Scheme design · sandbox

Move the line. Watch who falls the other side of it.

A state assistance scheme with four rules. Drag any of them and the coverage, the annual cost to the exchequer and the district spread recompute across all thirty districts, along with the names of fourteen households sitting close enough to the boundary to be moved by the decision.

Where it lands

Share of each district’s households that qualify

Fewer More

A cartogram, not a map: every district is drawn the same size because the question is about people, and drawing Mayurbhanj eight times the size of Jharsuguda answers a question nobody asked. East is coastal, south-west is the KBK belt.

District by district

Who is on the line

What this is and is not. The household distributions are shaped to the broad pattern of the state: coastal districts earn more and hold less land, the KBK belt the reverse, tribal share highest in the south and west. But they are illustrative and must not be quoted as census or survey figures. The arithmetic on top of them is the real thing: exact sums over a 96,000-cell table, not a sample, so the same rules always give the same answer. A live deployment reads the actual survey, and the four axes are assumed independent here, which a real study would not do. That assumption is named rather than hidden.