Peter Temin’s passing was announced only a few hours ago. An economic historian of extraordinary breadth, he wrote on a wide range of topics, from the Great Depression to the Industrial Revolution, nineteenth century Britain, and many more, including the subject of this book: ancient Rome. I have only read a fraction of his work, but for someone with my research interests his research on the Great Depression is essential.
I was thinking about his book The Roman Market Economy only yesterday, as I was finishing composing my thoughts on Chris Wickham’s Framing the Early Middles Ages for next week’s review. It is one of the books which I think about the most, not only for its ‘actual arguments’, but for Temin’s balanced and fair perspective on the role of quantitative methods in history.
It is also only the second book I reviewed on this blog, back in October 2023. To mark the occasion (and since I think about five people read this then) I will re-post my unedited review of this book below. I hope it encourages some to read the book itself, which is not too long and well worth your time.
Please forgive the poor writing — I like to think I have improved since I wrote it. My take on the book is positive but caveated: I disagree with Temin that serfdom did not “gradually emerge” in the third to fifth centuries, citing Anderson. Now I could not agree with Temin more! Wickham has convinced me, as I wrote in the last review, that such a transition is pure myth. So I suppose if I was to rewrite this review today it would only be more positive.
Without further ado,
In the last post, I roughly outlined the picture, drawn by Perry Anderson in Passages from Antiquity to Feudalism, of a Roman economy very foreign from our own. An economy without recognisable labour markets, dominated by the state, slavery, and technological stagnation, where east and west increasingly operated under different economic logics. But is this how it actually was? Economist Peter Temin would disagree. In The Roman Market Economy, Temin outlines a vision of a Roman economy integrated from Egypt to Spain; where labour markets function smoothly, large estates are of negligible significance, and slavery operates in the periphery. Further still, Temin claims this system functioned up until the collapse of the Western Empire. In fact, his vision contradicts Passages on almost every point - Temin even argues against Roman technological stagnation, citing evidence that pushes back the adoption of the water-wheel well into the first century AD. In short, Temin presents a Roman economy sufficiently market-oriented to be considered a “market economy” in the modern sense, charges of presentism be damned.
It is worth asking what the implications are if Temin is correct. On the one hand, if the Roman economy operated under an economic logic sufficiently similar to our own, then the dialectical process implied within Passages is, in a significant sense, invalidated. Feudalism is less clearly a necessary step on the way to capitalism, but instead a setback for a classical economy well on its way to escaping the Malthusian trap anyway. On the other, the actual origins of feudalism itself become contested. If feudalism was was not the product of a “slave mode of production”, slowly collapsing under the weight of its own contradictions, then where did it come from?
Discussion of the grain trade concerns much of the book, and is one of the first topics Temin broaches. This is important, because Temin’s claim that the Roman economy was a “market economy” largely rests on his argument that an integrated, well-functioning grain market existed across the empire, where price signals could equilibrate prices, sans transport costs, from Gibraltar to the Dardanelles. Marshalling a “thin but nicely random” dataset, Temin finds that the price of grain in a given region is highly correlated with the distance of that region from Rome (where grain is the most expensive). In an integrated market, where differences in prices are largely explained by differences in distance, this is to be expected: grain will be sold in Sicily at the expected price in Rome, minus the cost of shipping it there. In the absence of an integrated market, however, there is no reason to expect this result: the price of grain will be a result of local production costs interacting with local supply and demand. Temin presents more evidence for an integrated grain market later in the book, which I won’t go into here. Suffice to say, I think he presents enough evidence to shift the burden of proof onto those who would argue against the existence of an integrated Roman market for grain. However, I don’t think I am ready to call the Roman economy a “market economy” on this basis alone - and I don’t think Temin would expect me to, either. Integrated commodity markets do not a market economy make. For a market economy to be a plausible characterisation, Temin must show three things: the existence of functioning markets for land, labour, and capital. All three are important; however, and partially in the interest of concision, I would like to focus on the labour market. In particular, this is because Temin argues two things: that slavery was not structurally integral to the Roman economy, and that there was no gradual transition to serfdom from the third century onwards. Both run directly contra to what Anderson puts forth in Passages.
For Temin, the Roman labour market is comparable to the labour market in the modern United States. Slave labour was not insignificant, but the majority of the labour force was waged, and both groups were compensated commensurate with their productivity, at least to an extent. When Anderson wrote Passages, it was generally thought that slaves represented one-third of the Italian population by the time of the Principate (27 BC). Since then, however, it had become accepted that no more than one-in-six Italians were slaves at that time, a statistic which drops to less than 10% by the fourth century. In the face of this evidence, I think it’s pretty hard to disagree with Temin. Under these conditions, slavery could not have been structurally crucial. In a society with pre-modern agricultural productivity, it’s inconceivable that, with these demographics, slavery it could have formed the backbone of the Roman agricultural system.
However, I don’t think Temin presents enough evidence for the second argument; that a gradual transition to serfdom did not take place between the third and fifth centuries, and that instead, the labour market only broke down with the collapse of the Western Empire in the fifth century. As he himself admits, there is an absence of reasonable numbers for the proportion of slaves in this period. Yet Temin makes the assumption that slave labour in the West ceased declining by the end of the third century; remaining at around 10% of the total population up until the imperial collapse. From this, he asserts that the Roman labour market must have been structurally stable across this entire period. Of course, it is not implausible that the fall of the Western Empire induced a shock so significant it broke the labour market completely. But in the absence of more compelling evidence, I struggle to accept this as the likely scenario, especially given evidence of labour market restrictions emerging earlier (such as under the Diocletian Edicts). If labour mobility and institutionalised slavery collapsed overnight, one would think it would be more obvious. In fact, by making such a strong case for labour mobility in the early imperial period, I think Temin makes it even harder to accept that serfdom did not gradually emerge in late antiquity. Whether a crisis in slave agricultural was the pivot feels somewhat secondary. Whatever cause, a systemic labour-market crisis most likely developed from the late third century onwards, at least to my mind.
So where do we stand? Personally, I think Temin successfully argues that markets played a much larger role (and slavery a far smaller one) in Roman society than I had hitherto thought, and that books of the era of Passages would argue. Nonetheless, I find myself in disagreement on two significant points. Even if substantially market-oriented, the Roman economy could not have been a “market economy” in the modern sense. Accepting this would involve papering over the chasm between then and now; between an economy which stagnated technologically, demographically, and economically for over five-hundred years, and the dynamic-if-unstable market economies of today. Secondly, I still think Passages provides a better account of the transition to feudalism in late antiquity, although, in the light of modern scholarship, one that is in need of significant updating.
All things considered, The Roman Market Economy is a well-researched, well-written book which made me seriously question my priors about the Roman economy, and without doubt has shifted many too. One can no longer accept older notions, still all too often repeated, of what the Roman economy looked like. I would also like to note that the first section of the book provides the most compelling, accessible, and lucid argument that historians (and for that matter, economists!) should take econometrics - in particular, regression analysis - seriously that I have ever read. For someone like myself, who, perhaps to a fault, thinks through the lens of historical materialism, The Roman Market Economy is essential reading.