Three regression models fitted to NFHS-3, NFHS-4, and NFHS-5 data for all 36 states. Projections extend to 2045 with 95% bootstrap confidence intervals. Select any state to explore its trajectory.
Default sort: year hitting severe decline (soonest first). Click any column header to re-sort. Click any row to load that state in the explorer above.
| # | State | NFHS-5 | 2026 est. | Severe decline ↑ | Replacement | Model | Status |
|---|
Polynomial regression — which best fits both states — projects their TFR rising back toward 1.8–2.0 by 2026. The curvature coefficient (a > 0) confirms a genuine U-shaped trajectory, not noise. No other state shows this pattern.
Severe decline is defined as TFR below 1.5 — the level South Korea crossed in the mid-1990s before its demographic crisis became irreversible. More than two-thirds of India's states are projected to cross this threshold within this decade.
With 130 million people and the lowest GDP in the dataset, Bihar has the least infrastructure to absorb an aging population. Its projected crossing of the replacement threshold in 2031 starts a dependency clock with no safety net.
The dominance of polynomial over linear confirms that India's fertility transition is not happening at a constant rate. Most states show deceleration (slowing decline) or acceleration — a straight line systematically misrepresents their trajectory.
Bihar, UP, and Rajasthan are declining fast enough that by 2035 their projected TFR will be within 0.3 of Kerala and Tamil Nadu. The convergence is not because the South recovered — it is because the North is falling faster than expected.
States with the widest 95% confidence intervals — Manipur, Meghalaya, Nagaland — are those where the trajectory changed direction between survey rounds. Three data points cannot reliably constrain a model for states mid-transition.