Levy, Amnon and Romm, Tsilia, Optimal timing of migration: expected returns, risk aversion and assimilation costs, Department of Economics, University of Wollongong, Working Paper 91-2, 1991, 13.
Most empirical studies on migration are implicitly based on the assumption that there are only comer solutions to the individual's migration-timing problem. A switch model, which combines the concepts of expected earning differential, risk-bearing and assimilation costs, indicates that comer solutions are adopted only when individuals perceive no risk differential between the places of destination and origin, or when they have no time preference. In any other case, individuals delay migration to future dates within their lifespan. Disregarding these internal solutions understates the true number of migrants and hence leads to biased estimates of the determinants of place-to-place current migration.