Point-To-Point Migration from the 1999 Kenya Census: A Methodological Look at Push and Pull Factors in Space and Time

Michael Levin, Harvard University

Kenyan censuses have traditionally collected information on place of birth, place of enumeration, and place of residence 12 months before the census. Usually, two-point migration (birth or previous residence to current residence) is tabulated but that fails to track people at three different temporal points in space, so misses people who migrated but moved back to their place of birth before the enumeration date, and others who moved onward. In this study, we look at province to province migration flows, using numbers and characteristics of migrants by age and sex for the 1999 Kenya census. We discuss the methodology using the census data to construct the matrices. The Kenyan data show that migrants increasingly flow to urban areas and at higher rates than previously, tend to be more educated, have higher labor force participation rates, but also have higher unemployment rates because of raised, unfulfilled expectations.