Election
About & Data

An analytical layer over election data — not a newsroom, not a portal.

Election Mandate is India’s historical election database and analysis engine. It turns questions about Indian elections into structured analysis: comparisons, swings, close races, and hypothetical alliances — across every state and Lok Sabha, from the first general election to the latest state poll.

Principles

  • Official election data is the source of truth for every number.
  • A language model never invents seat counts, vote totals, vote shares or margins.
  • Actual results and simulations are always visually distinct.
  • Every scenario exposes its assumptions and its per-seat arithmetic.
  • A simulation is not a prediction, and is never presented as one.
  • The platform stays politically neutral; party colours appear only in data visualisation.

What’s real

  • 417 real elections, 1951–2026 — 18 Lok Sabha general elections and 399 state Assembly elections across 31 states & UTs, all mirrored from ECI results via the community sources listed below. Every dataset is tagged "community data" with its exact source shown on the page it appears on.
  • Constituency maps are real geography, not a schematic placeholder — Assembly-boundary shapes for 30 states (all but Uttarakhand, see Partial), sourced from public ECI shapefile mirrors, plus a Lok Sabha view that dissolves those same Assembly shapes into their real parliamentary-constituency boundaries.
  • Explore any election: house composition, vote share, seat change, two-party swings, closest races, constituency grid — with the previous election of the same house for comparison, or side by side on the Compare page.
  • Scenario Lab on any election: re-run it with a different alliance, or project a future one from the latest result plus an assumed vote swing. Alliance choices are the strongest parties of that election.
  • The deterministic scenario engine (anchor + per-party vote transfer), shared by the Lab and by AskVoter.
  • Responsive editorial design system, light and dark.

What’s partial

  • Andhra Pradesh 2024 lists only the winner and runner-up per seat — the compiled ECI workbook it comes from doesn't carry the rest of the field, so third-place-and-below shares are absent (not estimated).
  • Candidate names are thin in the older imports (the source often carries only party + votes). A few seats deferred by the ECI (Patkura 2019, Ramgarh 2018) are missing entirely, matching the source.
  • Uttarakhand has no Lok Sabha map view — its source shapefile doesn't carry the Assembly→parliamentary-constituency link the other 30 states' shapes do, so there's nothing to dissolve into real PC boundaries yet.

No figure shown anywhere on this site is an official Election Commission of India result — every number traces back to one of the community mirrors listed under Sources below, each mirroring ECI-published results.

Sources

None of these publish official Election Commission of India figures directly — they mirror or re-host ECI-published results and shapefiles. The exact source for any single number is shown on the page it appears on.

Election results

Constituency boundaries

  • datameet/maps — Assembly-constituency shapes for every state except Andhra Pradesh and Telangana
  • datta07/INDIAN-SHAPEFILES — current-delimitation shapes for Andhra Pradesh and Telangana, formed in the 2014 bifurcation

Next

  • Match constituencies across delimitation properly (2008 boundary redraw) instead of by name
  • Fill the deferred-poll seats and thicken candidate coverage in the pre-2010 imports
  • Source a Lok Sabha crosswalk for Uttarakhand's constituency shapes
  • Swap the local AskVoter classifier for a real query planner that emits the same AskAnswer contract
Try AskVoter Open Scenario Lab