Nonlinearities in the institutions-growth relationship in a dynamic panel data framework: Evidence from China’s provinces

It is by now a well-established fact that good institutions are key to long-run economic performance. However, there is also a growing body of literature that argues that the nature of the interaction between institutional quality and economic growth is complex and might not always be positive and linear. Our paper contributes to this discussion by focusing on China, which has achieved remarkable growth over three decades despite relatively low institutional quality. In particular, we analyze potential nonlinearities in the institutions-growth relationship across Chinese provinces within a dynamic panel data framework using Fan et al.’s NERI marketization index and system GMM estimation techniques. Our findings show that improvements in institutions are harmful to economic growth at very low levels of institutional quality. This attenuating-growth effect persists until a critical threshold is reached; beyond this threshold, the positive effect of institutional quality on growth starts to unfold and continues rising up to a second threshold after which diminishing marginal returns of institutional quality start to manifest. Importantly, our findings are robust when using an alternative measure of economic activity, namely satellite night-time lights. Our findings have important policy implications for China and other developing countries that are seeking to enhance their economic growth prospects through institutional reform. • We analyze nonlinearities in the institutions-growth relationship in China. • At low levels of institutional quality, institutional improvements hinder growth. • Beyond a first threshold, institutional improvements begin to enhance growth. • Diminishing returns on institutional quality manifest after a second threshold. • Our findings are robust when using satellite NTL data to measure economic activity.