquo 160) Harris ldquoDemographyrdquo Paul R C Weaver Familia

The Egyptian census listed no male slaves older than thirty-two Since the census counted household slaves only this agetruncation suggests widespread manumission rather than excep-tionally high slave mortality Female slaves generally were freed ifthey had more than three children which may not have been un-common in an age without family planning Manumission on thisscale must have been apparent to all slaves certainly to all urbanslaves and a powerful incentive for them to cooperate with theirowners and to excel at their work37

Slave conditions in the southern United States were com-pletely different Manumission was the exception rather than therule American slaves could not anticipate freedom with anycon dence Manumission required court action in Louisiana anonerous process that left traces in the historical record An exhaus-tive count of Louisianarsquos manumissions showed that the rate in theearly nineteenth-century was about 1 percent in each ve-yearperiod an order of magnitude less than Scheidel assumed forthe early Roman empire Many of those freed were childrenunder ten and the majority of the adults freed were womenmdashpresumably the childrenrsquos mothers Fogel and Engerman champi-ons of positive incentives in American slavery reported evenlower manumission rates at mid-century ldquoCensus data indicatethat in 1850 the rate of manumission was just 045 per thousandslavesrdquomdashthat is 045 per 100 slaves or 02 percent in a ve-yearperiod two orders of magnitude lower than Scheidelrsquos reasonableguess for Rome American slaves and particularly male slaves hadlittle anticipation of freedom and little incentive to cooperate inthe hope of freedom38

532 | PETER TEMIN

Caesaris A Social Study of the Emperorrsquos Freedmen and Slaves (Cambridge 1972) 1 Alan Wat-son Roman Slave Law (Baltimore 1987) 2337 Bagnall and Frier Demography 71 342ndash343 Lucius Junius Moderatus Columella (transH B Ash) On Agriculture (Cambridge Mass 1934 orig pub 1st century ad) I 1819 Ap-parently slave women had to have undergone either three live births or had to have three liv-ing children at the time of the next birth The stipulation is clearer in Theodor Mommsen andPaul Krueger (eds) The Digest of Justinian (Philadelphia 1985 orig pub 553) 1515 whichdeals with the disposition of triplets under a will that freed the mother at the birth of the thirdchild38 Slaves in Baltimore had slightly more hope than others they were freed with more fre-quency although the interval between the decision to manumit and actual freedom was oftenlong Stephen Whitman ldquoDiverse Good Causes Manumission and the Transformation ofUrban Slavery Social Science History XIX (1995) 333ndash370 Gwendolyn Hall Databases for theStudy of Afro-Louisiana History and Genealogy 1699ndash1860 (New Orleans 2000) Shawn Cole

In the absence of evidence that the prospects for slaves in an-cient Greece were as bleak as those in the United States the as-sumption herein is that the Greeks freed slaves with someregularity Slavery in the Caribbean by contrast was more likethat in the United States In fact given the tiny number of slaveowners in the Caribbean the probability of manumission mayhave been even lower than in the United States

In Brazil manumission began roughly at the outset of slaveryalthough many legal and circumstantial barriers prevented it frombecoming a matter of course Its pace was slow before the nine-teenth century but it accelerated rapidly during the last decades ofBrazilian slavery Rio de Janeiro contained 80000 freed slaves in atotal urban population of 200000 in 1849 Brazil as a whole con-tained 11 million slaves and 28 million ldquofreemenrdquo in 1823 and15 million slaves and 84 ldquofreemenrdquo in 1872 Non-white free per-sons had become a majority of the population in Salvador by 1872Brazilian slaves often could earn enough to purchase their wivesrsquofreedom although they frequently did not have enough to obtaintheir own As in Louisiana two-thirds of the freed slaves in Braziland in Rio de Janeiro were women A recent study of early nine-teenth-century censuses in Sao Paulo con rmed the Brazilianpredilection to manumit women rather than menmdash125 men foreach 100 women among Brazilian slaves in 1836 but only 87 menfor each 100 women among free coloreds Any effect that manu-mission might have had on Brazilian slave workers as an incentivewas diminished by the clear Brazilian pattern of freeing slavewomen rather than slave men39

Successful freedmen intensify the incentive for manumissionthat merges the work of slaves and free workers Even freedmenliving a marginal existence can serve as models for slaves since


ldquoCapitalism and Freedom Slavery and Manumission in Louisiana 1770ndash1820rdquo paper pre-sented at the National Bureau of Economic Research Summer Institute Cambridge MassJuly 17 2003 Robert W Fogel and Stanley L Engerman Time on the Cross The Economics ofAmerican Negro Slavery (Boston 1974) I 15039 Stuart B Schwartz ldquoThe Manumission of Slaves in Colonial Brazil 1684ndash1745rdquo His-panic American Historical Review LIV (1974) 603ndash635 Katia M de Queiros Mattoso To Be aSlave In Brazil 1550ndash1888 (New Brunswick NJ 1986) 50 164 Mary C Karasch Slave Lifein Rio de Janeiro 1808ndash1850 (Princeton 1987) 66 346 Meikoi Nishida ldquoManumission andEthnicity in Urban Slavery Salvador Brazil 1808ndash1888rdquo Hispanic American Historical ReviewLXXIII (1993) 365 376 Francisco Vidal Luna and Herbert S Klein Slavery and the Economyof Sao Paulo 1750ndash1850 (Stanford 2003) 162ndash163

freedom is desirable whatever the economic cost But its attrac-tion undoubtedly increases to the extent that freedmen are ac-cepted even prominent in free society Unlike in other slavesocieties freedmen in the early Roman empire were citizens Infact they were ubiquitous in the late republic and early empireengaged in all kinds of activities including administration andeconomic enterprise The number of men who identi ed them-selves as freed on the tombstones during this period is astonishingThey may not have ascended to high Roman society but theirchildren bore little or no stigma Their success was commonknowledge Seneca ridiculed a rich man by remarking that he hadthe bank account and brains of a freedman In Finleyrsquos wordsldquoThe contrast with the modern free Negro is evidentrdquo40

Why were freedmen so prominent The process of manumis-sion separated the more able from the others The prospect ofmanumission was an incentive for all slaves but the most activeambitious and educated slaves were more like to gain their free-dom as a reward for good behavior or by purchase The system didnot work perfectly many slaves were freed for eleemosynary mo-tives or at their ownerrsquos death But for the most part freedmenwere accomplished individuals It was good policy to deal withand hire them and it makes sense to say so only because Romehad a functioning labor market Contrast this scenario with that offreed slaves in the antebellum United States where the infamousDred Scott decision of the Supreme Court decreed in 1857 thatfreed slaves could not be citizens and ldquohad no rights which thewhite man was bound to respectrdquo41

Freed slaves in Brazil lived a similarly marginal existence notbound but not fully free either Known as libertos they and theirchildren were clearly isolated from the main society and were notprosperous Census material and related data always indicated towhich group a free person belonged Even though freed slaveswere Brazilian citizens their legal rights were ldquoquite limitedrdquoLibertos ldquocontinued to owe obedience humility and loyalty to thepowerfulrdquo The physical appearance of freed slaves in Brazil made

534 | PETER TEMIN

40 Arnold M Duff Freedmen in the Early Roman Empire (Oxford 1928) Susan TreggiariRoman Freedmen during the Late Republic (Oxford 1969) Lucius Annaeus Seneca (trans JohnW Basore) Epistulae Morales (London 1932 orig pub 1st century) 275 Finley Ancient Slav-ery 9841 Society became more rigid during the late empire As opportunities for advancement inurban activities diminished so did the incentive of manumission Slavery began to evolve intoa different institution Dred Scott v Sanford 60 US 393 407 (1857)

them easy to distinguish The marginalization of freed persons inNorth and South America demonstrates that slavery in these areaswas a largely closed systemmdashalthough Brazil was not as closedas the United Statesmdashin contrast to the open system of the earlyRoman empire42

Education is a key to the nature of Roman servitude Ameri-can slave owners relied on negative incentives and discouraged theeducation of slaves because they were afraid of slave revolts led byeducated slaves Ancient slave owners used positive incentives al-lowing and even encouraging slaves to be educated and performresponsible economic roles Education increased the value of slavelabor to the owner and it increased the probability that a slaversquoschildren would be freed Educated slaves had the skills to accumu-late a peculium and they would be good business associates of theirformer owners Most freedmen worked in commercial centerswhich provided an opportunity for advancement

Educated slaves are markedly associated with positive incen-tives and uneducated slaves with negative incentives Many edu-cated Roman slaves were administrators agents and authorsmdashforexample Q Remmius Palaemon who was educated in the rstcentury ce ostensibly ldquoas a result of escorting his ownerrsquos son toand from schoolrdquo but probably had more direct exposure thansimply acting as a paedagogus In the Republic Cato educatedslaves for a year in a sort of primitive business school and thensold them Anyone enacting such a plan with American slaveswould not have been celebrated he would have been ostracizedjailed and ned The Virginia Code of 1848 extended to freedmenas well as slaves ldquoEvery assemblage of Negroes for the purpose ofinstruction in reading or writing shall be an unlawful assembly If a white person assemble with Negroes for the purpose ofinstructing them to read or write he shall be con ned to jailnot exceeding six months and ned not exceeding one hundreddollarsrdquo43


42 Mattoso To Be a Slave 179ndash183 See also Schwartz ldquoManumissionrdquo Karasch Slave Life362 Sidney Chalhoub ldquoSlaves Freedmen and the Politics of Freedom in Brazil The Experi-ence of Blacks in the City of Riordquo Slavery and Abolition X (1989) 64ndash84 Nishido ldquoManu-missionrdquo Douglas Cole Libby and Clotilde Andrade Paiva ldquoManumission Practices in a LateEighteenth-Century Brazilian Slave Parish Satildeo Joseacute drsquoEl Rey in 1795rdquo Slavery and AbolitionXXI (2000) 96ndash127 Brazilian freedmen were represented in many occupations but theywere concentrated at the lower end of the economic and social scale Vidal Luna and KleinSlavery 17243 Bradley Slavery and Society 35 Plutarch (trans Bernadotte Perrin) ldquoMarcus Catordquo Par-

Many Roman slaves educated or not competed with freed-men and other free workers in a uni ed labor market Various oc-cupations emerged to meet the demands of urban residentsparticularly rich ones Skilled slaves were valuable to merchantsand wealthy citizens because they could serve as their agents inthe much the same way as their sons could ldquoWhatever children inour power and slaves in our possession receive by manicipatio orobtain by delivery and whatever rights they stipulate for or ac-quire by any other title they acquire for usrdquo Watson expressedsurprise that the Romans did not develop a law of agency but theRomans did have a law of agencymdashthe law of slavery (and sons)As Hicks noted slavery was the most common formal legally en-forceable long-term labor contract in the early Roman empire Aperson with a long-term relation to a principal would be his or hermost responsible representative Slaves were more valuable thanfree men in that respect Witness the frequent references to liter-ate skilled slave agents in the surviving sources44

Columella (181-2) aptly exposed the difference betweenancient and modern slavery ldquoSo my advice at the start is not toappoint an overseer from that sort of slaves who are physically at-tractive and certainly not from that class which has busied itselfwith the voluptuous occupations of the cityrdquo This warning wouldnot and could not apply to modern slavery both because modernslaves could not indulge in ldquovoluptuous occupationsrdquo likeColumellarsquos list of theater gambling restaurants etc and becausea modern slave could not have been appointed as manager of asubstantial estate45

Implicit in Columellarsquos advice is the ease with which slavescould change jobs For example when Horace was given an estate

536 | PETER TEMIN

allel Lives (London 1914 orig pub early 2d century) II 21 Va Code [1848] 747ndash748 Edu-cation does not even appear in the index to Fogel and Engerman Time on the Cross So fewBrazilians of any sort were educated that no contrast between slaves and free workers in thiscontext is possible44 Gaius (trans W M Gordon and O F Robinson) Institutiones (Ithaca 1988 orig pubc 161) 287 Watson Roman Slave Law 107 Hicks Theory See Andrew Lintott ldquoFreedmenand Slaves in the Light of Legal Documents from First-Century AD Campaniardquo ClassicalQuarterly LII (2002) 555ndash565 for a vivid description of slaves and freedmen as agents in therecords of a commercial house in Puteoli Free people were also agents Roman jurists beganto correct the legal discrepancy between them and slave agents in the Principate ( JohnstonRoman Law 106)45 Columella 181-2 ldquoIgitur praemoneo ne vilicum ex eo genere servorum qui corporeplacuerunt instituamus ne ex eo quidem ordine qui urbanes ac delicates artes exercueritrdquo

on which he employed ve free tenants and nine household slaveshe chose a vilicus from an urban household with no apparent train-ing in agriculture The mobility of labor must have been evenmore pronounced for free labor The demand for unskilled andsemiskilled labor for particular tasks varied widely over time inboth the country and the city Agricultural demand varied season-ally in the late republic and undoubtedly at other times the peakrural demand for labor was satis ed by the temporary employmentof free workers Urban labor demand varied less frequently butpossibly more widely Public building activity in the Principatewas sporadic workers must have been attracted to these projectsin one way or another The presumption among classicists is thatfree workers were hired for them lured by the wages offered Ifso they also must have had ways to support themselves and theirfamilies when public building activity was low46

Slave wages are not widely documented despite the fact thatsome slaves must have earned wages to accumulate a peculium Thepreceding discussion however indicates that slaves were inter-changeable with free wage laborers in many situations Althoughthe evidence for monthly and annual wages comes largely fromEgypt and the information about slaves comes mostly from ItalyRoman slaves appear to t Hicksrsquo view of slaves as long-term em-ployees The analysis of slave motivation and the wide distributionof slave occupations suggest that slaves were part of an integratedlabor force in the early Roman empire47

Workers in the uni ed labor market of the early Roman empirecould change jobs in response to market-driven rewards As in allagricultural economies the labor market worked better in citiesthan in the countryside Slaves participated in this system to a largeextent The restrictions on labor mobility may have been no moresevere than the restrictions on labor mobility in early modernEurope Education was the key to the good life in the earlyRoman empire as it is today Roman workers appear to havereceived wages and other payments commensurate with theirproductivity and they were able to respond at least as fully as in


46 Jean-Jacques Aubert Business Managers in Early Roman Empire (Leiden 1994) 133Garnsey Cities Peasants and Food 143ndash145 Brunt ldquoFree Laborrdquo M K Thornton and R LThornton Julio-Claudian Building Programs (Wauconda Il 1989)47 Hicks Theory

more modern agrarian societies to the incentives created by thesepayments

Marx and Toumlnnies may have been half right They were notcorrect to assert that functioning labor markets the hallmark ofGesellschaft were new in the nineteenth century Not only werelabor markets present earlier in modern European history theyalso were present in the early Roman empire But as they sur-mised functioning labor markets were associated with urbaniza-tion Rome was an urban society to an extent not duplicated againuntil the nineteenth century Its connection with a labor markethowever can only be suggested speculatively herein though theevidence is compelling

ldquoThe Roman lawyer Gaius wrote that the fundamental socialdivision was that between slave and freerdquo The fundamental eco-nomic division in the early Roman empire however was be-tween educated and uneducatedmdashskilled and unskilledmdashnotbetween slave and free Saller summarized this view succinctlyldquoThe disproportionately high representation of freedmen amongthe funerary inscriptions from Italian cities re ects the fact that ex-slaves were better placed to make a success of themselves in the ur-ban economy than the freeborn poor upon manumission many ofthe ex-slaves started with skills and a businessrdquo48

538 | PETER TEMIN

48 Garnsey Cities Peasants and Food 134 citing Gaius Institutiones 19 Saller ldquoStatus andPatronagerdquo in Alan K Bowman Garnsey and Rathbone (eds) The Cambridge Ancient His-tory XI The High Empire AD 70ndash192 (Cambridge 2000) 835

the presence of a labor market in which the services of labor canbe bought and sold which Polanyi argued emerged only twocenturies ago2

The widespread use of slavery in Rome is taken as a sign thatGemeinschaft dominated the life of the Roman republic and theearly Roman empire Finley and others following his lead arguedthat ancient economies were not market economies but an alter-nate even primitive form of organization Finley stated ldquoIn earlysocieties free hired labor (though widely documented) was spas-modic casual marginalrdquo According to Hopkins the early Ro-man empire was ldquoa society which had no labor marketrdquo Hopkinsspeculated as a result ldquoIn a society without a market in free laborrecruitment by force (ie slavery) was probably the only methodof securing large numbers of full-time dependents with particularskillsrdquo3

This view is mistaken A variety of evidence indicates thatRome had a functioning labor market and a uni ed labor forceWage dispersion in the early Roman empire to the extent that weknow it is indistinguishable from that in pre-industrial EuropeRoman labor contracts have a distinctly modern allocation of risksand rewards In addition Roman slavery was so different frommodern slavery that it did not indicate the presence of non-market traditional actions Instead ancient Roman slavery was anintegral part of a labor force that shares many characteristics withlabor forces in other advanced agricultural societies Contrary toFinley who asserted ldquo[A]ncient slavery co-existed with other

514 | PETER TEMIN

2 Karl Polanyi The Great Transformation (New York 1944) Max Weber (trans Talcott Par-sons) The Protestant Ethic and the Spirit of Capitalism (New York 1930 orig pub in German1905) More recent views reject the opposition of Gemeinschaft and Gesellschaft in favor of amore complex view that sees both ideal types as present in a wide variety of societies Thisview which also derives from Toumlnnies questions which form of society is dominant notubiquitous interpreting history as the shifting balance between Gemeinschaft and Gesellschaftrather than a total transition from one to the other A society might be regarded as dominatedby one mode or the other but the sense that progress has eliminated Gemeinschaft is gone SeeThomas Bender Community and Social Change in America (New Brunswick NJ 1978)3 Ian Morris ldquoForewordrdquo in Moses I Finley The Ancient Economy (Berkeley 1999 origpub 1973) 1ndash37 68 Finley Ancient Slavery 68 In ldquoFurther Thoughtsrdquo Finley reaf rmed hispositions with ldquonuancingrdquo that ldquofree hired labor was casual and seasonalrdquo and that ldquothere wasno genuine competition between slave and free laborersrdquo (185ndash186) Following Brunt heacknowledged abundant free laborers in the largest cities but he insisted that their employ-ment was ldquostrictly speaking casualrdquo (Finley ldquoFurther Thoughtsrdquo in idem Ancient Economy177ndash207) P A Brunt ldquoFree Labor and Public Works at Romerdquo Journal of Roman StudiesLXX (1980) 81ndash100 Hopkins Conquerors and Slaves 14 111










516 | PETER TEMIN









518 | PETER TEMIN












520 | PETER TEMIN













522 | PETER TEMIN











524 | PETER TEMIN














526 | PETER TEMIN











528 | PETER TEMIN












530 | PETER TEMIN









532 | PETER TEMIN










534 | PETER TEMIN










536 | PETER TEMIN










538 | PETER TEMIN









532 | PETER TEMIN










534 | PETER TEMIN










536 | PETER TEMIN










538 | PETER TEMIN










534 | PETER TEMIN










536 | PETER TEMIN










538 | PETER TEMIN










536 | PETER TEMIN










538 | PETER TEMIN










538 | PETER TEMIN


The Labor Market of the Early Roman...

Documents

Transcript of The Labor Market of the Early Roman...