Abstract

The paper reconsiders the contemporaneous relationship between real gross capital formation and real GDP growth in India by using 26 annual WDI measurements for the period 2000-2025. The bivariate OLS specification should be read as a baseline relationship, not as a growth regression in the spirit of a structural or causal growth model. The slope is positive (β̂ = 0.2628) but not statistically significant at the usual 5% level (p = 0.071) and the model accounts for 12.9% of the variation in growth. Diagnostics indicate that 2020 is highly influential (Cook’s D = 0.916) and results in significant non-normality of the residuals. Dropping 2020 or 2020-2021 in a lagged-GCF model with HC1, HC3 or Newey-West standard errors, the coefficient is positive but statistically insignificant. Adding a 2020 shock dummy brings the slope down to 0.1130 (p=0.206). The estimates for the sub-periods 2000-2007, 2008-2019 and 2020-2025 are highly volatile. When the sample is extended to 2025, the slope is a little steeper, but the message does not change: the relationship is positive but weak, and not a recognized benefit of investment-led growth.