A fundamental question in governance concerns the merits of centralized versus decentralized policymaking. While top-down, centralized policymaking may streamline adoption, enhance efficiency, and improve coordination between regions, it often sacrifices the local suitability of bottom-up, decentralized policy initiatives. This tension is especially relevant when governing large nations with many regional differences. However, measuring centralization in policymaking is difficult because it requires tracing the origin and spread of all policies across layers of government hierarchy; as a result, most studies of centralization and decentralization focus on only one policy. Assessing centralization’s impact on policy outcomes is even more demanding because it involves linking policies to both local conditions and intended outcomes.

Our research examines the centralization of policymaking in China and how it has affected the local suitability of policies across all domains over the past two decades. We investigated two questions: First, what share of local governments’ policy portfolios is shaped by the central government’s direct involvement? Second, does the central government’s direct involvement undermine policy suitability and effectiveness at the local level?

We compiled a comprehensive dataset of 422,000 central government policy documents and 3.3 million local government policy documents and work reports. From this dataset, we identified 115,679 distinct policies implemented from 2004 to 2020 and traced their origins and spread (from the central government to local governments and between local governments). We also measured the suitability and effectiveness of policies aimed at promoting industrial growth.

Our study reveals four main findings. First, policymaking in China is highly decentralized. Over the past two decades, 82 percent of policies appearing in local governments’ portfolios originated as local initiatives; of these, 74 percent spread solely between local governments and never involved explicit central government action. Further analysis of the turnover of local bureaucrats reveals that these bureaucrats were the primary drivers of decentralized policy initiation and spread and that they were rewarded with political promotion for local policy innovation.

Second, since 2013, decentralized policymaking has declined dramatically: The share of top-down policies in local governments’ portfolios has increased by 40 percent, the adoption rate of top-down initiatives has nearly tripled, and local replication of central policy details has more than doubled. Among other factors, changes in local bureaucrats’ career incentives likely played an important role: Before 2013, policy innovation was rewarded with political promotion; after 2013, promotions were granted to bureaucrats who most actively implemented top-down policies rather than those who pioneered new policies. The timing of this centralization aligns with the central government’s phased rollout of tighter top-down control across different ministries and policy domains.

Third, locally initiated policies that spread between local governments tended to be associated with higher local suitability and better economic outcomes. Our research focuses on policies aimed at promoting sector-specific industrial growth and innovation and finds that those initiated or adopted by local governments without central involvement were better aligned with local conditions, as measured by preexisting regional supply chains and private firms’ investment preferences. In contrast, top-down industrial policies initiated by the central government showed weaker alignment with local conditions. Given that locally suitable industrial policies proved significantly more effective in achieving policy objectives—including increased industrial output, patenting, and exports—the transition toward centralization has entailed a real cost to economic development.

Fourth, centralized policymaking suppresses strategic competition among local bureaucrats that would otherwise impede learning from peer jurisdictions. When policies are initiated and spread locally without the central government’s explicit involvement, local bureaucrats competing for the same political promotions may be reluctant to adopt policies from one another for fear of boosting their competitors’ credentials. This effect is exacerbated by the tendency to assign bureaucrats with similar promotion prospects to economically comparable localities that might benefit from similar policies. Thus, our research finds that bureaucratic rivalry stifles the spread of local policy innovations—especially those most suitable for a given locality—and dampens economic performance. Centralization alleviates these problems, since local bureaucrats no longer worry about political competitors receiving credit for implementing top-down policies. Nevertheless, our calculations suggest that centralization’s harms to local policy suitability have exceeded its benefits.

China’s extraordinary growth since its 1979 departure from a centrally planned economy has been widely credited to locally driven initiatives within a decentralized framework. Our research builds on this view by demonstrating that, under well-structured political incentives, an autocracy can also function as a vibrant laboratory for policy innovation, generating new, locally suited ideas, much like the “laboratories of democracy” found in federalist systems. Over the past decade, however, China’s decentralization trend reversed as political incentives for policy innovation were removed. This shift demands rigorous scrutiny, given China’s history of heavy-handed central-planning failures, its vast regional differences, and the mounting complexity of governance.

Note
This research brief is based on Kaicheng Luo et al., “Laboratories of Autocracy: Landscape of Central–Local Dynamics in China’s Policy Universe,” Becker Friedman Institute for Economics Working Paper no. 2025-121, September 2025.