Democracy, transparency, and the production of official statistics

Abstract

We examine whether democracy causally affects the quality of official statistics by focusing on historically colonized countries. To address endogeneity, we use historical settler mortality as an instrument, which shifted the probability that colonies developed democratic institutions. This approach identifies the causal effect among complier — colonies whose democratic trajectories were influenced by settler mortality. This group is substantively important, as many post-colonial states continue to face challenges in state capacity and data quality. Using cross-sectional data from 2023, combining the Economist Intelligence Unit’s democracy index and the World Bank’s Statistical Performance Indicators, we find that one unit increase in democracy raises the quality of official statistics by 5.3 to 6.2 percentage points. We further show that the effect is heterogeneous: it is about 3.7 points larger in common-law countries and 3.6 points smaller in civil-law countries, revealing that colonial legal traditions make democracy’s impact on statistical quality stronger in some countries and weaker in others. (JEL O11, P16, H83)