Does democracy die in recessions? A descriptive analysis of aggregate demand shortfalls and regime transition.

Can a central bank accidentally provoke changes to a country's political regime? Sumner (2015) proposes that macroeconomic mismanagement on the part of central bank authorities, leading to declines in nominal output, can cause voters to respond with populism at the ballot box. Murphy and Smith (2018) have found evidence for this hypothesis, with populism interpreted as movements away from liberal market institutions. This article extends the hypothesis to macroeconomic mismanagement and its relationship with democratic political institutions. It finds both that recessions foreshadow a lower probability of a dictatorship becoming a democracy, and that a democracy is more likely to become a dictatorship. The relationship with autocratisation is most visible after 15 years, while the negative relationship with democratisation is far more persistent over time. These findings, it must be stressed, are historical and descriptive, not necessarily causal. [ABSTRACT FROM AUTHOR]