In the Midst of Plenty. Marybeth Shinn

13 Maycock and Bretherton (2016) and the chapter by Pleace (2016) in particular describe the ways that women's homelessness in Europe is often hidden.

      14 14 A U.S. Census Bureau classification.

      15 15 This figure is based on estimates from point-in-time counts. Based on revised HMIS data standards and reporting protocols, HUD expects to start using administrative data to estimate the number of people homeless over the course of a year who have chronic patterns of homelessness.

      16 16 The way programs that serve homeless people are labeled may result in undercounts, both of people homeless on a particular night and of people homeless over the course of a year. HUD considers people homeless if they stay either in emergency shelters or in transitional housing programs, which do not provide permanent housing but instead temporary shelter for up to 2 years. There is no bright line between emergency shelters and transitional housing, either in actual lengths of stay or in the types of services provided. After some attempt to create definitions (Jill was involved), HUD gave up and accepts whatever programs choose to call themselves.With encouragement from HUD and advocacy organizations, and based on evidence that transitional housing was not effective (see Chapter 4), many communities have shifted their emphasis from transitional housing beds, whose occupants are counted as homeless, to short-term rental subsidies (“rapid re-housing”), whose occupants are deemed permanently housed. The logic of considering people in rapid re-housing programs permanently housed is that, although the subsidy is short-term, people with the rapid re-housing form of assistance live in housing they control and in which they could remain if they could find a way to pay the rent after the subsidy ends. When reporting data to HUD, communities appear to have changed the labels of some programs from transitional housing to emergency shelter (which does not affect the counts of sheltered homeless people) and the labels of other programs from transitional housing to rapid re-housing (which does). Such relabeling may account for some of the drop in sheltered homelessness—e.g., the 3,700 fewer families counted in shelters and transitional housing programs in January 2018 compared to January 2017, at the same time the inventory of transitional housing dropped by about 20,000 units (Henry, Bishop, et al., 2018; Henry, Mahathey, et al., 2018). But we don't really know, so we do not make an adjustment to the number of people using shelters over the course of a year.

      17 17 In the point-in-time counts in Nashville in 2018 and 2019, about half of unsheltered people who were interviewed said that they had used a shelter at some time during the past year (Bernard, 2019).

      18 18 Broader definitions yield larger numbers. For example, HUD counted 270,301 children aged 6 to 17 in shelters and transitional housing programs from October, 2015 to September, 2016 (Solari, Shivji, de Sousa, Watt, & Silverbush, 2017). The U.S. Department of Education counted 1,300,957 children enrolled in public schools (including preschool programs) that met its broader definition of homelessness in the 2015–2016 school year (Henry, Bishop, et al., 2018).

      19 19 The margin of error for the estimate of 7.4% was from 5.7 to 9.1%. Estimates of the percentages of people with different characteristics have much greater uncertainty, so we do not report them here.

      20 20 Of 1,012 respondents who answered this question in 1988–1991 and 1,332 who answered in 2014—a significant increase (Smith, Davern, Freese, & Hout, 2017, p. 2833).

      21 21 This detail is collected in the HMIS administrative data, but only for HUD-funded shelters and transitional housing programs, and not for the large number of shelter beds that are privately funded or funded by cities without using federal money.