Economic history is increasingly able to provide us with evidence and inform pressing questions at the intersection of research and real-world decision-making. This column uses natural language processing and network analysis of articles from five leading journals over the past 25 years to identify a shift towards a more global, data-driven, and methodologically advanced field. It maps changing thematic priorities, institutional collaborations, and author networks, highlighting both a generational turnover and growing geographical diversity. It also reveals a strong move towards causal identification-based econometrics alongside a sharp decline in qualitative research, signalling both convergence and trade-offs in the discipline’s integration with mainstream economics.
Back in 1976, Deirdre McCloskey posed a provocative question: “Does the past have useful economics?” Her answer was an unequivocal “yes”. McCloskey argued that economic history is a vital component of economics. It provides context, enables us to test theories against real-world events, and uncovers the human complexities that abstract models often sweep under the rug (McCloskey 1976).
Today, economic history stands increasingly ready to provide evidence and inform pressing questions at the intersection of research and real-world decision-making (Abramitzky 2015, Diebolt and Haupert 2019). In this context, it is both necessary and timely to examine how its central debates and publishing patterns have evolved, as the field continues to adapt to contemporary academic demands and emerging policy challenges.
This column leverages natural language processing and network analysis in the articles published in the five leading economic history journals between 2000 and 2024 to analyse the main debates enlivening the discipline, trace citation linkages among authors, institutions, and journals, as well as shifts in methodological approaches. 1 Our findings reveal a field that has become markedly more global, methodologically sophisticated, and intellectually integrated (Galofré-Vilà and Gómez-Blanco 2025).
From Anglo-American hegemony to a multipolar landscape
One of the initial findings of our analysis is the marked decline in Anglo-American dominance within top economic history journals. In the early 2000s, scholars based in the US and the UK accounted for over half of all these publications. Today, their share has fallen below 30%.
To further map where economic history is being written, we pool the data for the articles published in the last 25 years. Figure 1 presents a network where node size represents citation impact (with bigger bubbles corresponding to a larger number of citations), and the colour of the bubble indicates average publication year. Clusters comprise universities that display some degree of affinity, as indicated by the frequency of co-citations among their published works.
Figure 1 Bibliometric network of institutions
Anglo-American universities (such as Harvard, Oxford, LSE, Cambridge, UC Davis, or Vanderbilt) are central in the network and lead in citation impact, with prominent policy-research organisations like NBER and CEPR serving as vital connectors, linking authors and intellectual communities. Nevertheless, in recent times this dominance has become less absolute. Universities in Spain (e.g. Carlos III de Madrid and Valencia), Sweden (Lund), Netherlands (Utrecht and Groningen), and Italy (Bocconi) have emerged as relevant nodes in the global collaboration network. However, while European institutions have gained prominence within the network, the map continues to expose persistent structural imbalances, with universities in Africa, Asia, and Latin America occupying marginal positions, due to different channels to disseminate research, barriers of funding, and editorial preferences.
The architects of the discipline
To continue with the network analysis, Figure 2 maps the relatedness among economic historians, based on the assumption that authors cited together share some degree of intellectual affinity. This network captures how authors (and, consequently, the ideas and debates associated with them) are positioned relative to one another within the field. Robert C. Allen, Jan Luiten van Zanden, Jeffrey G. Williamson, Sheilagh Ogilvie, Stephen Broadberry, Gregory Clark, Kevin H. O’Rourke, and Leandro Prados de la Escosura stand at the forefront of economic history, having produced influential work on the origins of economic growth, living standards, the role of institutional, and historical national accounts.
Interestingly, around this core, a younger cohort of scholars including David S. Jacks, Guido Alfani, Ewout Frankema, Jutta Bolt, Johan Fourie, Nuno Palma, and Sergio Correia has emerged, pioneering new cliometric-type research. Their nodes, tinted yellow in the network, also betray more recent publications, signifying a generational turnover that is broadening both geography and methodology.
Figure 2 Bibliometric network of authors
When interpreting our networks, one must bear in mind that our analysis is confined to articles published in five journals since 2000, which inevitably leads to the omission or underrepresentation of several influential figures in the field. Scholars like Joel Mokyr, Jane Humphries, and the late Nicholas Crafts appear less visible in our networks simply because their seminal contributions are found in monographs. Likewise, authors like Hans-Joachim Voth, Sascha O. Becker or Thomas Piketty, who frequently publish historical papers in general economics journals, fall outside our journal-based citation maps.
Journal preferences
Using language processing we also show how each journal has cultivated a distinct editorial identity. Clustering algorithms help to categorise articles based on their subject matter and historical period. Figure 3 presents the conditional probabilities of each journal publishing articles on specific topics (left panel) or historical periods (right panel). Darker shades in the heatmaps indicate a greater likelihood of coverage.
Figure 3 Conditional probabilities
These figures reveal notable editorial leanings, with the Journal of Economic History having an orientation towards institutions and human capital (two longstanding pillars of the discipline), Explorations in Economic History an emphasis on demography and inequality, the European Review of Economic History towards trade studies, Economic History Review on industry and technology (publishing many local studies tied to the British Industrial Revolution), and Cliometrica on economic growth and historical statistics. Chronologically, there are also clear distinctions by journals: the Economic History Review (and to a lesser extent European Review of Economic History) favours pre-industrial topics, whereas the Journal of Economic History, Explorations in Economic History and Cliometrica concentrate predominantly on the 19th and 20th centuries.
Shifting methods
Finally, as part of the cliometrics revolution, over the last 25 years and especially over the last decade, economic history has undergone a profound and perhaps radical transformation in terms of how research is being conducted. In Figure 4, for some benchmark years, we classify each paper according to its methodological toolkit. We observe that in 2000 around 70% of the papers used time-series analysis, but today this has reduced to less than 40%. We also note that over time, descriptive time-series studies have largely been supplanted by more sophisticated econometric methods, including panel data models, instrumental variables, and difference-in-differences designs.
Figure 4 Methodological evolution
The last 25 years have also witnessed the near-extinction of purely qualitative research, declining from over 15% of published articles to under 3%, and the surge of machine-learning and text-mining approaches, signalling the discipline’s growing interest in harnessing big data and new methods for historical inquiry. These methodological shifts echo broader transformations within the discipline, and make a case, as discussed by Margo (2021) and others, for the increasing convergence of economic history within mainstream economics.
Concluding remarks
Our survey of the literature shows how over the last 25 years economic history has matured into a global, data-driven discipline. It has decisively outgrown its early Anglo-American core to embrace a multinational scholarly community. Indeed, not only have the centres of scholarly output shifted, but a new generation of economic historians is now driving the field, with an interest in new thematic areas such as inequality and demography, as well as the use of new methods.
Clearly, the embrace of causal identification-based econometrics, alongside new computational tools such as machine learning, reveals a discipline increasingly aligned with the methodological frontiers of economics, likely signalling a new step in the cliometrics revolution. This progress, however, has not come without trade-offs: the marginalisation of qualitative work may raise important concerns about the erosion of contextual depth.
Source : VOXeu
