Where Is India on the Feminisation U? Turning Points, Fieller Sets and a Decomposition of Women's Labour-Force Participation, 1991–2024

The feminisation U hypothesis says that women's labour-force participation first falls and then rises as income grows. I test it on World Bank and ILO-modelled data for 171 to 177 countries, 1991 to 2024, using quadratics in log GDP per head at constant 2021 PPP prices. The 2024 cross-section has a minimum at 12,569 dollars, with a 95% Fieller interval of 8,076 to 16,200 dollars. With country and year fixed effects and clustered errors (5,951 observations) the minimum is 11,445 dollars, interval 7,317 to 19,922. I prove the conditions under which a Fieller set for a turning point is a bounded interval and show by simulation that it has correct coverage (0.9480 against a nominal 0.95), while the delta method over-covers (0.9705). India, at 9,416 dollars in 2024, lies below the estimated minimum but inside the Fieller intervals, so the data cannot say whether it is still on the falling arm. Its participation rate of 32.41% is 15.08 points below the curve; only 10.5% of countries are lower. The within-country curve explains India's fall between 1991 and 2017 (curve-implied change of minus 5.89 points against an actual minus 3.78), but not the 2005 to 2017 collapse (minus 8.27 actual against minus 2.26 implied) nor the 2017 to 2024 recovery (plus 5.84 against minus 0.42). India's participation is mostly a story that income does not tell.