| This exploratory research attempts to fill a gap in existing knowledge by seeking to identify the empirical determinants of state-level Foreign Direct Investment (FDI) in India. In doing so, it differentiates itself from previous studies on India and the vast body of literature on the empirical determinants of FDI in general, which use the country as a unit of analysis. National level studies in general have ignored intra-national differences, which are quite extreme in countries like India.The period of study is the 'reform period' 1991--2002, which corresponds to the most ambitious economic reform program in independent India's history. Financial liberalization was an important feature of this program and FDI inflows were encouraged for the first time since 1947. Inflows to India have increased, but are concentrated in a handful of states, with the rest of the country largely ignored. We select a combination of economic, political, and social variables, measured at both the state and national levels to explain the variation in state-level FDI inflows. In doing so, we hope to provide policy advice to states that have not been particularly successful in attracting FDI, on what enabling conditions need to be created within their respective states, to make them attractive to investors, both domestic and foreign.We use a pooled time-series, cross section Ordinary Least Squares (OLS) model, corrected for heteroskedasticity, to regress state level FDI inflows on the combination of state and national level variables mentioned above. We also include a lagged dependent variable as an exogenous variable to correct for autocorrelation and mitigate specification bias. The control variables, which were macroeconomic indicators were highly correlated, which meant their influence on inflows could not be simultaneously measured.Quality of roads, agglomeration economies, incentives provided by states, and the lagged variable were statistically significant predictors of state-level FDI in the full model. We used backward elimination to obtain a parsimonious model, which included all the aforesaid variables, plus GDP as significant predictors, besides providing superior explanatory power. The results establish the importance of both gravity variables and policy intervention in determining FDI inflows to states. |
