Regional Incentives and Location Choice of New Firms in India: A Nested Logit Model

Abstract

This research paper attempts to develop an appropriate econometric model for regional incentives and location choice of new firms in India. The alternative econometric models of location choice include ordinary least squares, multinomial logit model, Poisson (count) model, conditional logit model and nested logit model. The present study utilised two-level Nested logit model, a discrete choice regression technique, to estimate the influence of regional incentives on the location choice of new firms in India. In considering what groupings of location choices to use for the upper nest of the tree structure, we simply categorised areas into designated and non-designated status, as defined by policy. We divided the area in two nests or subgroups, or regions: the Special Category States or the Beneficiary States (Jammu and Kashmir, Himachal Pradesh and Uttarakhand) which have been granted special investment incentives package and the Control Group or the Non-Beneficiary Neighbouring States (Haryana, Punjab and Uttar Pradesh) which were given no regional incentives. We have considered only firms established between 2002–2003 and 2010–2011. The independent variables considered for the analysis include investment incentives (dummy variable), manufacturing agglomeration, market potential and land cost. The nested logit model was estimated with full information maximum likelihood estimation using Stata software. The estimation validated the importance of regional fiscal incentives and manufacturing agglomeration as the most important determinants of industrial location.