China's remarkable economic growth, often termed the "China miracle," has been attributed to market-oriented economic policies and the government's governance approach. Since the 1980s, influenced by the New Public Management movement, China adopted performance evaluation (PE) as an administrative tool. This led to a shift in criteria for promoting local government officials, emphasizing economic performance (measured by GDP growth and productivity) over political loyalty. This performance-based promotion system incentivized performance-oriented governmental behavior (POGB), contributing significantly to economic growth. However, this system also resulted in negative consequences, including income inequality, environmental damage, and redundant infrastructure construction. The paper focuses on this last issue: repeated and superfluous projects across various sectors (textiles in the 1980s, airports in the 1990s, and more recently, new energy vehicles and subway systems). This phenomenon represents non-performance-oriented governmental behavior (NPOGB), where local governments engage in actions that are inefficient and waste resources, despite aiming for economic growth. The research aims to define POGB and NPOGB clearly, develop a game theory model explaining NPOGBs in China, and propose policy recommendations to mitigate these behaviors.

Existing literature on China's NPOGBs offers explanations focusing on the principal-agent relationship between central and local governments. Several studies attribute NPOGBs to fiscal decentralization, arguing that the tax-sharing system doesn't sufficiently incentivize local governments to prioritize overall social welfare. Others highlight the distortion of the principal-agent relationship due to inadequate incentive mechanisms. Promotion based on economic performance prioritizes short-term gains over long-term sustainable development. Inter-jurisdictional competition is also seen as a factor, pushing local governments to engage in a zero-sum game, leading to actions that harm overall efficiency. While these studies offer valuable insights, they have limitations. They lack a micro-level perspective on local political dynamics and individual motivations. They also overlook the influence of the political and administrative systems. Finally, while acknowledging the competitive aspects of local government actions, they do not fully explain why officials choose behaviors that harm their constituents.

This study employs a theoretical deduction method, using a two-step strategy. First, it constructs a conceptual model distinguishing POGB and NPOGB. POGBs are defined as actions that improve conditions in one jurisdiction without negatively affecting others, representing Pareto optimality. NPOGBs are actions that harm government performance or public welfare. The authors argue that contrary to previous research suggesting NPOGBs as unintended consequences, they are often deliberate choices by local officials. Second, the study applies game theory to model the causes of NPOGBs. A model with one central government and two local governments is developed. The model incorporates assumptions of risk-averse officials, information scarcity, and intergovernmental competition. The expected utility functions for both local officials are formulated, considering the impact of their actions on their own performance and the performance of their competitors. The model analyzes the Nash equilibrium, representing the point where neither official has an incentive to unilaterally change their behavior. The analysis examines the reaction functions of the local governments and shows how imitation strategies lead to NPOGBs.

The study's key findings are: 1. **POGBs represent Pareto optimality but are a small portion of all government behaviors.** The majority are NPOGBs, which are detrimental to long-term welfare. 2. **NPOGBs are rational choices for local officials under current conditions.** The competitive promotion system, characterized by information constraints and risk aversion, incentivizes imitation strategies among officials. This leads to a proliferation of NPOGBs, even though these behaviors undermine long-term performance and social welfare. 3. **The game theory model demonstrates a positive feedback loop.** When one local government engages in NPOGB (e.g., redundant construction), others tend to follow suit, resulting in a self-reinforcing cycle. 4. **NPOGBs are 'man-made mistakes,' not merely unintended consequences.** The model shows that local officials deliberately choose these behaviors to enhance their chances of promotion in a highly competitive environment. The model's equations illustrate how a local official's decision is dependent on the actions of other officials. The reaction functions show positive slopes, indicating mutual influence and the tendency toward NPOGBs at the Nash equilibrium.

The findings address the research question by demonstrating that NPOGBs in China's local governments are not accidental but rational responses to the institutional incentives created by the performance-based promotion system. The model shows that inter-jurisdictional competition, combined with information scarcity and risk aversion, pushes officials to adopt imitation strategies, leading to widespread NPOGBs. This challenges the perspective that these are merely unintended consequences or systemic errors. The significance of these results lies in highlighting the limitations of performance evaluation systems that do not adequately account for the strategic behavior of actors within a competitive environment. The study's relevance to the field lies in its contribution to the understanding of government behavior in decentralized systems and the importance of considering behavioral aspects in the design of performance management systems.

This study contributes to the understanding of non-performance-oriented governmental behavior (NPOGB) by demonstrating that such behaviors are not accidental but rational choices made by local officials in a highly competitive environment. Future research could focus on empirical testing of the model in different administrative contexts and exploring alternative incentive mechanisms that encourage cooperation and discourage wasteful competition. The suggested policy recommendations include promoting cooperative decentralization, involving neighboring jurisdictions in policymaking and implementation, and improving mechanisms for intergovernmental communication and coordination.

The study's limitations include the simplification of the game theory model to two local governments. A more complex model incorporating multiple jurisdictions could provide richer insights. The model's assumptions regarding risk aversion and information scarcity might need further refinement. Empirical validation of the model's predictions would strengthen its conclusions. Finally, the applicability of the findings to other countries with different political and administrative systems requires further investigation.