Association for Public Policy Analysis & Management

Policy experimentation has often been marked as one of the defining characteristics of China’s policy-making and cited as a factor accounting for China’s startling development. Though not something unique to China, government-led policy experimentation of such a large scale as observed in China, especially at the local level, is definitely something phenomenal. What ignites policy experimentation in an authoritarian regime for over 30 years? Why are the local leaders, -all appointed from above-, so passionate about policy experimentation, which often means to challenge the current norm and to be risky? In this paper, we try to answer the question by exploring the political incentive for provincial leaders regarding policy experimentation. Specifically, we aim to test whether cadre promotion system in China has systematically favored those provincial leaders who bring about more policy experimentation thus encouraged the latter to present even more. In particular, provincial leaders might take three different roles in policy experimentation, i.e. implementers of top-down experiments; supervisors of bottom-up experiments; initiators of province-led experiments. Our hypotheses are that 1) Provincial leaders standing up in policy experimentation might have a better chance to be promoted; 2) while policy experimentation would likely be rewarded by promotion, provincial leaders’ role in policy experimentation matters when the reward is delivered.

For empirical test, we construct a first-ever large-N (n=5253) dataset of China’s policy experimentation by coding all reports on policy experimentation in Peoples’ Daily (renmin ribao, the most authoritative newspaper in China) from 1980-2003. Descriptive analysis shows the regional distribution, initiators and contents of policy experiments observed by the central authority.  A panel dataset is constructed consisting of the numbers, types and contents of policy experiments carried out in each province from 1980 to 2003. We also collect career data of all provincial party secretaries and governors in the same period, including their promotion/demotion records and other biographic attributes. Ordinal Logistic Model is used to test the aforementioned hypotheses with other provincial socio-economic variables controlled.

Findings of this research will help understand better the relation between the micro-level incentives of political leaders on one hand, and the scales, contents and processes of policy innovation as observed on the other hand. By teasing out the existence of top-down incentive as a distinct attribute of China’s policy process, it will also shed light on the current discussion on the comparison of policy process between China and its counterparts in a more democratic setting.