As a central agenda of China's vocational education reform, industry-education integration has established a comprehensive policy framework at the institutional level. However, practical implementation often faces the governance paradox of "existing...
As a central agenda of China's vocational education reform, industry-education integration has established a comprehensive policy framework at the institutional level. However, practical implementation often faces the governance paradox of "existing policies failing to yield substantive collaborative outcomes." This study aims to address the fundamental theoretical question behind the phenomenon of "hot policies but difficult implementation": Why does deep collaboration rarely emerge spontaneously within highly institutionalized policy environments? Existing research either focuses solely on improving institutional design or appeals to moral incentives for stakeholder engagement, yet largely overlooks actors' strategic choices and their cumulative effects over time. Building on this, the paper incorporates key insights from rational choice institutionalism, integrates the Three-Helix Theory with the theory of gradual institutional change, and constructs an analytical framework encompassing "institutional provision → institutional embedding → incentive structures → behavioral equilibrium → path locking." Through qualitative comparative studies of Yantai, Tianjin, and Tangshan cities in the Bohai Rim region along with their representative higher vocational colleges, the study systematically elucidates the evolutionary logic and differentiated operational mechanisms of industry-education integration systems across different governance structures. The study reveals that within a unified national institutional framework, disparities in local governance resources, administrative hierarchies, and industrial foundations give rise to three distinct forms of institutional embedding through different translation mechanisms, ultimately solidifying into stable behavioral equilibria. Yantai has developed an "administratively integrated deep embedding" structure leveraging the local government's strong administrative coordination capabilities and fiscal resource allocation: it reduces sunk costs associated with enterprises' specialized asset investments through shared financial burdens, redefines profit-sharing mechanisms between educational institutions and businesses via performance-based incentives, and achieves a "locked-in equilibrium" through significant internalization of exit costs. Although this equilibrium involves rhythmic mismatches between goal-oriented rationality and organizational rationality coupled with high integration costs, it is maintained through cross-departmental coordination mechanisms and periodic performance feedback. In contrast, Tianjin has established a "legislative platform-based embedding" model by utilizing its municipal autonomy to create regulatory certainty through policies like the *Tianjin Regulations on Promoting Industry-Education Integration in Vocational Education*, while reducing transaction costs via regional industry-education alliances, thereby forming a "coordinated equilibrium." However, cyclical fluctuations in project-based funding and the complexity of multi-tiered administrative coordination limit collaboration depth to immediate benefit calculations, prompting enterprises to control dedicated asset ratios for strategic flexibility and hindering the development of a high-commitment structure akin to Yantai's. In Tangshan, facing resource constraints and pressures of industrial transformation, institutional frameworks primarily manifest as a "responsive adoption" of provincial policies, exhibiting a "structurally loose, shallowly embedded" pattern: fiscal support is highly initiative-driven but lacks binding force; corporate participation is driven by short-term labor needs; educational institutions adopt project-based cooperation strategies aimed at risk mitigation; ultimately resulting in a "defensive equilibrium." The institutional tension primarily stems from the structural gap between resource capacity and institutional objectives. The theoretical contribution of this study lies in transcending the limitations of static institutional analysis by extending rational choice institutionalism into a temporal dimension, thereby revealing the micro-level mechanisms of institutional evolution. The research demonstrates that deepening industry-education integration cannot be achieved automatically through linear enhancement of institutional provisions alone; rather, it depends on the extent to which incentive structures reshape stakeholders 'benefit functions. Only when institutions can alter cost-benefit calculations through risk-sharing mechanisms, increase opportunity costs for non-participation via performance-linked incentives, and accumulate relationship-specific assets through repeated interactions do actors transition from "marginal exploration" to "deep commitment." Conversely, if institutional provisions remain confined to rule declarations without being integrated into core resource allocation frameworks, they may reinforce existing low-commitment equilibria through "layered" mechanisms, creating a path dependence characterized by "strong provision but weak deepening." Furthermore, this study revises the tri-helix theory's assumption of naturally blurred stakeholder boundaries, arguing that under administrative dominance, clearly defined boundaries and institutionalized risk-sharing mechanisms are prerequisites rather than outcomes for deep collaboration. It also distinguishes three equilibrium types— "lock-in," "cooperative," and "defensive" —providing a typological framework for understanding the operational logic of industry-education integration across different governance contexts. Based on the above findings, this paper argues that the key to overcoming the challenges in industry-education integration lies in moving beyond the simplistic approach of merely increasing institutional supply and shifting focus to the compatibility between institutional embedding quality and incentive structures. Policy design should aim to establish substantive mechanisms for risk sharing and benefit distribution, respect the governance capacities and industrial characteristics of different regions, and allow differentiated institutional evolution paths to achieve stable equilibrium under their respective constraints, rather than demanding uniform compliance with standardized frameworks.