Disaggregating Stakeholder Perceptions of Implementation Barriers in Rural Homestead Reform: Evidence from Residents, Village Cadres, and Township Officials in Henan, China

Land-reform monitoring often combines different barriers and respondent groups into broad summary categories. This study asks whether that practice masks systematic differences in which barriers respondents identify. Using a 2026 survey of 2659 rural residents, village cadres, and township officials in 64 county-level units in Henan, China, we compare 15 matched barrier-identification indicators. The analysis uses item-specific logit models, marginal standardization, county fixed effects, county-clustered inference, false-discovery-rate adjustment, and a pooled respondent-group-by-item test. Group contrasts vary strongly across items (Wald χ2(28) = 180.1, p < 0.001). In financing, township officials more often identify inadequate valuation systems and limited financial-institution participation than village cadres. Excluding the broad mortgage item reduces the financing-domain interaction test to Wald χ2(6) = 9.85 (p = 0.131), while the valuation and lender-participation item-specific contrasts remain. Voluntary compensated withdrawal also produces distinct group profiles for compensation, willingness, resettlement, and institutional mechanisms. The all-item result remains stable across geographic, covariate-harmonization, respondent-dependence, and sparse-township sensitivity checks. Implementation surveys can yield different diagnoses depending on the specific barrier measured and the respondent group surveyed. Monitoring systems should retain item-level and respondent-group-level information before aggregation and use large, persistent differences as signals for targeted follow-up verification.