1Peter Temin is a renowned economic historian of the modern world, whose major works include Lessons from the Great Depression (1989) and the co-edited volume The World Economy between the World Wars (2007). Since the start of the millennium, Temin has applied his expertise as an economist and economic historian of the modern world to Roman society, thereby adding his voice to on-going debates about the economy of the Roman world. The book under review covers the Roman market economy under the three main headings ‘prices’, ‘markets’ and ‘macroeconomics’, and may be regarded as the synthesis of Temin’s work on the Roman world. Although some ancient historians writing on the Roman economy, like myself, dabble with economics and toy with the analytical tools of the discipline, none of us can claim an expertise in this field that even comes near that of Temin’s. His interest in ‘our’ field of study, and his contribution to the debate, is therefore very much appreciated.
2In the introduction, Temin justifies the application of modern economics to the ancient world—a justification that in my opinion is hardly necessary—by rightly pointing out that the Roman economy did not differ substantially from that of subsequent eras. I agree: if one can apply modern economic theory to Medieval France, Golden Age Holland or eighteenth-century England, one can apply it to the Roman world. Of course, Roman society was not the same as that of any later society, and the differences mean that not all concepts and theories devised for modern society can simply be projected on to the Roman world. Temin appreciates this, arguing, for example, in relation to the use of labour in the Roman world that one has to adapt modern conceptual tools to a society that, unlike modern society, knew slavery.
3It is a different question whether Temin always succeeds in adequately taking into account the specific features and distinctive traits of Roman society. Although I wholeheartedly welcome the application of economic tools and models to the ancient world, I fear that the book does not always succeed in that respect. Temin sees the Roman Empire and Roman society as purely and wholly determined by commercial links, and in this sense he may have lost a necessary sense of nuance. He passionately and unequivocally rejects the views of Moses Finley cum suis on the Roman economy. Few, if any, ancient historians would nowadays unreservedly support their ‘primitivist’ stance. Arguing against the mercantile bourgeoisie of Rostovtzeff’s Roman world, Finley clearly overstated the underdevelopment of market channels, the aversion to commerce of the political and social elites, the backwardness of commercial institutions, and the absence of any potential for growth. Temin writes: “I hope to clarify the issues in this debate and even resolve the debate for the period of the early Roman Empire” (4). However, exposing the weaknesses and errors in Finley’s primitivism is not the same as establishing the truth of the opposite position. At the same time as ancient historians are realizing the complexity and achievements of the ancient economy, modern historians are appreciating the social, cultural, and political embedding of economic relationships and commercial links in early-modern Europe. In sum, in many respects Temin’s view is as un-nuanced as Finley’s, overstating his case and missing the necessary shades of grey.
4Temin’s assumption that all links in the Roman market economy were commercial means that he has a tendency to tread lightly over those details that do not fit the presumed market predominance. For example, he systematically infers trade when the evidence merely indicates distribution. One particular point here concerns his argument that the grain market in the entire empire was dominated by the city of Rome and that the grain supply of the capital were handled by the private market. I should point out here that I am an interested party in this matter, as Temin argues against views I have expressed in The Grain Market in the Roman Empire (2005). Although a review is clearly not the place to settle this debate, I wish to go into these issues in some detail as they are crucially important to show the methodological shortcomings of Temin’s argumentation.
- 1 The annona was the imperial organization that took care of the food supply, primarily grain, of the (...)
5To begin with, Temin writes: “The government transported the wheat for the annona1 privately. They let contracts to societates to provide wheat, and they offered inducements for private merchants to participate in this process” (32). He cites the measures taken by Claudius in support, but disregards the nature of their involvement. Claudius offered social privileges and advantageous conditions to ship-owners who engaged in contracts with the state to ship tax-grain to Rome. There was no case of ‘merchants’ ‘providing’ grain, only of transport-contracts. Transportation is big business, of course, but it is not trade. Temin argues furthermore that the grain dole fulfilled between one quarter and one half of the city’s needs. “At least half of the wheat imported to Rome at the time of Augustus, and probably more, therefore was imported privately” (ibidem). Again, this is a selective reading of ancient sources and modern literature. While the grain dole may have fulfilled between one quarter and one half of the city’s needs, this does not exhaust the involvement of non-market supply channels. Even a conservative estimate of the grain tax in Egypt—notably, a tax in kind—shows that this province alone provided the state with twice as much wheat as it needed for the grain dole. Moreover, the political elites of the Roman Empire, many of whom resided in the capital, consisted of large-landowners, whose income largely derived from landed estates in Italy and abroad. These landowners brought a significant part of the produce of their estates to Rome, in order to both feed their huge urban households and support their clientele. Hence, the private market’s contribution to provisioning the empire’s capital was much smaller than Temin implies. Finally, in the case of Sicily and Egypt—the two instances for which we have some quantitative evidence—it can be plausibly argued (though ignored by Temin) that the taxation levied by Rome actually more or less exhausted the potential contribution of both provinces to outside consumers. Taxation, as Temin observes, is part of the market. This is correct, but the situation was more nuanced than that. While monetary taxes mean that the state acquires the capital to pay for goods and services, taxes in kind rule out trade. Private businessmen might have been involved as collectors, contractors, and transporters, but that does not turn taxation into trade. This misrepresentation is unnecessary, as one does not need to try to hide that public channels had a place besides private trade in order to establish that the Roman economy was a market economy. Much of the complexity of Roman society, and of its spectacular achievements (such as having given rise to the largest city in Europe until nineteenth-century London) was built on the use for specific purposes of coercive, non-market channels within a market context.
6The main purpose of chapter two, ‘Wheat prices and trade in the early Roman Empire’, is to show that the Roman Empire constituted a single, large, integrated grain market (rather than an agglomeration of regional and local grain markets). This controversial hypothesis is based on the application of modern economic tools to analyse grain prices in the Roman Empire. To summarize Temin’s argument: as Rome dominated the integrated grain market of the Roman Empire, the price of wheat in Rome determined the price of wheat in all other markets. This is not to say that the price of wheat in all markets was the same. After all, there were also costs involved in shipping wheat to Rome and selling it there. Hence, in the integrated grain market of the Roman Empire, the local price of wheat was the price of wheat in Rome minus the costs of transporting the wheat to Rome and selling it there. Hence, the model predicts that the price of grain would be lower the further one goes from the capital. Temin produced a graph that shows precisely this. Conclusion: the Roman grain market constituted a single empire-wide integrated market.
7However, the prices, statistics and graphs presented here offer a false sense of confidence. To begin with, the entire statistical analysis is based on a mere six prices. The sources do provide more grain prices, but most are famine-prices or administrative prices, leaving just six prices that, more or less, constitute market prices. Crucially, the prices have to be random. Temin argues that the six prices are random, as they are taken from Rickman, and Rickman did not gather these prices in order to perform the kind of analysis that Temin offers. There is the possibility, Temin admits, that the outcome, supporting the statistical link between distance from Rome and grain price, is merely the result of chance. However, Temin assures us, economists can actually calculate this chance: if all figures are randomly selected, there is a possibility of one in twenty that the outcome is merely the result of chance. Now, are these six figures random, or are there other variables that determined the input and thereby distort the outcome? That Rickman did not pick these prices for this kind of analysis does not show their random nature. They were not picked for this purpose, but there might be other distortions stemming from our sources. Looking at the graphs on pages 42 and 43, some peculiarities are noticeable: first, the three cases on the left (i.e. closest to Rome) are all from the Republican period, the three cases on the right (i.e. farthest from Rome) are all imperial. Second, the three cases on the left are all from the West, the three cases on the right are all from the East. Even though they are arranged according to distance to Rome, Temin acknowledges, they are almost chronological (40). This suggests that there are other organizing variables that distort the outcome, which may result from a combination of geographical or chronological patterns and mere chance. Significantly, when Temin tries to employ a different set of grain prices—taken from Rathbone, of which only eight are useable—no significant pattern emerges. It should make one worry that it works with six prices, but not with eight! Temin realizes the shortcomings of the price data: he notes the “noise introduced by casual or distorted price observations” (48), and also acknowledges that “the seasonal price variation introduces another source of noise” (49). Remarkably this only strengthens his conclusion: if the picture is so clear with all this noise, there must be a strong relationship. In other words, the distance and price regression analysis is like a mythical beast with six heads: the more heads one chops off, the more lives it has!
8Unlike modern economic history, the ancient world offers us few figures that can be analyzed statistically, but it does offer us a wealth of qualitative evidence concerning the nature of the grain market in the Roman world. This qualitative data shows that the performance of the grain market should not be exaggerated. To name just one strand of evidence, central and local authorities continuously intervened in the grain market by creating institutions and mechanisms that ensured local supply when the market failed. The nature of the authorities’ involvement in the grain supply of the cities reflected the weakness of the market, which could not be counted on to overcome the obstacles that connected supply and demand. It is good to see that rhetorical trickery has a place in modern statistical research, as Temin has dubbed the use of the phrase “the weakness of the grain market” as more emotional than rational (49). The meaning of this phrase is quite clear and rational: it refers to the degree to which the market was able to compensate the unpredictable fluctuations of production in relation to the stability of consumer demand. A strong market is able to overcome the obstacles to a large extent, a weak market is not. Apart from any logistical obstacles to an integrated market, the qualitative evidence shows that there was no integrated grain market in the Roman Empire.
9In the next chapter, ‘Price behaviour in Hellenistic Babylon’, Temin uses formal tests to confirm the conclusions of Slotsky, Grainger, and Van der Spek: Babylon had a market economy. One may wonder about Hellenistic Babylon in a book on the Roman economy, but the purpose of this chapter is made clear in the next: “The fortunate survival of many Babylonian prices reveals to us that market prices were widespread in the ancient world. I infer that the scarce Roman observations that resemble market prices also were market prices” (72). This is quite revealing concerning the limitations of the Roman price evidence: far from providing a solid foundation for the postulated integrated grain market, Temin has to use data for Hellenistic Babylon to infer the nature of the economy in the Roman world.
10The rejection of Temin’s one-sided approach—which focuses solely on private market channels, and of his purposeful overemphasis of the performance of the Roman market—should not be taken to mean that the book does not make valuable contributions to our understanding of the Roman economy. Particularly helpful is Temin’s analysis of the extent and causes of inflation in the Roman imperial period, where he differentiates between, on the one hand, short-term inflation resulting from the great fire of Rome in Nero’s reign and the temporary crisis caused by the unbalancing impact of the Antonine Plague from, on the other hand, long-term inflation as a result of the destabilizing policies of third-century emperors. Also chapter eight, ‘Financial intermediation’, offers an interesting overview of the means that were available in financial transactions. It covers the whole spectrum of social and commercial relationships involved and the legal tools that were devised or adapted to the interests of the people who wished to invest, borrow or lend out capital on a small or large scale. My knowledge about early-modern banking is very limited, but as far as I can see, the comparison between Roman and later European financial systems, and hence the conclusion that the financial market in the Roman world functioned no worse or even better than in later European countries, is nuanced and well-informed.
11Temin, together with David Kessler, was also one of the earliest historians who let his study of the Roman grain trade be guided by the principles of the New Institutional Economics (Kessler and Temin 2007; chapter 5, ‘The grain trade’, is based on this earlier publication). Temin rightly stresses the formal and informal institutions that were available to merchants in the Roman world to reduce risks and costs. Unfortunately, a tendency towards one-sided positivity is again noticeable in this chapter. While Temin establishes which formal and informal institutions potentially could have contributed to lowering transaction costs and enhancing the performance of the market, the question remains as to what extent they actually did contribute to the functioning of the market. Assuming that institutions are beneficial because that is what they are for, Temin takes the constructive role of Roman ‘firms’ and collegia more or less for granted. A critical perspective, like that of Sheilagh Ogilvie (2007; not listed in the bibliography) who offers her ‘conflict model’ as an alternative to Douglass North’s ‘efficiency model’, would have been welcome. Collegia were not only institutions that enabled traders to base their trade relations on trust and reputation, but also social networks that helped those in power to impose their own interests—in politics just as much as in business—on less powerful local artisans, traders, and merchants. The annona was not so much an “information clearinghouse” (112) as it was an imperial organization that privileged the wellbeing of the capital’s populace over all other interests. Many people in business undoubtedly profited from their dealings with the annona, but there is precious little evidence in support of their supposed role in reducing transaction costs beyond the provisioning of Rome. Temin cites a third-century Egyptian receipt as evidence of the complex system of documentation used in Roman trade and by Roman merchants: “The public administration used a system of receipts to record important information about grain cargoes that were available for merchants buying grain or by other third parties” (107). In fact, the receipt merely functioned to assure state officials that the grain that left one state-controlled depot of tax-grain would enter another state-controlled depot of tax-grain. No trade involved, no information about price levels at different locations, available stocks, harvest predictions, or anything else that a trader might wish to know before embarking on a costly and uncertain enterprise.
12In short, the institutional environment of ancient trade and the conditions of the market are of crucial importance in understanding the economic performance of the Roman Empire. It is clear that the individuals who profited from the many opportunities in the more or less peaceful and ordered society of the Roman Empire found countless ways to circumvent problems and to make risks acceptable. However, we should avoid a biased and rosy view based on a-priori assumptions about Roman economic performance.
13Chapter six, ‘The labor market’, makes the noteworthy point that the price of slaves in the Roman world may have been higher in relation to wage labor than in the Greek world because the value of slave labor was also higher in the former than in the latter. Comparing the price and value of free and servile labor is only possible if both are part of the same labor market, and this idea is rightly stressed by Temin. Those seeking to employ labor had the choice between slaves and free laborers, and in many cases both worked side by side, and they could buy and sell slaves in response to market conditions. Temin emphasizes the nature of slavery in the Roman world, which was more open and incentive-based than Greek slavery, let alone slavery in the nineteenth-century American South. Manumission and peculium are crucial in understanding the value of Roman slave labor.
14Unfortunately, Temin again tends to go for the extremes in his desire to show that slaves functioned just like free workers and that both were part of one integrated labor market spanning the whole empire. To make the first point, he even employs Friedrich Engels’ criticism of labor conditions in the days of early industrialism: wage laborers in those days were just like slaves; hence, Temin implies, slaves were just like nineteenth-century wage-laborers (121). However, it does not follow from the idea that employers could choose between free and servile labor systems that “slaves were able to participate in the labor market of the early Roman Empire in almost the same way as free laborers” (134). There is still a substantial difference between, on the one hand, slaves who enjoyed much independence in their activities (and sometimes even appreciation of their expertise), but still could be coerced into unquestioning obedience, and, on the other hand, free workers, most of whom may have had few opportunities, but at least had the freedom to go and find other employment, if they did not like their current labor conditions. One should not forget that manumission and peculium were part of the lives of only a small minority of slaves. Slaves had the choice to work hard or avoid any exertion when nobody was looking; they could rebel, kill or commit suicide. But they did not have the choice to be manumitted or to seek work elsewhere. Slaves were very mobile, but not of their own choosing. Mobility was a basic right that slaves did not possess. The fact that you could whip, brand and chain your slave, and thus coerce him/her to remain in service even though employment could be found elsewhere, surely makes a difference in comparing the labor market of the ancient and early-modern worlds. It would have been far more interesting if Temin had taken the characteristics of the slave system and its economic consequences truly into account, rather than to try to argue the differences away.
15Also the notion that the Roman Empire had one integrated labor market should be rejected. Temin argues that it is not possible to reject the hypothesis that real wages were in the same in rural Egypt and the city of Rome. “This finding suggests that there was a functioning labor market that allowed workers to move to where wages were higher” (255). Temin argues that in an integrated labor market wages would be approximately equal. “Real wages for similar tasks might have varied by a factor of two or three, as real wages did in eighteenth-century Europe, but they were not different orders of magnitude. As just described, this presumption is consistent with the fragmentary evidence about wages in the Principate” (120). Yet the implied argument is not valid. Defining similarity (as Temin does) as varying by a factor of less than two or three, and expressed in silver, real wages were of the same magnitude in preindustrial Europe and China for a long time, but nobody would argue on those grounds that an integrated labor market characterized the entire Eurasian continent!
16The book ends with the issue of economic growth (chapter 10) and ‘per capita GDP’ (chapter 11). Temin argues that the Roman economy – as with all preindustrial economies – was subject to Malthusian restraints, meaning that population and material resources were in balance and that any shocks to either population (e.g. epidemic) or material resources (e.g. technological innovation) had only a temporary impact on living standards. If per capita income rose, an increase in population followed, in the long run evening out any per capita progress. However, Temin argues, this scenario does not rule out economic progress, as there are significant delays in this Malthusian process. Productivity growth as a result of trade and specialization may have led to higher living standards, but the adjustment may have taken several generations, during which higher income translated itself into higher fertility levels and population growth. Temin thus tries to reconcile the theoretical model of Malthus with the archaeologically noticeable rise in living standards in the Roman world. Temin still takes the validity of the Malthusian model for granted, even though its value as an effective tool to analyse economic developments over the past 1,000 years has come under increasing criticism by such eminent economic historians as Jack Goldstone, George Grantham, and Karl Gunnar Persson (whose publications on this matter are missing from the bibliography). Temin regularly uses comparison with later European history as a means to overcome the limitations of ancient sources, but I would have liked to have seen his defence of the Malthusian model against the serious objections that have been raised. Postulating long delays in the response between material wellbeing and population is one way of escaping the criticism that the Malthusian predictions are not empirically proven, but at some point this reasoning turns into special pleading. Letting go of Malthus would have been logical and also easier to reconcile with Temin’s predilection to stress the modernity of Rome’s economy.
17An estimate of Roman per capita Gross Domestic Product offers quantitative support of the idea that the Roman world experienced substantial economic growth. Temin emphasizes the “exceedingly narrow evidentiary base” of all GDP estimates for the Roman world, and hence he asks “do GDP calculations affect our views of the Roman economy or are they the result of our views?” (244). To a large extent, I would say the first, but these calculations do at least visualize the consequences of our assumptions. As long as we keep in mind the danger of circularity of argumentation, they are a valid and meaningful exercise. While evaluating and criticizing the calculations made by others, such as Hopkins, Goldsmith, and Scheidel and Friesen, Temin offers various noteworthy observations, such as the error of transferring all requirements into wheat equivalents and subsequently calculating seed and transport requirements on the basis of this volume of wheat (contra Hopkins and Bang; 247). On the complex issue of ‘subsistence’, Temin rightly observes that the question is not whether ordinary Romans lived at subsistence level, but rather the kind of subsistence they experienced (250). Finally, comparisons between the Roman Empire and Golden Age Holland are meaningless, Temin emphasizes (259), as these are not equivalents. If the most prosperous parts of the Roman Empire are compared to the prosperous parts of Europe, the wide gulf dividing their economies disappears. While real wages did not differ substantially within the Roman Empire (an element related to Temin’s integrated market hypothesis), it is the great income of a small minority that explains the differences in wealth between the various parts of the empire. If we take this one step further, we might argue that the huge incomes of the political and social elite (partly in the form of rents) provided the economic foundation of the towns and cities that are the hallmark of ancient society. Hence, we are back with Weber’s and Finley’s consumer city, but I am sure that Temin would emphatically disagree.
18In the first pages of his book, Temin expressed his ambition to solve the old debates concerning the Roman economy. I do not think that he achieved this goal. It is not that ‘we ancient historians’ are not yet ready for the ‘truth’ of economic theory as applied by a true economist, as Temin regularly supposes, but rather that solving the debate requires a more balanced and nuanced approach, allowing for the shades of grey that characterize all historic reality. Since modern theory can be applied to the Roman world, Temin seems to reason, the Roman world has to be as modern as the modern world, with integrated grain and labor markets, institutions that oil the wheels of commerce, and landowners and workers that think and act the same as modern capitalists and wage-earners. Sources and modern interpretations that do not accord with these assumptions are ignored or swept aside. This is a shame, because I can only agree with what might be regarded as the general motto of Temin’s work: “A model that structures our discussion allows us to place ancient economic history into the general study of economic history” (239). We should base the discussion of the ancient economy on theoretical foundations rather than intuitive interpretations, using conceptual tools and analytical instruments that we might seek in other disciplines. Models are inevitably simplifications of the truth, but that is something that a careful and balanced reading of the sources should be able to overcome.