Were people in Roman Egypt driven by economic rationality when they engaged in financial transactions? This is the central question of Paul V. Kelly’s book, a revised version of his 2020 PhD dissertation. K. offers a nuanced, pluralist answer, emphasising the complexity of the Roman market. There was no single, monolithic market (or bazaar) but rather a mosaic of distinct submarkets (housing, land, credit and others), each governed by its own rules and norms. Individuals navigated these markets differently, with behaviour ranging from profit-seeking rationality in some (such as the secured cash loan and agricultural land markets) to reciprocity or familial concerns in others (notably unsecured loans, in-kind loans and rural housing markets).
K.’s approach combines statistical analysis of over 4,000 financial transactions preserved in the papyrological evidence with a range of formal modelling techniques. The latter include a synthesis of disparate price series into a ‘blended’ inflation measure (to track inflation over time), probabilistic microsimulations (to assess the risk of ruin for tenant families), and estimates of default rates on loans (to calculate average annual returns on different types of investment). While K.’s choice to relegate much of the technical detail of these models to the appendices and other publications may limit the ability of specialists to verify his results, it also makes the book more accessible to a broader audience.
One minor issue concerns how K. presents some of his data. In several instances, polynomial trendlines are added to the graphs representing the datasets, even though there is no theoretical basis to assume the data follow a polynomial function. This can be misleading. In chapter 2, for example, K. analyses price series to assess trends in inflation over time. Some of these series might actually be better represented by for instance an exponential or linear trendline (e.g. fig. 4), which would lead to different conclusions about whether prices were rising, falling or staying stable over time. An even greater concern arises in cases where figures show only the polynomial trendlines, without even including the data points (e.g. fig. 29). In these figures, a reader cannot check the data to form their own opinion. A simpler method, such as a moving average, might have been a more transparent way to summarise the datasets.
To infer the motivations of those engaged in financial transactions, K. examines a wide range of factors, including the identities of the participants, the timing of transactions across the agricultural calendar, and estimated rates of return. K. is commendably prudent in his assessment of what his results can and cannot reveal about the intentions of actors.
Additionally, K.’s analysis of his new, expanded datasets enable him to revisit and revise many earlier scholarly claims, from the sources of elite income (62) to the ‘normal’ interest rates in specific periods (125). The concluding chapter has a similarly broad scope. It explores, in addition to addressing the central research question, the implications of his results for a wide range of debates, for example the economic impact of the Antonine Plague or social inequality. While the chapter convincingly demonstrates the relevance of the papyrological evidence to these discussions, the treatment of these topics is rather cursory. For example, K. argues that the Antonine Plague had a limited impact on the financial markets in Egypt (167–68), engaging with just two works from a substantial body of scholarship on the demographic and economic impact of this epidemic.
Overall, a question of scale lingers: How significant were these financial markets within the broader Egyptian economy? K. discusses the relative importance of different markets based on the number of surviving documents, but as he acknowledges, not all transactions were officially recorded. By focusing on papyrological evidence we risk missing both informal, non-monetised transactions and monetised transactions that did not require official documentation. A deeper exploration of these limitations would have provided more grasp on the importance of K.’s work for our understanding of the Roman economy.
An intriguing aside in the introduction concerns the finger scarring of individuals mentioned in the papyri (scars were used for the identification of people). K. argues that a high prevalence of scars on the left little finger — likely due to reaping accidents — implies that the participants in the transactions in the papyri were ‘very ordinary agrarian people’ (14). While his explanation is convincing, the fact that fewer than 5% of Arsinoite men with descriptions had such scars again raises questions about scale: How representative are these men with scars on their left little finger of all participants in financial transactions?
K.’s book is a significant scholarly achievement. It surpasses previous scholarship in its use of an up-to-date dataset of financial transactions in Roman Egypt combined with more formal modelling. His pluralist view is also a welcome addition to the debate on Roman economic rationality. To further amplify the work’s impact, I recommend publishing the underlying data according to FAIR open science principles. Doing so would make this work an even greater milestone in the collective effort of ancient historians to understand the Roman economy.
