This section provides provisional estimates and profiles of populations who are potentially stateless or potentially at risk of statelessness in the United States. Data limitations preclude definitive estimates of the populations set forth in CMS’s profiles, much less an aggregate estimate of the US stateless population. CMS’s provisional estimates come from two main sources, the American Community Survey and refugee- and immigration-related administrative data, particularly WRAPS data. This section describes these sources of data and offers a brief summary of UNHCR mapping statelessness studies in other nations. The ACS is an annual individual-level survey conducted by the US Census Bureau. Its large sample size, thorough geographic coverage, and high response rate allow CMS researchers to provide provisional estimates and profiles of several populations that are potentially stateless or potentially at risk of statelessness in the United States. Unlike administrative data, the ACS can be used to estimate the potentially stateless or those potentially at risk of statelessness, even if they have never applied for asylum, been granted refugee status, or come to the attention of US immigration enforcement agencies. It is thus likely to include individuals who became stateless following their entry in the United States, as well as persons who entered the United States without sufficient documents. The ACS is based on a representative sample of the US population; i.e., 3.5 million households in all US Census tracts across all 50 states and the District of Columbia.44 From 2013 to 2017, each round of the survey had a response rate of 90 percent or above (US Census Bureau 2019e), ensuring broad coverage of the US population. That said, exact response rates vary from year to year. Given the likely small size of many potentially stateless or potentially at risk of statelessness populations, CMS based its provisional estimates on 5-year combined ACS data spanning 20132017 (US Census Bureau 2019, Table S1501). The 5-year microdata contain more than 15 million individual records, permitting the identification of groups with a low incidence in the overall population, including stateless groups.45 To arrive at its estimates, CMS tabulated counts of individuals whose countries of birth, ethnic backgrounds, birth years, years of arrival, and language proficiencies match the CMS profiles of persons in the United States who are potentially stateless or potentially at risk of statelessness. For example, in the case of ethnic Nepali Bhutanese, CMS counted populations born in Bhutan that arrived in the United States between 1986 and 2017 (the most recent year for which data 44 The ACS sample includes populations residing in group quarters such as nursing homes, correctional facilities, mental hospitals, college dormitories, group homes, military barracks, and shelters (US Census Bureau 2017). Anyone residing for at least two months at a given address is included in the sample. Populations that are potentially stateless or potentially at risk of statelessness that are detained in correctional facilities or that reside in shelters may be included in the ACS. 45 The 5-year PUMS data combines data from 5 years of statistically independent ACS rounds. Each one-year survey is based on a sample of 3.5 million households and actual response rates vary from year to year. The 5-year PUMS microdata include about 15.6 million observations. For more information about how the 5-year PUMS sample is assembled, see US Census Bureau,”PUMS Technical Documentation,” https://www.census.gov/programs-surveys/ acs/technical-documentation/pums/documentation.html. 41