by Giacomo Benati and Federica Carugati
It is a productive and rewarding time for scholars doing historical political economy. Those who believe that the discipline has a robust intellectual identity have come out proudly to say so. Those who felt lost in the absence of that identity may have found a new home. But, as the field develops, important questions remain about what exactly historical political economy is and how it’s done.
Two field-defining pieces tell a somewhat different story about the what and the how. According to Volha Charnysh, Evgeny Finkel, and Scott Gehlbach, “[w]hat unites HPE is methodological approach… HPE is predominantly quantitative…typically seeks to isolate the causal effect of a specific factor using quasi-experimental methods. When HPE emphasizes theory, the theory is often articulated as a formal game-theoretic model.” In the introduction to the forthcoming Oxford Handbook of Historical Political Economy, Jeffery Jenkins and Jared Rubin take a broader view when they state that what distinguishes HPE is not a methodological approach but ”three distinct criteria”: the establishment of a falsifiable argument, an interest in understanding and explaining historical context, and the presence of a political economy element.
Jenkins and Rubin’s three criteria are a beacon for those who think that they do historical political economy, but fall ruinously in the crevices of Charnysh et al.’s adverbs. Yet, like the fire of Plato’s cave, which allowed the prisoners to see only the shadow of objects, Jenkins and Rubin’s beacon is producing a somewhat illusory image. A rapid glance at the handbook’s methodological chapters, in fact, suggests that HPE work is, in fact, predominantly quantitative, typically focused on causal identification, and often articulated in the language of game theory.
One reason for this convergence might be the training of those who do HPE. Both overviews are quick to posit the field’s aspiration toward interdisciplinarity. Charnysh, Finkel, and Gehlbach state that HPE “spans not only the traditional subfields of political science but also economics, history, and sociology.” In the first line of the handbook’s introduction, Jenkins and Rubin echo the sentiment, suggesting that “[a]t its core, HPE is an interdisciplinary endeavor.”
What, then, are we to make of the fact that, of the sixty-five authors of the handbook, all but four are either economists or political scientists, and only one is a historian?
The Benefits of Interdisciplinarity
One of us raised this question at the roundtable on Historical Political Economy organized at the annual meeting of the Society for Institutional and Organizational Economics (SIOE), which was held in Frankfurt in late August of this year. Scott Gehlbach – notably the only author of the handbook who co-authored his contribution with an historian – answered that collaborating with historians is difficult because social scientists and historians differ substantially in their approaches to evidence (sources vs. data), and preferred methods (descriptive vs. causal). These are real challenges. But the argument is worth debating further – perhaps with the larger audience of this blog. This is not, however, what we want to do here.
In this piece, we’d like to showcase some of the HPE work that might be less well-known to the crowd that gathers in this blog, and that is being done by scholars whose training is not in political science or economics, but in history, archaeology, or related disciplines traditionally labeled as “humanities.” Before we delve into this unorthodox lit review, however, we want to articulate why anyone should bother to continue reading.
We have all heard, at one point or many, some (often tenured) people decry the state of the fields of political science and economics as captured by the causal inference revolution. Some of us roll our eyes, others unconvincingly nod, a few wholeheartedly agree. But the tradeoffs of this methodological ascendancy are real. Social scientists are increasingly able to tease out causality among extremely complex phenomena. This is a significant achievement, fueled by sophisticated statistical approaches, clever designs, and massive amounts of data. This stuff is amazing and fun to read. And scholars should continue doing it. But if quantitative causal inference becomes the only, or the dominant, methodological approach, it is possible that political scientists and economists will end up losing the forest for the trees. While our big theories of democracy, state formation, and economic growth (among others) grow stale, new theorizing about the very changing nature of these phenomena in the 21st century is sparse.
The worry is that HPE might be moving in the same direction – just with much cooler data. This would be a missed opportunity. One way forward, we suggest here, might be to try a little harder to train, read, and collaborate with scholars who are interested in establishing falsifiable arguments, understanding and explaining historical context, and studying political economic processes – but whose background is not in economics or political science. Such collaboration is valuable precisely because these scholars ask different questions, use different methods, and have different – but potentially complementary – expertise. But, to be worthwhile, the collaboration should not relegate history, and historians, to the side-show – a passive source of data, variation, natural experiments, or contextual backdrop. Instead, we should work as peers. Such is the way we harness the power of diversity in teams, as organizations scholars tell us.
Social Science History
The first stop of our review is Stanford University, where a group of ancient historians and archaeologists – Ian Morris, Josh Ober, and Walter Scheidel – has spent the best part of the last two decades applying social scientific approaches to the study of pre-modern economies and polities. Their contribution to historical political economy is immense, but to avoid finishing our allotted word count listing their output, we limit ourselves to linking only a few of the latest books (link, link, link, link, link, link, link, link, link, link, and link).
Besides publishing books, these scholars have also trained students. And while many ancient historians continue to row their boat gently down the stream of descriptive historiography, those who have come into contact with the “Stanford school” can’t quite rub it off. Those of us who had the privilege to receive this training directly, have used it to explore phenomena such as the dynamics of authoritarian regimes, the collapse of large scale trade networks, the role of constitutions in promoting development, and the contribution of religion to state formation. Whereas the focus of these studies is the ancient Mediterranean, particularly Greece and Rome, other qualitative or mixed-methods work inspired by institutional approaches has moved beyond this region, investigating the structure and performance of states, empires, and fiscal regimes, as well as the evolution of collective action and inequality across the world. Although the temporal span varies, this scholarship often concentrates on, or includes, societies that existed centuries or even millennia before the early modern period, where most HPE work is concentrated.
Data-Driven History and Archaeology
A related but distinct branch of scholarship stems from the revolution in digital methods and computing power that has boosted the production, dissemination, and consumption of evidence in historical disciplines that are traditionally geared towards data production in the field (whether an archaeological excavation, an archive, or a lab). Data-driven history and archaeology make use of cutting-edge quantitative methods not only to reconstruct and simulate past dynamics, but also to explore drivers and impacts of crucial historical phenomena such as climate change, trade expansion, cultural diffusion, social evolution, inequality, and urbanization. In addition to a stream of increasingly sophisticated new proxies, these efforts produce insights that can be operationalized and integrated into formal inferential frameworks, easing the dialogue between scholars with different backgrounds. This work is mostly carried out outside the US – an area often overlooked in overviews like the handbook, where all contributors but four either work or got their PhD in the US – and it often focuses on places and historical periods that are still at the margins of HPE, such as prehistoric and pre-literate societies, Bronze Age South West Asia and Egypt, the pre-Columbian world, pre-Colonial Africa, and the Iron Age Mediterranean. Thus, the potential for cross-fertilization is significant.
Four separate strands of this literature are worth bringing to the attention of the Broadstreet audience.
GIS and networks. A particularly intriguing avenue for interdisciplinary work lies at the intersection between spatial analysis in archaeology/history and the historical turn in urban economics/economic geography. Understanding the rise and evolution of cities has long been a core question in both historical and social scientific disciplines, and historical statistics about urbanization are widely used as proxy for growth (and sometimes political institutions). Lately, a tighter integration between these disciplines can be seen in the application of structural trade models and economic theories to predict trade outcomes in the deep past, and in the increasing use of archaeological data for understanding locational factors shaping social interaction, market access, political integration, and the long-term effects of transportation networks. Moreover, the application of Settlement Scaling Theory and network science to the analysis of ancient communication networks and urban systems offers an additional methodological bridge for integrating the rich archaeological and historical record into formal analytical frameworks. Notably, in addition to rethinking models of growth and trade in pre-industrial societies, this convergence is also fueling the production of theory-based proxies and network data. Such proxies and data, in turn, open up new directions for testing theory, expanding the temporal reach of network approches in HPE beyond the transition to modernity.
Paleoscience, society & history. The convergence, or consilience, of natural scientific approaches and the humanistic study of the past is at the core of a research agenda that aims at understanding the causes and consequences of past environmental changes. This agenda builds upon recent advances in paleo-environmental research that have made available increasingly high-resolution (digital) archives of pollen, sediments, isotopic, geochemical, and paleoclimatic data. Scholars have focused, for example, on the human responses to climatic shocks, including demographic fluctuations, social collapse and resilience, state consolidation and breakups, and migrations. Some recent work by Broadstreet contributor Emily Sellars illustrates the gains that can be made from the systematic consideration of geographical fundamentals as determinants of economic and political outcomes. Further examples are studies that look at exogenous factors – especially climate – shaping state origins and development in the deep past, macro-economic and political cycles, and conflict and instability.
Natural science and humanities. A related sub-field studies the interplay between cultural, biological, and ecological dynamics in the past. This strand of research makes use of computational tools and statistical inference applied to archaeological proxies to reconstruct past landscapes and population dynamics, to study the origin and resilience of agriculture, and to simulate processes in human behavioral changes, people-plants co-evolutionary dynamics, or long-term trajectories of biodiversity and sustainability in prehistoric societies. Although the main goal of this literature is extracting lessons that can inform adaptation strategies in fragile agricultural systems today, there are several avenues for scholarly collaboration. One involves testing hypotheses about the links between crop dynamics and social and political evolution. Another consists of inferring patterns of cultural evolution from the distribution of archaeological artifacts and settlements. Finally, this body of research has begun to develop techniques for assessing the heterogeneity of fuzzy datasets, using probabilistic modeling and estimation techniques that can account for uncertainty and comparability to enhance confidence in measurements based on fragmentary evidence.
Bioarchaeology and long run development. At the intersection of economic history and archaeology/anthropology lies a strand of research that combines analytical tools routinely employed in historical economics to the analysis of excavated human osteological remains. In the past decades, economists interested in the historical evolution of standards of living have started to exploit proxies such as human stature, aDNA, mortality rates, bone traumas, enamel hypoplasia and strontium isotopes to study changes in human welfare across time and space. The convergence is facilitated by the growing familiarity of bioarchaeologists (and paleopathologists) with statistical inference, and by the digital revolution, which has led to the proliferation of standardized databases covering more and more historical periods and areas (link, link, link, link, link, link). The increasing availability of bioarchaeological proxies makes it possible to test hypotheses that are at the core of social scientific and historical inquiry, such as the nexus between violence, inequality and socio-political evolution, or the welfare effects of crucial historical phenomena, such as agricultural specialization, early state formation, or colonialism.
What can HPE scholars learn from all this? First, historical political economy need not be uniquely focused on causal inference. Second, collaboration across disciplines aside from economics and political science is possible and potentially advantageous to all. Third, through such collaborations, we can continue to develop clever proxies for missing data, think creatively about designs that are rigorous and falsifiable, ask big questions, and develop (and test) theory. We can also advance the sophistication of our behavioral models by looking for theories in a vast disciplinary field beyond economics, and by drawing on rich and fine-grained data of how historical actors actually behaved in various places and across time.
At the SIOE roundtable mentioned above, Steve Haber – a historian by training who, by his own admission, fled history for the purpose of making falsifiable claims – quipped that what HPE scholars are really doing is not founding a new and exciting discipline, but colonizing the old discipline of history. Historians, according to Haber, have ceded historical questions to economics and related social sciences a long time ago. Our worry is that an HPE uniquely focused on causal inference might follow political science and economics in ceding the big questions of our time to other fields: computer science is already in line. As this new and exciting discipline takes off, we should work together to avert this path.
Header Image: Plato’s Cave (Republic 514a–520a). Can we see HPE, or only a projection? Source: Wikipedia
 Inspired by Piketty’s work, archaeologists and historians have started to accumulate increasingly sophisticated measures of income and wealth in ancient societies and systematically compare them. In turn, the crucial insights about the origins and evolution of economic inequality drawn from archaeology have inspired a stream of interdisciplinary research that aims at testing more formally such ideas by combining historical econometrics with Gini indexes calculated on the basis of archaeological features, thus triggering a lively cross-disciplinary debate (link, link, link, link, link).