Are IMF rescue packages effective? A synthetic control analysis of macroeconomic crises

Abstract

Whether, and to what degree, IMF lending succeeds in stabilizing economies remains an open question. Here, a synthetic control analysis of macroeconomic crises with IMF intervention is performed—leveraging the existence of similar crises without intervention—that finds positive recovery effects. In the first five years following a crisis, output differences are, on average, nearly two percent of GDP per year. Consistent with a liquidity channel, effects are hump-shaped and fade in the medium run. An analysis of historical IMF forecasts provides evidence against selection as a spurious driver of this result, suggesting that these positive estimates are indeed causal.