Natural experiments are suddenly everywhere. Over the last decade, the number of published social-scientific studies that claim to use this methodology has more than tripled (Dunning 2008a). More than 100 articles published in major political-science and economics journals from 2000 to 2009 contained the phrase “natural experiment” in the title or abstract—compared to only 8 in the three decades from 1960 to 1989 and 37 between 1990 and 1999 (Figure 1.1). Searches for “natural experiment” using Internet search engines now routinely turn up several million hits. As the examples surveyed in this book will suggest, an impressive volume of unpublished, forthcoming, and recently published studies—many not yet picked up by standard electronic sources—also underscores the growing prevalence of natural experiments.

This style of research has also spread across various social science disciplines. Anthropologists, geographers, and historians have used natural experiments to study topics ranging from the effects of the African slave trade to the long-run consequences of colonialism. Political scientists have explored the causes and consequences of suffrage expansion, the political effects of military conscription, and the returns to campaign donations. Economists, the most prolific users of natural experiments to date, have scrutinized the workings of labor markets, the consequences of schooling reforms, and the impact of institutions on economic development.