Entrepreneurship literature (Parker 2004) has rarely considered spatial location as a micro-determinant of occupational choice, although there are compelling reasons to posit that spatial location influences economic behavior. Using Bayesian semiparametric methodologies and geoadditive techniques, we examine spatial location as a micro-determinant of self-employment choice of Indians, in addition to standard determinants such as age, gender and education. The empirical analysis suggests the presence of spatial occupational neighborhoods and a clear north-south divide in self-employment choice in India when individuals of agricultural and nonagricultural sectors are considered together; however, such spatial patterns are less pronounced when individuals in nonagriculture alone are considered in the analysis. These residual spatial patterns are found to be inversely related to the per-capita GDP of the region. The results further suggest nonlinear relationships between age, wealth and the probability of self-employment.
Entrepreneurship Self-employment Developing Countries Bayesian Semiparametric Methods Geoadditive Models