We investigate whether the occupational productivity and employment status of
individuals living in a household with migrants differ from those living in non-migrant
households using the sixth round of the Ghana Living Standards Survey (GLSS6) and the
Africa Sector Database (ASD). We find that rural households and households with a head
in more productive occupations are more likely to have migrant members, and that rural
households and households with a head who are waged-employed are more likely to have
a migrant than households with members who are self-employed. While these findings are
not suprising, we find some more unexpected results. For instance, migrants do not always
migrate to more productive occupations; migration can result in downward occupational
mobility. Migrants in our sample do not send back much remittances. Migrant-sending
households in Ghana are in fact more likely to send remittances to their relatives currently
away, than to receive remittances. In an attempt to explain these somewhat puzzling
findings, we argue that a motivation for rural households or households with a head in a
more productive occupation to send out relatives is to support younger household members
to pursue their education elsewhere. Migration is therefore a long(er)-run income-andoccupational
diversification strategy of the more productively employed rural households in
Ghana.