Unveiling the origins of non-performance-oriented behavior in China's local governments: a game theory perspective on the performance-based promotion system

China’s rapid economic development since the late 1970s has been linked to market-oriented reforms and a performance evaluation (PE) system that shifted cadre promotion incentives toward economic outcomes (e.g., GDP growth). While PE has driven performance-oriented government behaviors (POGBs), it has also produced negative consequences—income inequality, environmental degradation, government–business collusion, and extensive redundant construction (e.g., airports, smart cities, subways, and new energy industries), often misaligned with local needs. This creates a puzzle: despite heterogeneous local conditions, many jurisdictions adopt similar projects and policies that can harm overall performance. The study asks why local governments engage in non-performance-oriented behaviors (NPOGBs) that impede outcomes and waste resources. Existing accounts point to central–local relations, official–citizen relations, and rational self-interest, but comprehensive explanations are limited. This research aims to: (1) clearly define POGBs and NPOGBs; (2) build a game-theoretic model to uncover causes, characteristics, and incentives behind NPOGBs under PE; and (3) propose policy suggestions to reduce NPOGBs. Contrary to views treating negative outcomes as unintended by-products, the study argues many NPOGBs are deliberate, rational choices by officials maximizing personal promotion prospects within a performance-based promotion system.

The paper reviews three perspectives on performance in public administration: results (performance as outcomes), behavior (performance as a function of bureaucratic behaviors), and competency (performance driven by officials’ competencies). These strands collectively suggest PE can incentivize POGBs. However, a substantial dark side exists in China, notably redundant construction across regions and sectors (from 1980s consumer durables to 1990s heavy industry and recent smart cities and new energy), often ignoring local conditions and causing waste. Extant explanations of NPOGBs emphasize: (1) fiscal decentralization and tax-sharing reforms that encourage local revenue-maximizing behavior at the expense of broader welfare; (2) incentive mechanisms where promotion tied to economic metrics induces target distortion and pursuit of personal advancement; and (3) inter-jurisdictional competition (political championship and resource competition) that fosters zero-sum, short-sighted behaviors like redundant construction. The literature’s limits include insufficient micro-level explanations of local decision-making under fiscal decentralization, incomplete treatment of institutional constraints in incentive-based accounts, and unresolved questions about why officials knowingly undermine long-term welfare. This motivates a game-theoretic approach to model strategic interaction, risk aversion, and information asymmetry in promotion tournaments.

The study employs theoretical deduction in two steps. First, it constructs a conceptual model distinguishing POGBs and NPOGBs. POGBs are intentional actions that improve performance within a jurisdiction without harming others, reflecting Pareto optimality; NPOGBs are intentional actions contrary to citizens’ needs that can harm performance and welfare. A quadrant framework situates POGBs as a small positive subset within the broader domain of government behaviors, with most behaviors falling into NPOGB territory. Second, it develops a game-theoretic model of inter-jurisdictional competition under a performance-based promotion system. Assumptions: one central government and two local governments (α and β); homogeneity in conditions and officials; risk-averse officials with information constraints; promotion rewards comprise functional (F) and hierarchical (H) components; returns to government behaviors share a common rate when exogenous. Officials allocate limited fiscal resources across behaviors to maximize expected utility, where own payoff depends on absolute performance and relative performance vis-à-vis competitors. Best-response functions are derived for each official’s resource allocation, yielding identical, positive slopes (H/(F+H) > 0). The intersection of best responses (Nash equilibrium) implies that increases in spending by one jurisdiction induce the other to follow, rationalizing imitation strategies. Under information scarcity and risk aversion, officials adopt behavior-matching to protect relative standing and promotion prospects, leading to persistent NPOGB proliferation (e.g., repeated construction). The model formalizes deliberate herd behavior as an equilibrium outcome of promotion tournaments.

• POGBs are Pareto-optimal behaviors that improve a jurisdiction’s performance without harming others, but they account for only a small fraction of all governmental behaviors; the majority are NPOGBs.
• In a promotion-tournament environment, NPOGBs are rational choices for risk-averse officials facing information constraints. Imitation of peers’ fiscal allocations and project choices emerges as a dominant strategy to preserve relative gains, avoid risks, and compete for scarce promotions.
• The reaction functions of competing jurisdictions have identical positive slopes, and their Nash equilibrium supports mutual imitation; if one jurisdiction increases investment in a behavior, the other follows, fueling cycles of redundant construction and other NPOGBs.
• The zero-sum nature of promotions links short-term performance races to strategic mimicry, distorting long-term welfare and resource allocation.

The findings address the core question of why Chinese local governments engage in behaviors that undermine performance: under performance-based promotion, relative standing matters more than absolute outcomes, and risk-averse officials facing limited information rationally imitate peers. This strategic interdependence produces deliberate, repeated NPOGBs (e.g., redundant infrastructure) despite heterogeneity in local needs. The results reframe many negative outcomes as man-made, predictable products of institutional incentives rather than unintended errors. This has significant implications: correcting NPOGBs requires altering competitive incentives and introducing mechanisms that reward cooperation, cross-jurisdictional consideration, and longer-term welfare, not merely tightening performance metrics. The analysis illuminates how promotion tournaments, coupled with resource control and information limits, shape governance behaviors and can propagate inefficiencies across jurisdictions.

NPOGBs are widespread in unitary systems that encourage local competition. Using a conceptual framework and a game-theoretic model, the study shows that: (1) POGBs represent Pareto-optimal yet relatively rare behaviors; most actions are NPOGBs; (2) NPOGBs are rational responses to zero-sum promotion competitions under information constraints and risk aversion, where imitation strategies proliferate. Policy implications include: promoting cooperative decentralization that incentivizes inter-jurisdictional collaboration; instituting neighborhood governance and stakeholder participation across borders to internalize spillovers; and embedding cooperation and cross-boundary input into policy design, implementation, and performance evaluation. These steps can mitigate NPOGBs and steer competition toward Pareto improvements.

This is a preliminary theoretical study without empirical testing. The POGB/NPOGB terminology and boundaries may require refinement and validation. The model abstracts from heterogeneous administrative contexts and assumes homogeneity and common return structures; future research should empirically test the concepts across different levels of government and varied contexts, and examine robustness under alternative assumptions (e.g., asymmetric information, heterogeneous capacities). No datasets were generated or analyzed.