How Radical Administrative Reforms Unfold: Evidence from France’s “Rotating Capitals”

In most modern states, public administrations are organized in a pyramidal structure. National territories are divided into provinces, regions, or states, which are themselves subdivided into smaller and smaller administrative units, as in a game of nesting dolls. At these lower levels of administration, citizens can often expect to obtain all types of public services in their regional capital, from which the local state projects its power onto neighboring areas.

This particular way of organizing the state has not always been the norm historically, nor is it universal. Some countries still operate a dual legal system in which national institutions coexist alongside a range of informal or customary institutions. Even in Europe, the adoption of administrative systems of this kind is a recent phenomenon. In medieval times, European states were often governed through multiple layers of administration accumulated over centuries with no coherent design in mind. The complexity of these administrative arrangements reflected the long shadow of history and the intricate dynamics that led to the emergence of the continent’s first nation-states.

The problems of governance that resulted from this state of affairs were nowhere more salient than in France at the end of the Ancien Régime.  By 1789, at least three separate administrative hierarchies coexisted in the French Kingdom, each fulfilling different types of judicial, fiscal, and military functions. While some of these institutions were inherited from the Roman period, others had appeared under the feudal system (e.g. the bailiwicks, a type of judicial institution). Yet other administrative institutions had been created by the French monarchy as part of a broader endeavor to centralize power across the entire territory. The intendancy system, organized around administrative units known as généralités and later replicated in the Spanish Empire, best exemplified the latter type of institution. For citizens, the coexistence of these multiple institutions implied that a given territory could fall under distinct jurisdictions for different administrative matters.

In the 17th and the 18th century, philosophers of the Enlightenment began to question these institutional arrangements and promoted the idea that efficient government required easy and centralized access to the local administration for all citizens. These ideas would prove particularly influential during the American and the French Revolution. In revolutionary France, one of the first major reforms tackled by the Constituent Assembly was to overhaul administrative institutions inherited from the Ancien Régime and to divide the kingdom into a radically new administrative nomenclature inspired by principles of efficient government.

Exploring the historical debates that accompanied this reform delivers some fascinating insights. Indeed, the constitutional committee in charge of this reform considered a variety of radical designs, including a scheme to partition France into a perfect square grid (left panel below). Since such an artificial partition proved politically infeasible, the National Assembly settled on a new division of the territory into 83 equal-sized units known as the départements (right panel below), many of which cut across the boundaries of old French provinces. Inside these new administrative units, citizens would always live in close proximity to their nearest local capital, in which all types of administrative services would be available. In 1800, Napoleon consolidated the 1790 reform by creating inside these new capitals the function of prefect—a type of local governor concentrating all the local powers of the state.

New division of the French Kingdom in 1790: proposed and final designs

The rest of Europe watched with horror this radical departure from the old order, which involved tearing apart France’s existing provinces. In his Reflections on the Revolution in France published in 1790, the Anglo-Irish statesman Edmund Burke asked:  “Do you seriously think that the territory of France, upon the republican system of eighty-three independent municipalities (to say nothing of the parts that compose them), can ever be governed as one body or can ever be set in motion by the impulse of one mind? When the National Assembly has completed its work, it will have accomplished its ruin.”

Such a comprehensive reform provides a fantastic avenue to study how the Revolution shaped subsequent patterns of local development. In recent work, Cédric Chambru, Emeric Henry, and I exploit one particularly original feature of the 1790 reform, which allows us to study how the reallocation of administrative presence across space affected economic outcomes observed in the nineteenth and twentieth centuries. Facing tremendous pressure from local delegates in 1789–90, in about a third of French departments, the Constituent Assembly did not determine which city should become the new local capital. Instead, it allowed individual departments to establish a scheme called alternat whereby administrative functions would rotate across multiple candidate cities. In a handful of other cases, administrative functions were shared across multiple cities (as in the Mayenne department), or the choice of the final capital was left to a vote by local representatives (as in the Aisne department).

This compromise solution was obviously not viable, and it quickly appeared that these “wandering administrations” were a “ridiculous and expensive” policy design, as stated in a decree passed on September 11, 1791. On that date, the National Assembly voted to abolish rotations for good in all but one department (Cantal), and it ordered the local administration to remain where it was currently located. We show that this generated exogenous variation in the allocation of capital status, and we use this setting to compare the cities that were ultimately chosen (capital cities) to the ones that were initially considered (other candidate cities). Importantly, we provide evidence that both types of cities had similar economic potential before the Revolution: most notably, they were indistinguishable in terms of population levels, urban potential, and the likelihood to have hosted administrative functions under the Ancien Régime.

We explore how these capitals, once endowed with new administrative functions, diverged from other comparable candidate cities in the short and the long run. In the short run, interestingly, capitals did not grow faster than comparable candidate cities. Instead, the capitals concentrated investments in extractive and enforcement capacity. This allowed successive governments to raise taxes to fund and fight a series of external wars after 1792. We find that France’s new local capitals and their periphery were substantially more likely to have established their cadaster (a type of land registration system used for taxation purposes) by 1815, and that they paid higher levels of business tax in the country’s first industrial census conducted in 1839-47. Moreover, capitals sent more conscripts to Napoleon’s armies after 1802, and they concentrated investments in the construction of new prisons and tribunals. Consistent with these results, we find that capitals remained no more populated than other candidate cities up to five decades after the adoption of the 1790 administrative reform.

Over the long run, however, capitals dramatically outgrew their 1790 rivals. By the end of the nineteenth century, the population of capitals was 40 percent larger than that of comparable candidate cities. Based on complete census counts in a subset of departments, we estimate that civil servants and their families accounted for about a third of this difference, with the rest coming from economic growth and urban agglomeration. This demographic divergence, which continued in the twentieth century, is consistent with a large literature showing that public investments and market agglomeration forces promote economic growth around centers of administrative power. We represent the dynamic evolution of population in the following figure, where we plot the causal effect of the administrative reform on the population of capitals over time.

We also explore the dynamic effects of the reform on public goods provision and private sector activity. The reform’s impacts on these dimensions are substantial but take a long time to unfold. For example, the residents of capital cities begin to produce more patents in the second half of the 19th century, reflecting a higher pace of innovation from that point onwards. Two centuries after the reform, the local capitals and their periphery have much higher levels of public and private sector employment, but these effects have not yet fully materialized in the country’s first industrial census conducted between 1839 and 1847.

Our paper provides a unified framework to reconcile these contrasting effects of the 1790 reform in the short and the long run. The main insight behind this framework is that building a new administrative apparatus is an arduous endeavor whose dividends may only materialize in the longue durée. In the immediate aftermath of the 1790 reform, investments in the local coercive capacity of the state might have made capitals less attractive for firms and citizens alike. Empirically, we find that capital cities established their cadaster before other candidate cities and that they focused on building up enforcement capacity, in the form of tribunals and prisons. Over time, however, these investments allowed the local state to deliver local public goods such as schools, social services, and transportation and communication networks. Eventually, these investments in public goods and infrastructure stimulated private sector growth, which we measure in terms of innovation (patents) and financial activity (bank establishments).

France’s state-building experiment illustrates that it takes time and patience for effective states to emerge. Early investments in the state’s ability to extract resources require upfront efforts that may be costly for local residents. The dividends from such investments, while substantial in the long term, may not become apparent for several decades. These delayed benefits make the task of building effective states, which remains a key issue in many parts of the world, a particularly challenging enterprise.


  • Benjamin Marx is an Assistant Professor at the Department of Economics at Sciences Po in Paris. His research interests are mainly in political economy and development. One strand of his research studies the determinants of political accountability, state capacity, and voting behavior in developing countries. In connection with this research, he has worked in various countries, including Kenya, Senegal, and Uganda as well as, more recently, Indonesia. Another strand explores the political economy of religion, with the goal of better understanding the interplay between religious institutions, politics, and culture. He is also interested in the study of discrimination and the political economy of urban poverty.

  • Cédric Chambru is a post-doctoral researcher in Economic History at the University of Zurich. His research, which spans political economy and economic history, centers on understanding the role of institutions and state-building in long-run development. Another strand of his research studies the evolution of living standards during the Industrial Revolution. Other work investigates historical social conflict and colonial history, including the emergence and properties of the French imperial elite. Since November 2020, he is the co-maintainer of the Historical Social Conflict Database (HiSCoD).

  • Emeric Henry is a Professor at the Economics Department in Sciences Po. He is a microeconomist using theory, experiments, and empirical methods to study questions in law and economics.

Leave a Reply