From infrastructure completion to market outcomes: cold-chain commissioning and tomato price instability in India

Introduction Whether completed cold-chain infrastructure translates into measurable market outcomes depends on the step from commissioning to actual service delivery. Methods We link India's Ministry of Food Processing Industries registers to monthly district wholesale prices (2013–2023) and evaluate the first documented commissioning of fruit-and-vegetable cold-chain projects. The primary outcome is the absolute exact-calendar-year-on-year change in tomato log modal price relative to the mean of onion and potato. Because these comparison commodities may also use cold-chain services, the parameter captures a differential commodity effect rather than the program's total price effect. We compare 20 commissioning districts with 59 unique same-state matched controls across common pre/post-commissioning windows, using stack weights that preserve the intended comparison. The 20 districts are a subset of 34 candidates, limiting external validity. Results During months 0–23, the corrected estimate is 0.0112 (standard error 0.0206; 95% confidence interval −0.0298 to 0.0521; conventional p = 0.588; wild-cluster-bootstrap p = 0.581). The equal-weight mean across 20 district comparisons is 0.0099 (95% confidence interval −0.0300 to 0.0497). A matching sensitivity fixed before outcome estimation reduces the pooled tomato pretrend standardized difference from 0.323 to 0.039 while retaining 10 comparisons. Its estimate changes sign to −0.0198 (95% confidence interval −0.0721 to 0.0326). The relative mean log-price estimate is also imprecise (−0.0120; 95% confidence interval −0.0544 to 0.0304). Commodity-specific and alternative-outcome estimates are likewise imprecise. Six-month pre-commissioning coefficients are jointly insignificant. Sensitivity intervals widen materially when deviations from parallel trends are allowed. Discussion The analysis does not precisely detect an average differential change during the first 2 years after commissioning. Estimates remain compatible with modest changes of either sign. Food loss, welfare, energy use, and emissions are unmeasured. We conclude that commissioning records alone cannot establish tomato-specific services or price stabilization; evaluations should link infrastructure milestones to measured utilization and commodity-level services.