The Political Economy of Policy Experimentations in China

Abstract: Policy learning, often involving experimentations, is an essential component of government's decision making. In this project, we aim to describe and understand China's policy experimentations, which are speculated by many to have played a vital role in fostering China's reform and growth. We collect comprehensive data on policy experimentations conducted in China over the past 4 decades by 139 ministries and commissions. We document three main results. First, a substantial share of policy experimentations deviate from representative experimentation sites selection as they over-sample more developed jurisdictions. Second, while the observed deviation from representativeness cannot be justified by a range of models of optimal experimentation design, about half of such deviation can be attributed to misaligned incentives between the central and local governments. Third, deviation from representativeness results in implemented policies that systematically favor more developed regions, because the central government does not adjust for positive selection and discard irrelevant information when evaluating experimentation outcomes. Taken together, these results suggest that policy experimentations take place under various political and bureaucratic constraints, which could limit the scope and bias the direction of policy learning.