1 Maurice Lévy-Leboyer, “La ‘New Economic History,’” Annales ESC 24, no. 5 (1969): 1035 – 69; Jean Heffer, “Une histoire scientifique. La nouvelle histoire économique,” Annales ESC 32, no. 4 (1977): 824 – 42; Claudia Goldin, “Cliometrics and the Nobel,” Journal of Economic Perspectives 9, no. 2 (1995): 191 – 208. According to several recent, more or less critical essays, the natural experiments movement in history represents either the continuation or the potential successor of new economic history. Those texts include Peter Temin, “The Rise and Fall of Economic History at MIT,” History of Political Economy 46, no. 1 (2014): 337 – 50; Temin, “Economic History and Economic Development: New Economic History in Retrospect and Prospect,” in Handbook of Cliometrics, ed. Claude Diebolt and Michael Haupert (Berlin: Springer, 2016), 33 – 51; Francesco Boldizzoni, The Poverty of Clio: Resurrecting Economic History (Princeton: Princeton University Press, 2011). There has also been lively debate on the relevance of dubbing the many applications of natural experiments to African history “the new economic history of Africa.” For an introduction to this issue, see Morten Jerven, “A Clash of Disciplines? Economists and Historians Approaching the African Past,” Economic History of Developing Regions 26, no. 2 (2011): 111 – 24; Denis Cogneau, “The Economic History of Africa: Renaissance or False Dawn?” Annales HSS (English Edition) 71, no. 4 (2016): 539 – 56.

2 Ran Abramitzky, “Economics and the Modern Economic Historian,” Journal of Economic History 75, no. 4 (2015): 1240 – 51, here pp. 1245 – 47.

3 Jared Diamond and James A. Robinson, eds., Natural Experiments of History (Cambridge: Belknap Press of Harvard University Press, 2010).

4 Randolph Roth, “Scientific History and Experimental History,” Journal of Interdisciplinary History 43, no. 3 (2013): 443 – 58.

5 Joel Mokyr, review of Diamond and Robinson, Natural Experiments of History, American Historical Review 116, no. 3 (2011): 752 – 55. Mokyr concludes by stating that if the approach reined in its ambitions, it could in fact lead to better interdisciplinary use of comparative history, as advocated by Marc Bloch.

6 Abramitzky, “Economics.”

7 This obviously does not mean that any economics study based on this type of causality necessarily fits into the neo-institutionalist paradigm. Here we detail the specific problems involved in applying to history natural experiment methods that draw on an underlying theory of institutions, and explain how these problems differ from the ones raised in criticisms of this type of causal approach applied in other fields.

8 Jo Guldi and David Armitage, The History Manifesto (Cambridge: Cambridge University Press, 2014), 110; and the dossier entitled “Debating the Longue Durée,” Annales HSS (English Edition) 70, no. 2 (2015): 215 – 303.

9 Michel De Vroey and Luca Pensieroso, “The Rise of a Mainstream in Economics” (IRES discussion paper no. 26, Institute for Economic and Social Research, Université catholique de Louvain, 2016), 2 – 27; Matthew T. Panhans and John D. Singleton, “The Empirical Economist’s Toolkit: From Models to Methods” (working paper no. 3, Center for the History of Political Economy, Duke University, Durham, 2015), 1 – 27.

10 Joshua D. Angrist and Alan B. Krueger, “Instrumental Variables and the Search for Identification: From Supply and Demand to Natural Experiments,” Journal of Economic Perspectives 15, no. 4 (2001): 69 – 85; Joshua D. Angrist and Jörn-Steffen Pischke, “The Credibility Revolution in Empirical Economics: How Better Research Design Is Taking the Con out of Econometrics,” Journal of Economic Perspectives 24, no. 2 (2010): 3 – 30. For a discussion of these methods as applied in political science, see Allison J. Sovey and Donald P. Green, “Instrumental Variables Estimation in Political Science: A Readers’ Guide,” American Journal of Political Science 55, no. 1 (2011): 188 – 200; Jasjeet S. Sekhon and Rocío Titiunik, “When Natural Experiments are Neither Natural nor Experiments,” American Political Science Review 106, no. 1 (2012): 35 – 57.

11 Angrist and Pischke, “The Credibility Revolution,” 5.

12 Heffer, “Une histoire scientifique,” 824 – 25.

13 Abramitzky, “Economics.”

14 In Angrist and Pischke’s manifesto “The Credibility Revolution in Empirical Economics,” the possible applications of this method to macroeconomics are largely drawn from studies using natural experiments in economic history.

15 This connection is studied by Boldizzoni in The Poverty of Clio.

16 In Why Nations Fail: The Origins of Power, Prosperity, and Poverty (New York: Crown Business, 2013), Daron Acemoglu and James A. Robinson identify two types of institutions said to be observable in different periods and regions: “extractive” and “inclusive.”

17 For a comprehensive introduction to neo-institutionalist method and how it is applied in economic history, see Avner Greif, Institutions and the Path to the Modern Economy: Lessons from Medieval Trade (Cambridge: Cambridge University Press, 2006); and Robert Boyer’s critical assessment, “Historiens et économistes face à l’émergence des institutions du marché,” Annales HSS 64, no. 3 (2009): 665 – 93. See also Guillaume Calafat, “Familles, réseaux et confiance dans l’économie de l’époque moderne. Diasporas marchandes et commerce interculturel,” Annales HSS 66, no. 2 (2011): 513 – 31. In Why Nations Fail, Acemoglu and Robinson present both their neo-institutionalist notion and the distinction between extractive and inclusive institutions, a distinction which builds on that made by Douglass North in his last writings; see especially Douglass C. North, John Joseph Wallis, and Barry R. Weingast, Violence and Social Orders: A Conceptual Framework for Interpreting Recorded Human History (Cambridge: Cambridge University Press, 2009). For a critical discussion of the filiation between North, Wallis, and Weingast and their references, see Martin Daunton, “Rationality and Institutions: Reflections on Douglass North,” Structural Change and Economic Dynamics 21, no. 2 (2010): 147 – 56.

18 Technically, exogeneity is ensured when the independent variable in an econometric model is not correlated with the residuals. In addition to the two conditions stated above, measurement errors in the independent variable can also lead to correlation with the residual.

19 An experiment is understood to be “controlled” when the researcher designs it and intervenes. An example might be comparing a group of students who benefited from an education reform with another group who were not exposed to it.

20 Angrist and Pischke, “The Credibility Revolution.”

21 Ibid. See also Mark R. Rosenzweig and Kenneth I. Wolpin, “Natural ‘Natural Experiments’ in Economics,” Journal of Economic Literature 38, no. 4 (2000): 827 – 74. For an epistemological discussion of instrumental variables, see Julian Reiss, “Causal Instrumental Variables and Interventions,” Philosophy of Science 72, no. 5 (2005): 964 – 76.

22 Angrist and Pischke, “The Credibility Revolution.”

23 It goes without saying that total control of outside parameters is seldom possible, even in so-called “randomized controlled” experiments: see Agnès Labrousse, “Learning From Randomized Controlled Experiments: The Narrative of Scientificity, Practical Complications, Historical Experience,” La vie des idées 2016, https://booksandideas.net/Learning-from-Randomized-Controlled-Experiments.html; Angus Deaton and Nancy Cartwright, “Understanding and Misunderstanding Randomized Controlled Trials” (NBER working paper no. 22595, National Bureau of Economic Research, Cambridge, 2016). Experimental economics makes extensive use of laboratory experiments with individuals to test economic theories about behavior but, paradoxically, seldom uses the notion of causality as employed by applied economics in studies on economic policy realized outside of a laboratory setting. More clarifications are needed on the sources of the notion that laboratory experiments are the ideal study situation; here we simply analyze the presuppositions underlying the definition of causality associated with the laboratory reference.

24 Diamond and Robinson, Natural Experiments, 1.

25 Ibid.

26 Adopting a temporal notion of causality (cause precedes effect) does not mean assuming that researchers can fully account for temporality itself (the succession of events and whole chains of cause and effect). See the second half of the present article.

27 For a deeper analysis of this postulate of the temporal asymmetry of causes and effects and the conceptual possibility of reverse causality, see Bourgeois-Gironde, Temps et causalité (Paris: Presses universitaires de France, 2002).

28 The paradigmatic example is price determination in the neoclassical model.

29 For a particularly well-formulated version of this definition, see James Woodward, Making Things Happen: A Theory of Causal Explanation (Oxford: Oxford University Press, 2005). For a discussion of instrumental variables in that framework, see Reiss, “Causal Instrumental Variables.”

30 Judea Pearl, “Causal Inference,” in “Causality: Objectives and Assessment,” ed. Isabelle Guyon, Dominik Janzing, and Bernhard Schölkopf, Proceedings of Machine Learning Research, vol. 6 (2010): 39 – 58.

31 As can be seen in the articles discussed here, specifically in those by Angrist and Pischke (“design-based studies”) and Acemoglu etal. (“designed institutional change”). The more general notion of “research design” is widely used to describe the methods of applied economics, while “institutional design” is recurrent in the work of neo-institutionalists.

32 Diamond and Robinson, Natural Experiments, 271 – 74.

33 Nancy Cartwright, “Are RCTs the Gold Standard?” BioSocieties 2, no. 1 (2007): 11 – 20; Deaton and Cartwright, “Understanding and Misunderstanding.”

34 François Simiand, “Méthode historique et science sociale. Étude critique d’après les ouvrages récents de M. Lacombe et de M. Seignobos,” Revue de synthèse historique 16 (1903): 1 – 22; Simiand, “La causalité en histoire,” Bulletin de la Société française de philosophie 6 (1906): 245 – 90. On the context of Simiand’s position on causality and the debates it sparked, see Jacques Revel, “Histoire et sciences sociales. Lectures d’un débat français autour de 1900,” Mil neuf cent. Revue d’histoire intellectuelle 25, no. 1 (2007): 101 – 26.

35 Diamond and Robinson, Natural Experiments, 271 – 74. From the outset, the editors explain that they think of “natural experiments” as similar to the “comparative method,” but define them only in reference to laboratory experiments, not to other comparative strategies in the discipline of history.

36 Ibid. So little is offered here on the status of generalization, on relations to theory, and therefore on the question of laws that we do not know, for instance, if the authors would go so far as to conclude that “laws of history” exist and that causal analysis is intended to reveal them.

37 Daron Acemoglu etal., “From Ancien Régime to Capitalism: The French Revolution as a Natural Experiment,” in Diamond and Robinson, Natural Experiments, 221 – 56; Daron Acemoglu etal., “The Consequences of Radical Reform: The French Revolution,” American Economic Review 101, no. 7 (2011): 3286 – 307.

38 Acemoglu etal., “From Ancien Régime to Capitalism,” 249 – 50.

39 Acemoglu etal., “The Consequences of Radical Reform,” 3303.

40 Daron Acemoglu, “Oligarchic versus Democratic Societies,” Journal of the European Economic Association 6, no. 1 (2008): 1 – 44.

41 On applying similar methods and reasoning in other contexts, see Daron Acemoglu, Tarek A. Hassan, and James A. Robinson, “Social Structure and Development: A Legacy of the Holocaust in Russia,” Quarterly Journal of Economics 126, no. 2 (2010): 895 – 946; Sara Lowes etal., “The Evolution of Culture and Institutions: Evidence from the Kuba Kingdom,” Econometrica 85, no. 4 (2017): 1065 – 91. In both these cases, it is again institutionalist theory that links actions, rules, and human behavior and allows for interpreting as causal a correlation between two observations separated by several centuries.

42 Réka Juhász, “Temporary Protection and Technology Adoption: Evidence from the Napoleonic Blockade” (Centre for Economic Performance discussion paper no. 1322, London School of Economics, 2014); Peter Koudijs, “The Boats That Did Not Sail: Asset Price Volatility in a Natural Experiment,” Journal of Finance 71, no. 3 (2016): 1185 – 226. The criticisms presented in the second half of the present article also apply to this type of study.

43 See Daniel M. Hausman, ed., The Philosophy of Economics: An Anthology (Cambridge: Cambridge University Press, 1984; repr. 1994); Julian Reiss, Error in Economics: Towards a More Evidence-Based Methodology (London: Routledge, 2008; repr. 2016).

44 Two renowned articles come to mind here: Rajnish Mehra and Edward C. Prescott, “The Equity Premium: A Puzzle,” Journal of Monetary Economics 15, no. 2 (1985): 145 – 61; Karl E. Case and Robert J. Shiller, “The Efficiency of the Market for Single-Family Homes,” American Economic Review 79, no. 1 (1989): 125 – 37.

45 In addition to the aforementioned texts by Diamond and Robinson, and Acemoglu and Robinson’s Why Nations Fail, the following articles are particularly clear on this point: Nathan Nunn, “The Importance of History for Economic Development,” Annual Review of Economics 1, no. 1 (2009): 65 – 92; James Fenske, “The Causal History of Africa: A Response to Hopkins,” Economic History of Developing Regions 25, no. 2 (2010): 177 – 212.

46 Daron Acemoglu, Simon Johnson, and James A. Robinson, “The Colonial Origins of Comparative Development: An Empirical Investigation,” American Economic Review 91, no. 5 (2001): 1369 – 401; Acemoglu, Johnson, and Robinson, “Reversal of Fortune: Geography and Institutions in the Making of the Modern World Income Distribution,” Quarterly Journal of Economics 117, no. 4 (2002): 1231 – 94.

47 Gareth Austin, “The ‘Reversal of Fortune’ Thesis and the Compression of History: Perspectives from African and Comparative Economic History,” Journal of International Development 20, no. 8 (2008): 996 – 1027; Antony G. Hopkins, “The New Economic History of Africa,” Journal of African History 50, no. 2 (2009): 155 – 77; Hopkins, “Causes and Confusions in African History,” Economic History of Developing Regions 26, no. 2 (2011): 107 – 10; Jerven, “A Clash of Disciplines.”

48 For a recent review of this literature, see Sascha O. Becker, Steven Pfaff, and Jared Rubin, “Causes and Consequences of the Protestant Reformation,” Explorations in Economic History 62, no. 3 (2016): 1 – 25. The authors define the research procedures in this new field as identifying causality using econometric procedures and estimating long-term effects from an institutionalist perspective. An exception is Davide Cantoni’s “test” of the relation Weber established between the Protestant ethic and development; see Cantoni, “The Economic Effects of the Protestant Reformation: Testing the Weber Hypothesis in the German Lands,” Journal of the European Economic Association 13, no. 4 (2015): 561 – 98. Cantoni finds no positive effect on economic growth in regions that converted to Protestantism and concludes that Weber’s theory has been invalidated. Although, strictly speaking, he only assesses links between the differentiated impacts of the Peace of Augsburg (presented as the source of a natural experiment) and population growth in particular German cities, the initial ambition of producing a generalization leads the author to present his work as an econometric test of a major theory. He is ultimately compelled to downsize that claim.

49 Max Weber, The Protestant Ethic and the Spirit of Capitalism [1906], trans. Talcott Parsons (London: Routledge, 1930; repr. 2005), 124: “Since asceticism undertook to remodel the world and to work out its ideals in the world, material goods have gained an increasing and finally inexorable power over the lives of men as at no previous period in history. Today the spirit of religious asceticism—whether finally, who knows?—has escaped from the cage. But victorious capitalism, since it rests on mechanical foundations, needs its support no longer.” To our knowledge, this crucial comment by Weber does not seem to have had any resonance in economic studies of the long-term link between the Protestant religion and economic growth.

50 Greif, Institutions; Robert D. Putnam, Making Democracy Work: Civic Traditions in Modern Italy (Princeton: Princeton University Press, 1993). For an introduction to how economists have appropriated these two schools of thought, see Guido Tabellini, “Presidential Address: Institutions and Culture,” Journal of the European Economic Association 6, no. 3 (2008): 255 – 94. For a critique of how economists have used Putnam’s work and the historical determinism that follows from it, see Nicolas Delalande, “Is a History of Trust Possible? Remarks on the Historic Imagination of Two Economists” [2008], La vie des idées, 2011, https://booksandideas.net/Is-a-History-of-Trust-Possible.html.

51 Our way of proceeding here resembles that of Quentin Deluermoz and Pierre Singaravélou in Pour une histoire des possibles (Paris: Éd. du Seuil, 2016), 219. These authors call for examining the political uses of counterfactuals in history and criticize the claims to scientificity of mechanical conceptions of causality that fail to apply counterfactual reasoning to their own “analytical framework.”

52 In addition to the references cited above, see Boldizzoni, The Poverty of Clio.

53 Rosenzweig and Wolpin, “Natural ‘Natural Experiments’”; Cartwright, “Are RCTs the Gold Standard?”; Deaton and Cartwright, “Understanding and Misunderstanding”; Sekhon and Titiunik, “When Natural Experiments.”

54 In public policy, too, it can be ill-advised and potentially dangerous to predicate policy implementation on the ability to assess it using statistical methods; see Stephen T. Ziliak and Deirdre N. McCloskey, The Cult of Statistical Significance: How the Standard Error Costs Us Jobs, Justice, and Lives (Ann Arbor: University of Michigan Press, 2008).

55 Naomi Lamoreaux, “The Future of Economic History Must Be Interdisciplinary,” Journal of Economic History 75, no. 4 (2015): 1251 – 57, here p. 1255. Lamoreaux refers explicitly to Nunn’s studies of African history cited in n. 64 below.

56 This criticism could already have been made of attempts by economists—including the first generation of new economic historians—to use statistical and theoretical tools to study history. However, Lamoreaux is right that, over time, some historian economists have either surrendered or adjusted their neoclassical assumptions and managed to tailor their quantitative procedures to more context-sensitive approaches; see Clément Dherbécourt and Éric Monnet, “Les angles morts de The Poverty of Clio,” Tracés 16 (2016): 137 – 50.

57 Witold Kula, “Histoire et économie. La longue durée,” Annales ESC 15, no. 2 (1960): 294 – 313; Jean-Yves Grenier and Bernard Lepetit, “L’expérience historique. À propos de C.-E. Labrousse,” Annales ESC 44, no. 6 (1989): 1337 – 60; Claire Lemercier, “A History Without the Social Sciences?” Annales HSS (English Edition) 70, no. 2 (2015): 271 – 83.

58 Nathan Nunn and Nancy Qian, “The Potato’s Contribution to Population and Urbanization: Evidence from an Historical Experiment,” Quarterly Journal of Economics 126, no. 2 (2011): 593 – 650.

59 See Hopkins’ critique of using population data as a proxy for economic growth in Africa in “The New Economic History.”

60 Rafael La Porta etal., “Law and Finance,” Journal of Political Economy 106, no. 6 (1998): 1113 – 55.

61 Claire Lemercier, “Napoléon contre la croissance ? À propos de droit, d’économie et d’histoire,” La vie des idées, 2008, http://www.laviedesidees.fr/Napoleon-contre-la-croissance.html; Jérôme Sgard, “Do Legal Origins Matter? The Case of Bankruptcy Laws in Europe, 1808 – 1914,” European Review of Economic History 10, no. 3 (2006): 389 – 419; Aldo Musacchio and John D. Turner, “Does the Law and Finance Hypothesis Pass the Test of History?” Business History 55, no. 4 (2013): 524 – 42.

62 See the thematic dossier “The Economics of Contemporary Africa,” in Annales HSS (English Edition) 71, no. 4 (2016): 503 – 79; Jerven, “A Clash of Disciplines”; Denis Cogneau and Yannick Dupraz, “Institutions historiques et développement économique en Afrique,” Histoire et mesure 30, no. 1 (2015): 103 – 34.

63 Austin, “The ‘Reversal of Fortune.’”

64 Nathan Nunn, “The Long-Term Effects of Africa’s Slave Trades,” Quarterly Journal of Economics 123, no. 1 (2008): 139 – 76. See also Nathan Nunn and Leonard Wantchekon, “The Slave Trade and the Origins of Mistrust in Africa,” American Economic Review 101, no. 7 (2011): 3221 – 52.

65 Ewout Frankema and Marlous van Waijenburg, “Structural Impediments to African Growth? New Evidence from Real Wages in British Africa, 1880 – 1965” (Centre for Global Economic History working paper no. 24, Utrecht, 2011).

66 Edward E. Leamer, “Tantalus on the Road to Asymptopia,” Journal of Economic Perspectives 24, no. 2 (2010): 31 – 46, here p. 44.

67 Christopher A. Sims, “But Economics Is Not an Experimental Science,” Journal of Economic Perspectives 24, no. 2 (2010): 59 – 68, here p. 59.

68 It is beyond the scope of this article to present all the alternative conceptions of causality that have been developed in macroeconomics, but we can cite three—while noting that none is without problems or better than the others. Granger’s probabilistic causality is based on much weaker postulates than interventionist-type causality, stating simply that A causes B when A chronologically precedes B and when variations of A allow us to (partially) predict variations of B. Another procedure often used in macroeconomics is to estimate a stylized model of the economy and simulate exogenous shocks in that model in order to discuss how those shocks may correspond to observed variations. This type of causality is both interventionist and structural in that all interactions resulting from an exogenous shock are modeled. The third procedure, closer to process theory, is to document common statistical trends, or regularities, and study them in relation to general historical developments.

69 Wesley C. Salmon, Causality and Explanation (Oxford: Oxford University Press, 1998); Phil Dowe, “Process Causality and Asymmetry,” Erkenntnis 37, no. 2 (1992): 179 – 96.

70 Phil Dowe, “On the Reduction of Process Causality to Statistical Relations,” British Journal for the Philosophy of Science 44, no. 2 (1993): 325 – 27.

71 On the analogy between models and thought experiments, see Mary S. Morgan, The World in the Model: How Economists Work and Think (Cambridge: Cambridge University Press, 2012). This analogy is often used by economists themselves; see Leamer, “Tantalus on the Road to Asymptopia,” 44.

72 This is because microeconomic models are models of partial rather than general equilibrium. See Steven D. Levitt’s test of a standard economic model of crime, which uses an exogenous change in the probability of the crime being identified: “Testing the Economic Model of Crime: The National Hockey League’s Two-Referee Experiment,” Contributions to Economic Analysis and Policy 1, no. 1 (2002): 1 – 21.

73 Hume, “Of the Balance of Trade,” Political Discourses II, here 5.11. Hume argued that the balance of payments would always achieve equilibrium due to the link between money supply, prices, and trade. After constructing a model interlinking these variables, he devised the following thought experiment: “Suppose four-fifths of all the money in GREAT BRITAIN to be annihilated in one night, and the nation reduced to the same condition, with regard to specie, as in the reigns of the HARRYS and EDWARDS, what would be the consequence?” (Political Discourses II, 5.9). Once he had used this thought experiment to isolate and describe the theoretical mechanisms at work, he clarified that what he imagined could not happen in reality: “Now, it is evident, that the same causes, which would correct these exorbitant inequalities, were they to happen miraculously, must prevent their happening in the common course of nature, and must for ever, in all neighbouring nations, preserve money nearly proportionable to the art and industry of each nation” (Political Discourses II, 5.11).

74 In “Chronicle of a Deflation Unforetold,” Journal of Political Economy 117, no. 4 (2009): 591 – 634, François R. Velde studied the effect of an arbitrary reduction of the money supply in France in 1724, an event that might seem to correspond to Hume’s thought experiment. But he refused to identify this as a natural experiment because, in eighteenth-century France, it was customary for the king to make this type of decision and economic actors took that into account.

75 Judea Pearl, Causality: Models, Reasoning, and Inference (Cambridge: Cambridge University Press, 2009). Pearl’s conception of causality as it relates to econometrics is based on older works by Trygve Haavelmo. Specifically, the aim is to distinguish clearly between a condition and a cause, the latter involving fixing a parameter in the model (and therefore exogenizing it from the set of equations). By contrast, a condition is defined as a purely statistical concept. Pearl’s approach is a reminder that what statisticians and econometrists estimate is only ever a statistical relation and therefore a condition. To have causality, researchers must make an additional hypothesis; that is, they must explain why the variable can be thought of as “fixed.” For Pearl’s own critique of the confusion economists perpetuate on the statistical handling of causality, see Bryant Chen and Judea Pearl, “Regression and Causation: A Critical Examination of Six Econometrics Textbooks,” Real-World Economics Review 65 (2013): 2 – 20.

76 John Leslie Mackie, “Causes and Conditions,” American Philosophical Quarterly 2, no. 4 (1965): 245 – 64. INUS is the term used in the international literature to designate Mackie’s definition of causality: “Insufficient but Non-redundant parts of a condition which is itself Unnecessary but Sufficient.”

77 Benoît Rihoux and Charles C. Ragin, eds., Configurational Comparative Methods: Qualitative Comparative Analysis (QCA) and Related Techniques (Thousand Oaks: Sage Publications, 2008). We are much beholden to one of the anonymous peer reviewers of this article for suggesting how Ragin’s analyses could help imagine alternative solutions to the type of causality envisioned by proponents of natural experiments.

78 In his review of Diamond and Robinson, Natural Experiments of History (p. 274), Mokyr recalls that it was possible to adopt “progressive” institutions even where there was no French conquest and that Spain instated no reforms despite the French invasion.

79 In “Causation in the Social Sciences: Evidence, Inference, and Purpose,” Philosophy of the Social Sciences 39, no. 1 (2009): 20 – 40, Julian Reiss explains that the natural experiment method is based on an interventionist conception of causality.

80 See the references listed in n. 1 above.

81 This is probably the case for some studies on the effects of Protestantism.

82 Lemercier, “A History Without the Social Sciences?”