Migrant stock vs. lagged migrant flow as a determinant of migrant settlement. A comment on Anjomani and Hariri

In a recent contribution to this Journal Anjomani and Hariri present an interesting study of United States interstate migration which explicitly incorporates so-called flow diversion and flow creation effects. Their discussion and evaluation of the model, however, are marred by several factors. This paper contrasts the roles of migrant stock and lagged migration in migration analysis and then addresses (a) the problems encountered when the family-friends effect is proxied with measures of lagged migrant flows, (b) the problem of using a two-period lagged value of earlier migrant flow as an explanatory variable, and (c) this paper suggests an alternative method of correcting the Anjomani-Hariri model's problems with multicollinearity.The author is grateful to Walter J. Wadycki and to David A. Macpherson for helpful comments and suggestions. The author, of course, bears full responsibility for all errors that remain. This research benefited from a grant from the Miami University Committee on Faculty Research.