Big reforms, big returns? Evidence from structural reform shocks

Following a series of disappointing outcomes in Latin America and Sub-Saharan Africa, traditional structural reform shocks, of the type advocated under the ‘Washington Consensus’, came to be widely viewed as unsuccessful. This paper revisits that conclusion by applying a novel generalised use of the non-parametric Synthetic Control Method with multiple treated units to estimate the impact of 23 policy reform shocks (spanning both real and financial sector measures) implemented globally between 1961 and 2000. Our results suggest that, notwithstanding a muted short-term impact, wide-reaching reforms on average raised GDP per capita by around 6 percentage points over a decade. These findings are robust across alternative specifications, placebo and falsification tests, and different reform indicators. While outcomes were heterogeneous, the results indicate that broad liberalising reforms have more often than not delivered medium-term growth improvements, underscoring the importance of understanding the conditions under which they succeed. • This paper analyses whether broad reform packages boost economic growth • A generalised synthetic control method is used across 23 global reform episodes • Reforms show muted short-term effects but positive medium-term GDP gains • A decade after reform, GDP per capita is on average 6 p.p. above the counterfactual • Results reveal strong heterogeneity, highlighting the role of reform design