Social stability risk assessment and economic competitiveness in China

China’s rapid economic growth since 1978 has been driven by a guided market economy, industrial policy, and significant capital accumulation, but has also entailed large-scale infrastructure development and social costs, including land confiscation and displacement. To pre-empt social unrest that could jeopardize growth, China introduced social stability risk assessment (SSRA) policies nationwide in 2012 for major fixed-asset investments. SSRA aims to identify and mitigate risks to social stability prior to project approval, focusing on risks acceptable to local governments and developers. This study asks whether and how SSRA policies contribute to provincial economic competitiveness, adopting the World Economic Forum’s definition of competitiveness as the set of institutions, policies, and factors determining productivity. Using qualitative comparative analysis (QCA), the study investigates which configurations of SSRA policy features are associated with high versus low provincial economic competitiveness and develops a typology of SSRA policies to guide analysis.

The research context highlights the social costs accompanying China’s growth: sharp increases in income inequality across households, regions, and urban–rural divides; rent seeking and corruption; environmental pollution; and unequal access to opportunities. Public concern about inequality has been high, with China’s Gini coefficient approaching levels associated with social instability. Prominent incidents such as the 2004 Hanyuan protests, linked to displacement for hydropower development, galvanized government attention to maintaining a harmonious society. SSRA emerged as a policy tool to analyze stability risks and plan responses. Prior literature underscores two broad approaches within SSRA practice: a problem-solving orientation that identifies solvable social risks through participation and stakeholder analysis, and a risk orientation emphasizing ex-ante risk grading and checklists. Additional orientations include performance assessment of officials and specific field (sector/issue) targeting. The literature also notes tensions between stability preservation and genuine problem resolution, and the potential for risk management frameworks to overlook harms that do not trigger overt unrest.

The study applies fuzzy-set qualitative comparative analysis (fsQCA) to 173 SSRA policy cases (2003–2020) across China, linking policy configurations to provincial economic competitiveness outcomes measured using the Reports on China’s Provincial Economic Competitiveness (2019–2020) from Fujian Normal University and Social Science Literature Publishing. Outcome variable: high vs. low provincial economic competitiveness. Conditional variables derived from a typology of SSRA policies: (P1) problem-solving orientation; (P2) risk level identification orientation; (P3) performance assessment orientation; (P4) specific field orientation. Procedure: (1) Data calibration using the direct method with thresholds at the 75th percentile (full membership), median (crossover), and 25th percentile (full non-membership); fsQCA 3.0 computed fuzzy membership scores for outcome and conditions. (2) Necessary condition analysis tested each condition and its negation against the outcome, using a consistency threshold of 0.90 and considering coverage. (3) Truth table construction with thresholds: consistency ≥ 0.80, PRI ≥ 0.70, and case frequency ≥ 1; analysis of intermediate and parsimonious solutions to identify core vs. peripheral conditions. (4) Collinearity testing to ensure conditional variables are not collinear and to assess statistical significance of variables in a regression check. Data sources for SSRA policies included official government portals, news media, academic literature, and provincial SSRA reports; the dataset spans five national and 168 provincial policies; Hong Kong, Macao, and Taiwan were excluded due to data paucity.

• No single SSRA policy type is individually necessary for high or low economic competitiveness (all necessary condition consistencies < 0.90).
• High economic competitiveness configuration (S1): requires presence of problem-solving orientation (core) and specific field orientation (also present). Solution coverage 0.4387 (≈43.87%), consistency 0.9444. Interpretation: about 43.87% of high competitiveness across provinces is explained by SSRA policy configuration S1; the remainder reflects other factors.
• Low economic competitiveness configurations: three distinct solutions (NS1, NS2, NS3a/NS3b) collectively cover ≈50.77% of low competitiveness cases. • NS1 (blind pursuit of performance assessment): absence of problem-solving (core), presence of performance assessment (peripheral), irrespective of risk orientation or specific field; consistency 0.9302; raw coverage 0.2322; unique coverage 0.1316. • NS2 (only focusing on a specific field): absence of problem-solving (core), presence of specific field (peripheral), absence of performance assessment (peripheral); consistency 0.9002; raw coverage 0.2038; unique coverage 0.1406. • NS3a/NS3b (simply identifying risk level): absence of problem-solving (core), presence of risk identification (peripheral); NS3a further shows absence of specific field (peripheral) and NS3b absence of performance assessment (peripheral). Both have consistency 0.9331; raw coverage 0.2341; unique coverage 0.1141 each.
• Collinearity analysis indicates no collinearity among explanatory variables; the variable “problem-solving” is statistically significant (p = 0.006) with regression coefficient −8.611, interpreted by the authors as more problem-solving-oriented policies being associated with higher economic competitiveness, consistent with the configuration results.

Findings indicate that SSRA policies contribute positively to economic competitiveness only when they are structured to solve social stability problems and are targeted to specific industries/issues. Configurations emphasizing performance assessment, singular field focus without comprehensive problem-solving, or mere risk grading without mitigation are associated with low competitiveness. This underscores that stability-oriented administrative control or checklist-based risk grading does not substitute for genuine problem resolution that incorporates stakeholder engagement and addresses social risks experienced by affected communities. The results also illuminate the broader policy tension: SSRA’s emphasis on protecting government and developer interests can neglect social harms that may not manifest as immediate unrest but can seed future instability and economic costs (rebound dynamics). Aligning SSRA with international risk and impact assessment practices and expanding its scope to include social risk alongside social stability risk could improve both governance outcomes and economic performance.

The study introduces a typology of SSRA policies and employs fsQCA to link policy configurations to provincial economic competitiveness. It finds a single effective configuration for high competitiveness: combining problem-solving orientation with specific field targeting. In contrast, configurations centered on performance assessment, narrow field focus alone, or risk identification without mitigation are associated with low competitiveness. Policy implication: reorient SSRA frameworks toward genuine problem-solving and sector-specific application, with stronger public participation and mitigation mechanisms. Future research should examine how to operationalize such reorientation across diverse provincial contexts, address structural barriers to participation and accountability, and extend SSRA to systematically assess social risks, aligning with international standards (e.g., World Bank, IFC).

• Data limitations: SSRA dataset relies on publicly available documents; private/internal government policies were inaccessible, potentially omitting relevant policies or details.
• Scope: Analysis excludes Hong Kong, Macao, and Taiwan due to limited data; SSRA policies span 2003–2020 while competitiveness data are 2019–2020, which may introduce temporal mismatch.
• Causal complexity: Economic competitiveness is influenced by many factors (resource endowments, geography, demographics, central support), so SSRA explains only a portion of variance (e.g., S1 coverage ≈43.87% for high competitiveness).
• Limited diversity: Unobserved configurations (remainders) constrain inference; QCA relies on threshold choices for calibration and consistency/PRI that may affect solutions.
• Institutional/implementation variability: Differences in provincial governance and resistance to public participation may affect how policy orientations translate into practice, limiting generalizability.