Facets of Democratic Change and their Causal Effect on Economic Growth

• In Event: Democracy and Development: Insights from new Data Collection Efforts

Abstract

What are the economic consequences of changes to the system of governance? The existing empirical literature on the growth effect of democratic change covers a wide range of methods and specifications. Three important aspects, however, have not been adequately accounted for: First, an empirical specification which allows the democracy-growth relationship to differ across countries. Such heterogeneities can be motivated with arguments for democratic legacy or thresholds in economic or human development as necessary conditions for positive democracy effects. Second, the adoption of methods accounting for the distorting effect of time-varying unobserved heterogeneity, which can arise from common shocks and/or from spillover effects. And third, flexibility in the functional form of the democracy-growth nexus: existing studies either assume a linear effect relationship or adopt ad hoc single thresholds to distinguish `democratic' from `autocratic' regimes. Our specifications allow for (i) a linear, (ii) single threshold, or (iii) dual threshold relationship, for which we adopt a novel IV estimator (Norkute et al., 2019), a difference-in-difference estimator (Chan and Kwok, 2018) and a simple extension of the latter, respectively --- all implementations are heterogeneous panel estimators and account for strong cross-sectional correlation. Afforded by the internal coherence of the Varieties of Democracy (V-Dem) data we are able to provide a closer mapping between an encompassing theory of democracy and its constituent components on the one hand, and our empirical analysis covering up to 159 countries (1960-2017) on the other. Employing high-, mid-, and low-level democracy indices we can trace the positive democratic dividend we establish to the most relevant constituent components which appear to drive these results.

Author

• Vanessa Alexandra Boese