Output-based monitoring has dominated India's rural infrastructure schemes. Evaluate the significance of longitudinal impact assessments, with reference to PMGSY-IV and PM-JANMAN.

Output-based monitoring counts kilometres sanctioned and completed. PMGSY tracks these on OMMAS and reports about 95% physical progress [7]. Building a road does not by itself show that lives improved. In September 2026, NRIDA and IIM Kolkata signed an MoU for a four-year longitudinal study of PMGSY-IV and PM-JANMAN roads [1]. The study moves the focus to outcomes. It is significant, but it cannot do the job alone.

Why output metrics fall short

• Completed does not mean lasting: the CAG found 21% of completed PMGSY work poorly maintained. Only 15 States had a maintenance policy [5].
• Poor quality stays hidden: the Standing Committee (2023) found 41% of roads in poor condition because States did not fund upkeep. It also flagged low-tendering, where contractors bid too low and cut corners [4].

Significance of longitudinal assessment

• Tracks change over time: a 14-month baseline, about 12 months of transition and a 22-month endline. It covers 10 States/UTs, 250 roads and about 5,000 households [1].
• Measures outcomes: livelihoods, market access, travel costs, health, education and women's empowerment [1].
• Tests equity for PVTGs: PM-JANMAN has an 8,000 km road component for 75 PVTGs [3]. The study reports results separately by region and vulnerable group [1].
• Tests the eligibility rules: PMGSY-IV relaxes population thresholds to 250+ for hill, NE and Fifth Schedule areas and 100+ for LWE districts [2]. The sample can compare outcomes across these terrains.
• Accountability: an independent evaluator can flag implementation gaps [1].

Limitations

• Hard to isolate the road's effect: PM-JANMAN delivers 11 interventions through 9 ministries [3]. No comparison group has been announced [1].
• Results come late: they will arrive around 2030, after PMGSY-IV ends in FY 2028-29 [2].
• Misses the real maintenance test: the endline falls inside the contractor's five-year e-MARG upkeep period [6]. It will not see what happens after States take over.

Longitudinal assessment should add to output monitoring, not replace it. NRIDA should add comparison villages from later PMGSY-IV sanctions, link survey data with OMMAS and e-MARG records, and publish the data for NITI Aayog's DMEO. With these fixes, the study would be a major step towards evidence-based policy. It would help ensure that roads deliver Antyodaya and SDG-9's goal of fair access to infrastructure.