The Setu Index
What gets published gets attention. So the measures have to be chosen carefully, reported honestly, and designed to be hard to game.
What a unit is judged on
Responsiveness
Time to resolve ordinary requests, share of standards met, escalation rate, repeat failures on the same issue.
Livelihood
People entering paid work, and — the number that matters — people moving out of community employment into better-paid or skilled work.
Learning
Participation and completion at the Learning Park, adult literacy and digital literacy, student civic-service hours verified.
Care
Reach of support and referral, published alongside privacy complaints and grievance rates — never one without the other.
Environment
The Jal-Van Ledger: groundwater, surviving tree cover, waste diverted, water treated, air, and local heat.
Trust
Independently surveyed resident trust, reported with the grievance rate and the share of residents who know how to complain.
A distribution, not a league table
Ranking fifty thousand units against each other would be irresistible to publish and ruinous in practice.
A ranking rewards the units that were already prosperous, punishes the ones facing the hardest conditions, and creates an immediate incentive to manipulate whichever number is most visible. Within two years the measurement becomes the activity.
- Report the spread, not the winner. Every unit publishes its own distribution — including how the worst-served decile of its own residents is doing, which is the number that actually reveals whether a unit serves everybody.
- Compare like with like. A drought-prone block is compared against similar blocks, never against a metro ward.
- Publish the denominators. Percentages without the underlying counts are how bad results are hidden.
- Independent survey. Trust is never measured by the body being trusted.
- Open data. Machine-readable, freely licensed, so anyone can check the arithmetic and recompute it differently.
- Publish failures explicitly. Units under administration appear in the Index, labelled, rather than being quietly omitted.
Money spent. Projects announced. Events held. Volunteers registered. Attendance at meetings. App downloads. Every one of these is easy to produce, easy to inflate, and tells a resident nothing about whether their drain was fixed.