interviews on the difficulties and challenges facing stateless persons in the United States.17 The semi-structured interview instrument included questions on demographic background, causes of statelessness, immigration status, migration history, conditions before and after coming to the United States, experience with US immigration authorities, and support needs. As necessary, interpreters facilitated the interviews, which took place between May and December 2018. The interviews began with a description of the project, and review and signing of the consent forms. These steps allowed CMS to underscore the benefits, risks and voluntary nature of the project, and the fact that participants could withdraw consent at any time and their identity would remain protected. After reviewing multiple immigration-related data sources, CMS selected the ACS as the most comprehensive source of data on US residents who are potentially stateless or potentially at risk of statelessness. The ACS is a nationally representative survey that gathers detailed social and economic data on approximately 3.5 million US households each year. CMS distilled its population profiles of those potentially stateless or potentially at risk of statelessness into discrete characteristics that aligned with variables contained in the ACS. These characteristics included country of birth, year of entry, ancestry, and language. It then used these (profile) characteristics to query ACS data on foreign-born US residents who had not naturalized. To supplement ACS data, CMS sought immigration and refugee-related administrative data on non-citizens who shared some of the characteristics of CMS’s stateless profiles and of those designated as “stateless” or of “unknown” nationality. In particular, it gathered publicly-available data on refugee admissions from WRAPS, an interactive computer system employed by DOS’s Refugee Processing Center that “process[es] and track[s] the movement of refugees … to the US for resettlement under the US Refugee Admissions Program” (RPC n.d.). This interactive data tool contains data on refugees’ country of birth, admission date, religion, ethnicity, age, education, and native language. These variables match (in part) the characteristics identified in CMS’s stateless profiles. The WRAPS system does not permit cross-tabulations of admitted refugees by ethnicity, nationality, language, and other characteristics. CMS researchers requested detailed crosstabulations of WRAPS data, but did not receive them. In addition, CMS compiled data—collected by TRAC—from the Executive Office for Immigration Review (EOIR) on asylum seekers and others in removal proceedings. This data includes information on custody status, relief from removal, case outcome, country of birth and case year. However, it does not include year of entry and other important characteristics identified in the profiles. In addition, CMS reviewed USCIS data on “affirmative” asylum seekers (those not in removal proceedings). It also obtained data, compiled by TRAC, of every person in immigrant detention in 2018. CMS also submitted a far more extensive data request on potentially stateless or potentially at risk of statelessness US residents that the US Department of Homeland Security (DHS) had 17 Some participating organizations self-selected due to their willingness to participate in the study and their familiarity with one or more stateless population. In addition, the project advisory groups pre-identified select geographic areas. In order to reduce this inherent bias, CMS reached out to entities with diverse missions, geographic locations, and target populations, including CBOs, legal service providers, community organizers, and places of worship. 8