KERALA RISING · FROM EXPORTING PEOPLE TO EXPORTING IDEAS How we would know it is working The decisive question is whether recognition actually changes behaviour where enforcement alone stalled, and that is answerable against hard data the state already collects. The program’s leading indicators, opt-in rates among auto and bus operators, verified safe-driving segments per participant, tell the institution whether the loop is forming. The lagging indicators are the ones that matter to the public: corridor-level change in helmet and seatbelt compliance and in junction-blocking, and, over a longer horizon, the correlation with the Motor Vehicles Department’s own accident and citation data on participating corridors. Because the MVD already publishes accident statistics by district and corridor, the program can be evaluated against an external, pre-existing series rather than its own self-report. A far stronger test. If participating corridors do not show a divergence in the right direction over time against comparable non-participating ones, the program has not earned its place, and the published assessment will say so. Tying the evaluation to data the enforcement system already produces is what keeps the recognition program honest about whether it adds anything beyond the cameras. 133