Determinants of variations in state per capita personal income: a panel data approach

This article employs a panel data analysis to examine factors that affect variations in state real per capita personal income (PCPI) by using a panel of 48 contiguous US states over the period 1950-2000. The preferred Panel-Corrected Standard Errors (PCSE) approach reveals that the important explanatory variables are, respectively, the percentage of the population with a college education (COLLEGE), the percentage of the population with a high school education (HIGHSCHOOL), the percentage of the population employed in the manufacturing sector (MANUFACTURING), the percentage of population employed in the service sector (SERVICE) and the percentage of the population that was African American (BLACK). The results are found to be sensitive to estimation technique, suggesting formal model misspecification tests are required in order to choose an approach with better statistical properties.