TN 2026Election Analyser

What changed between 2021 and 2026 — across 234 Assembly constituencies, in numbers.

The TVK wave →

Method & data sources

🌊

TVK surged +33.7 pp

Median swing across all 234 ACs. From non-existent in 2021 to a major force in 2026.

📉

DMK lost −15.3 pp, AIADMK −13.6 pp

Both established Dravidian majors lost roughly equal ground. The losses fed almost entirely into TVK.

🧪

Religion didn't predict the swing

At AC level, religion-share correlations with party swing are all weak (|r| < 0.2). The story isn't communal.

🏛️

BJP overperformed in reserved seats

+28.5 pp swing in SC/ST-reserved ACs vs +12.5 pp in General. Real, narrow, statistically suggestive.

🏙️

TVK's strongest gains — Chennai metro

9 of the top-10 TVK-gain ACs are in Chennai + Thiruvallur. Urban metropolitan concentration.

🛠️

Reproducible

Every row links a source URL. Re-run with `uv run python pipelines/path_a_build.py`.

What this is

A post-poll AC-level analyser for Tamil Nadu's 2026 Legislative Assembly Election. Built from public data: ECI results, Census 2011 religion data at sub-district, and the kracekumar 2021 detailed-results CSV (open on GitHub).

234 Assembly Constituencies × ~18 candidates per AC × ~22 counting rounds
4,257 candidate-result rows for 2026
4,232 candidate-result rows for 2021
32 districts of religion mix from the 2011 Census

The analyser joins these into per-AC swing-by-demographic tables and surfaces the headline numbers below.

Scope

Path A (this site) is an Assembly Constituency-level analysis. Path B — booth-level — is documented in pipelines/SOURCES_BUY_VS_DIY.md and would need ~$2K of cloud OCR to execute. The data pipeline supports it; this site does not yet present those findings.

The five-minute summary

If you only read one paragraph: TVK ran solo in 2026 and gained roughly 34 percentage points on average across all 234 ACs. DMK lost 15 pp and AIADMK lost 14 pp — both established Dravidian parties surrendered similar ground. The gain wasn't concentrated by religion (correlations are weak), nor by reservation status (TVK gained ~33 pp in both SC/ST and General seats). The clearest geographic concentration: Chennai metropolitan urban belt, where TVK gains routinely exceeded +50 pp. BJP's swing of +13.5 pp (where contested) was over 2× larger in reserved seats than in General — a narrow but real subpattern.