When does local representation matter? The case of Ancien Regime France

By Anne Degrave (Institute for Advanced Study in Toulouse, Université Toulouse)

Over the course of state-building, rulers face a well-known choice between centralizing governance or delegating power to local elites. A large literature explores factors influencing this choice and long-run implications on political and economic development. However, whether the general population fares better under these institutional arrangements is poorly understood. For most citizens, does it make a difference to be ruled indirectly by local elites as opposed to being ruled directly by an authoritarian central state?

Theoretically, whether the general population benefits from local (or indirect) rule is unclear. As compared with central rule, institutions of local (or indirect) rule have better information on local conditions, and they might also be more accountable to citizens. Many of such historical institutions are considered proto-democratic, implying positive outcomes. However, unless the population is broadly represented at the local level, local rule also means lower oversight of elites by the central state. In other words, when left unchecked, elites may use their informational advantage to extract resources from local populations, leading to worse outcomes for most citizens.

Provincial representation and absolute monarchy in Ancien Regime France

In a paper published in May 2023 in the Journal of Historical Political Economy, I explore this question in the context of Ancien Regime France. This historical case is typically associated with absolute monarchy and direct rule, but in fact, around a third of the territory was governed indirectly through provincial representative institutions, called provincial estates.

Figure 1: Consent to taxation in French provinces (1789)

Their main role was to “consent to taxation”, which involved bargaining with the crown over royal taxation amounts and levying it. In some instances, they borrowed on the crown’s behalf and paid the debt back with taxes they raised. They also retained some revenue to fund local projects, including agricultural improvement, or building roads and canals. In contrast, in provinces without provincial estates, taxation was levied by central agents without formal or informal consent of the governed.

Clearly, these institutions were not representative in a modern sense. They were assemblies of the three social orders of the time — Clergy (0.5% of the population), Nobility (1.5%), and Third Estate (the commoners, 98%). Typically, each order had one vote, despite the Third Estate representing 98% of the population. In the provincial estates voting “by head”, the votes of the Third Estate never exceeded 50%. Further, Third Estate representation was narrow: delegates were the unelected mayors of the main provincial cities.  As a result, the interests of peasants were represented only indirectly through their noble or ecclesiastical landlords.

Figure 2: Meeting of the Languedoc Estates, December 4, 1704

Prominent thinkers such as Tocqueville and Montesquieu praised the governance of the provincial estates. Later, historians were very critical of their actions and saw them as purely oligarchical, though the current consensus is more nuanced. Given that the alternative to this weak form of representation was no representation at all, was there a positive effect for peasants? I address this question by conducting an empirical comparison of outcomes in the provinces with or without representation. As we will see, I do not find much evidence that the provincial estates had substantial positive effects for the general population.

Table 1: Characteristics of regions with or without local rule

Empirics: Popular rebellions, grievances, and living standards

Measuring historical welfare level is challenging, especially in France where the archives of the Finance ministry burned down in a fire during the 1871 Commune. As an alternative, I use detailed information on popular grievances, and whether they are driven by central taxation or the rent-seeking actions of local elites. I mainly use two sources, drawing on the work of two groups of historians:

  • Popular rebellions between 1661 and 1789, from Nicolas (2002). Historian Jean Nicolas and 58 collaborators looked for references to “rebellions” in local archives, defined as “when a group of at least four individuals, not belonging to the same family, directly perpetuates violence against one or more representatives of a political, religious, economic power, etc., or attacks property, buildings, furniture, papers, various signs symbolizing these powers.” (Nicolas, 2002, p. 75).[1]
  • Grievance lists drafted for the 1789 Estates-General, from Shapiro et al. (1998). In January 1789, the monarchy convened the Estates-General, a national assembly of the three orders (the first one since 1614), a process which ended up triggering the French Revolution. 384 districts elected delegates to communicate their grievances to the King. I focus on the Third Estate grievances. [2]
  • As a complement, I also use partial data on living standards and economic development, during and after the Ancien Regime, available from various sources.
Figure 3: Popular grievances: rebellions and grievance lists

I compare locations with and without provincial estates. To address compound treatment issues, models include a number of geographic, administrative, and political controls. I also use a geographic regression discontinuity approach to minimize bias from the largest potential confounds, under the assumption that these are not homogeneously distributed within provinces. There could still be unobserved confounding variables, although the direction of the bias is unclear. The estimates should be interpreted with caution.

Results: consent to taxation made little difference

First, I find that local representation had no substantial impact on either living standards or economic growth in the short or long run (Table 2).

Table 2: Living standards and economic development (RD in a 50 km bandwidth)

Second, both rebellions and grievances about taxation are less frequent under local rule, suggesting greater satisfaction with royal taxation (Figure 5). Further evidence suggests that peasants benefited from elites mitigating the tax burden for themselves, but that these benefits were relatively limited.

Figure 4: dissatisfaction with central taxation (RD results)

Third, local rule had negative consequences for the general population on areas other than tax collection. Overall, popular rebellions caused by the actions of local elites are more frequent under local rule. Grievances about the seigneurial system are also more frequent, although only for a subset of provincial estates where the influence of privileged orders was the greatest.

Figure 5: Dissatisfaction with local elites (RD results)

Local elites in regions with provincial estates exploited the lack of oversight by the central state to extract more resources from peasants. This suggests that for peasants, local rule made little difference as the lower pressure from the central state was counterbalanced by increased pressure from local elites.


Local rule and so-called representative institutions, because they were dominated by elites, had no major positive effect for the general population. While it improved overall satisfaction with taxation, its benefits were limited. Further, it made it easier to exploit peasants through other channels than state taxation, notably through the infamous seigneurial regime.

A limitation is that I cannot measure the net result of the countervailing mechanisms of rent-seeking by state agents and rent-seeking by local elites. Still, these results contribute to our understanding of when representation might or might not matter for welfare. Unless they are structured to include non-elite interests, institutions of local rule by themselves are unlikely to improve outcomes for the general population.


Nicolas, Jean (2002). La rébellion française : mouvements populaires et conscience sociale, 1661-1789. Editions du Seuil.

Shapiro, G., Markoff, J., Tackett, T., & Dawson, P. (1998). Revolutionary demands: A content analysis of the Cahiers de Doléances of 1789.

[1] The Jean Nicolas data was made available by the HiSCoD project: https://www.unicaen.fr/hiscod/hiscod.html.

[2] The full dataset of the content analysis of 1789 grievances from Shapiro et al. (1998) is available on my homepage: https://annedegrave.com/data/.


  • Anne Degrave

    I am a postdoctoral research fellow at the Institute for Advanced Study in Toulouse. I obtained my PhD in Politics at New York University in Spring 2022. I study comparative politics, historical political economy and state formation. My work investigates the implications of state formation for citizens and their relations with the state and elites, using quantitative analyses of historical data.

Leave a Reply