Agricultural Productivity and Urbanization: A Smooth Coefficient Regression Analysis

The study investigates the role of agriculture in the process of urbanization. According to Matsuyama (1992), the relationship between agricultural productivity and urbanization depends on the economic openness. In closed economies the relation is positive while for open economies it is negative. This paper empirically tests the role of agricultural productivity on a country’s level of urbanization through a spatial econometric regression model that allows for smooth coefficients. Spatial autoregressive errors are included in the model to account for unobservable spatially correlated natural, social or historical factors that may determine urbanization levels. The slope coefficient of agricultural productivity on urbanization is modeled as a smooth transition function of countries’ degree of openness to allow for variation by openness. The models are estimated using cross section samples of countries, separately for 1990 and 2005. Our results show that the effects of agricultural productivity on urbanization are positive for closed economies and decrease with the degree of openness. Moreover, we find a significant coefficient differential between closed and open economies.