April 24, 2018

Sub-Saharan African Immigrants in the U.S. Are Often More Educated Than Those in Top European Destinations


To provide comparable data across countries, population estimates for countries of residence and countries and territories of birth for sub-Saharan migrants rely on global migrant stock by their origins and destinations as estimated by the United Nations (UN) for 2015. Estimates for the number of sub-Saharan immigrants living in each destination country differ slightly from totals provided by destination governments featured in this report because of UN estimation procedures.

Demographic and economic findings, such as education, employment, time in country, sex, age, marital status and household composition, were based on two large-scale household surveys – the U.S. Census Bureau’s 2015 American Community Survey (ACS) and Eurostat’s 2015 Labor Force Survey, a data compilation of common variables from labor force surveys across European Union countries.

Survey sampling designs, populations of analysis and data access

Survey countries’ data collection methods vary. For the surveys’ complete sampling and data collection methodologies, see the ACS’s methodology website and Eurostat’s Labor Force Survey annual methodological report. All surveys are considered nationally representative.

The year 2015 was the common year for migrant population estimates from the UN as well as surveys in the U.S. and European countries. The United States and top European countries of residence for sub-Saharan immigrants (United Kingdom, France, Italy and Portugal) as estimated by the UN, were selected for analysis.

Many sub-Saharan immigrants in the U.S. and selected European countries were born in African countries where the European-surveyed country’s language is also widely spoken. Consequently, it is expected that many sub-Saharan African immigrants surveyed could respond to the surveys. Italy is an exception; in most of the origin countries for sub-Saharan immigrants, Italian is not an official language.

Survey findings presented in this report exclude the small share of respondents living in institutions (e.g., schools, hospitals, prisons). Immigrants are those born outside of the survey country of residence, even if they were citizens of the survey country when they were born.14 Native-born respondents were born in the survey country, even if their parents were foreign born. Sub-Saharan African-born respondents were born in any of the sub-Saharan African countries listed in Appendix D.15 The European data do not permit further breakdowns for specific sub-Saharan countries of birth, as Eurostat categorizes all of sub-Saharan Africa as a single group.

The American Community Survey data were accessed via Integrated Public Use Microdata Series or IPUMS-USA (University of Minnesota). Eurostat’s Labor Force Surveys was provided by Eurostat to use microdata for scientific use. Data in both surveys were anonymized to protect the identity of respondents.

Variables used in each survey

Comparable variables between the U.S. and European countries are reported. The following table lists the variables, universes, variable names and categories.16


  1. Respondents living in France’s overseas departments were considered part of France, even when these territories were in sub-Saharan Africa (for example, Reunion). One exception is Mayotte, which Eurostat considers a sub-Saharan African territory.
  2. Sudan is included as a sub-Saharan African country in this report. This differs from previous reports, in which Sudan was included as a North African country. This modification is due to the classification of sub-Saharan African countries by Eurostat. Eurostat defines “Other Africa” in the Labor Force Survey to be these countries outside of North Africa.
  3. Missing cases in the Eurostat data representing don’t knows or refusals are not reported, but were included in the overall analyses. Very low shares of responses among sub-Saharan immigrants were missing for any particular variable. But, a large share (20%) of the UK-born was missing for the education variable; thus, this estimate in particular should be treated with some caution.