“Let them eat cake”: drought, peasant uprisings, and demand for institutional change in the French Revolution

Waldinger, Maria Article — Published Version “Let them eat cake”: drought, peasant uprisings, and demand for institutional change in the French Revolution Journal of Economic Growth Provided in Cooperation with: Springer Nature Suggested Citation: Waldinger, Maria (2023) : “Let them eat cake”: drought, peasant uprisings, and demand for institutional change in the French Revolution, Journal of Economic Growth, ISSN 1573-7020, Springer US, New York, NY, Vol. 29, Iss. 1, pp. 41-77, https://doi.org/10.1007/s10887-023-09230-y This Version is available at: https://hdl.handle.net/10419/313129 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Vol.:(0123456789) Journal of Economic Growth (2024) 29:41–77 https://doi.org/10.1007/s10887-023-09230-y 1 3 “Let them eat cake”: drought, peasant uprisings, anddemand forinstitutional change intheFrench Revolution MariaWaldinger1 Accepted: 24 May 2023 / Published online: 8 July 2023 © The Author(s) 2023 Abstract The paper studies whether a drought in 1788 affected the outbreak of peasant revolts during the French Revolution. I construct a community-level data set with information on local drought severity and peasant uprisings in 1789. Communities with severe drought conditions more often experienced peasant revolts against the feudal system. Then, I investigate a mechanism through which drought may have affected peasant revolts. Those more affected by the drought had higher demand for institutional change as expressed in the lists of grievances. The results provide evidence on specific ways in which the drought of 1788 impacted the French Revolution, a milestone in the democratization of Western Europe. Keywords Democratization· Economic history· Weather shock JEL classification N0· Q0· P0 1 Introduction In 1788, a drought hit France and caused severe crop failure. Grain prices soared. A growing part of the population went hungry. At this point, France was already in deep financial, economic and political crisis, the result of numerous longand medium-term developments in French politics, the French economy and society.1 As a result, King Louis XVI had limited means to import grain from abroad or to quell public discontent using military means. While grain prices continued to rise, public discontent grew. It was the time of the year, the so-called lean season, when last year’s harvest had been depleted, and before the new The sentence “Let them eat cake” (fr. “Qu’ils mangent de la brioche” ) was attributed to MarieAntoinette (1755–1793), wife of King Louis XVI, to illustrate the selfishness and ignorance of the French upper class. In reality, Jean-Jacques Rousseau wrote this sentence already in 1765 in Confessions: “At length I remembered the last resort of a great princess who, when told that the peasants had no bread, replied: “Then let them eat brioches.” * Maria Waldinger [email protected] 1 ifo Institute, University ofMunich, Munich, Germany 1 See Sect.2.1 for a discussion of the long-term causes of the French Revolution. 42 Journal of Economic Growth (2024) 29:41–77 1 3 harvest was brought in. By summer 1789, open revolt against the feudal Ancien Régime had broken out. Several historians have hypothesized that the drought of 1788 and ensuing harvest failure affected the outbreak of revolts during the French Revolution (Le Roy Ladurie, 1971; Lefebvre, 1973; Neumann, 1977; Neumann & Dettwiller, 1990). In this paper, I provide an econometric test of this hypothesis. I first examine whether the drought affected the French Revolution by impacting the outbreak of peasant revolts in the summer of 1789. Then, I investigate a mechanism through which this impact may have operated and that has so far been little documented, drought’s impact on popular demand for institutional change. For this purpose, I construct a canton-level data set for all 3500 French cantons with information on local drought severity in the growing season of 1788, on local peasant uprisings in the summer of 1789, and on local demand for institutional change as documented in the cahiers de doléances (lists of grievances) in May 1789 (i.e. this is after the drought but before the outbreak of the revolution). In the first part of the analysis, I test whether peasants in areas more affected by the drought in 1788 were more likely to participate in uprisings. In the summer of 1789, shortly before the new harvest was to be brought in, famished peasants rose in revolt against their landlords. The peasants were enraged by the fact that they—many close to starvation— owed excessive taxes to their landlords. Their landlords, in contrast, were exempt from all taxation and also enjoyed numerous other political and economic privileges.2 The peasants attacked their landlords’ castles and burned charters enshrining these privileges. It is plausible that peasants more affected by the drought faced higher price increases and were more likely to rise in revolt against their landlords. Using data on the exact locations of around 300 peasant revolts (Lefebvre, 1973), I show that areas with more severe drought were significantly more likely to experience peasant revolts.3 I also show that the drought had a particularly large effect on early revolts from where protests spread to other parts of France. Historical evidence documents that these peasant revolts mattered greatly for the course of the French Revolution. They were one of the factors that finally coerced the National Assembly in Paris into committing to comprehensive institutional reform. In August 1789, “[...] in an effort to appease the peasants and to forestall further rural disorders”, the National Assembly formally abolished the feudal regime (Merriman, 2010, 447; Neely, 2008, 80). Then, I investigate a possible mechanism through which the drought may have impacted the outbreak of revolts: the drought’s impact on popular demand for institutional change. This demand is a key prerequisite for democratization (Acemoglu & Robinson, 2006, 2001; Lipset, 1959; Besley & Persson, 2019).4 The drought may have affected the content of the lists of grievances, for example, if those more affected by the drought had better 2 These privileges (privilèges) were a hallmark of the French feudal system. 3 To measure drought, I obtain information on local growing-season temperature and precipitation from Luterbacher etal. (2004) and Pauling etal. (2006). These data have been reconstructed from instrumentallymeasured weather records, historical sources, and geological archives such as tree ring series, ice core data and lake sediments. 4 A recent strand of literature has set out to identify specific factors that influence this demand—such as previous experience with democracy at the individual, local, or national level (Fuchs-Schündeln & Schündeln, 2015; Giuliano & Nunn, 2013; Persson & Tabellini, 2009; Welzel, 2007); state indoctrination through education (Cantoni etal., 2017); and socialization during childhood (Sapiro, 2004). 43 Journal of Economic Growth (2024) 29:41–77 1 3 information on the inadequacy of the existing regime to protect the common population from such calamities. For this purpose, I use information from the cahiers de doléances (lists of grievances). These contain the results of a nationwide, locality-level survey, commissioned by King Louis XVI, on the French population’s demands for change, including demands for institutional change. The survey was undertaken in the spring of 1789, this is after the drought but before the storming of the Bastille on July 14th, 1789, which is often viewed as the beginning of the revolution, in preparation of the Estates General.5 Using information on more than 1100 demands for institutional changes from more than 600 lists of grievances, I document that commoners in areas hit more severely by drought conditions had significantly higher demand for institutional change, e.g., for eliminating certain privileges of the nobility, for establishing free municipal elections, and for creating a free press. Demands were chosen that represent hallmarks of democracies: equal political and economic rights irrespective of class, equality before the law, the demand for elections, and for a free press. They represent a clear departure from the class-based, highly unequal Ancien Régime. During the French Revolution, growing demand for institutional change in general and the content of the lists of grievances in particular gave the representatives of the Third Estate at the Estates General in Paris, later at the National Assembly, a clear mandate to fight for more participatory institutions and a more equitable distribution of political and economic rights.6 The paper contributes to the literature examining the relationship between the drought of 1788 and the French Revolution (Le Roy Ladurie, 1971; Lefebvre, 1973; Neumann, 1977; Neumann & Dettwiller, 1990; Labrousse, 1939, 1944). It provides the first econometric test of the hypothesis that the drought influenced the French Revolution, the outbreak of peasant revolts and popular demand for institutional change in the lists of grievances. The paper shows that demand for institutional change, which had already developed very gradually over many decades, was further affected by this short-term event. In addition, I show that the effect of the drought was particularly large for the crucial first revolts that later-on spread across France. It also contributes to work that uses information from the lists of grievances to gauge public sentiment before the Revolution, such as Hyslop (1936, 1968), Shapiro and Markoff (1998), and Johnson (2015). It also contributes to more recent literature addressing important questions in economics using rich French historical data from the French Revolution or from later time periods (Franck, 2016; Squicciarini & Voigtlaender, 2016; Franck & Michalopoulos, 2017; Squicciarini, 2020). This paper contributes to this literature by examining the role of a weather shock in an especially important moment in French history. The paper also adds to work on the relevance of historical climatic and geographic conditions for human development. Galor and Özak (2016) show that pre-industrial agro-climatic characteristics shaped economic behaviour today. It is also related to papers examining the political impacts of shocks to geographical conditions, such as short-term weather shocks (Acemoglu etal., 2020; Dell, 2012; Brückner & Ciccone, 2011; Spoerer & Berger, 2001; Belloc etal., 2016; Zhang etal., 2006; Burke & Leigh, 2010; Miguel etal., 2004; 5 The Estates General was an assembly of representatives of the three estates (the First Estate representing the clergy, the Second Estate representing the nobility, and the Third Estate representing all commoners) that met in Paris in May 1789 to debate ways out of the country’s deep political and financial crisis. 6 This is consistent with economic theory, where manifestations of demand for democracy by the disenfranchised increase incentives of the ruling elite to agree to institutional reform (Acemoglu & Robinson, 2001, 2006). 44 Journal of Economic Growth (2024) 29:41–77 1 3 Burke etal., 2009; Bai & Kung, 2011; Chaney, 2013). Consistent with Acemoglu etal. (2001; 2006), this literature shows that such short-term shocks can have both short-term political impacts and long-term consequences for development if they appear at a pivotal moment in history. So far, this literature has focused on the impact of weather shocks on civil wars, uprisings, and other forms of violence.7 This paper contributes to this strand of literature by documenting that the weather shock not only affected peasant uprisings but also demand for institutional change in the population. This could be interpreted as an additional channel through which an economic shock leads to political unrest. It may, for example, be a relevant mechanism underlying the relationship between weather shocks and democratic transitions documented in Brückner and Ciccone (2011). Historians generally agree that both the peasant revolts and the demands for institutional change mattered greatly for developments during the French Revolution that had farreaching consequences, e.g. the insistence of the Third Estate on political reform during the National Assembly and the abolition of the feudal system. Hence, this paper sheds light on one of the factors that shaped this landmark event in the democratization of Western Europe. 2 Historical background 2.1 Long‑term causes oftheFrench Revolution In the 1780s, France was in deep economic and political crisis. After years of warfare and an exuberant lifestyle at the French court, France’s financial resources were exhausted.8 Unlike the English monarchy—the French monarchy lacked a financial institution such as the Bank of England from where it could have borrowed at lower interest rates (Norberg, 1994; Bossenga, 2011; Hoffman & Rosenthal, 2000). A related problem was the split of authorities over expenditures (the government) and taxation (the Parlement de Paris). Increased spending was not financed by increasing taxes but had to be financed by debt (White, 1995). The highly inefficient tax system exempted the rich and overburdened the poor. It was firmly in the hands of the nobility and did not generate sufficient income for the state. When the French king sought to reform the rigid and inequitable tax system, which would have reduced the nobility’s income from taxation, a lengthy dispute between the king and the French nobility ensued. The members of French nobility were not willing to agree to reform unless the king considerably increased their political power (Hoffman & Rosenthal, 2000) One of the reasons for the inefficiency of the French tax system was that the right to raise taxes had been sold to individuals in return for short-term loans. Many powerful groups within the French society, especially the aristocracy and the church, had also been exempt from taxes in return for their loyalty, and their political and military support. In the French political tradition, these privileges were irrevocable. Each new king was accepted by the provincial nobility based on his promise to respect the existing privileges (Neely, 2008). In the long-term, the money that French kings had raised in return for tax 8 The wars include the War of the Austrian Succession (1740–48), the Seven Years’War (1756–63), and the American War of Independence (1775–1783). 7 Hsiang and Burke (2014) provide a useful overview. 45 Journal of Economic Growth (2024) 29:41–77 1 3 exemptions and other privileges did not generate as much income as taxes. “Fiscal immunities, whether inherited or purchased, undermined French finances and put the king in the apparently ridiculous situation of taxing those who had nothing to tax. They also made it impossible for the Crown to reform the fiscal system and render it more equitable and more efficient,” (Norberg, 1994). The manifold privileges and tax exemptions did not only render the French tax system inefficient and unresponsive they also lead to a highly unequal distribution of the tax burden across the French population. The poorest part of the population paid taxes to the King and had feudal obligations toward their local landlords (Neely, 2008). This strained economic situation had been further aggravated in the 1780s by a crisis of the manufacturing sector. This crisis caused unemployment of textile workers and decreased tax revenues. Finally, intellectual developments also played “an essential role in transforming a fiscal crisis into a revolution,”(Kaiser & Van Kley, 2010). New ways of thinking about political rights of all people influenced the behaviour of the Third Estate in the Estates-General. 2.2 Lists ofgrievances By spring 1789, the French king Louis XVI was under immense political and economic pressure. To avert state bankruptcy he had to solve the stalemate with the French nobility over the reform of the tax system. In spring 1789, he felt forced to convene the Estates General. It was the first time since 1614. The Estates General was an assembly of representatives of the three estates (clergy, nobility, and commoners) with a purely consultatory function. The plan of both the king and the nobility was to win the clergy and commoners over to take their respective side. In preparation of the Estates General, the King asked the nobility, the clergy, and the male tax-paying population of the Third Estate9 to hold town hall meetings at the parish or town level, and to produce lists of grievances (cahiers de doléances) (Merriman, 2010, 443). These grievances could include any topic concerning all spheres of French public life. The estates then elected representatives at the bailliage (electoral-district) level who took these lists to the Estates General in Paris.10 Starting in May 1789, representatives of all Estates advocated for these requests at the Estates General in Paris. The lists of grievances are a unique historical source. An important question is how reliable they are and how their content can be interpreted. First of all, as researchers have pointed out, the lists of grievances are not reports on living conditions in France. Instead, they are an expression of the population’s sense of these living conditions. For example, a demand to reduce a certain tax is not a measure of the actual tax burden but a measure of the population’s sense that the tax burden was too high or unjust. This sense is exactly what I am interested in studying. Assemblies were held to draw up the lists of grievances. It is thinkable that power dynamics at these assemblies influenced their content. Shapiro and Markoff (1998) find that the presence of outsiders at these assemblies, such as seigneurial representatives or lawyers, influenced the form but not the content of the lists. They also show that the lists of grievances were taken very seriously and produced with great care, despite the limited time availability. Shapiro and Markoff (1998) also find that more educated commoners or 9 That is the male tax-paying population above 25 years of age. 10 The lists were also summarized at the electoral district level. I include the original locality-level lists and lists at the electoral district. 46 Journal of Economic Growth (2024) 29:41–77 1 3 commoners with more economic or political power exercised greater influence on the content of the lists compared to the uneducated and illiterate, even though the latter were also present at these assemblies (Hyslop, 1936, 1968). It is important to keep this in mind when interpreting results. 2.3 Peasant revolts Hopes had been high among commoners that the Estates General would establish a more equitable and participatory political and economic system. Yet, the nobility and the king completely disregarded their demands for institutional reform and these hopes were soon disappointed. In July 1789, any sort of major political reform remained elusive, and frustration grew among the common population. Besides, there were rumours that the aristocracy planned to starve the rural communities into submission. Finally, the population was under enormous economic pressure due to the failed harvest of 1788 (Lefebvre, 1973). By the early summer of 1789, the rural population was at their most vulnerable because any reserves were depleted, the so-called lean season. “As the price of bread rose to its highest points [...], the interaction between subsistence problems and the stimulated popular awareness of politics became explosive,” (Campbell, 2006, 32). As mentioned before, a large part of the agricultural sector consisted of small peasant cultivators who lived just above self-sufficiency, with little in the way of a financial cushion. In July 1789, shortly after the storming of the Bastille, bread prices rose to new heights, and peasants organized armed protests against their landlords, against the heavy tax that they owed to them and the inequitable distribution of economic and political rights (Lefebvre, 1973, 118). The financial crisis limited the king’s means to either appease or subdue these uprisings. No longer able to pay for grain imports and unable to pay military troops, the king could no longer rely on the two main strategies he had used in the past to avert social unrest. By August 1789, the majority of the National Assembly, nobility, clergy, and the Third Estate alike, wished to restore public order. Many sympathized with the peasants’ request to abolish both the peasants’ feudal obligations and the nobles’ economic and political privileges; others, whose properties were attacked, were willing to make concessions to limit further damage (Neely, 2008, 80). “On August 4, 1789, in an effort to appease the peasants and to forestall further rural disorders, the National Assembly formally abolished the “feudal regime,” including seigneurial rights [and] renounced privilege, the fundamental organizing principle of French society,” (Merriman, 2010, 447). 3 Data andvariables To examine whether the drought of 1788 impacted political outcomes in 1789, I construct a cross-section data set for 3,666 French cantons11 with information on local drought exposure, on the outbreak of peasant revolts, on local demand for institutional change as expressed in the lists of grievances, and on control variables. 11 The canton is a small administrative division of the French state, one level below the départment. The 3,666 cantons in the data set are all cantons in mainland France. Other territories, e.g. Corsica and French overseas territories (France d’outre-mer) were excluded because most data were not available for these areas. 47 Journal of Economic Growth (2024) 29:41–77 1 3 3.1 Drought in1788 In historical contexts, drought has commonly been measured as deviation in precipitation from the long-term mean (Acemoglu etal., 2020; Dell, 2012) or as deviation in temperature from the long-term mean (Chambru, 2019). The European Drought Observatory (European Drought Observatory, 2012) defines drought as both a lack in precipitation or high temperature or a combination of both that leads to a lack in water supply.12 Based on previous work and following this definition, I measure drought severity as deviations in local growing-season temperature and in precipitation in 1788 (from the long-term mean) and the interaction of the two variables. The interaction effect of temperature and precipitation on crop yields is well-documented (Leng etal., 2016; Matiu etal., 2017; Ray etal., 2015). I use deviations from local long-term means in temperature and precipitation as explanatory variables to capture the effect of the local weather shock rather than the effect of a canton having a warmer and drier climate in general. I obtain information on temperature and precipitation in the growing season of 1788 from Luterbacher etal. (2004) and Pauling etal. (2006).13 These are grid data with grid cells measuring about 50 by 50kms. The data have been reconstructed by paleoclimatologists based on instrumentally-measured weather data, historical records, and geological archives, such as tree-ring series and ice cores from all over Europe. France was among the earliest adopters of weather stations and already had several in the second half of the 18th century (Garnier, 1974). Verification checks of the authors also show high predictive power for France in the 18th century (Pauling etal., 2006). These data are unique in their high geographical and temporal resolution and are regularly used including by Sequeira etal. (2020) and Tabellini (2020). Following Dell (2012)’s definition of drought, I measure drought as the ratio of growing-season temperature in 1788 over the long-run growing-season temperature mean (1750 to 1800).14 Deviations from the long-run precipitation mean in the growing season of 1788 are defined accordingly (see Eqs.1 and 2). I exploit the fact that—while long-term average temperature and precipitation remain relatively stable over time—the deviation from longterm mean temperature and precipitation is exogenous. Table1 shows summary statistics. Growing-season temperatures in 1788 were up by between 2.5 and 15 percent. Growing-season precipitation levels in 1788 were between 25 (1) Temperature cr = Temperature 1788 1 51 1750 ∑ 1800 Temperature t (2) Precipitation cr = Precipitation 1788 1 51 1750 ∑ 1800 Precipitationt 12 High temperatures may lead to a lack in water supply by increasing evapotranspiration. 13 The growing season is defined as the spring and summer season spanning the months March to August. 14 Table11 shows that results are robust to the use of other definitions of long-term temperature and precipitation. 48 Journal of Economic Growth (2024) 29:41–77 1 3 percent below the long-term mean and 3 percent above the long-term mean. Probability of peasant revolt and higher demand for institutional change were higher in warmer areas. These outcomes were also higher where precipitation was higher. In Sect.4, I will explore the contributions of each weather component and, importantly, of their interaction to outcomes. Figure1 shows a map of the distribution of temperature and precipitation deviations in the growing season of 1788. Different parts of France were differently affected by the drought. Areas in central France and Eastern France were especially affected (in red and orange). Northwestern France was especially affected by low precipitation (in beige), whereas Southeastern France was especially affected by high temperatures (in yellow). Table A.1 in the online Appendix shows local characteristics by how strongly a canton was affected by the drought. When comparing those most affected by the drought (with relatively warm and dry climate) to those less affected (with relatively cool and wet climate)15 we see that population density is lower in more affected areas. Literacy is also lower. Population density is generally related to more popular unrest. A random correlation between severe drought conditions and population density is unlikely to be a concern for omitted variable bias here. 3.2 Peasant uprisings To measure whether a canton experienced a peasant revolt in the summer of 1789, I collect data on the exact locations of 299 peasant revolts during the summer of 1789 using information from Lefebvre (1973, 4).16 Based on place names, I identify geographic coordinates for each peasant revolt. I then create an indicator variable, PeasantRevolt, that is one if canton c in region r experienced a peasant revolt in 1789, 0 otherwise. Figure1 provides a map of peasant revolts. They touched most parts of France, except for Brittany in northwest France. Panel 2 in Table1 shows that the share of cantons that experienced a peasant revolt is higher among relatively warm cantons (with above-median growing-season temperature deviation) compared to relatively cool cantons (with below-median growing-season temperature deviation). 3.3 Lists ofgrievances I collect information on demands for institutional change from the lists of grievances (cahiers de doléances). I base my analysis on one of the most comprehensive collections of lists of grievances compiled by the French Parliamentary Archives (Mavidal & Colombey, 1870). The French Parliamentary Archives are a chronologically-ordered edited collection of archival and published sources on the French Revolution. It was conceived in 1862 as a definitive record of parliamentary deliberations at the request of the French legislature. They started as a curatorial endeavour spearheaded by government archivists. It later became a scholarly project under the direction of leading historians. The French Archives parlementaires (Mavidal & Colombey, 1875, 1870) were an effort to parallel government publications in other countries such as the Journals of the House of Commons, the formal records of House of Commons business in the UK. They contain about 600 lists of grievances, including the name of the place where the list was produced, and the estate (nobility, clergy, Third Estate). 15 Note that all cantons had above average temperature in the growing season of 1788. 16 This constitutes the most comprehensive database on peasant revolts in France in the summer of 1789. 55 Journal of Economic Growth (2024) 29:41–77 1 3 4 Empirical analysis andresults In this section, I examine the impact of drought in 1788 on two political outcomes: the outbreak of peasant revolts in the summer of 1789 and the demand for institutional change as expressed in the lists of grievances in the spring of 1789 in Sects.4.1 and 4.2. Section4.3 explores heterogeneous impacts of the drought in Pays d’Élection and Pays d’État. 4.1 Peasant uprisings First, I examine the relationship between the drought of 1788 and peasant uprisings in the summer of 1789. To estimate whether the drought impacted the outbreak of peasant revolts I estimate the following equation: where PeasantRevoltcr is an indicator variable that is one for canton c if an uprising took place in this canton, and zero otherwise. Drought conditions are captured by the interaction of the two following weather variables: growing-season temperature (deviation of the growing-season temperature of 1788 from the long-term mean), and growing-season (3) PeasantRevolt cr = 𝛼 + 𝛽Temperature cr + 𝛾Precipitation cr +𝛿Temperature ∗Precipitationcr + RegionFEr + Xcr + 𝜖cr 0 .1 .2 .3 .4 .5 Share of cantons 123456 Fig. 2 Share of cantons with demands for institutional change. Notes: The graph shows the share of cantons for which a cahier exists (608 in total) that demand a specific institutional change. 1 = the share of cantons that demand at least one institutional change; 2 = the share of cantons that demand better political representation of the Third Estate and access to all professions; 3 = the share of cantons that demands a free press and the end of censorship; 4 = the share of cantons that demand the introduction of municipal elections; 5 = the share of cantons that demand to abolish feudal privileges of the nobility; 6 = the share of cantons that demand to abolish Lettres de Cachet 56 Journal of Economic Growth (2024) 29:41–77 1 3 precipitation (deviation of the growing-season precipitation of 1788 from the long-term mean, see Sect.3.1). RegionFEr is a full set of region fixed effects; Xcr is a set of control variables (see Sect.3.4 for a detailed description of control variables) and 𝜖cr is the error term. Standard errors are adjusted for 299 clusters at the grid level of the underlying temperature data set. Fig. 3 Cantons with list of grievances. Notes: This map is a bivariate map of the geographic distribution of temperature and precipitation in the growing season of 1788 in France. Areas with high temperature and low precipitation are shown in red, areas with low temperatures and high precipitation are shown in green. Areas with low temprature and low precipitation are shown in yellow and areas with high temperature and high precipitation are shown in bright yellow. The map also shows cantons with a list of grievances in grey 57 Journal of Economic Growth (2024) 29:41–77 1 3 Table 2 Drought and peasant uprisings (1) (2) (3) (4) (5) (6) (7) (8) Growing-season 24.62*** 23.28*** 28.67*** 30.38*** 30.59*** 31.02*** 31.86*** 24.64*** Temperature (4.892) (5.125) (5.697) (5.677) (5.864) (5.768) (5.996) (5.658) Growing-season 26.54*** 25.19*** 31.04*** 32.76*** 33.01*** 33.59*** 34.28*** 26.84*** Precipitation (5.387) (5.603) (6.164) (6.126) (6.358) (6.235) (6.478) (6.141) GS temperature* −25.38*** −24.09*** −29.59*** −31.22*** −31.47*** −31.98*** −32.71*** −25.49*** GS precipitation (5.158) (5.356) (5.874) (5.841) (6.066) (5.955) (6.189) (5.873) Control variables Regions FE Yes Yes Yes Yes Yes Yes Yes Pop. density 0.00** 0.00* 0.00* 0.00** 0.00* 0.00** (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) Canton area −0.00 −0.00 −0.00 −0.00** −0.00*** (0.00) (0.00) (0.00) (0.00) (0.00) Commun. 0.03 0.04 0.04 0.07 (0.11) (0.11) (0.11) (0.11) Literacy −0.00 −0.00 −0.00 (0.00) (0.00) (0.00) Paris −0.06** −0.08*** (0.03) (0.03) Salt tax 0.05*** (0.01) Marginal effect of increase in Temperature from p25 to p75 0.047*** Marginal effect of decrease in Precipitation from p75 to p25 −0.006 Observations 3666 3666 3666 3666 3666 3666 3666 3666 R-squared 0.006 0.007 0.008 0.009 0.010 0.010 0.012 0.016 58 Journal of Economic Growth (2024) 29:41–77 1 3 The table presents OLS estimates. Observations are at the canton level covering mainland France. Growing-season temperature is the growing-season temperature deviation in 1788 from the long-term mean. It measures average growing-season temperature in 1788 (March to August) as a percentage of long-run average growing-season temperature. Growing-season precipitation is defined accordingly. GS temperature*GS precipitation is an interaction term of the two weather variables. Peasant Revolt is an indicator variable that is 1 if a canton experienced peasant revolt in the summer of 1789. Control variables are population density, canton area, distance to closest road, whether a canton is subject to the great salt tax, literacy in 1690, whether a canton is part of the Paris region Ile-de-France, and region fixed effects. See Sect.3.4 for detailed information on control variables. Robust standard errors are clustered at the grid level of the underlying temperature data. Significance levels are indicated as *** for p < 0.01 , ** for p < 0.05 , * for p < 0.1 Table 2 (continued) 59 Journal of Economic Growth (2024) 29:41–77 1 3 I show results in Table2. In column 1, I estimate the specification without any controls, and then introduce controls one by one. Results show a significant relationship between drought conditions in the growing season 1788 and the outbreak of peasant revolts. The coefficient sizes and significance levels remain similar with the introduction of the control variables. Figure 4 depicts the relationship between temperature and precipitation and outcomes in a contour plot. The contour plot shows the predicted effect of temperature and precipitation on the outcome. It illustrates that the combination of high temperature deviations with low precipitation deviations predicts an increase in revolt incidence. The coefficients indicate that an increase in growing-season temperature from the 25th percentile (relatively cool) to the 75th percentile (relatively warm) where precipitation is low (at the 25th percentile) raises the probability of peasant revolt by 4.7 percentage points (an increase of 58 percent from the sample mean of 8 percent, significant at 1 percent). The marginal effect of a decrease in precipitation from the 75th percentile (relatively wet) to the 25th percentile (relatively dry) where temperature is high (at the 75th percentile) does not significantly affect the probability of revolt. This indicates that the effect I observe in the main results is mainly driven by the effect of temperature on revolt. These results are consistent with historical evidence that link the outbreak of peasant revolts to the drought in 1788. The timing of the revolts in July is also consistent with a causal relationship between the two events because July is in the lean season, the time of the agricultural year when last year’s resources are depleted and destitution and anxiety about the outcome of the coming harvest are on the rise. In 1789, anxiety was heightened by rumours that the aristocracy planned to quell any popular discontent and demands for reforms. Cantons most affected by the drought were more likely to face 1 1.05 1.1 1.15 Temperature .7 .8 .9 1 1.1 Precipitation 0 .1 .2 .3 .4 .5 .6 .7 Prob. Peasant Revolt Probability of Peasant Revolt Fig. 4 Marginal effects of temperature and precipitation on peasant uprisings. Notes: This contour plot visualises the estimated marginal effects of growing-season temperature deviation and growing-season precipitation deviation on the probability of peasant revolts. Yellow and orange indicate higher effect sizes, green shades indicate lower effect sizes 60 Journal of Economic Growth (2024) 29:41–77 1 3 higher prices, more destitution and anxiety and were therefore more likely to rise in revolt. Consistent with the description of control variables in Sect.3.4, signs on the coefficients of control variables show that areas with higher population density and those in smaller cantons are more likely to revolt. Lefebvre (1973) describe that revolts first broke out in a small number of places and then spread from there all over France. The data on revolt outbreaks in Lefebvre (1973) include information on the relative timing of the revolt (the authors use symbols to indicate the locations of the first revolts and arrows to indicate the geographic spread of these revolts, see Figure A.1 in the online Appendix). In Table3, I examine whether the drought had a role in triggering these early revolts. I create two indicator variables. EarlyPeasantRevoltcr is an indicator variable that is one for canton c if it experienced one of the 50 percent earliest revolts. LatePeasantRevoltcr is an indicator variable that is one for canton c if it experienced one of the 50 percent later revolts. I measure the timing of the revolt based on the number of arrows that lie between each revolt location and the closest location marked as “original revolt.” Table3 shows that the estimated effect of drought on early uprising is notably larger than its effect on later revolts. The result that the drought impacted peasant uprisings is consistent with findings from other settings on the relationship between weather shocks and political violence (Dell, 2012; Spoerer & Berger, 2001; Belloc etal., 2016; Zhang etal., 2006; Brückner & Ciccone, 2011; Burke & Leigh, 2010; Miguel etal., 2004; Burke etal., 2009; Bai & Kung, 2011; Chaney, 2013) and with economic theory. Acemoglu and Robinson (2001, 2006) show that a negative economic shock increases the probability of revolt among the disenfranchised by decreasing opportunity costs of revolt. Moreover, the revolts began in July, which is a crucial moment for peasants as this is typically the time when the resources from last year’s harvests are exhausted and when the peasants have not yet collected the crops from this year. Peasants therefore have to buy cereals in high amounts at this time of the year. In areas where the drought of 1788 was harsher, peasants certainly faced higher prices in July and could more easily fear that the nobility would corner part of the harvests to come. The peasant revolts were a significant event in the course of the French Revolution. They were one of the factors that finally—after months of discussions—coerced the political elite at the National Assembly into committing to comprehensive institutional reform, notably the abolition of the feudal system (see Sect.2.3, Merriman, 2010, 447; Neely, 2008, 80). 4.2 The Cahiers de Doléances anddemand forinstitutional change Then, I analyze a possible mechanism through which the drought may have increased peasant revolt, that is the impact of the drought on demand for institutional change among commoners as expressed in the lists of grievances. Those more affected by the drought may have had better information on the inadequacy of existing institutions to avert the adverse consequences of such calamities and may therefore have had greater demand for institutional change.20 20 Demands for institutional change in a broader set of areas captures a stronger will for more comprehensive institutional changes. 61 Journal of Economic Growth (2024) 29:41–77 1 3 I examine the relationship between drought and demands for institutional change in the lists of grievances by estimating the following specification: where Demandscr represents six outcome variables. The first variable is the number of demands for institutional change in canton c. The other five indicator variables capture five types of demands for institutional change: (1) to improve the political representation of the Third Estate and grant access of commoners to all professions, (2) to establish a free press and end royal censorship, (3) to introduce municipal elections, (4) to abolish feudal privileges of the nobility, (5) to abolish the King’s right to issue Lettres de Cachet (see Sect.3.3 for a description of these demands). I include these demands because they represent hallmarks of a democratic system: elections in a specific context (the election of municipal officers), more equal rights regardless of rank (a better political representation of the Third Estate, the abolition of the political, economic, and judicial privileges of the nobility, and the abolition of the King’s right to pass arbitrary judgment in the Lettres de Cachet) and (4) Demands cr = 𝛼 + 𝛽Temperature cr + 𝛾Precipitationcr +𝛿Temperature ∗Precipitationcr +RegionFE r +X cr +𝜖 cr Table 3 Early peasant uprisings The table presents OLS estimates. Observations are at the canton level covering mainland France. Growing-season temperature is the growing-season temperature deviation in 1788 from the long-term mean. It measures average growing-season temperature in 1788 (March to August) as a percentage of long-run average growing-season temperature. Growing-season precipitation is defined accordingly. GS temperature*GS precipitation is an interaction term of the two weather variables. Peasant Revolt is an indicator variable that is 1 if a canton experienced peasant revolt in the summer of 1789. ’Early Peasant Revolt’ is an indicator variable that is 1 if a canton was among the 50 percent earliest places who experienced peasant revolt. ’Late Peasant Revolt’ is an indicator variable that is 1 if a canton was among the 50 percent latest places who experienced peasant revolt. In column 2, I only include cantons in the sample with either early revolts or without revolts. In column 3, I only include cantons in the sample with either late revolt or without revolts. Control variables are population density, canton area, distance to closest road, whether a canton is subject to the great salt tax, literacy in 1690, whether a canton is part of the Paris region Ile-de-France, and region fixed effects. See Sect.3.4 for detailed information on control variables. Robust standard errors are clustered at the grid level of the underlying temperature data. Significance levels are indicated as *** for p < 0.01 , ** for p < 0.05 , * for p < 0.1 Peasant revolt Early peasant revolt Late peasant revolt (1) (2) (3) Growing-season 24.41*** 16.40*** 10.16** temperature (5.622) (5.517) (4.405) Growing-season 26.54*** 18.32*** 10.55** precipitation (6.093) (6.023) (4.834) GS temperature* −25.20*** −17.51*** −9.914** GS precipitation (5.827) (5.750) (4.625) Region fixed effects Yes Yes Yes Control variables Yes Yes Yes Observations 3666 3560 3473 R-squared 0.016 0.024 0.033 62 Journal of Economic Growth (2024) 29:41–77 1 3 a free press. While these features alone do not constitute a democracy, their introduction constituted first steps towards a democratic system in France. The indicator variable Political Representation takes a value of one for canton c if a location in canton c demanded better political representation of the Third Estate and access of commoners to all professions in their list of grievance, and zero otherwise. The indicator variable for demands 2) to 5) are defined accordingly. I base the analysis only on lists produced by the Third Estate. Except for the outcome variable, the specification is identical to specification 3 in Sect.4.1.21 Column 1 of Table4 shows a significant relationship between drought conditions and the number of demands for institutional change. Areas with more severe drought conditions listed more demands for democratic change in their lists of grievances. In Fig.5, I visualize the estimated effects of temperature and precipitation deviations in a contour plot. It shows that the interaction between high temperature deviations and low precipitation deviations predicts a higher number of demands for institutional change to be included in the lists of grievances. This is consistent with results in the previous section showing that the drought increased peasant revolts. A smaller increase in the number of demands is also predicted for areas with high precipitation and low temperatures, which seems surprising. On the other hand, this could reflect the effects of a hail storm that hit France on July 13, 1788. Hail is a form of precipitation and is accompanied by low temperatures. While historians have argued that the effects of the hailstorm were much less detrimental than the effect of the drought (Le Roy Ladurie, 1971), it could explain why areas with relatively high precipitation and low temperatures saw a rise in demands for institutional change. The coefficients in Table 4 indicate that an increase in growing-season temperature from the 25th percentile (relatively cool) to the 75th percentile (relatively warm) where precipitation is low (at the 25th percentile) is associated with an increase in demands by.4 demands (an increase of 32 percent from the sample mean of 1.28 demands significant at 10 percent). The marginal effect of a decrease in precipitation from the 75th percentile (relatively wet) to the 25th percentile (relatively dry) where temperature is high (at the 75th percentile) does not significantly affect the number of demands. Consistent with estimates in Sect.2.3, this indicates that the effect of drought on the number of demands is driven by the effect of temperature on revolt. I then examine the relationship between drought and the frequency of demands for change in different types of institutions. Results in Table 6 show that areas with more severe drought conditions had significantly higher demand for change in the political rights of the Third Estate, in particular for better political representation and access of commoners to all professions (column 1), for a free press (column 2) and for the introduction of elections at the municipal level (column 3). Demands for abolishing privileges of the nobility are also significantly more frequent in areas more affected by the drought. There is no significant effect of weather conditions on the king’s right to issue Lettres de Cachet. These results are consistent with historical accounts that, in the spring of 1789, the common population blamed primarily the nobility, the excessive amounts of taxes that they owed to them, and their lack of political participatory rights, and not so much the king, for their dire economic situation. 21 I estimate results at the electoral-district level as part of the lists were produced at that level in Sect.5.4. Effect sizes are very similar to the one in my main specification. Significance level are somewhat lower, which is not surprising given the much lower number of observations. 63 Journal of Economic Growth (2024) 29:41–77 1 3 Table 4 Drought and demands for institutional change from lists of grievances The table presents OLS estimates. Observations are at the canton level covering mainland France. Growing-season temperature is the growing-season temperature deviation in 1788 from the long-term mean. It measures average growing-season temperature in 1788 (March to August) as a percentage of long-run average growing-season temperature. Growing-season precipitation is defined accordingly. GS temperature*GS precipitation is an interaction term of the two weather variables. The dependent variable is the number of demands for institutional change included in the list of grievances in canton c. Control variables are population density, canton area, distance to closest road, whether a canton is subject to the great salt tax, literacy in 1690, whether a canton is part of the Paris region Ile-de-France, and region fixed effects. See Sect.3.4 for detailed information on control variables. Robust standard errors are clustered at the grid level of the underlying temperature data. Significance levels are indicated as *** for p < 0.01 , ** for p < 0.05 , * for p < 0.1 Total number of demands for institutional change (1) (2) (3) (4) (5) (6) (7) (8) Growing-season 252.0** 214.8* 308.8** 332.4*** 315.0** 315.1** 331.1** 327.0** Temperature (107.1) (113.5) (122.4) (124.1) (128.5) (128.9) (127.7) (132.2) Growing-season 314.6*** 273.1** 372.5*** 397.0*** 372.1** 372.2** 383.1*** 379.0** Precipitation (119.9) (126.5) (135.9) (137.6) (143.3) (144.0) (142.0) (145.8) GS temperature* −297.0** −258.4** −352.4*** −375.5*** −351.7** −351.8** −364.7*** −360.8** GS precipitation (114.5) (121.1) (129.8) (131.5) (136.9) (137.5) (135.6) (139.5) Control variables Region FE Yes Yes Yes Yes Yes Yes Yes Population density Yes Yes Yes Yes Yes Yes Canton area Yes Yes Yes Yes Yes Routes of commun. Yes Yes Yes Yes Literacy in 1690 Yes Yes Yes Paris Yes Yes Great salt tax Yes Marginal effect of increase in temperature from p25 to p75 40.56* Marginal effect of decrease in precipitation from p75 to p25 0.09 Observations 608 608 608 608 608 608 608 608 R-squared 0.040 0.051 0.058 0.061 0.065 0.065 0.095 0.095 64 Journal of Economic Growth (2024) 29:41–77 1 3 The question arises to what extent cantons with lists of grievances exist are comparable to the others. Summary statistics in Table A.5 in the online Appendix show that cantons with lists of grievances are on average more densely populated, they are located closer to routes of communication and a higher share of them is located in the French region of high salt taxes. These factors indicate that they might be on average economically stronger, more densely populated areas. How representative are these results for cantons without a list of grievance? Is it likely, for example, that the effects I estimate are smaller in less densely populated areas because the population there are less well informed of French politics? To explore this concern, I test whether the estimated effect differs when excluding what are likely the most urbanized cantons in the sample, cantons in the Paris and Aix-en-Provence region. In Table A.2 in the online Appendix, I show that point estimates and significance levels remain very similar and are somewhat higher when excluding all lists of grievances from these two regions. These results provide evidence that the drought, possibly in combination with the hail storm, increased demand among French commoners for more participatory political institutions. This effect may have operated through an information channel: Those more affected by the drought had better information on the inadequacy of existing institutions to protect citizens from such calamities as the drought and its consequences.22 I further explore whether the drought affected demands at the intensive or extensive margin. In Table5, I investigate whether the estimated effect stems from locations without 1 1.05 1.1 1.15 Temperature .7 .8 .9 1 1.1 Precipitation 0 1 2 3 4 5 6 No. Demands No. Demands for Institutional Change Fig. 5 Marginal effects of temperature and precipitation on demands for institutional change. Notes: This contour plot visualises the estimated marginal effects of growing-season temperature deviation and growing-season precipitation deviation on the number of demands for institutional change as expressed in the lists of grievances. Yellow to red shades indicate higher effect sizes, green shades indicate lower effect sizes 22 Those more affected by the drought may in addition have had lower opportunity costs of contributing to the lists of grievances. The opportunity cost channel, however, is likely to be less relevant here because including demands for institutional reform in the lists of grievances was of low cost. It was not punishable by law nor are there any historical accounts of commoners being punished for including such demands in the lists of grievances. 71 Journal of Economic Growth (2024) 29:41–77 1 3 Table 10 Placebo test for an effect of weather deviations in later years The table presents placebo estimates. Estimates in column one show the main results to facilitate comparison. Columns 2 to 6 show results for the estimated relationship between weather in 1790, 1810, 1830, 1850, 1870, and 1890, and outcomes in 1789. Growing-season temperature is the growing-season temperature deviation in 1788 from the long-term mean. It measures average growing-season temperature in 1788 (March to August) as a percentage of long-run average growing-season temperature. Growing-season precipitation is defined accordingly. GS temperature*GS precipitation is an interaction term of the two weather variables. Demands for Institutional Change is the number of demands for institutional change included in the list of grievances in canton c. Control variables are population density, canton area, distance to closest road, whether a canton is subject to the great salt tax, literacy in 1690, whether a canton is part of the Paris region Ile-de-France, and region fixed effects. See Sect.3.4 for detailed information on control variables. Robust standard errors are clustered at the grid level of the underlying temperature data. Significance levels are indicated as *** for p < 0.01 , ** for p < 0.05 , * for p < 0.1 1788 1790 1810 1830 1850 1870 1890 Demands for institutional change (1) (2) (3) (4) (5) (6) (7) Growing-season 327.05** 760.11** 85.71 −71.39* −11.11 11.87 203.50 Temperature (132.19) (303.78) (56.38) (43.20) (82.06) (56.45) (180.67) Growing-season 379.03** 772.81*** 83.70 −59.24 −72.49 35.56 132.58 Precipitation (145.80) (297.11) (58.68) (45.97) (75.28) (81.29) (127.07) GS temperature* −360.76** −757.18** −83.96 56.41 80.43 −36.44 −141.17 GS precipitation (139.52) (292.97) (58.83) (43.98) (80.19) (79.50) (137.55) Observations 608 608 608 608 608 608 608 R-squared 0.09 0.09 0.08 0.09 0.10 0.09 0.10 6 Conclusion The paper shows evidence that the drought of 1788 had political impacts during the French Revolution. Those more affected by the drought more often participated in peasant revolts against the feudal system. The drought was especially relevant to the outbreak of relatively early revolts that—historians have suggested—helped spark later revolts. I investigate a possible mechanism through which drought may have affected the outbreak of revolts: the drought’s effect on popular demand for institutional change. I find that demand for institutional change, a key prerequisite for regime change, was higher in areas more affected by the drought. The results provide evidence on specific ways in which the drought impacted the French Revolution, a milestone in the democratization of Western Europe. They also contribute to our understanding of the political impacts of weather shocks. It is also important to note, that the estimated effect of weather shocks on uprisings has to be interpreted as an effect conditional on underlying circumstances. Drought was a regularly occurring phenomenon (see Fig.6). The same weather conditions may have had different effects if the underlying political, economic and social conditions in France had been different.26 26 England, for example, was also affected by adverse weather conditions in the 1780s. However, as England’s economy depended less on agriculture and already had a more developed manufacturing sector the general population was less affected by these weather conditions. 72 Journal of Economic Growth (2024) 29:41–77 1 3 Table 11 Alternative weather variables The table presents OLS estimates. Observations are at the canton level covering mainland France. Growingseason temperature is the growing-season temperature deviation in 1788 from different long-term means: 1750 to 1800 (as in the main specification), 1600 to 1700m, and 1700 to 1800. Peasant Revolt is an indicator variable that is 1 if a canton experienced peasant revolt in the summer of 1789. Demands for Institutional Change is the number of demands for institutional change included in the list of grievances in canton c. Control variables are population density, canton area, distance to closest road, whether a canton is subject to the great salt tax, literacy in 1690, whether a canton is part of the Paris region Ile-de-France, and region fixed effects. See Sect.3.4 for detailed information on control variables. Robust standard errors are clustered at the grid level of the underlying temperature data. Significance levels are indicated as *** for p < 0.01 , ** for p < 0.05 , * for p < 0.1 1750 to 1800 1600 to 1800 1700 to 1800 1750 to 1800 1600 to 1800 1700 to 1800 Main specification Peasant revolt Main specification Demands for institutional change (1) (2) (3) (4) (5) (6) Growingseason 24.41*** 17.72*** 14.42*** 327.05** 273.11** 353.76*** Temperature (5.62) (5.64) (4.32) (132.19) (116.59) (114.06) Growingseason 26.54*** 19.92*** 15.85*** 379.03** 315.31** 401.71*** Precipitation (6.09) (6.35) (4.79) (145.80) (128.35) (123.78) GS temperature* −25.20*** −18.25*** −14.69*** −360.76** −292.83** −377.36*** GS precipitation (5.83) (5.91) (4.51) (139.52) (120.26) (117.16) Observations 3666 3666 3666 608 608 608 R-squared 0.02 0.02 0.02 0.09 0.09 0.10 73 Journal of Economic Growth (2024) 29:41–77 1 3 Table 12 Drought and demands for institutional change at the electoral-district level The table presents OLS estimates. Observations are at the electoral district (bailliage) level. Growing-season temperature is the growing-season temperature deviation in 1788 from the long-termmean. It measures average growing-season temperature in 1788 (March to August) as a percentage of long-run average growing-season temperature. Growing-season precipitation is defined accordingly. GS temperature*GS precipitation is an interaction term of the two weather variables. The dependent variable is the number of demands for institutional change included in the list of grievances in canton c. Control variables are population density, distance to closest road, whether an electoral district is subject to the great salt tax, literacy in 1690, whether an electoral district is part of the Paris region Ile-de-France, and region fixed effects. See Sect.3.4 for detailed information on control variables. Robust standard errors are clustered at the grid level of the underlying temperature data. Significance levels are indicated as *** for p < 0.01 , ** for p < 0.05 , * for p < 0.1 Total number of demands for institutional change (1) (2) (3) (4) (5) (6) (7) Growing-season 274.8** 302.7** 375.3** 378.3** 369.1** 373.0** 290.1* Temperature (108.1) (124.3) (154.4) (156.2) (159.3) (159.4) (169.0) Growing-season 307.3** 345.7** 424.7** 428.7** 417.3** 422.0** 338.8* Precipitation (121.4) (141.2) (173.2) (176.2) (180.1) (180.1) (188.0) GS temperature* −294.5** −332.1** −406.8** −410.5** −400.0** −404.3** −323.3* GS precipitation (116.3) (135.5) (165.6) (168.3) (171.7) (171.7) (179.7) Control variables Regions FE Yes Yes Yes Yes Yes Yes Population density Yes Yes Yes Yes Yes Routes of commun Yes Yes Yes Yes Literacy in 1690 Yes Yes Yes Paris Yes Yes Great salt tax Yes Observations 282 282 282 282 282 282 282 R-squared 0.015 0.043 0.045 0.046 0.046 0.055 0.063 74 Journal of Economic Growth (2024) 29:41–77 1 3 Supplementary Information The online version contains supplementary material available at https:// doi. org/ 10. 1007/ s1088702309230-y. Acknowledgements I thank participants of the Berkeley Climate Lunch,at the ifo Christmas Conference 2019, the Conference of the Economic History Society 2021, at the University of Bayreuth, and Davide Cantoni, Raphael Franck,Helmut Rainer and Fabian Waldingerfor very valuable comments and suggestions. I thank Dominik Ammon, Max Müller, Annalisa Tassi, Franca Schirmer, Patricio Urtaza Suarez, and Franziska Wintersteller for excellent research assistance. Funding Open Access funding enabled and organized by Projekt DEAL. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. References Acemoglu, D., De Feo, G., & De Luca, G. D. (2020). Weak states: Causes and consequences of the Sicilian Mafia. The Review of Economic Studies, 87(2), 537–581. Acemoglu, D., & Robinson, J. (2001). A theory of political transition. American Economic Review, 91, 938–963. Acemoglu, D., & Robinson, J. (2006). Economic origins of dictatorship and democracy. Cambridge University Press. Ager, P., Bursztyn, L., Leucht, L., & Voth, H.-J. (2022). Killer incentives: Rivalry, performance and risktaking among German fighter pilots, 1939–45. The Review of Economic Studies, 89(5), 2257–2292. .6 .8 1 1.2 1.4 1750 1760 1770 1780 1790 1800 year Drought GS Precipitation deviation GS Temperature deviation Fig. 6 Temperature and precipitation deviations over the long-run. Notes: The plot visualizes temperature and precipitation deviations over the period 1750 to 1800 in France 75 Journal of Economic Growth (2024) 29:41–77 1 3 Bai, Y., & Kung, J. (2011). Climate shocks and sino-nomadic conflict. The Review of Economics and Statistics, 93(3), 970–981. Belloc, M., Drago, F., & Galbiati, R. (2016). Earthquakes, religion, and transition to self-government in Italian cities. The Quarterly Journal of Economics, 131(4), 1875–1926. Besley, T., & Persson, T. (2019). Democratic values and institutions. American Economic Review: Insights, 1(1), 59–76. Bonin, S. & Langlois, C. (1987). Atlas de la Révolution française, Nr. 1, Routes et communications. Paris: École des Hautes Études en Sciences Sociales. Bossenga, G. (2011). Financial origins of the French Revolution (pp. 37–66). Stanford University Press. Brückner, M., & Ciccone, A. (2011). Rain and the democratic window of opportunity. Econometrica, 79(3), 923–947. Burke, P. J., & Leigh, A. (2010). Do output contractions trigger democratic change? American Economic Journal: Macroeconomics, 2(4), 124–157. Burke, M., Miguel, E., Satyanath, S., Dykema, J., & Lobell, B. (2009). Warming increases the risk of civil war in Africa. Proceedings of the National Academy of Sciences of the United States of America, 106(49), 20670–20674. Campbell, P. R. (2006). The origins of the French Revolution. Palgrave Macmillan. Cantoni, D., Chen, Y., Yang, D. Y., Yuchtman, N., & Zhang, Y. J. (2017). Curriculum and ideology. Journal of Political Economy, 125(2), 338–392. Caprettini, B., & Voth, H.-J. (2022). New deal, new patriots: How 1930s government spending boosted patriotism during world war II. The Quarterly Journal of Economics, 138, 465–513. Chambru, C. (2019). Do the right thing! leaders, weather shocks and social conflicts in pre-industrial France. EHES Working Paper. Chaney, E. (2013). Revolt on the Nile: Economic shocks, religion, and political power. Econometrica, 81(5), 2033–2053. Clout, H. D. (1977). Themes in the historical geography of France. Academic Press. Colella, F., Lalive, R., Sakalli, S. O., & Thoenig, M. (2019). Inference with arbitrary clustering. IZA Discussion paper: Technical report. Dell, M. (2012). Path dependence in development: Evidence from the Mexican revolution. Harvard University. Drelichman, M., Vidal-Robert, J., & Voth, H.-J. (2021). The long-run effects of religious persecution: Evidence from the Spanish inquisition. Proceedings of the National Academy of Sciences, 118(33), e2022881118. Ducruet, C., Juhász, R., Nagy, D. K., & Steinwender, C. (2020). All aboard: The effects of port development. National Bureau of Economic Research: Technical report. European Drought Observatory. (2012). Drought and drought observation. Fouka, V., & Voth, H.-J. (2022). Collective remembrance and private choice: German-Greek conflict and behavior in times of crisis. American Political Science Review, 8, 1–20. Franck, R. & Michalopoulos, S. (2017). Emigration during the French Revolution: Consequences in the short and longue Durée. Working Paper 23936. National Bureau of Economic Research. Franck, R. (2016). The political consequences of income shocks: Explaining the consolidation of democracy in France. The Review of Economics and Statistics, 98(1), 57–82. Fuchs-Schündeln, N., & Schündeln, M. (2015). On the endogeneity of political preferences: Evidence From individual experience with democracy. Science, 347(6226), 1145–1148. Furet, F. & Ozouf, J. (1977). Lire et écrire: l’alphabétisation des français de Calvin à Jules Ferry. Editions de Minuit. Galor, O., & Özak, Ö. (2016). The agricultural origins of time preference. American Economic Review, 106(10), 3064–3103. Garnier, M. (1974). Longues séries de mesures de précipitations en France, no 53. In Mémorial de la météorologie nationale (pp. 1–109). Ministère des Travaux Publics des Transports et du Tourisme. Giuliano, P. & Nunn, N. (2013). The transmission of democracy: From the village to the nation-state. Technical Report 3. National Bureau of Economic Research. Hoffman, P. T., & Rosenthal, J.-L. (2000). New work in French economic history. French Historical Studies, 23(3), 439–453. Hsiang, S., & Burke, M. (2014). Climate, conflict, and social stability: What does the evidence say? Climatic Change, 123(1), 39–55. Hyslop, B. F. (1936). A guide to the general cahiers of 1789 with the texts of unedited cahiers. Octagon Books. Hyslop, B. F. (1968). French nationalism in 1789. According to the general cahiers. Octagon Books. 76 Journal of Economic Growth (2024) 29:41–77 1 3 Johnson, N. D. (2015). Taxes, national identity, and nation building: Evidence from France. GMU Working Paper in Economics. Kaiser, T. E. & Van Kley, D. K. (2010). From deficit to deluge: The origins of the French Revolution. Stanford University Press. Labrousse, C.-E. (1939). Les Prix: Prix et Structure Régionale Le froment dans les régions françaises 1782–1790. In Annales d’histoire sociale (Vol. 1, pp. 382–400). Cambridge: Cambridge University Press. Labrousse, C.-E. (1944). La crise de l’économie française à la fin de l’Ancien régime et au début de la Révolution. Number 1 in La crise de l’économie française à la fin de l’ancien régime et au début de la Révolution. Presses Universitaires de France. Le Roy Ladurie, E. (1971). Times of feast, times of famine: A History of climate since the year 1000. Doubleday. Lefebvre, G. (1973). The great fear of 1789: Rural panic in revolutionary France. Princeton University Press. Leng, G., Zhang, X., Huang, M., Asrar, G. R., & Leung, L. R. (2016). The role of climate covariability on crop yields in the conterminous United States. Scientific Reports, 6(1), 1–11. Lipset, S. M. (1959). Some social requisites of democracy: Economic development and political legitimacy. The American Political Science Review, 53(1), 69–105. Luterbacher, J., Dietrich, D., Xoplaki, E., Grosjean, M., & Wanner, H. (2004). European seasonal and annual temperature variability, trends, and extremes since 1500. Science, 303(5663), 1499–1503. Matiu, M., Ankerst, D. P., & Menzel, A. (2017). Interactions between temperature and drought in global and regional crop yield variability during 1961–2014. PLOS ONE, 12(5), 1–23. Mavidal, J. & Colombey, E. (1870). Archives parlementaires de 1787 à 1860: Recueil Complet des Débats Législatifs & Politiques des Chambres Françaises (vol. 18). Librairie administrative de Paul Dupont. Mavidal, J., & Colombey, E. (1875). Archives parlementaires de 1787 à 1860: Tome VII—Table des cahiers des Etats généraux. Librairie administrative de Paul Dupont. McGuirk, E. F., & Nunn, N. (2020). Transhumant pastoralism, climate change, and conflict in Africa. Technical report. National Bureau of Economic Research. Merriman, J. (2010). A history of modern Europe: From the renaissance to the present. W. W. Norton & Company. Miguel, E., Satyanath, S., & Sergenti, E. (2004). Economic shocks and civil conflict: An instrumental variables approach. Journal of Political Economy, 112, 725–753. Neely, S. (2008). A concise history of the French Revolution. Rowman & Littlefield Publishers. Neumann, J. (1977). Great historical events that were significantly affected by the weather: 2, the year leading to the revolution of 1789 in France. Bulletin of the American Meteorological Society, 58(2), 163–168. Neumann, J., & Dettwiller, J. (1990). Great historical events that were significantly affected by the weather: Part 9, the year leading to the revolution of 1789 in France (II). Bulletin of the American Meteorological Society, 71(1), 33–41. Norberg, K. (1994). The French fiscal crisis of 1788 and the financial origins of the revolution of 1789. In Fiscal Crises, Liberty, and Representative Government, 1450–1789, 253–298. Pauling, A., Luterbacher, J., Casty, C., & Wanner, H. (2006). Five hundred years of gridded high-resolution precipitation reconstructions over Europe and the connection to large-scale circulation. Climate Dynamics, 26(4), 387–405. Persson, T., & Tabellini, G. (2009). Democratic capital: The Nexus of political and economic change. American Economic Journal: Macroeconomics, 1(2), 88–126. Ray, D. K., Gerber, J. S., MacDonald, G. K., & West, P. C. (2015). Climate variation explains a third of global crop yield variability. Nature Communications, 6(1), 1–9. Sapiro, V. (2004). Not your parent’s political socialization: Introduction for a new generation. Annual Review of Political Science, 7, 1–23. Sequeira, S., Nunn, N., & Qian, N. (2020). Immigrants and the making of America. The Review of Economic Studies, 87(1), 382–419. Shapiro, G., & Markoff, J. (1998). Revolutionary demands: A content analysis of the cahiers de Doléances of 1789. Stanford University Press. Shepherd, W. R. (1926). Historical atlas. Henry Holt & Company. Spoerer, M., & Berger, H. (2001). Economic crises and the European revolutions of 1848. The Journal of Economic History, 61, 293–326. Squicciarini, M. P. & Voigtlaender, N. (2016). Knowledge elites and modernization: Evidence from revolutionary France. NBER Working Paper 22779. National Bureau of Economic Research, Inc. 77 Journal of Economic Growth (2024) 29:41–77 1 3 Squicciarini, M. P. (2020). Devotion and development: Religiosity, education, and economic progress in nineteenth-century France. American Economic Review, 110(11), 3454–3491. Tabellini, M. (2020). Gifts of the immigrants, woes of the natives: Lessons from the age of mass migration. The Review of Economic Studies, 87(1), 454–486. Welzel, C. (2007). Are levels of democracy affected by mass attitudes? Testing attainment and sustainment effects on democracy. International Political Science Review, 28(4), 397–424. White, E. N. (1995). The French Revolution and the politics of government finance, 1770–1815. The Journal of Economic History, 55(2), 227–255. Zhang, D. D., Jim, C. Y., Lin, G. C., He, Y.-Q., Wang, J. J., & Lee, H. F. (2006). Climatic change, wars and dynastic cycles in china over the last millennium. Climatic Change, 76(3–4), 459–477. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.