Migration decision making and migration behavior in rural Kenya

Micro theories of migration have generally been tested with data which do not allow one to distinguish between the decision to move and the act of moving. The latter is usually treated as a proxy for the former. Using data from a two round survey of youth in rural Kenya we classify respondents as decision processors/nondecision processors from round one data, and determine who migrates over a nine month interval from round two data. Thirty-four percent of the sample are classified as decision processors and 25% as migrants, but decision processors were only slightly more likely to migrate than nondecision processors and more migrants actually came from the latter group.A simple model is developed to determine the extent to which the same factors that influence decision-processing influence who migrates. The model is tested using regression procedures by first treating decision-processing as the dependent variable then migration. The model explains 26% of the variance in who decides, but only 10% of the variance in who actually moved. Finally, separate analyses are done of the determinants of migration for decision processors and nondecision processors.