The Economics of State Fragmentation: Assessing the Economic Impact of Secession
91 Pages Posted: 24 Feb 2016 Last revised: 18 May 2018
Date Written: May 15, 2018
This paper provides empirical evidence that declaring independence significantly lowers per capita GDP based on a large panel of countries covering the period 1950-2016. To do so, we rely on a semi-parametric identification strategy that controls for the confounding effects of past GDP dynamics, anticipation effects, unobserved heterogeneity, model uncertainty and effect heterogeneity. Our baseline results indicate that declaring independence reduces per capita GDP by around 20% in the long run. We subsequently propose a quadruple-difference procedure to demonstrate that the results are not driven by simulation and matching inaccuracies or spillover effects. A second methodological novelty consists of the development of a two-step estimator that relies on the control function approach to control for the potential endogeneity of the estimated independence payoffs and their potential determinants, to shed some light on the primary channels driving our results. We find tentative evidence that the adverse effects of independence decrease in territorial size, pointing to the presence of economies of scale, but that they are mitigated when newly independent states liberalize their trade regime or use their new-found political autonomy to democratize.
Addendum can be found at https://ssrn.com/abstract=2736848
Keywords: Independence dividend; panel data; dynamic model; synthetic control method; difference-in-difference; triple-difference; quadruple-difference; control function approach
JEL Classification: C14, C32, H77, O47
Suggested Citation: Suggested Citation