Loading...
HomeMy WebLinkAboutach-homelessness-prevention-framework-appendix-d-0113251 Homelessness Prevention Framework — Appendix E Targeting requires identifying the factors that create the greatest likelihood that a household in crisis will become literally homelessness, and then prioritizing available homelessness prevention services to households where those factors are present. The more accurate the targeting, the lower the cost of each successful prevention. The most current research suggests that households at highest risk for becoming homeless are those with a recent history of homelessness, extremely low-incomes, high rent burdens, not on a lease, living in zip codes with high rates of poverty and housing instability, and who lack a network of social supports. Of those factors, a recent history of homelessness appears to have the greatest predictive power.2,3 In addition to targeting accuracy, the cost of each successful homelessness prevention intervention has to account for the fact that a percentage of people who are helped with homelessness prevention services nonetheless end up losing their housing. Whatever target number of preventions the system sets out to achieve must be adjusted upward to account for this failure rate, and that rate grows as the number of risk factors increases.4 2 Phillips, David et. al., Do homelessness prevention programs prevent homelessness? Evidence from a randomized controlled trial, The Review of Economics and Statistics, 2023, 1-30, and interviews with All Home and BACS. 3 While the benefits of targeting are clear, how great they are — how significantly targeting can drive down the ratio of households served to actual homelessness preventions — is an area of considerable uncertainty. Not only does it depend on the predictive power of the multiple factors considered, it also depends on the threshold risk level at which households become eligible for homelessness prevention services. For purposes of this Framework, therefore, the targeting accuracy factors used should be treated as considered estimates and in need of regular review and updating in light of ongoing research and program experience. 4 Shinn, et. al. This appendix to the Alameda County Homelessness Prevention Framework estimates the costs of achieving reductions in inflow into literal homelessness through targeted homelessness prevention. There are multiple changes in policy, as well as changes in practice among public systems that engage with vulnerable households, that could significantly reduce the number of people at imminent risk of homelessness and, in turn, the inflow into homelessness. To the extent that we rely on targeted homelessness prevention services to keep people out of literal homelessness, the costs are significant for the reasons outlined below. However, in assessing these costs it is essential to keep in mind that there are numerous benefits to providing prevention services to households in crisis, even if they would not have become literally homeless in the absence of those supports.1 Determining the amount of funding needed to prevent one household from becoming homeless is complicated by the fact that only a small percentage of people who find themselves at imminent risk of homelessness will immediately become homeless if they do not receive prevention services. This means that the system will have to provide prevention services to multiple households for each household it prevents from falling into homelessness. How many households must be served in order to significantly reduce inflow into homelessness depends on how effectively prevention services are targeted. 1 Shinn, et. al., Efficient Testing of Homelessness Prevention Services for Families, American Journal of Public Health, 2013, pg. S329. Estimating the Cost for the Community-Wide Homelessness Prevention Response Appendix D 2 Homelessness Prevention Framework — Appendix E Taking these factors into consideration, the cost equation of reducing inflow into homeless by some number of households — “N” — looks like this: Number Prevented (N) + (N x Failure Rate) x Targeting Accuracy Factor (TAF) x Cost Per Prevention (CPP) = Total Cost For example, if the goal is to reduce inflow into homelessness by 100 households, the estimated failure rate of prevention interventions is 15%, the estimated targeting accuracy factor is 5 for1, and the projected cost per prevention is $7,000,5 the cost of achieving a 100 household reduction in inflow is: 100 + (100 x. 0.15) x 5 x $7,000 = $4,025,000 In this scenario, the $4 million provides prevention services to 575 households. All of them receive substantial benefits from the increased housing stability that results from these services, and 100 of them would have been literally homeless but for the intervention. While the cost to the system of preventing a household from experiencing literal homelessness in some cases may not be lower than helping a household return to housing from homelessness, the additional households served and the numerous benefits to a household and other systems from avoiding homelessness make targeted homelessness prevention investments cost effective.6 5 This estimated cost per prevention is based on current prevention programming in the community. Home Together’s original modelling used a more conservative estimate of $4,500. 6 Phillips, David et. al., Do homelessness prevention programs prevent homelessness? Evidence from a randomized controlled trial, The Review of Economics and Statistics, 2023, 1-30, pg. 17. A | Estimating the cost of reducing returns to homelessness in Alameda County Home Together sets as a goal reducing returns to homelessness from the current level of 15% to 9% by the final year of the plan. Determining how many successful preventions are needed each year to achieve this is dependent on the change in the number of households housed. As the number of people housed increases, the number of preventions needed to achieve and maintain a decrease from 15% to 9% grows. This is consequential because Home Together projects a significant annual increase in housing placements. The following table sets out the projected additional successful preventions that will be needed to reach a 9% rate of return to homelessness assuming the current FY 22–23 15% rate is the baseline and using Home Together’s projected increases in households housed.7 7 For purposes of determining the number of successful preventions needed to reach 9%, the percentage rate of return is calculated based on the number of housing placements two fiscal years earlier. This is consistent with the HUD “Returns to Homelessness” system performance measure. Table 1 | Additional successful preventions among returnsa FY22–23b FY23–24 FY24–25 FY25–26 Households Housed 4,263 6,923c 7,778c 8,633c 15% Rate of Return 441 452 639 1,038 9% Rate of Return 271 384 623 Additional Preventions Needed 181 255 415 a) This analysis uses only Adult Only households because families are already experiencing an 8% return rate, which is below the Home Together goal. b) FY 22–23 households housed and returns are actuals. c) Home Together projections from pg. 60. 3 Homelessness Prevention Framework — Appendix E Through FY25–26, a total projected additional 850 successful preventions will be needed to achieve a 9% rate of return to homelessness. Because of the continued increase in housing placements, that number will grow larger, ultimately to an additional 518 preventions per year once the projected annual housing placement rate reaches the projected 8,633 in FY 26. Because these households all have a recent experience of homelessness, are extremely low-income, and required support to obtain housing, it is reasonable to estimate a very low targeting accuracy factor (TAF) of 3 for 1. These households are also likely well known to providers, allowing for better tailoring of prevention services, so the calculations estimate that only10% of households receiving prevention services will nonetheless end up returning to homelessness. While some households may need higher than average levels of services, others will likely need lower, so this analysis uses the current average system cost of a unit of prevention, approximately $7,000.8 Based on these estimates, the projected costs of achieving and then maintaining the Home Together goal of a 9% rate of return to homelessness are as shown in Table 2: In summary, using the assumptions of the Home Together model, achieving the objective of a 9% rate of return to homelessness will require serving an estimated 2,807 formerly homeless imminently at-risk households through 2026, at a projected total cost of just under $20 million. Thereafter, if new housing placements reach 8,633 per year, the system will need to serve an additional 1,554 formerly homeless households per year at an estimated cost of approximately $10.9 million a year. 8 Local eviction prevention providers estimate that they spend between $6,000 and $8,000 per household served, including the cost of administration and staffing for the direct service provider. This is higher than the $4,500 per prevention estimated in Home Together. B | Estimating the number of prevention resources needed to reduce inflow of newly homeless households. Home Together’s goal is ultimately to have the resources and capacity to resolve homelessness for everyone in need. The Home Together model projects a drastic reduction in homelessness by both reducing the number of households becoming homeless and adding more than 24,000 new housing opportunities for people experiencing homelessness. The Home Together model projects that the total number of households entering homelessness would increase by 20% the first year of the Plan (FY21–22), slow to 10% in the second year (FY22–23), be unchanged in the third year (FY23–24), and then decline by 10% each of the final two years of the plan.9 These reductions from the originally projected 20% annual increase in inflow are projected to result from the increased focus on homelessness prevention.10 Using these inflow reduction targets, and assuming these will be achieved entirely through increased targeted prevention efforts, we are able to estimate the number of new successful preventions that will be needed each of the next three years. Unlike returns to homelessness, those newly at risk of homelessness who will actually become literally homeless without prevention assistance are harder to identify accurately in advance and the percentage of those highest-risk households that will nonetheless lose their housing is also expected to be higher. As a result, a significantly larger number of households are projected to need prevention assistance for each household that is prevented from entering literal homelessness. 9 Based on Home Together modeling analysis. 10 Home Together, pg. 58. Table 2 | Estimated annual cost of achieving a 9% rate of return to homelessness Fiscal Year Target Reduction in Homeless HH Failure Rate Targeting Factor Total HH Served Cost Per Prevention Total Cost FY 23–24 181 10%3 to 1 597 $7,000 $4,179,000 FY 24–25 255 10%3 to 1 841 $7,000 $5,890,500 FY 25–26 415 10%3 to 1 1,369 $7,000 $9,586,500 TOTAL 850 N/A N/A 2,807 N/A $19,656,000 4 Homelessness Prevention Framework — Appendix E Based on this analysis, the system will need to keep 1,070 new households from literal homelessness in FY 24–25 and 2,701 in FY 25–16. Assuming other structural factors do not change to reduce inflow of new households into homelessness, maintaining inflow at projected FY 25–26 levels will require significant additional annual increases in successful homelessness preventions. Unlike those returning to homelessness, households imminently at risk of homelessness for the first time who will actually become immediately homeless in the absence of prevention services are more difficult to identify. While research into the most effective targeting criteria is ongoing, available studies and local subject matter experts estimate that even with effective outreach and prioritization of those who have the highest risk level, the Targeting Factor is very unlikely to be better than 5 for 1 and could exceed that significantly. The analysis here provides cost estimates based on the 5 for 1 TAF and the results should therefore be taken as a conservative estimate of the number of preventions that would have to be provided to achieve the Home Together targets. Because less will likely be known about the newly at-risk households served, and there will be a focus on serving those households assessed to be most at risk in order to improve the Targeting Factor, the following analysis assumes a failure rate of 15%. This is also a conservative estimate, and the actual number could be higher.11 11 Shinn, et al. Based on the above assumptions, the community wide homelessness prevention response would need to serve more than an estimated 15,000 households in FY 25–26 and invest an estimated $109 million in the effort in order to meet the Home Together projected reductions in inflow of new households into literal homelessness. Preventing first 1,070 and then 2,701 households from entering homelessness will require a significant investment of new resources. However, this Framework outlines a series of additional recommendations, including important policy changes, that could reduce the number of households that find themselves at imminent risk of falling into homelessness each year, and thereby significantly reduce the number of necessary successful preventions. The above model assumes no change in the amount of inflow that will occur in the absence of homelessness prevention investments. As the recent eviction moratorium in Alameda County demonstrated, policy changes can significantly alter the actual rate of inflow into homelessness even in the absence of targeted homelessness prevention investments. Based on the modeling for Home Together 2026, and assuming that targeted homelessness prevention is responsible for achieving all the necessary reductions in newly homeless households below what they otherwise would have been, the community wide homelessness prevention response is estimated to need the capacity to serve approximately 15,550 households in FY 25–26 and invest approximately $109 million in the targeted homelessness prevention effort. Table 3 | Projected number of reductions in newly homeless households needed FY22–23 FY23–24 FY24–25 FY25–26 Newly Homeless w/o additional preventions 4,024a 4,829 5,795 6,954 Home Together goal number based on % change for newly homeless HHb 5,250 5,250 (0%) 4,725 (-10%) 4,253 (-10%) Successful Preventions Needed 0 0 1,070 2,701 a) Actual FY22–23 as compared to 5,250 projected for FY 22–23 in Home Together. b) This includes both adult only households and households with children. 5 Homelessness Prevention Framework — Appendix E © 2025 Alameda County. All rights reserved. (01-25)