Basis for fulfilling responsibilities, behavior, and professionalism of government agencies and effectiveness in public–public collaboration for food safety risk management

The study addresses how to improve the effectiveness of public–public collaboration in managing food safety risks, a complex, long-chain issue with diversified and evolving risks. Given WHO’s estimate of substantial global mortality from foodborne and waterborne diseases and the accelerating emergence of new risks, governments typically rely on multi-agency regulation, which can suffer from fragmentation and coordination challenges. The research gap lies in limited systematic examination of the effectiveness of public–public collaboration in food safety risk management, particularly in transition countries like China. The authors aim to identify the main factors influencing governance effectiveness in such collaboration. China is chosen due to extensive reforms of its food safety regulatory system (1983–2023) and notable improvement in the Global Food Security Index. The study proposes 20 hypotheses grouped into five dimensions: (1) basic characteristics of agencies (legal person, hierarchy, subordination, professionalism); (2) basis for fulfilling responsibilities (laws and regulations, normative documents, informal rules, path dependence); (3) functions (comprehensive dominant, single-function dominant, auxiliary agencies); (4) behavior and capabilities (administrative regulation and enforcement, judicial enforcement, legislative, environmental improvement); and (5) infrastructure and culture (information sharing, information technology application, coordination of resources, organizational culture, social concern).

The paper reviews cross-boundary governance theory, emphasizing goals (shared objectives via information and resource sharing), actors (public and private sectors, NGOs, citizens), and methods (exchange and cooperation to overcome bureaucratic fragmentation). It highlights public–public collaboration as a core form, occurring horizontally (across agencies at the same level) and vertically (across levels), aiming for holistic governance to mitigate government failures caused by fragmented responsibilities. In food safety, multi-agency systems are common (e.g., the U.S. system with numerous federal and subnational regulators), but they often face decentralization, overlap, and fragmentation, underscoring the need for cross-boundary collaboration. The study defines public–public collaboration for food safety risk management as horizontal, vertical, or combined collaboration to break barriers to information/resource sharing and clarify powers and responsibilities across agencies. It contextualizes China’s system with comprehensive dominant, single-function dominant, and auxiliary agencies, and notes that agencies can undertake multiple roles (e.g., market regulation agencies also legislate and enforce). The review sets the foundation for hypothesizing 20 factors across five dimensions that may affect collaboration effectiveness.

Design: A structured questionnaire measured 20 hypothesized influencing factors (X1–X20), each mapped to the five dimensions and rated on a 5-point Likert scale (importance). A pre-survey (Wanxiu District, Wuzhou City, Guangxi) validated and refined items; technical terms (hierarchy, path dependence) were clarified in plain language. Sampling and setting: Focused on grassroots agencies at county/district and township/subdistrict levels in China, where public–public collaboration is operationalized. Survey sites included Jiangyin City and Runzhou District (Jiangsu Province) and Wanxiu District and Lingchuan County (Guangxi Zhuang Autonomous Region). Questionnaires were distributed via online working groups of local administrations for market regulation; trained personnel supported respondents; straight-line patterns were excluded. Sample: 1202 valid responses covering 82.18% of eligible staff in the selected jurisdictions. Demographics: 45.84% male; 54.16% female. Job titles: staff (73.63%), section chiefs (25.12%), division chiefs (1.25%). Education: bachelor’s degree (64.97%) most common. Age and tenure broadly distributed. Agency types included market regulation (41.76%), agriculture and rural affairs (3.66%), health (6.07%), customs (1.58%), justice system agencies, education/science/industry agencies, and others. Measures: Variables X1–X20 covered basic characteristics (legal person, hierarchy, subordination, professionalism), basis for responsibilities (laws and regulations, normative documents, informal rules, path dependency), functions (comprehensive dominant, single-function dominant, auxiliary), behavior and capabilities (administrative regulation and enforcement, judicial enforcement, legislative, environmental improvement), and infrastructure/culture (information sharing, IT application, resource coordination, organizational culture, social concern). Outcome Y measured perceived governance effectiveness in public–public collaboration for food safety risk management. Analysis: Data suitability confirmed by KMO=0.932 and Bartlett’s test χ²≈12,727.812, p<0.001. Principal Component Analysis (PCA) reduced 20 correlated variables to nine orthogonal principal components (PCs) with cumulative variance explained=80.258%: F1 eigenvalue 8.526 (42.630%), F2 1.776 (8.882%), F3 1.537 (7.684%). Varimax rotation aided interpretability. PCs were classified: behavior and capabilities (F1), basis for fulfilling responsibilities (F2: informal rules, path dependence; F5: laws and regulations, normative documents), functions (F3: single-function dominant, auxiliary; F8: comprehensive dominant), infrastructure and culture (F4: information sharing, IT application; F7: social concern), and basic characteristics (F5: legal person, hierarchy; F9: professionalism). PC score coefficients were estimated via regression in SPSS. Regression modeling: A multivariable linear regression of Y on the nine PCs was estimated: Y = b0 + Σ(bi·Fi). Model fit: Adjusted R²=0.748; F=397.375; p<0.001; VIFs ~1, indicating no multicollinearity among PCs. Using the derived PC score coefficients, an equivalent regression in terms of the original 20 variables X1–X20 was obtained. Robustness: Despite PCA’s typical robustness, an additional check removed five variables (normative documents X6, auxiliary agencies X11, judicial enforcement X13, organizational culture X19, social concern X20). Regression on the remaining 15 variables yielded consistent identification of key factors (professionalism X4, laws and regulations X5, administrative enforcement X12, legislative X14, environmental improvement X15).

- Model performance: The PCA-based regression explained 74.8% of the variance in governance effectiveness (Adjusted R²=0.748; F=397.375; p<0.001), with orthogonal PCs (VIF≈1). - Principal components: F1 (behavior and capabilities) had the strongest standardized effect (Beta≈0.599; B=0.531; p<0.001), followed by other PCs representing functions, infrastructure/culture, and basic characteristics. - Variable-level effects (from the back-transformed regression on X1–X20): Key positive coefficients were: • Legislative agencies (X14): 0.215 • Professionalism (X4): 0.195 • Laws and regulations (X5): 0.161 • Administrative regulation and enforcement (X12): 0.137 • Environmental improvement agencies (X15): 0.119 Smaller magnitude coefficients for other factors indicated limited influence on Y under current conditions. - Classification of key drivers: The most influential elements span behavior and capabilities (legislative, administrative enforcement, environmental improvement), the professional capacity of agencies, and formal legal bases (laws and regulations). Informal rules (X7) and organizational culture (X19) showed comparatively weak effects in this context. - Robustness check confirmed the same five key factors using an alternative specification (largest coefficients again for X4, X5, X12, X14, X15).

The findings directly address the research question by identifying which dimensions most strongly enhance the effectiveness of public–public collaboration in food safety risk management. Strength in legislative capacity and reliance on formal legal frameworks (laws and regulations) appear foundational in the Chinese context to align and coordinate multi-agency actions across boundaries. High professionalism within agencies further equips them to manage complex, technical risks, while administrative enforcement and environmental improvement agencies’ behaviors and capabilities translate legal mandates into effective implementation, public communication, and supportive governance environments. Comparatively, factors often emphasized in Western contexts—such as informal rules and organizational culture—were not key drivers here, likely reflecting differences in institutional arrangements, legal oversight, and administrative traditions. This suggests that in China’s current stage of system development, formalization through law and the professionalization of agencies are the primary levers for improving collaborative governance. The results imply that developing countries adopting multi-agency regulatory models may benefit from prioritizing legal coherence, legislative capacity, professional skills, and coordinated enforcement to overcome fragmentation and enhance collaboration outcomes.

This study contributes by empirically identifying the principal drivers of effective public–public collaboration for food safety risk management in China: (1) the behavior and capabilities of legislative agencies; (2) administrative regulation and enforcement; (3) environmental improvement functions; (4) agency professionalism; and (5) laws and regulations as the formal basis for action. Policy implications include: - Strengthen legislative bodies and improve legal coherence to address legislative fragmentation and close gaps or ambiguities; employ normative documents to bridge needs while comprehensive legislation evolves. - Build capable administrative enforcement systems with clear powers, responsibilities, and complementary functions across levels to reduce fragmentation and buck-passing; leverage environmental improvement agencies (education, publicity, online governance) to support compliance and public awareness. - Consolidate a stable, interconnected legal framework and conducive enforcement environment that aligns mandates across agencies and mitigates local protectionism. - Invest in professional capacity: talent development, technical infrastructure, and specialized tools aligned with agency functions to fulfill responsibilities and enable cross-boundary collaboration. While other factors were not identified as key in this study, they may become more influential as institutions mature. Future research should expand geographic coverage, refine measurement (multi-item constructs), and examine causal mechanisms to deepen understanding of how these factors shape collaborative effectiveness.

- External validity: Data come from four county/district jurisdictions in two provinces (Jiangsu, Guangxi); results may not generalize across China’s 2844 counties/cities/districts. - Construct overlap: The 20 hypothesized factors may contain conceptual overlaps; mechanisms by which factors affect effectiveness were not analyzed. - Measurement limitations: Single-item Likert measures per factor constrain reliability and depth; respondents’ knowledge and potential agency self-interest could bias ratings. - Design limitations: The analysis establishes associations, not causality, between factors (X) and governance effectiveness (Y). - Context dependence: Findings reflect China’s institutional setting where formal legal rules dominate; informal rules and culture may gain importance as systems evolve.