Competence Without Conversion: Organizational Conditions Linking Frontline Competence and Responsiveness in Indonesian Local Food Governance

Background Merit-based personnel systems are widely expected to raise bureaucratic quality by improving the competence of public officials, but this expectation says less about whether organizations convert competence into implementation that citizens experience as responsive. This article examines that conversion problem in Sidenreng Rappang (Sidrap), South Sulawesi, Indonesia, selected as a most-likely case because food production is economically and administratively salient. Methods The study combines item-level perception data from 60 farmers across six farmer associations with role-differentiated interviews with four actors positioned at different points in the food-governance system. Survey analysis uses exact Wilcoxon signed-rank tests for four within-respondent competence × responsiveness contrasts, Holm correction, matched-pairs rank-biserial effect sizes, and bootstrap confidence intervals. Interview summaries were analysed thematically and integrated with survey evidence using joint-display logic. Results Farmers evaluated frontline officials favourably on competence and interactional quality (M = 3.60–4.10) but evaluated the administrative system less favourably on responsiveness (M = 2.82–3.10), programme fit (M = 3.20–3.40), procedural accessibility, and targeting. All four paired contrasts remained supported after Holm correction, with rank-biserial effect sizes of.53–.76. The weakest positively worded indicator concerned tangible follow-up to farmers’ suggestions. Conclusions The pattern is consistent with a competence–responsiveness gap rather than a simple competence deficit. We specify four organizational conditions on which conversion depends—feedback conversion, vertical integration, procedural accessibility, and contextual adaptability—with feedback conversion appearing to be the clearest constraint in this case. The contribution is analytical and contextual rather than causal, and the findings do not support population-level generalization.