Differences in the determinants of intentions to migrate versus actual migration: an analysis based on Ecuadorian survey evidence.

Studying migration decision making may be approached in a number of ways. In this study intentions data were used 1) to examine factors influencing intentions and then 2) to test whether the same mechanisms affecting intentions also affected actual migration behavior. Individual household and areal factors were used to examine intentions and rural to urban migration behavior. Data were obtained from a detailed 1977-78 Survey of Internal Migration conducted in 52 cantons of the Sierra in Ecuador. Households with out-migrants and those with large landholdings were oversampled. Migrants were those 12 years and older who left within 5 years preceding the survey. Migration intentions were expressed intentions to migrate within 2 years of the survey date. Independent variables were worktime household size by number of adults and number of children agricultural labor absorptive capacity and access to roads. The findings reflect a weak migration intentions model and little correspondence with the variables associated with actual behavior. Several of the household and areal variables have opposite effects. The conclusion was that factors affecting intentions are different in important ways from those influencing migration intentions. Migration intentions are influenced by individual perceptions of conditions necessary for migration. Actual migration is influenced by the objective conditions. The behavioral models were similar to earlier models by Bilsborrow et al. Differences were in the inclusion of additional community level factors and the change to the age group 12-59 years. In contrast to earlier findings education shows a positive impact and age shows a negative impact on migration. There was also a negative relationship between the Gini land distribution coefficient and out-migration of those from small farms (<5 hectares). Out-migration is affected both by "relative deprivation" and "absolute status." Other statistically significant factors were road access distance and local labor market opportunities.