Natural Experiments and Causality in Economic History: On Their Relations to Theory and Temporality

Abstract

A recent and influential research methodology, mainly endorsed by economists, proposes to renew historical analysis based on the notions of natural experiment and causality. It has the dual ambition of unifying various disciplines around a common understanding of causality in order to tackle major historical questions (such as the role of colonization, political regimes, or religion in economic development) and of making the analysis of history more scientific. The definition of causality it promotes—of the “interventionist” type—tends to liken historical events to laboratory experiments. This is articulated with a neo-institutionalist perspective aimed at measuring the long-term effects of past institutional changes, which are considered exogenous. In the first part of this article, we present the ambitions, contributions, methods, and hypotheses (implicit and explicit) of this approach, showing how it differs from more traditional quantitative economic history and placing it in the context of the recent empirical and neo-institutionalist “turns” of the economic discipline. In a second stage, we consider the criticism—often scathing—voiced by historians or economists against this method and its objectives. Finally, we emphasize the many difficulties posed by this approach when it comes to taking into account the historicity of phenomena, to producing general statements based on particular cases, and to providing a complete and coherent definition of causality in history.