Loading...
HomeMy WebLinkAbout06/28/2016 03 Homelessness in the Yakima ValleyBUSINESS OF THE CITY COUNCIL YAKIMA, WASHINGTON AGENDASTATEMENT Item No. 3. For Meeting of: June 28, 2016 ITEM TITLE: Homelessness in the Yakima Valley SUBMITTED BY: Jeff Cutter, Interim City Manager Sara Watkins, Senior City Attorney Joan Davenport, Community Development Director SUMMARY EXPLANATION: The purpose of this study session is to introduce several prominent Yakima homeless service providers and to present a summary of significant current issues related to homelessness in Yakima. The tools and strategies used by local service providers to reduce the length of time and incidence of the occurrence of homelessness will be discussed. The entire Yakima community is impacted by homelessness. Many groups and agencies are working on this broad and complex issue and this is an opportunity to hear from several of them about the roles they are playing to meet needs in our community. The 2016 "Point in Time" Survey is provided to the City Council for background information. This survey is a HUD requirement for the County programs to qualify for Federal and State homeless funds. The survey must be conducted in the last week of January each year. The report is produced by the staff at Yakima Valley Conference of Governments. ITEM BUDGETED: STRATEGIC PRIORITY: APPROVED FOR SUBMITTAL: STAFF RECOMMENDATION: 0 Neighborhood and Community Building Interim City Manager Study session and community listening opportunity requested by the City Council BOARD /COMMITTEE RECOMMENDATION: ATTACHMENTS: Description Upload Date Type 2016 Point in Time Survey 6/22/2016 Backup Material Homelessness in Yakima County 2016 Point in Time Stakeholder Report Introduction The Point in Time count, also referred to as `PIT' or simply `the count', is conducted annually throughout Yakima County to estimate the number of people experiencing homelessness on a single night in our communities. The local PIT count is part of a nationwide data collection effort required by the Department of Housing and Urban Development (HUD). Data collection for the count comes from two sources: a Sheltered Count covering the homeless population staying in housing of various types that is dedicated to serving the homeless and an Outreach Count that attempts to reach the homeless or at risk wherever they may be located within the community. The Sheltered Count is conducted with the assistance of area service providers who house and serve homeless populations. A two page survey is completed by each household engaged in housing services by specially trained data collectors. Whenever possible, case managers with existing relationships with their homeless clients are trained to complete data collection. Virtually all local housing providers participate in this count on some level, with the exception of a single transitional housing project refusing to participate in 2016. This allows reliable data collection for the homeless population that is engaged with a housing provider, and cooperation during deduplication and analysis allows for a full population count of those sheltered in participating programs. The Sheltered Count is generally composed of homeless persons staying in emergency shelters (ES), transitional housing (TH), and permanent supportive housing (PSH). The Outreach Count data collection survey is identical to the sheltered data collection tool, but does not have a defined population to count and targets the homeless who are unsheltered or otherwise scattered across our communities. Data is gathered by volunteer and professional outreach teams, either in the field, at other partner social service or mainstream agencies such as the Department of Social and Health Services, or on site at concurrent service fairs known as Project Homeless Connect events. Each field team is lead and trained by professional outreach workers and homeless or formerly homeless advocates. Field teams target known locations where the homeless congregate based on input from service providers, outreach workers, current and formerly homeless advocates, and past survey results. It is important to understand that the Outreach Count represents only a subset of the homeless not engaged by housing providers, and as a result is not directly comparable to the shelter count in many respects. The total number of homeless individuals in our county on the night of the count is certainly higher than captured by the Outreach Count, and some subpopulations are likely notably undercounted due to an avoidance of known locations, mistrust or hesitance regarding service providers, unwillingness to respond, and many other factors. Because of this the generaliz ability of the Outreach Count to the larger unsheltered and couch surfing population is imperfect. Descriptions of the Outreach Count participants can still provide insight into the characteristics of the unsheltered populations and how it may reflect or contrast with the priorities of the housing services system, but comparisons do involve a level of uncertainty that may not be easily quantifiable. Overview The total number of those identified as homeless during the 2016 PIT Count can be summarized by the number of unduplicated individuals and households. Chart 1.1 shows the number of homeless individuals counted since 2010, grouped by the type of housing in which they were counted. 500 — 400 300 200 100 2016 data shows an overall decrease of 16 %, falling below 600 0 individuals for the first time since 2010 2011 2012 2013 2014 2015 2016 2010. The largest proportional decrease came in the ES /TH Unsheltered Sheltered (ES /TH) Sheltered (PSH) literally homeless sheltered category and is believed to be related at least in part to the shift of resources into permanent supportive housing models — the only area of the count to see a year- over -year increase. Full data regarding homeless individuals may be referenced in Table 1.1 below. Chart 1.1 2016 Homeless Individuals Notable results from the current count include year -over- year decreases in individuals counted as unsheltered (down 11% from Table 1.1 Homeless Individuals 2010 2011 2012 2013 2014 2015 2016 Unsheltered 83 61 53 47 47 72 64 Sheltered (ES/TH) 424 399 472 516 486 466 348 Sheltered (PSH) 115 150 178 132 168 150 168 2015) and the 622 literally homeless sheltered population staying transitional housing placements (down 25 %). Chart 1.2 2016 Homeless Households 400 1 300 200 100 0 2013 2014 2015 2016 • Unsheltered ■ Sheltered (ES /TH) • Sheltered (PSH) 610 703 695 701 688 580 in temporary emergency shelter or A total of 435 homeless households were identified during the 2016 Point in Time count. This represents a 10% decrease from the prior year. Household data Data prior to 2013 is only available as an aggregated total, a problematic measure due to the differences in the outreach and shelter counts. Available data broken down appropriately by housing type since 2013 is included in Chart 1.2. As is generally the case, changes in the count of individuals are largely consistent at the household level. Sheltered Count The homeless in Yakima County may find housing assistance through a variety of programs and housing models. Typically, we discuss three categories of shelter provided to the homeless. Emergency shelter (ES) is intended as a short term intervention; clients are typically not expected or allowed to stay for periods longer than 90 days, generally target around a month long stay per client, and may or may not allow clients to return during a subsequent time period. Transitional housing (TH) models provide housing to the homeless for a longer period and are intended to enable those served to address the root causes of their homelessness. Housing in transitional housing models is generally available for 12 -24 months, and most homeless families served in transitional housing also receive in depth housing case management and referral to other mainstream services. Finally, permanent supportive housing (PSH) projects provide housing indefinitely to those with the most serious barriers to stable housing. Typically this housing is utilized for clients with an extensive history of homelessness and serious physical or mental health disabilities who would be projected to remain homeless indefinitely without integrated housing and supportive services. Clients served in these programs are not considered homeless by most jurisdictions or funders, but as a critical response to the hardest to serve homeless populations it has historically been included in local data. Chart 2.1— Sheltered Count Age Distribution 30 25 It 20 0 U Total Shelter Count individuals and households are available as part of the Overview data. Demographic data on the shelter count population is provided below, beginning with the age distribution in Chart 2.1. 15 Of note in the age distribution is the prominence of children, who make up 10 38% of the total shelter count It population. All of the 5 most 5 frequently reported ages are children, under the age 0 �d Jd I I III I III ■ of t5.4 Th se seemsetog indicate that 0 10 20 30 40 50 60 70 80 families with children are being Age targeted for housing interventions, particularly households with very young children. Counts of households served by family type shows that approximately a third of shelter count households and nearly 60% of total shelter count individuals were part of a family with children. Chart 2.2 on the following page details the reported race of individuals counted in the 2016 Sheltered Count. Because individuals may consider themselves to be of more than one race, this is not equal to the unduplicated number of individuals counted. Chart 2.2 - Sheltered Count Individuals by Racial Identification ■ 2016 ■ 2015 White American Indian /Alaska Native Refused Black/African American Nat. Hawaiin /Pacific Islander I Asian 0 100 Gender data shows that 270 individuals identified as female, 246 as male, and no individuals identified as transgendered. In absolute terms this is a very minor shift, but does put females in the majority which was not the case in 2015. 200 300 400 2015 data is also included for comparison of year- over -year changes, which show small shifts but little variation of the overall pattern, with the largest segment continuing to identify as white by a substantial margin. Note that clients who identified with none of the available racial options were recorded as `Refused'; of the 68 refused cases, 61 (or 90 %) identified as being of Hispanic ethnicity. Full data regarding reported ethnicity since 2013 is available below in Table 2.1. Table 2.1 - Ethnicity of Sheltered Count Individuals Ethnicitv 2013 2014 2015 2016 Hispanic 245 255 234 211 Not Hispanic 393 385 379 288 Refused 10 14 3 17 648 654 616 516 In addition to demographic markers, data is also collected on geographic location, frequency and duration of homeless episodes, and background information such as reported causes of homeless, service needs, and income resources. Table 2.2 illustrates the location of shelter count participants on the night of the count. This shows the vast majority of individuals, over 85 %, staying within the city of Yakima on the night of the count. This is largely determined by the allocation of housing services, and as would be expected changed very little; prior counts showed 83% and 86% in 2014 and 2015, respectively. Chart 2.3 on the following page illustrates the duration of current homeless episode for shelter count individuals counted in transitional housing and emergency shelter; Table 2.2 — Sheltered Count T"'ChrichiclIs by Lnrrrti,nn, City 2015 Yakima 432 Wapato 35 Toppenish 22 Granger 9 Sunn side 9 Grandview 6 Selah 3 84% 7% 4% 2% 2% 1% 1% permanent supportive housing has been omitted, since it is intended to be of indefinite duration by design. Here, and in general throughout this report unless otherwise noted, color has been used to designate the smallest number of categories to encompass a majority of responses. It is important to note that duration of homelessness includes not just the time spent in a housing program, but also the (sometimes substantial) length of time spent homeless and unsheltered or couch surfing prior to entry into a housing service. In spite of this factor, nearly a third of the emergency shelter and transitional housing population has been homeless for less than 6 months (31 %). Unlike prior years, more than half of those counted as part of the outreach count has been homeless for a year or more (55 %). Chart 2.3 — Sheltered Count Individuals by Duration of Homelessness Chart 2.4 summarizes participants in the 1 month or less ■ 1 -6 months ■ 6 mos - 1 year shelter count by the number of homeless . 1 -2 years ■ 2 years or more episodes they reported within the past 3 years; those continuously homeless over that period recorded only a single episode. Nearly two thirds of those surveyed (65 %) had experienced only a single episode of homelessness during the relevant period. Chart 2.4 — Shelter Count Individuals By Number of Homeless Episodes 300 200 ioo social security Dental care shows an almost Mental Health Care identical drop of 60 %. While it is Child Care Dental difficult to provide a complete 0 Ism Clothes /Blankets this reduction in demand for health Food services is tied to the expansion of Education health insurance through the Job Training /Placement Transportation Affordable Care Act and the 0 Participants were also asked about the top needs of their household, aside from housing, and directed to select up to 5 responses. The top ten most frequently selected additional service needs are summarized in Chart 2.5 below. Comparisons from 2015 Point in Time data have been included for reference. The most frequently selected household needs have remained similar over time, with a notable exception in the area of health and dental care needs. Reductions in reported need for dental and health care services were first reported in 2015 and continued in the 2016 survey. Since 2013, when 37% of households in the sheltered count reported a healthcare need, the prevalence has dropped by 59 %; in 2016 only 15% of Shelter Count Chart 2.5 — Shelter Count Households Top 10 Reported Needs households reported such a need. social security Dental care shows an almost Mental Health Care identical drop of 60 %. While it is Child Care Dental difficult to provide a complete Health Care explanation, providers feel it is likely Clothes /Blankets this reduction in demand for health Food services is tied to the expansion of Education health insurance through the Job Training /Placement Transportation Affordable Care Act and the 0 ■ 2014 50 2015 100 150 a expansion of health care options provided specifically for homeless clients locally. Chart 2.6 describes the number of sheltered count households indicating various causes of their homelessness. Again, households were allowed to provide multiple responses but were limited to the five selections they felt were most relevant to causing their homelessness. While a large number of Chart 2.6 — Shelter Count Households options were available, the Reported Causes of Homelessness majority of responses fell into just four categories, as Alcohol /drug use i 11 a stra to d i n the chart. Unable to pay rent /mortgage Family crisis/break -up Job loss Temp. living situation ended Domestic violence Medical problems Refused Mental illness Evicted (other reasons) Lack of job skills Evicted (non - payment) Poor credit rating Medical costs Discharged from institution /jail Convicted of felony Convicted of midemeanor Failed job drug screen Language barrier Lack of child care Aged out of foster care ■ 0 25 50 75 Chart 2.7— Shelter Count Households Reported Sources of Income Social security benefits None TANF Other Public Assistance Per capita Part time work Low wage job Refused Child Support Day laborer work Relatives /friends Pension from former job Alimony /spousal support Panhandling 0 25 50 75 These four primary causes account for more than half of all responses. Two of the top four causes, accounting for more than a quarter of all reported causes, relate solely to economic conditions of the household. Another, a family break up, is also often associated with economic distress. This is clearly reflected in the data on household 100 income sources (summarized in Chart 2.7) which show the majority of participating households indicating either no income whatsoever, or what is typically very low income from public benefits. Combined, these account for nearly two thirds of all reported income sources. 100 Outreach Count The Outreach Count is conducted by community volunteers, professional outreach workers and case managers, homeless and formerly homeless advocates, and local homeless and mainstream service providers. In addition to those literally homeless (sleeping outside, in vehicles, or in other places not suitable for human habitation) the Outreach Count also collects some data regarding the number of households who are temporarily staying with family or friends due to housing need. This segment of the population is often referred to as `couch surfing', and data for this group will be presented separately as a distinct subpopulation. Data is collected via survey; this restricts the sample to those who can be located by surveyors, are able to consent to participate (which means minors cannot complete the survey for their household), and are willing to respond. When reviewing the resulting data, it is important to understand that unlike the Shelter Count, the Outreach Count cannot reach its full target population. The numbers reported here represent some subset of the unsheltered homeless population. Estimates are frequently based on the idea that for each homeless person counted two are missed, and the disparity is likely to be larger for some subsets of the homeless population. Specifically homeless families and unaccompanied youth, who typically avoid known locations where the adult homeless population congregates and are frequently reluctant to self- identify as homeless, are likely to be even further undercounted. Note that because the Sheltered Count captures a picture of a full homeless population (those sheltered in housing programs) while the Outreach Count captures a non - random subset of the homeless population not receiving housing support, the two counts are not directly comparable, and the generalizability of the Outreach Count to the larger unsheltered and couch surfing population is imperfect. Descriptions of the Outreach Count participants can still provide insight into the characteristics of the unsheltered populations and how it may reflect or contrast with the priorities of the housing services system, but comparisons do involve a level of uncertainty that may not be easily quantifiable. Review of the 2016 data begins with a demographic overview, specifically the age distribution presented in Chart 3.1. Notice that in contrast to the Sheltered Count, children do not make up a significant portion of participants; children make up only It 5% of those counted, and none of the top ten most frequently observed U° ages are under 18. This likely indicates an over prioritization of t families with children within the housing service system, but is also ~ almost certainly influenced by the systemic undercount of homeless families mentioned above. Chart 3.1— Outreach Count Age Distribution 6 4 2 - 0 LR I II I I I 0 10 20 30 40 50 60 70 7 Chart 3.2 presents the reported race of individuals counted as part of the 2015 Count. As a reminder, participants can identify as members of more than one racial group, and responded with `Refused' if they identified with none of the available options (of those who selected Refused over 70% identified as being of Hispanic ethnicity). Full ethnicity data is available in Table 3.1 below. Chart 3.2 - Outreach Count Individuals by Racial Identification White American Indian /Alaska Native Refused Black /African American Nat. Hawaiin/Pacific The Outreach Count racial Islander demographics have tended to be more volatile than the sheltered 0 10 20 30 count, and that remains true in ■ 2016 ■ 2015 2016. This is closely tied to the variable success of community- specific Project Homeless Connect events. In the 2015 report, it was clear that the increase in Native Americans counted was tied to greater participation and leadership on the part of the Yakama Nation and that effect continues in 2016 with Native Americans remaining the second most populous group within the unsheltered population. Table 3.1 - Ethnicity of Outreach Count Individuals 2013 2014 2015 2016 Hispanic 6 15 12 15 Not Hispanic 40 31 55 43 Refused 1 1 5 6 TOTAL 47 47 72 64 Table 3.2 - Gender of Outreach Count Individuals Gender 2013 2014 2015 2016 Female 13 15 30 20 Male 34 32 42 44 Transgender 0 0 0 0 TOTAL 47 47 72 64 Table 3.2 presents the Outreach Count participant gender rates since 2013. The overall gender distribution shows very low variance, with the percentage of those counted identifying as female changing by only 1.3 percentage points. Table 3.3 details the location of the participants counted. Note that one response confirmed a location in Yakima County but refused to specific a community. While this data is not necessarily an exact reflection of the overall geographic distribution of the larger homeless population, it does contrast starkly with the Sheltered Count Table 3.3 — Outreach Count Individuals by Location City 2016 Yakima 33 Wapato 11 Toppenish 9 Sunn side 7 Buena 1 Grandview 1 Gran er 1 Refused 1 geographic distribution. Based on the allocation of housing resources, 84% of the sheltered homeless were counted within the city of Yakima. 52% 17% 14% 11% 2% 2% 2% 2% 8_7 Chart 3.3 - Geographic Distribution of Unsheltered Individuals 100% 75% 50% 25% 0% 2013 2014 2015 2016 ■ Yakima ■ Yakama Nation ■ Sunnyside ■ Other Chart 3.4- Outreach Count Individuals by Number of Homeless Episodes 50 40 30 20 10 0 — 1 2 3 4 5 6 7 8 910+ Chart 3.5 - Outreach Count Individuals by Duration of Homelessness ■ 1 month or less ■ 1 -6 months ■ 6 mos - 1 year ■ 1 -2 years ■ 2 years or more However, the total distribution of unsheltered homelessness seems to show two major populations within the county; 52% of the unsheltered population was counted in Yakima, as mentioned above, and another 31% within the Yakama Nation in Toppenish and Wapato. These two populations have tended to dominate the geographic distribution of the unsheltered count year to year, and this data is summarized fully in Chart 3.3. Note that unlike most charts in this report, both the Yakima and Yakima Nation data are in color despite the city of Yakima constituting a majority of the unsheltered population in most years. Of interest this year is the possible presence of a third distinct unsheltered population in Sunnyside, which exceeds the rest of the `Other' areas for the first time since 2013 and makes up 11% of the total unsheltered count. It is unclear if this is a new development, a better picture provided by the improved count tied to the Sunnyside Project Homeless event, or a single year anomaly on the data, but it bears watching in subsequent counts and could potentially impact the distribution of housing resources if additional data does reinforce the existence of a third distinct location for the unsheltered population. Chart 3.4 illustrates the number of reported instances of homelessness within the past three years for outreach count individuals. Single instances of homelessness constitute a majority or responses, and have accounted for the majority of responses every year since 2013 to varying degrees. In 2016 nearly three quarters of all outreach count participants had been homeless only once in the past three years. Note that this includes those who have been continuously homeless for the entire three year period, which does account for a majority of those single reported episodes. Chart 3.5 shows the duration of homelessness for outreach count participants. The majority of individuals surveyed as part of the outreach count were homeless for a year or more. This distribution is largely consistent with the shelter count data, but does show a bias towards very long duration of homelessness. Individuals in the outreach count were more than twice as likely to have been homeless for 3 years or more as a proportion of those counted than those included in the sheltered count. Taken together, this data on duration and recurrence shows an unsheltered population that is very heavily composed of individuals with a long, and frequently uninterrupted, history of homelessness. Chart 3.6 compares the top ten reported needs of households participating in the outreach count. As has commonly been the case in the outreach count, many of the most common responses dealt with meeting basic needs such as food, clothing, and transportation. More than half of all outreach count households (56 %) reported needing assistance with food, and 53% requested help with clothing and blankets. In the 2015 report, the Outreach Count households lagged significantly behind those in the Shelter Count in increased access to health and dental services. While this gap has not disappeared in the 2016 count, it has narrowed significantly, with the proportion of outreach count households requesting healthcare services down 41% from 2013 numbers and dental service requests down 25 %. While this is an improvement, access to these services still lags the reports from the shelter count. Chart 3.7 presents the causes of homelessness reported by households participating in the outreach count. The top four Chart 3.6 — Outreach Count Households Top 10 Reported Needs Counseling Education Mental Health Care Other Health Care Dental Job Training /Placement Transportation Clothes /Blankets Food IM 0 10 20 30 0 201 02015 Chart 3.7— Outreach Count Households by Reported Causes of Homelessness Family crisis/break -up Job loss Alcohol /drug use Unable to pay rent /mortgage Domestic violence Medical problems Poor credit rating Refused Temp. living situation ended Convicted of felony Evicted (other reasons) Discharged from institution /jail Mental illness Lack of job skills Failed job drug screen Medical costs Language barrier Convicted of midemeanor Evicted (non - payment) 0 5 10 15 10 Tor 20 selections represent the majority of all responses, as is common across almost all sub - groupings, and indeed are the same categories that make up the majority of responses in the sheltered count. Chart 3.8 illustrates reported income sources. As is historically the case, `None' (le, being completely without income) is the most common response for households participating in the outreach count with 44% of households reporting no income from any source. This has been the most commonly source of reported income in every year with full data available. No income together with per Chart 3.8 — Outreach Count Households capita income, available to Reported Income Sources some Native American households including None members of the Yakama Per eapita Nation, constitutes a Social security benefits majority of the responses. Other Public Assistance Households counted as part Panhandling of the Outreach Count are L &I /Workman's Comp nearly twice as likely to Private disability insurance report having no source of Relatives /friends income as those counted Unemployment insurance within housing programs. Day laborer work Pension from former job 0 10 20 11 4 30 Homeless Sub - Populations & Addenda In addition to the overall totals reflecting the Outreach and Sheltered counts, data on specific sub groups may be useful in decision making. This portion of the report will provide some summary of the various subgroups across both the sheltered and outreach counts. Note that this is not necessarily representative or generalizable to the entire homeless population or larger relevant subgroups than the data set itself, because the combination of the sheltered and outreach counts is almost certainly not a representative sample of the overall homeless population. Chronically Homeless HUD defines a Chronically Homeless Individual as a homeless adult who meets all of the following criteria: 1) Is currently staying in an emergency shelter or an unsheltered state (outside, in a vehicle, or other locations not intended for habitation). 2) Has been homeless continuously for at least one year OR has experienced at least four homeless episodes within the past three years totaling at least one year in combined duration 3) Has a qualifying permanent disability that substantially impacts their ability to gain and maintain stable housing. Households of more than one person who include at least one chronically homeless adult are referred to as `Chronically homeless families'; for the purposes of this report, Chronically Homeless Individuals and individuals who are part of Chronically Homeless Families are considered together unless otherwise noted. In 2016, HUD expanded the qualification for repeated instances of homelessness to include a total combined duration minimum of one year. In 2016 there were no households who would have qualified under prior definitions that did not under the new rule, so comparison to prior years excludes this new filter. In 2016 a total of 72 individuals were identified representing 11% of those counted, down 3 percentage points from 2015. A breakdown of individuals by chronic homelessness status is available in Table CHI Table CHI -All Individuals by Chronic Homelessness Status and summarized in Chart 2013 2014 2015 2016 CHL Not Chronically Homeless 624 615 599 508 Chronically Homeless Individual 69 80 74 70 Chronically Homeless Family 2 6 15 2 TOTAL 695 701 688 580 Chart CHI —All Individuals by Chronic Homelessness Status 2016 2015 2014 2013 0 100 200 300 400 500 600 700 ■ Not Chronically Homeless ■ Chronically Homeless Individual ■ Chronically Homeless Family In 2016 roughly 12% of all individuals counted were part of a chronically homeless household, with the vast majority single adults. This is typical for the data collected since 2013, with most chronic homelessness concentrated in single adult households every year and the prevalence of chronic homeless consistently falling between 10 -12% of the overall homeless population. 12 While the portion of the overall count qualifying as chronically homeless has remained low and quite consistent, note that a large portion of those counted cannot possibly be chronically homeless due simply to the type of housing in which they are counted. As a result there has been some speculation that it might be instructive to look at the rate of chronic homelessness among those with a housing type that could potentially be chronically homeless (i.e., those in shelters, transitional housing, or unsheltered, sometimes referred to as the literally homeless). Chart CH2 shows the rate of chronic homelessness among the literally homeless population reported at Point in Time annually since 2008. While the increased incidence is not surprising, the high variability is not necessarily expected - variance in the rate of homelessness among the literally homeless ranges from only 9% in 2010 to 22% in 2012. No immediate explanation for the increased variability of chronic homelessness in this sub- group is entirely convincing. Indeed, it's possible the larger range is simply a product of the smaller size of the literally homeless group; in general, the literally homeless make up roughly two thirds of the total in any given year since 2013. Chart CH2 — Incidence of Chronic Homelessness Among the Literally Homeless 20% — 15% 10% 5% 0% °� ° N° N° NZ^ ti°Nry ti°N� ti°N� tiZN leb N°N° Chart CH3 shows the location of last permanent residence of those counted as chronically homeless, including individuals in chronically homeless families. This is used as a proxy for a point of origin, and corresponds to the last location the responding household lived when they were NOT homeless. This is an imperfect method, but does provide an estimate regarding origin. In 2016 90% of the chronically homeless indicated that their last permanent address was within Yakima County. All recorded rates are over 80% with a local origin. A more detailed look at point of origin across sub - populations is also included separately below. Chart CH3 — Chronically Homeless by Location of Last Permanent Housing ■ In Yakima County ■ Not in Yakima County As discussed in more detail in the section dedicated to origin, must of the interest is also tied to beliefs regarding institutional utilization. That is of particular relevance among the chronically homeless population, a group that is frequently associated with very high demand for mainstream institutional support via everything from emergency room use to jail bed nights. Chart CH4 on the following page shows the institutional releases reported by each chronically homeless household. Households could select multiple release types, unless they specified `None' Chart CH4 - Chronically Homeless Households Reported Institutional Releases None or refused to respond. In 2016 the majority of chronically homeless households, 64 %, reported no institutional utilization. Jail /prison On the surface this seems to Refused contradict conventional wisdom regarding the high institutional Hospital demands of the chronically homeless. Substance abuse treatment It is important to point out, therefore, that this data does not encapsulate Phsychiatric hospital I usage rates — a chronically homeless individual with a single hospital 0 10 20 30 40 5o admission is indistinguishable from one who was seen and admitted a dozen times. This is significant, since available data tends to show a minority of households driving the majority of interactions; this is true in general, and has been born out locally in other research, such as the Winter Shelter project that sees a small number of daily utilizers driving a disproportionate share of the demand for bed nights. In other words, it's likely that among the minority of chronically homeless individuals who do report institutional involvement there are a small handful of very heavy service utilizers. It is not possible to verify the existence of such an effect within this area using the Point in Time data, but it is consistent with the data that is available and other local experience. The chronically homeless population is generally older than the general homeless population counted as part of the 2016 Point in Time survey, with none of those counted being children. Chart CH5 shows the age distribution of the chronically homeless relative to the general homeless population surveyed in 2016. The chronic homeless population includes no children in the 2016 count, and children in chronically homeless households peaked in 2015 at roughly 5% of those chronically homeless. By comparison, children make up nearly a third of all persons counted in the combined count. Chart CH5 Age Distribution 30% — 20% 10% 0% ti� ti� Nti N� b �b y� 6 ■ CH ■ All As might be expected, this implies that chronically homeless individuals are far more likely to be adults, especially older adults — an individual in a chronically homeless household is two and a half time as likely to be over the age of 55 as a random individual from the general count. 14 Veterans Homeless veterans are often a focal point for communities, and have been targeted recently by several HUD and VA initiatives meant to end unsheltered homelessness among veterans. 26 participants self - identified as veterans during the 2016 count across both the sheltered and outreach surveys. Total for adults by veteran status since are available in Table VL Chart V1 -Homeless Veterans by Housing Type 30 - 20 10 n L L 2013 2014 2015 2016 • Sheltered (ES /TH) ■ Sheltered (PSH) • Unsheltered Table V1 - Homeless Adults by Veteran Status 2013 2014 2015 2016 Veteran 36 44 35 26 Not a Veteran 404 412 412 350 Refused 4 0 5 6 TOTAL 444 456 452 382 It is important for this total to point out that one program dedicated to providing transitional housing to homeless veterans refused to participate for the first time in 2016. Chart Vl shows veterans by the type of housing veterans were staying in at the time of the count; the decrease in the ES /TH sheltered category can be almost entirely explained by the lack of data from this program, making year over year comparisons for veterans problematic at best. Additionally, many interventions specific to veterans are provided via housing vouchers, in which homeless veterans hold their own lease. These vouchers are not counted as part of the Point in Time survey. Chart V2- Homeless Veterans Many services available to veterans are accessed Veteran Benefit Rates through veteran specific providers rather than traditional housing providers. As a result, this report 40% has typically tracked the engagement with these 30% veteran specific resources by asking homeless veterans if they receive any veteran's benefits. Chart V2 shows 2o�% o the rate at which veterans have been receiving benefits 10% ' since 2010; data from before 2013 is taken from the 0% 2012 report. oti° otiti oti oti oti` oti' oti° ti ti ti ti N N N Access to benefits dropped sharply after 2010, when 42% of veteran counted were receiving some kind of veteran benefit, and continued to decrease steadily through 2013. Although this did improve in 2014, it has consistently decreased since. This may be because those who are closely tied to veteran's services are able to receive assistance through housing vouchers not captured here to exit homelessness entirely. Gender data is included in Table V3; as has historically been the case, veteran gender distribution skews starkly towards males. No transgender veterans have been counted to date. Table V3 - Veteran by Gender 2013 2014 2015 2016 Female 1 3 3 4 Male 35 41 32 22 TOTAL 36 44 35 26 16 C' Point of Orimin One of the common questions from decision makers regards the location of origin of the local homeless population. This is not directly asked on the standard survey data collection tool, but in recent reporting cycles the point of origin has been estimated using the location of last permanent housing as a proxy. This is not a perfect analog; a lifelong resident of the area who moved away for employment or another reason might very reasonably return to the area to connect with informal support networks such as family if falling upon hard times. However, these exceptions are in some sense edge cases, and the location of last permanent housing will provide the best estimates available regarding the location of origin for the survey group until any changes to the survey can be incorporated in the next cycle. Chart 01 - Homeless Individuals by Point of Origin 2016 2015 2014 2013 0% 20% 40% 60% 80% 100% ■ In Yakima County Not in Yakima County ■ DK/Refused For the 2016 year, 93% of the participants in the count reported a last permanent address that was within Yakima County (Chart 01). This is not unusual when looking at the historical data. Since 2013, more than 85% of those surveyed have listed an origin within the county every year, and this has remained very stable. Rates of local origin range from 86 -93% during this period. Discussion about a hypothetical out of area origin for the homeless population often involves a parallel discussion about what would attract homeless individuals to the area. This often takes the form of postulating that perhaps local homelessness is driven by out of area homeless individuals being released locally from institutions (notably prisons and treatment facilities) into the community. However, nearly three quarters (74 %) of the households counted that did show an out of area origin reported no exits from institutional facilities. See Chart 02 for a full breakdown. In short, accessing these services does not seem to be in any way driving the population of households with a last permanent address outside the area observed as part of the count. Chart 02 - Institutional Releases for Out of Area Households None Hospital Jail /prison Refused Phsychiatric hospital Work release Foster care Substance abuse treatment 0% 25% 50% 75% 100% Available data is also not supportive of any claim that demand for these mainstream services among the homeless is attributable to the presence of homeless households from outside the local area. If households that reported no institutional involvement whatsoever are discarded, we can compare institutional usage between the local and out of area groups. This 17 ,:0�;. data is presented in Chart 03; Chart 03 - Institutional Utilization by Household Origin* 93% of reported institutional involvement by homeless Substance abuse treatment households comes from those Hospital with a local origin. Jail /prison Refused - Phsyehiatrie hospital Foster care Work release 0 10 20 30 ■ Local Households ■ Out of Area Housheolds *Households with no institutional utilization omitted 18 CW Substance Abuse The prevalence of substance abuse issues among homeless populations is frequently a topic of discussion, often a discussion based around stereotype. As shown earlier in this report, substance abuse is generally among the most commonly cited causes of homeless episodes by households who participate in the survey, but this can be somewhat misleading. Historically, although it is indeed one of the most commonly cited causes, substance abuse is still cited as a primary cause of homeless by a minority of the households involved in the count. In the current data, only 30% of households identified drug or alcohol abuse as a primary cause of their homelessness, and since 2013 the value has not exceeded 36 %. Chart SA1 - Proportion of Homeless Adults Reporting a Substance Abuse Disabling Condition 2016 2015 2014 2013 0% 25% 50% 75% 100% • Individuals with Substance Abuse Disability • Individuals with no Substance Abuse Disability Chart SA2 - Proportion of Homeless Individuals Reporting a Substance Abuse Disability, by Housing Type 100% 75% 50% - 25% 11 .E. 2013 2014 2015 2016 ■ Sheltered (PSI) ■ Sheltered (ES/TH) ■ Unsheltered Data collection surveys also ask all individuals about their disability status, including an option for reporting a disabling drug or alcohol abuse condition. Reports of substance abuse by this measure also represent a minority of participating homeless adults (see Chart SA1). The proportion of adults reporting a substance abuse disability in 2016 was 10 %, the lowest level on record. The data also does not seem to show a consistent type of housing in which substance abuse is more prevalent. Although there is occasionally speculation that substance users are excluded from, or alternatively exclusively make up the population of, a given type of homeless housing the proportion of homeless individuals reporting a substance abuse disability does not seem to be consistent by housing type. Some years, notably 2013, show a particularly wide variance in the incidence of substance abuse disabilities (with the highest recorded rate being four times the lowest), while in 2014 the rate is relatively similar across all housing types. This would seem to indicate that type of housing is not predictive of substance abuse status. Also of note, substance abuse disabilities remain a minority in all types of housing and have decreased very dramatically in permanent supportive housing programs. This is likely tied 19 r to the broadening of the PSH model to new sub - populations. In 2013 nearly all of the PSH beds covered by the survey were provided to clients in recovery from substance abuse, many with long term sobriety requirements. Chart SA3 — Reported Causes for Individuals with a Substance Abuse Disability Alcohol /drug use Family crisis /break -up Mental illness Medical problems - Job loss Unable to a rent /mort a e Perhaps unsurprisingly, alcohol and p y g g Refused � drug abuse is the leading reported Discharged from institution /jail cause of homelessness for individuals Domestic violence with a Substance abuse disability. Convicted of felony Temp. living situation ended Chart SA3 details the reported needs lack of job skills for individuals with substance abuse Evicted (other reasons) disabilities. As comparison of the Medical costs Poor credit rating totals might indicate, however, alcohol Convicted of midemeanor or drug use 1S not universally cited as Evicted (non - payment) a primary cause of homelessness by 0 10 20 this population. This is of particular interest because the reported cause is necessarily a prior event to current state at the time of data collection, potentially supporting the idea that for at least a Chart SA4 - Rate of Substance Abuse Cause of subset of homeless substance abusers Homelessness their substance abuse is symptomatic 100% of their homelessness rather than a 75% causal factor. 50% Chart SA4 shows this in greater detail. 25% Note that the reported rate has 0% remained stable for the general 2013 2014 2015 2016 population, sitting consistently around ■Adults w /Substance Abuse Disabilitv ■All Adults 33 %. However, reports of substance abuse as a causal factor have decreased generally over time for the group of individuals reporting a disabling substance abuse disorder. Adults who identified as having a substance abuse disability were also far more likely to Chart SA5 - Rate of Substance Abuse Cause of Homelessness report a mental health disability than other participating homeless 100% individuals. Chart SA5 shows the 75% relative rate of mental health disability between the two groups for 50% data since 2013; adults with a substance abuse disability have been 25% more than twice as likely to have a mental health disability as those 0% SA Non- SA Non- SA Non- SA Non - without in every year with data sA SA SA SA available. 2013 1 2014 2015 2016 ■ MH Disabilitv No Mental Health Disabilitv While this does not directly support the idea that homeless substance abusers are self - medicating untreated mental health issues, it is certainly the case that mental health issues are much more prevalent among substance abusers within the available dataset. Families with Children Families with children (sometimes abbreviated FWC) make up a substantial portion of the overall count, as shown in Chart Fl. Individuals in such households have made up a small majority of those counted every year since 2013. 53 -55% of all individuals counted each year have been a part of families with children; this stability is probably tied to the overwhelming majority of families with children being counted within the sheltered count. Individuals in families with children tend to skew younger on average than the general population, which should come as no surprise given that the group is defined by the presence of children. The full age distribution is presented in Chart F2. Of more interest is the distribution of very young children, defined as those age five or under. Households in at -risk groups with such a young child have sometimes been linked to a higher risk of homelessness. Prevalence of such households counted as homeless is illustrated in Chart F3, and have represented a majority of households with children counted in every year since 2013. 2016 2015 2014 2013 750 600 450 300 150 0 Chart F1 — Homeless Individuals by Household Type 2013 2014 2015 2016 ■ Family w /Children ■ Adults Only ■ Child Only 30 20 10 0 Chart F2 — FWC Individuals by Age 1.1 ... I . . 0 10 20 30 40 50 60 Chart F3 — FWC Households by Presence of Very Young Children 0% 25% 50% 75% 100% ■ Young Child ■ No Young Child Families with young children were even more likely than the general population to show a local origin. For 2016, 98% of individuals in families with children had a last permanent residence within Yakima County, compared with 93% of the general homeless population. Full origin data in presented in Chart F4. 22 r 2016 2015 2014 2013 Chart F4 — FWC Individuals by Origin 0% 25% 50% 75% 100% ■ In Yakima County ■ Not in Yakima County ■ DK/Refused Income source data for FWC households shows a smaller range of reported income types than for most subpopulations. Full data is presented in Chart F5. The condensed distribution is largely attributable to TANF (Temporary Assistance for Needy Families), which nearly 40% of all FWC households reported as an income source. Data for TANF income benefit rates for FWC households is detailed in Chart F6. Regrettably data at this level goes back only to 2013, after a fiscally motivated administrative rule change rendered large numbers of families unable to receive benefits. As a result the most useful comparison is not available, and TANF income rates for FWC households have remained fairly stable over the period with available data ranging from 40 -55 %. Chart F6 - FWC Households by Chart F5 — FWC Households by Income Source TANFBenefits TANF 100% Social security... Part time work I 75% — Low wage job Child Support 50% None Per capita Refused 25% Relatives /friends Alimony /spousal... 0% 0 10 20 30 40 2013 2014 2015 2016 ■ Receiving TANF ■ Not Receiving TANF As a result of the TANF benefits available specifically for (some) homeless families with children, these households are much less likely to report having no income. Notice that unlike the general trend for most groups considered in this report, `None' is not listed among the income sources comprising a majority of responses in Chart F5. Chart F6 — Households with No Income by Household Type Adults Only Family w /Children 0% 25% 50% 75% 100% ■ No Income ■ At Least 1 Income Source Households without children are four times more likely to report having no source of income, as illustrated in Chart F6. 23 FWC households receiving TANF tend to be counted in transitional housing; TH placements have accounted for more than two thirds of TANF recipients in all years on record. Full data is presented in Chart F7. Chart F5 - TANF Households by Housing Type 50 - 40 30 20 - 10 -� 0 .= 2013 2014 2015 ■ ES ■ PSH ■ TH 2016 Unstably Housed & At- Risk /Couch Surfing' Data collection surveys allow respondents to indicate that they are homeless and staying temporarily with friends or family, a situation commonly referred to as `couch surfing'. This housing type is not generally recognized as homeless by most funders, but more importantly is exceedingly unlikely to be a true representation of the couch surfing population - it requires both self - identification and engagement with homeless service fairs or participating providers in most cases. It is included largely as a data quality measure, since these requirements mean the subset of data collected is almost certainly not generalizable to the larger at -risk, couch surfing population. Data presented here describes only those participants who provided data, and should not be taken to be representative of the larger unknown couch surfing group. Chart CS] — Couch Surfing Individuals 250 200 150 100 50 2013 2014 2015 2016 Chart CS2 — Couch Surfing Households by Household Type 150 100 2013 2014 2015 2016 ■ Adults Only ■ Family w /Children Child Only Total numbers of couch surfing individuals counted annually since 2013 are available in Chart CS]. This figure has been volatile over the history of the count, as it is very sensitive to external factors such as the success of local Project Homeless Connect service fairs, community resources and the number of home visitors and case managers able to conduct the count with known households. Chart CS2 shows total households over the same period broken out by household composition. While couch surfing is often associated with families with children, households with no children have made up the majority of households counted every year since 2013 and accounted for 79% of those who reported couch surfing during the 2016 count. However, families with children still make up a significant portion (46 %) of all individuals counted, and children account for nearly a quarter of the total. A full age distribution for couch surfers is available in Chart CS3 — Couch Surfing Individuals Age Distribution 1 111 10 20 30 40 50 60 70 80 25 Chart CS3. Of the 5 most commonly reported ages, only one is a child; three represent youth over the age of 18 but under 24. Race data for couch surfers is presented in Chart CS4. The largest single group is those who refused to respond; 97% of these individuals identified as being of Hispanic ethnicity. Together this group of Hispanic identified refusals and Native Americans encompassed 75% of all respondents. Chart CS4 - Couch Surfing Individuals by Race Refused American Indian /Alaska Native White Nat. Hawaiin /Pacific Islander Black/African American Asian ■ ■ 0 25 50 75 100 ■ 2015 ■ 2016 Reported needs for couch surfing households are presented in Chart CS5. As is generally the case, the top responses are related to basic needs. Half of all couch surfing households reported a need for food assistance. Chart CS5 - Couch Surfing Households by Needs Social Security Counseling Other Health Care Education Dental Transportation Clothes /Blankets Job Training /Placement Food 0 25 50 75 ■ 2015 ■ 2016 Chart CS6 - Couch Surfing Households by Cause of Homelessness Family crisis/break -up Job loss Unable to pay rent /mortgage Alcohol /drug use Temp. living situation ended Poor credit rating Refused Medical problems Evicted (other reasons) Lack of job skills Domestic violence Convicted of felony Lack of child care Mental illness Evicted (non - payment) Discharged from.. Medical costs Failed job drug screen Language barrier Aged out of foster care Convicted of midemeanor 0 10 20 30 40 50 Reported causes of homelessness are included in Chart CS6. The top four causes are consistent with other populations in the count, and constitute a majority of responses. Notice also that the first response after the typical causes refers to the end of a temporary living situation similar to that in which these households were counted. Initially this appears redundant — households are couch surfing because they were couch surfing at some earlier date. However, this could be taken simply as an indication that these households have had prolonged periods of housing instability and even when they did not consider themselves homeless were still under housed and cohabitating. Indeed, the data on duration of homelessness for couch surfing households shows that the single largest group, 36% of those surveyed, has been homeless for two years or more. Full details on the duration of homelessness for couch surfing households can be found in Chart CS7 on the following page. More than half of all couch surfing households had been homeless for a year or more. Chart CS7- Couch Surfing Households by Duration of Homelessness ■ 1 month or less ■ 1 -6 months ■ 6 mos - 1 year ■ 1 -2 years Chart CS8 - Couch Surfing Individuals by Origin 100% 75% 50% 25% 0% 2013 2014 2015 2016 • DK/Refused ■ Not in Yakima County • In Yakima County As expected, most couch surfers have an origin within Yakima County. In 2016 77% of couch surfing individuals reported a local origin, and this is actually a historic low, with the proportion trending down slightly since 2014. Full data is presented in Chart CS8. 27 For more information on this report contact: Avery Zoglman Yakima Valley Conference of Governments 311 North 4th Street, Suite 204 Yakima, WA 98901 Office: 509 - 574 -1550 avery.zoglman @yvcog.org