Empirical Evidence for the Inefficiency of ModeratePolitical Coercion

Does the efficacy of political coercion in building state legitimacy vary by intensity?Using Freedom House, V-Dem, and Polity-5 indices as proxies for coercion, combinedwith data on legitimacy and regime survival, we identify a mid-level coercion range withnegative efficacy. Significantly fewer regimes operate in this range, and those that do ex-hibit lower legitimacy and survival rates. Time-series analysis reveals authoritarian regimesoften reduce coercion into this range, only to increase it abruptly. This suggests applyingcoercion is costly and regimes’ delayed feedback on its impact on legitimacy results in adynamic overshoot. We present historical evidence supporting the view that the data aredriven by an intrinsic negative response to coercion, namely coercion resentment. We ex-amine policy implications for how democracies can reduce the risk of slipping into softauthoritarianism. JEL Classification: D74, P16, P47