Harmonized Host and Refugee Labor Market Survey 2022

The sampling design included 265 initial enumeration areas selected using probability proportionate to size with the number of households used as a measure of size. The enumeration areas were select-ed based on the sample frame constructed during the 2014 population census of Uganda. The Ugan-dan Bureau of Statistics (UBOS) conducted the selection of the EAs and provided their list along with detailed maps of the areas.

Using maps of the selected enumeration areas provided by UBOS, the study team conducted the list-ing of all households in the selected EAs with door-to-door visits. The listing exercise was carried out during November 2021-January 2022 by a team of local field workers recruited and trained for this purpose.

In Kampala city, in addition to a traditional PPS sample, we employed adaptive cluster sampling (ACS) (Thomson 1997; Thomson and Seber 1996) to capture a sufficient number of refugee house-holds. Using the listing of households in the initial 150 clusters in Kampala, the survey team identified those EAs that have 10 percent or more refugee households and conducted the listing of all of their neighbors. This resulted in listing additional 49 clusters that are identified as neighbors to these initial clusters. The exercise served as a basis for selection of both refugees and national households in Kampala.

In general, the sample design is a two-stage sample, with EAs first selected randomly for listing, followed by random selection of households from the listing. There is an extra third stage of choosing individuals randomly selected in households (RSI). Within each household, one person is selected at random (RSI) from the list of eligible members: persons aged between 18 and 65 years old in a national household; or refugees aged between 18 and 65 years old in a non-national household.