Attracting New Maine Residents: The Effects of Educational Attainment and Age on Interstate Mobility

Abstract

MAINE POLICY REVIEW • Vol. 28, No. 2 • 2019 Appendix 1 Attracting New Maine Residents: The Effects of Educational Attainment and Age on Interstate Mobility by Paul Leparulo APPENDIX I use a regression model to estimate the economic and statistical significance between educational attainment and interstate mobility. The particular type of regression (probit model) is designed for situations where the outcome variable is binary rather than continuous. In this case, the outcome variable represents whether an individual made an interstate move and the regressions yield the predicted probability of a move. More specifically, the regression model examines the potential relationships between moving and such variables as bachelor’s degree holder, age and homeownership. Age and homeownership are included as control variables as they are correlated with both the likelihood of moving and educational attainment. To more accurately model these relationships, I allow for nonlinearities in the relationship between moving and age and include interaction terms. The regression output of primary importance is the marginal effect of having at least a four-year degree on the propensity to make an interstate move. The marginal effect quantifies the causal relationship between these variables. I focus on the average marginal effect (AME), which is the average of all marginal effects across all observations, (i.e., the average effect of having at least a four-year degree on the probability of an interstate move for all 23to 55-year-olds in the restricted sample), and the marginal effect for persons of a representative age (e.g., the marginal effect of having at least a four-year degree for the average 25-year-old). From this regression model I am also able to produce the predicted probabilities of an interstate move for an average individual of a representative age.