Homelessness is solvable; we have learned a lot about what works. (United States Interagency Council on Homelessness 2010)

We don’t know how to end homelessness. Not in the aggregate, anyway. (O’Flaherty 2019)


Much of the discussion in Mission Economy (Mazzucato 2021)—and from other contributors to this book—focus on the relationship of the state to for-profit companies as they collectively engage with major social objectives. However, it is also instructive to consider how the principles of state-directed, mission-oriented thinking have fared in contexts where third sector-state relations dominate. Indeed, in the United States and elsewhere, government and nonprofits (rather than companies) are the key players in a broad array of initiatives aimed at social problems like poverty, education, healthcare, and homelessness, among others. In addition, the focal target for Mission Economy prescriptions are the UN’s Sustainable Development Goals (SDGs). Many of the SDGs involve contexts where extensive (and expensive) public-private partnerships already abound, wherein governments provide public services via grants, contracts, and policies that are primarily channeled through nonprofit organizations (Smith and Lipsky 2009). As such, it is useful to consider the prospects of mission-focused state leadership in these contexts.

This chapter provides such an analysis. Specifically, we provide a case study of the US federal government’s (mis)adventures in its mission to prevent and end homelessness, focusing on the period from 2008 to 2022. Homelessness in the United States is an illuminating case to consider in relation to the Mission Economy’s core principles and focus. First, homelessness looms large as an enduring and pervasive social issue—stubbornly persisting across impoverished nations and wealthy, market-oriented economies alike (Toro et al. 2007). Second, homelessness has been referred to as “intractable” (Eide 2020), a “wicked problem” (Brown et al. 2013), and a “grand challenge” (Henwood et al. 2015)—i.e., exactly the kind of social issue that proponents of the Mission Economy would see amenable to an innovative, collaborative, government-led response. Third, and most importantly, the US government’s engagement with homelessness in the twenty-first century is distinctive in its almost-prescient alignment with Mazzucato’s prescriptions for successful “moonshot” initiatives (Mazzucato 2021). Rather than a passive remediator of market failures, the federal government played a catalyzing leadership role in defining and pursuing a bold and innovative mission against homelessness. Spearheaded by the United States Interagency Council on Homelessness (USICH), a “federal agency with a sole mission on preventing and ending homelessness in America” (USICH 2023), the government worked throughout this period to facilitate broad coordination, dramatically expand funding, and mobilize a shift toward innovative, evidence-based approaches to homeless services—all while engaging diverse stakeholders across areas and levels of government, sectors, and communities to pursue a suite of clearly defined goals related to the prevention and eradication of homelessness.

But despite a clear mission, good intentions, bipartisan political support, evidence-based innovations, major funding increases, thorough stakeholder engagement, and unequivocal state leadership, the results during this period were underwhelming at best. A more-than-doubling of federal expenditures and the widespread diffusion of evidence-based practices saw a mere 9 percent reduction in total homelessness; in fact, the downward trend stalled early, with no single year-over-year decline in homelessness since 2016. Not a single one of the four objectives initially outlined in 2010 were met, and each one was eventually delayed, revised, or dramatically curtailed. Even those supportive of the government’s efforts to date acknowledge that “most plans to end homelessness have fallen far short of the mark” (Lee et al. 2021, p. 15).

In what follows, this chapter will analyze the government’s “rescue mission” of homelessness. We first outline the historical conditions that gave rise to a new era of homeless response in the wake of the Great Recession, documenting the emergence of a government-led, evidence-based mission unlike previous efforts. We then review the results during the ensuing period, marking the divergence of realized outcomes from the original objectives. Next, we highlight several unintended consequences generated by the government’s interventions during this period. Finally, we assess the repetition of history with the most recent federal strategic plan for homelessness, which features a suite of updated goals and a renewed state-led strategy to achieve them. The case study reveals the limits associated with a government-coordinated mission to address a particular form of human suffering—even when that mission is designed to promote evidence-based policies and engage “bottom-up” solutions.

A Very Brief History

Homelessness is not a new phenomenon, nor is a government-led response to engage with it. As part of the New Deal programs enacted in response to the Great Depression, a program called the Federal Transient Service was created specifically to address the apparent rise of what were then referred to as “tramps” or “hoboes”—mostly White, older men with highly migratory lifestyles. While the program only lasted for 2 years, it exemplified an “era” of homelessness lasting throughout the early twentieth century that was characterized by this demographic.

By the 1980s, however, there was a growing consensus that the problem of homelessness had taken on a new chapter. Homelessness was affecting broader swathes of society, including greater numbers of people, more families, and more people of color. Activists emerged as prominent national voices to elevate homelessness into a cultural crisis. Highly politicized debates about the extent of homelessness raged; one activist claimed 2–3 million, while a 1984 report from the Reagan administration estimated roughly 350,000 (United States Department of Housing and Urban Development 1984). Whatever the number, the sense was that this new homelessness also required a new approach—one where the federal government played a key role in funding and mobilizing homeless responses. In 1987, the Stewart B. McKinney Act (later updated to the McKinney-Vento Act) emerged as the centerpiece of federal homelessness legislation, ushering in what many refer to as the “modern era” of homelessness.

Through the end of the twentieth century, the nation’s homeless response infrastructure matured and reached a relatively stable equilibrium. By the 1990s, homeless services began to be coordinated via Continuums of Care (CoCs), networks of nonprofit, and local government organizations serving the homeless in a particular geography. The federal government was a major source of funding, channeling resources through the CoC system. Notably, this infrastructure of homeless service providers evolved distinctly from (and often disconnected from) other programs related to affordable housing (e.g., the Low Income Housing Tax Credit), public housing, and most housing subsidies (e.g., Sect. 8 Housing Vouchers).

From the Linear Model to Housing First

The modern homeless services infrastructure consists of organizations providing four main forms of housing programs: emergency shelter, transitional housing, rapid rehousing, and permanent supportive housing. Emergency shelter fits with the layperson’s understanding of a homeless shelter, typically characterized by many beds in larger, mostly public sleeping areas. These exploded in the 1980s as a stopgap meant to keep individuals in crisis from “sleeping rough.” As such, they are the lowest common denominator, intendedly meant to address bouts of housing displacement as manifested emergency.

To understand transitional housing, however, requires a brief foray into the notion of service models—the programmatic strategies for facilitating exits from homelessness. Table 1 illustrates the two main alternative service models. The early modern homeless era was characterized by what scholars call the “linear model” of homeless services. Because homelessness was viewed as a problem of individual behaviors and responsibility, providers sought to cultivate and ensure “housing readiness” in their clients—through e.g., some combination of mental health stability, work, and control of substance abuse. Typically ranging from 6 to 18 months, transitional housing programs provided more private housing units but also targeted case management and support services. In the linear model of homeless services, transitional housing was viewed as a key intermediate step out of emergency shelter and toward an exit into stable housing. For some, this meant moving back into traditional, independent housing. For others, the ultimate destination was permanent supportive housing—fully or heavily subsidized housing on a long-term basis, often with case management.

Table 1 Alternative homeless services models

Eventually, a second service model arose in stark contrast to the linear model: Housing First. While the next section will discuss the political circumstances that propelled Housing First to the front-and-center of the government’s mission to end homelessness, we will first review the approach and philosophy. Housing First is an approach that emphasizes rapid and unconditional placement into supportive housing. As its proponents indicate, the original “Pathways Housing First” program, originating in New York City, featured a “(1) program philosophy and practice values emphasizing consumer choice; (2) community based, mobile support services; and (3) permanent scatter-site housing” as well as “harm reduction” (Padgett et al. 2016, p. 3). As such, the Housing First model’s main program is permanent supportive housing—long-term subsidized units for individuals or families with (optional) case management. Alongside this, however, was the introduction of “rapid rehousing” a medium-term option following the Housing First philosophy. As such, rapid rehousing effectively replaces transitional housing in the Housing First model, because it offers a similar timeframe but removes all preconditions and notions of readiness.

The Emergence of a Mission for Homelessness

The Preconditions

As hinted above, the linear and Housing First models stand in stark philosophical and practical contrast. In the 1980s, when the homeless services industry emerged alongside targeted federal legislation, the linear model was the dominant paradigm. However, Housing First was introduced in the 1990s by a pioneering organization in New York City, Pathways Housing First. Led by Sam Tsemberis, Pathways focused on the long-term, hard-to-house subset of the homeless population—arguing that the behavioral and psychiatric expectations of the linear model programs functioned as unnecessary barriers to housing for this group. Seeing housing as a human right, Tsemberis developed Pathways Housing First to remove these barriers—and, in doing so, to pursue a harm reduction rather than behavioral change focus.

Importantly, Tsemberis also developed randomized trial studies of Pathways’ programs. Tsemberis and a team of researchers published a series of papers demonstrating much higher rates of housing retention for those served by his program than the traditional services for this particularly hard-to-house population (Tsemberis et al. 2004; Tsemberis and Eisenberg 2000). These studies proved pivotal for an emerging body of academic research that would lay the foundation for Housing First to become one of the most prominent examples of evidence-based policy in the United States (Lucas 2018).

Lucas (2018) elaborates on this to describe how policy entrepreneurs, including Tsemberis, leveraged academic evidence to elevate Housing First as a centerpiece of a renewed policy discussion around homelessness. Importantly, the housing retention success of Tsemberis’ New York City program dovetailed nicely with other researchers’ findings about the taxpayer costs of homelessness. Specifically, Dennis Culhane’s pioneering scholarship showed that the costs of homelessness in terms of public services were essentially power-law distributed: a very small subset of homeless individuals utilized a very large share of shelter services and also imposed high costs in terms of hospitals, jails, and police expenses (Culhane et al. 2002; Poulin et al. 2010). The vast majority, by contrast, utilized shelter services in a manner that was brief and nonrecurring.

Taken together, the two components of the evidence showed that (i) a small group of homeless individuals the “chronically homeless”—accounted for a disproportionate share of the economic burden of homelessness, and (ii) the Housing First approach could dramatically improve housing retention for this group—at a lower cost than that imposed through these individuals’ use of public services in the status quo. As such, a narrative emerged around Housing First as both a fiscally responsible use of taxpayer dollars and a compassionate alternative to the linear model. Championed both by journalists (e.g., Gladwell 2006) and policy advocates alike, Housing First steadily garnered increasing bipartisan support through the early 2000s (Stanhope and Dunn 2011). Housing First was championed within the George W. Bush administration by Phil Mangano, Executive Director of USICH. In a telling profile in The Atlantic, Mangano is quoted as saying, “Research is the new advocacy” (McGray 2004). Indeed, evidence proved pivotal to mobilize political energy around a renewed vision and mission for homelessness policy (Lucas 2018). Given the ideologically divisive nature of entitlement spending in the United States, stakeholders in the homeless services ecosystem recognized a profound opportunity to transform homelessness policy for the first time in decades.

But it was not only advocates’ efforts to elevate Housing First that initiated the emergence of a national policy mission for homelessness. Rather, the 2009 global financial crisis created fertile conditions for dramatic policy change. It was evident that the Obama administration’s policy response was intendedly revolutionary. As Obama’s then-Chief of Staff, Rahm Emanuel, famously asserted in November 2008, “Never let a serious crisis go to waste. What I mean by that is it’s an opportunity to do things you couldn’t do before” (Wall Street Journal 2009). Homeless services proved a suitable target for this opportunity, given the housing market’s role in the Great Recession. Therefore, alongside the federal government’s sweeping stimulus package came the HEARTH Act of 2009, a major revamp of the McKinney Vento Act that established the scope of federal homelessness response. The HEARTH Act redefined homelessness policy, promised a dramatic increase in federal funding, and set forth a mandate for the federal government to develop a strategic plan aimed at ending homelessness (Lucas 2017).

Opening Doors: The Federal Plan to Prevent and End Homelessness

Under the mandate of the HEARTH Act, USICH produced a revolutionary strategic plan that outlined a federal mission for homelessness: “Opening Doors: The Federal Plan to Prevent and End Homelessness.” Opening Doors introduced four ambitious goals:

  • Finish the job of ending chronic homelessness in 5 years.

  • Prevent and end homelessness among Veterans in 5 years.

  • Prevent and end homelessness for families, youth, and children in 10 years.

  • Set a path to ending all types of homelessness.

Of these four goals, the first three have a well-defined success criterion. It is worth noting that USICH does note that these goals could be seen as “aspirational” (p. 52). Yet, the plan also exuded significant optimism about the effectiveness of the strategies recommended: “Solutions exist. New collaborative leadership, more coordination, and wise investments in proven strategies…will lead to major reductions in homelessness” (USICH 2010, p. 24, italics added).

The plan also provided ten objectives related to these goals. Notably, the last objective falls under the theme of “Retool the Homeless Crisis Response System.” Hence, although “collaboration” is emphasized throughout the plan, this functionally translated to top-down pressure to conform to “expert” policy prescriptions: “The Plan also proposes the re-alignment of existing programs based on what we have learned and the best practices that are occurring at the local level, so that resources focus on what works” (p. 4, emphasis added). As then-HUD Secretary and USICH Chair Shaun Donovan writes, “The Council members and the Administration are fully committed to taking these best practices and proven solutions to scale across the federal government” (p. 3). The plan makes clear that the “proven solutions” being considered were based on the presumed superiority of Housing First. Discussing the goal of ending chronic homelessness, the plan states: “Permanent supportive housing using Housing First is a proven solution that leads to improvements in health and well-being. Supportive housing also has been shown to be a cost-effective solution in communities across the country” (p. 38). Elsewhere, the plan declares the “documented success” of Housing First over the linear model (p. 49), chiding communities that have not yet emphasized a transition toward a Housing First-based system of care.

In sum, the plan was equal parts optimistic, focused, and evidence-based, articulating a confidence in the ability to make significant impact by leveraging evidence to initiate systems change. In other words, the federal government set forth a top-down mission to alleviate homelessness. Furthermore, this mission focused on federal leadership, cross-agency and cross-sector collaboration, provider coordination, and rigorous applications of evidence—all hallmark prescriptions offered by the advocates of centralized policy missions to address grand challenges.


The Four Goals

We begin our assessment of the results with respect to the 2010 strategic plan’s four overarching goals. Of the four goals outlined in USICH’s 2010 strategic plan, zero had been met by 2020. Figure 1 details these results. The dashed lines report the government’s goal projections; for simplicity, we take the target deadline for eliminating homelessness among each subpopulation and linearly interpolate the trajectory. The solid lines are the observed trends as reported by the Department of Housing and Urban Development in its Annual Homeless Assessment Report to Congress.

Fig. 1
Four line graphs. 1. Chronic homelessness. The actual line has an increasing trend after 2016, reaching 2022. The 2010 goal line reaches to zero by 2015. 2. Veteran homelessness. The actual line has a decreasing trend after 2010, reaching 2022. The 2010 goal line reaches to zero by 2015. 3. Family and Youth homelessness. The actual line has decreasing trend after 2012, reaching 2022. The 2010 goal line drops to zero by 2020. 4. Total homelessness. The actual line has decreasing trend up to 2016 after which it rises to 2022.

US government goals vs. observed trends in key homeless populations, 2008–2022. Source: Point in Time Counts of Homelessness, United States Department of Housing and Urban Development

Chronic homelessness. Among the four goals, eliminating chronic homelessness appeared to offer the greatest certainty of success ex ante. This is because the chronically homeless were the group for whom the evidence supporting the government’s preferred service model, Housing First, was deemed to be strongest. As the 2010 strategic plan boldly declares, “For people experiencing chronic homelessness, the research is clear that permanent supportive housing using a Housing First approach is the solution” (p. 18, emphasis added). By many accounts, this goal also seemed positioned to do the “most good,” in that it focused on housing a group acutely suffering from the cooccurrence of long-term or repeated bouts of homelessness and the presence of disabilities. Although a small fraction of the homeless population, chronically homeless persons are perhaps the most visible subpopulation and most aligned with the typical lay notion of a person experiencing homelessness.

It is important to recognize that chronic homelessness was also central to the “economic” arguments for Housing First. Individuals facing long-term or recurrent bouts of homelessness have long been identified as heavy users of public services like hospitals and jails—imposing disproportionate costs on taxpayers. The federal government projected Housing First would save taxpayers money and provide long-term housing to those whom had otherwise been unable to secure it. At the heart of this claim was the elimination of chronic homelessness. In turn, much of the bipartisan support for the federal homelessness strategy was grounded in the dual claim that Housing First was not only compassionate but also cost-effective (Stanhope and Dunn 2011).

However, despite a dramatic increase in permanent supportive housing beds and a widespread commitment from communities across the country to implement Housing First, chronic homelessness persisted and even worsened. Not only was the end of chronic homelessness far from being realized in 2015, but the chronic homeless population was estimated to have increased by roughly 30 percent from 2010 to 2022. Furthermore, chronic homelessness had not only increased but also become a more pervasive experience among the homeless, increasing from 16 to nearly 24 percent of the homeless population.

Anecdotal evidence also affirms that this particularly harsh, long-term, and visible form of homelessness is indeed worsening—especially in major urban cities. National media regularly covers the “crisis” of homelessness in San Francisco, a city increasingly associated with long-term tent encampments and open-air drug markets. Despite the Bay Area’s affluence and prominent reputation for cultivating tech giants—along with USD 680 million in direct programmatic homeless spending by the city each year—chronic homelessness nearly doubled from 2015 to 2019 (1629 to 3030). By the end of the period, roughly 1 in 100 residents was counted as homeless (Enzinna 2021). Other California cities like Los Angeles have seen similarly bleak trends in chronic homelessness and an increasingly visible “street homeless” population, transforming the safety profile of many neighborhoods.

The abject failure to eliminate or even reduce chronic homelessness is striking, and we will consider some of the factors for these trends later in this chapter. In sum, the raw data make clear that chronic homelessness proved not to be the “golden goose” of the evidence-based policy as had been anticipated, but rather the “canary in the coal mine”—indicative of a profound disconnect between the intentions and outcomes of this state-led mission.

Veteran homelessness. The second goal, ending veteran homelessness by 2015, appeared to fare better than the chronic homelessness goal. Indeed, veteran homelessness saw the largest reduction of any subgroup throughout the period. While still far short of the initial goal, veteran homelessness did decline by roughly 50 percent from 2010 to 2020. Understandably, then, proponents of the US policy interventions typically hold veteran homelessness as the proof of success. It is typically contended that this proves the efficacy of the government’s leadership and the validity of its evidence-based initiatives.

Yet, inquiry into this trend casts significant doubt on the lessons that can be drawn for other homeless populations. One key reason is a dramatic secular decline in the number of veterans at risk of homelessness during the period. O’Flaherty (2019) documents how the number of adult veterans aged 18–65 dropped by 26.7 percent from 2010 to 2016. Because the total number of veterans at risk of homelessness declined dramatically, O’Flaherty estimates that the trends in the incidence or rate of homelessness among this group mirrored overall trends. Recent work has also shown that exposure to combat is an important predictor of homelessness among veterans (Ackerman et al. 2020), indicating that the antecedents and experience of veteran homelessness is likely atypical.

Furthermore, the veteran homeless population receives far more targeted homelessness funding than the rest of the homeless population (Perl 2023). From 2010 to 2020, while moving down in the share of the total homeless population from 11 to 6 percent, veterans moved up in the share of targeted homelessness funding from the federal government from 16 to 26 percent. On a per capita basis, the spending difference is particularly stark. Back-of-the-envelope calculations indicate that each homeless veteran is allocated about USD 47,000 in federal spending compared to USD 9100 for each nonveteran. To increase nonveteran per capita funding to spending parity would require an additional USD 20.5 billion per year (over a 300 percent increase from 2020 levels). And even this would not account for the myriad other government programs to support veterans beyond homelessness.

Taken together, the secular decline in the population of veterans at risk of homelessness along with the massive and targeted investment into this group reveal that the “strongest” evidence of any margin of success in the state’s mission to end homelessness is, in fact, dubious. O’Flaherty (2019, p. 21) summarizes the results of the federal efforts toward veteran homelessness in the early 2010s: “This great initiative appears to have accomplished little.”

Family and youth homelessness. Family and youth homelessness trends fall between veterans and the chronically homeless, with modest but notable declines. USICH’s 2010 goal was to end family and youth homelessness by 2020. Over that decade, family and youth homelessness declined by roughly 30 percent. It is worth noting that families are typically said to experience homelessness as a brief and nonrecurring experience—and rarely as “unsheltered” homeless outside of the established shelter system. This has implications for the long-term potential for this downward trend to persist as will be discussed below.

Total homelessness. Total homelessness declined only modestly during the period. In fact, no single year-over-year reduction in total homelessness was observed from 2016 to 2022, such that the second half of the period unraveled the vast majority of the modest gains made during the initial years of the mission. The lackluster trend has been difficult to ignore. In fact, scholarly consensus has also shifted significantly away from the confidence of a decade prior. As noted in the epitaph quote from a leading homelessness economist, “We don’t know how to end homelessness. Not in the aggregate, anyway” (O’Flaherty 2019). Overall, the goal to “set a path to end all types of homelessness” has not been met.

Federal Strategies Through the Period

The federal government did not sit idly by while these underwhelming trends unfolded. Its mission engagement and leadership continued throughout the period, revolving around three main areas: funding increases, Housing First prioritization, and strategic revisions.

Federal Homelessness Funding

First, although some pundits blame limited funding availability for the results above, the federal government dramatically increased the fiscal footprint of its homeless services spending throughout the period. Figure 2 presents the nominal expenditures for targeted federal homelessness programs from 2010 through 2022. Funding increased steadily throughout the period; by 2022, funding had more than doubled from 2010 levels to USD 7.9 Billion annually. This amounts to roughly USD 13,500 per person counted as homeless in the 2022 point-in-time count—greater than the annual median gross rent in the United States.Footnote 1 It should also be noted that these trends were unaffected by the election of Donald Trump in 2016; funding continued to increase at a similar rate as in the Obama era. This casts significant doubt on the claim that the new administration’s policy shifts slowed progress toward the homelessness goals in the later end of the period. Overall, there is little evidence that funding constraints hindered the plan’s success, even though this is a relatively common assertion.

Fig. 2
A vertical bar graph of Federal targeted homelessness expenditures between 2010 and 2022 in billion U S dollars. There is a steady increase from 3.8 in 2010 to 7.9 in 2022. Values are estimated.

Federal targeted homelessness expenditures, 2010–2022 (billion USD). Source: President’s Budget: Fact Sheet on Homelessness Assistance, United States Interagency Council on Homelessness

Advancing Housing First

The second role of the government to pursue this mission was in its “leadership” role in pressuring homeless service providers to implement Housing First and its associated programs. As the 2010 strategic plan clearly indicated, the federal government aimed to lead an evidence-based “retooling” of the homeless services industry—away from the linear model and toward Housing First. While the federal government did not explicitly require these changes, the competitive funding model prioritized Housing First programs both within and across communities. The largest grant under this mission, the CoC grant, was structured such that organizations in a community would apply jointly for funding. The representative organization for that community was tasked with ranking each program on a set of criteria provided by HUD. As of 2013, existing programs that aimed to revise away from linear model practices toward Housing First received extra points and a high likelihood of funding (USICH 2020, p. 9). By contrast, new programs were very unlikely to be competitive for federal funding if they did not commit to following Housing First practices; these projects were scored to rank near the bottom in each community’s annual grant request. Similarly, existing programs that did not fit the Housing First ideology—most notably, transitional housing—faced a high risk of defunding.

Figure 3 shows how these funding priorities corresponded to a rather successful “retooling” of local homeless providers’ programs. Each line indicates the total number of year-round beds associated with shelter programs of each main type. The two hallmarks of the Housing First model, permanent supportive housing and rapid rehousing, exploded over the period. Permanent supportive housing more than doubled to nearly 400,000 beds. Rapid rehousing, which was not tracked until 2013, also increased rapidly, reaching 150,000 beds by 2022. It should be noted that there is limited data on how faithfully these two programs have been implemented with respect to Housing First principles. However, it is clear that the reason for the bed increases in these two program types was a federal emphasis on Housing First. We will revisit this point in a later section.

Fig. 3
A multi-line graph of the number of homelessness beds in 5 shelter programs versus years from 2008 to 2022. Permanent supportive housing, emergency shelter, and rapid rehousing lines have rising trends. The transitional housing line declines. The line for others remains flat until 2013, then rises.

Retooling the homeless shelter system. Source: Housing Inventory Counts, United States Department of Housing and Urban Development

By contrast to permanent supportive housing and rapid rehousing, linear model programs dwindled precipitously. The clearest evidence of this is in transitional housing. The supply of transitional housing beds was halved over the decade. While some providers did retool their transitional housing programs to a rapid rehousing (the most direct “substitute” in the Housing First model) or other approach, the data also suggest that many providers simply closed these programs as funding decreased. Little is known about how different transitional housing providers navigated this period, and this is an important area for organizational scholarship.

Overall, the federal government was highly effective at mobilizing local public and nonprofit providers to abandon what the government deemed to be outdated practices and replace them with the “evidence-based” innovations they believed to be effective at preventing and ending homelessness. This is particularly notable in that transitional housing had been a bedrock component of homeless service provision across the United States since the 1980s. In this sense, a crucial component of the strategic plan to prevent and end homelessness was carried out with remarkable precision and rapidity.

Another important point is that the trends do not appear subject to changes in political power. As with the funding continuations during the Trump administration, the trajectory away from transitional housing programs and toward the Housing-First-friendly programs continued with no visible evidence of disruption.

Revisions to the Federal Mission

It was evident long before 2020 that the nation was not on pace to complete the mission set out in the initial Opening Doors plan. As such, in addition to the funding increases and institutional pressures imposed on providers during the period, the federal government also actively revised the strategic plan in response to the lackluster trends. In 2015, the federal government introduced an updated version of the plan and revised the goal of preventing and ending chronic homelessness—moving this deadline from 2015 to 2017. Other goals remained in place.

In the 2015 update, the revised strategic plan also doubled down on Housing First commitments and strategies. For instance, even while updating the goal of ending chronic homelessness due to the limited results to date, the plan boldly reiterates: “For people experiencing chronic homelessness, the research is overwhelmingly clear that permanent supportive housing using a Housing First approach is the solution” (USICH 2015, p. 25, emphasis added). The plan also indicates that the mission’s strategies were, in fact, being pursued: “Consistent with Opening Doors, communities are increasingly adopting evidence-based practices and replicating promising program models that incorporate a Housing First approach, leveraging resource commitments from the public and private sectors and from homeless assistance and mainstream systems” (ibid., p. 11). According to the revise plan, the ostensible reason for limited successes to date was due to “a lack of Congressional support for the expansion of permanent supportive housing” (ibid., p. 8), even though funding allocations and permanent supportive housing inventory had increased steadily and significantly, as noted above.

The 2015 strategic plan update doubled down on efforts to prioritize, fund, and spread Housing First practices: “This crisis response system involves the coordination and reorientation of programs and services to a Housing First approach that emphasizes rapid connection to permanent housing, while mitigating the negative and traumatic effects of homelessness” (ibid., p. 55). Notably, rather than grappling with the growing number of questions about the ability of the federal government to achieve its stated mission via these strategies, the plan stated an unwillingness to consider alternative solutions: “To attain value for money, agencies and communities alike must direct resources towards evidence-based and cost-effective solutions like permanent supportive housing, Housing First, and rapid re-housing, and away from models and programs that are outdated, unsupported by evidence, or are not cost-effective” (ibid., p. 61).

In 2018, USICH put out a new strategic plan entitled “Home, Together” (USICH 2018). This strategic plan update was notable in that it departed from offering “deadlines” for the goals articulated in prior plans—which, as noted above, were either past or far short of the stated targets. Still, the plan reaffirmed efforts to end homelessness among the same groups as before (ibid., p. 6). In addition, the plan reiterated the prioritization of Housing First once again, indicating that the new administration was continuing in the mission of its predecessor. When looking retrospectively over the period, the plan even appears to assert that past efforts had been little short of a great success. For instance, the document confidently states that “evidence-based Housing First approaches have helped serve more people with better results” (ibid., p. 12). Such statements, paired with the trends in funding and program evolution above, affirm that the shortcomings in the federal plan are not well explained by shifts in political power or by a lack of “political will.”

Another important update in the 2018 was the first recorded instance in the strategic plan efforts to define “success.” Recall that the prior plans had set out to “set a path to end all types of homelessness.” In Home, Together, USICH offered a working definition thereof: “Achieving these goals is grounded in a shared vision of what it means to end homelessness: that every community must have a systemic response in place that ensures homelessness is prevented whenever possible, or if it can’t be prevented, it is a rare, brief, and one-time experience.” This is notable because the plan clarifies that “ending homelessness” functionally means having a “systemic response” in place—i.e., a permanent, federally funded homeless services industry (Lucas 2017).

A Brief Tide-Turning

However, as the end of a decade of the federal homelessness mission loomed, the tides began to turn. Executive branch leaders offered a striking series of reports that offered a rigorous and contemplative assessment of the past decade’s efforts, while revisiting these strategies for the future. In October 2020, prior to the presidential election, USICH put out a remarkably different plan called “Expanding the Toolbox: The Whole-of-Government Response to Homelessness” (USICH 2020). This plan dovetailed with a report issued by the Council of Economic Advisors just 1 year prior: “The State of Homelessness in America” (CEA 2019).

The CEA report offered a series of careful critiques of the federal mission. After affirming that federal policy and funding was responsible for the surge in permanent support housing, rapid rehousing, and the corresponding Housing First approach, the report offered two major critiques. First, the report reviewed academic literature studying the trends of the period, concluding that “research suggests that previous Federal policy is not capable of explaining a large portion of the reported decline in homelessness between 2007 and 2018.” Second, the report questioned whether homelessness was in fact decreasing at all—even at the minimal rate reported in annual estimates. Although the raw numbers indicated a 15 percent reduction in total homelessness from 2007 to 2018, the report noted that definitional inconsistencies may have driven this change. Specifically, persons in transitional housing programs—the intermediate-length, linear model program disfavored by the Housing First hegemony—are counted as homeless. But persons in rapid rehousing programs—the Housing First substitute—are not counted as homeless. Because the inventory of transitional housing beds plummeted by over 100,000 beds and was replaced by rapid rehousing programs, a similar number of people could be served by the homeless services industry while numbers were improving.

The remarkable CEA report garnered considerable debate in academia (O’Flaherty 2020) as well as opposition from many homelessness advocates (National Alliance to End Homelessness 2019). Nonetheless, when USICH introduced a revised strategic plan in 2020, it was evident that the report reflected a substantive shift in the administration’s homelessness mission. “Expanding the Toolbox” cited the CEA report on page one. The plan also acknowledged that “despite significant increases in funding and beds, overall homelessness has been increasing in the United States” (USICH 2020, p. 1).

In addition, for the first time in a decade, the plan indicated a significant reconsideration of the supremacy of Housing First: “prioritizing housing first as a one-size-fits-all approach has not worked to reduce homelessness for all populations and communities” (ibid., p. 1). In the plan, an entire section devoted to Housing First provided a potent critique of the federal emphasis on Housing First. The plan asserted that the government’s mission effectively imposed housing first as a “one-size-fits-all” approach that “has produced concerning results” (p. 11). In turn, “Housing First should be considered as one tool in the toolbox, but not the only tool in the toolbox” (p. 11).

In contrast to the prior plans, this approach called for a more heterogeneous set of locally driven solutions. Some of the most notable deviations from prior strategies included support for programs that emphasized job training and related work requirements, trauma-informed care, and flexible programming for various homeless subpopulations. Furthermore, the plan also removed the specific goals of the earlier plans entirely. Instead of targeting the end of homelessness for particular subpopulations, the plan emphasized the development of solutions that would facilitate innovation in program delivery and self-sufficiency for clients. In this, the plan asserted (p. 12):

Our aspirational goals should expand our thinking to move beyond the basic goal of providing subsidized housing assistance. As Congress has suggested, we must optimize self-sufficiency through the reduction of reliance on public assistance and implement policies that pursue this as an end goal. Communities should prioritize projects that increase self-sufficiency. Regulatory constraints should be removed, and innovation should be encouraged. Program quality should be measured by reductions in homelessness and by increases in exits from any kind of subsidized housing to unsubsidized market rate housing.

Another goal revision was a call for an emphasis on “outcomes” instead of “outputs,” where the latter relates to the programs themselves (e.g., success as implementing Housing First), while the former relates to substantive, long-term changes in the experience and prevalence of homelessness. In sum, the end of the Trump administration marked a thoughtful and honest assessment of the prior decade’s homelessness mission considering the observed results.

Lessons (Not) Learned

Unfortunately, the revised vision of “Expanding the Toolbox” was short-lived. Following the 2020 election, the Biden administration’s USICH returned with vigor to the earlier period’s policies. Given that then-Vice President Biden was a part of the administration that had so strongly championed Housing First, this return to the Obama-era federal policies was somewhat unsurprising. However, the confidence with which these strategies were reasserted stood in stark contrast to their observable results. In December 2022, USICH published “All In: The Federal Strategic Plan to Prevent and End Homelessness” (USICH 2022). The new plan “restores the importance of Housing First” (ibid., p. 5). Notably, while prior plans were cautious in the application of Housing First evidence to the chronic homeless population, the new plan aspires “To truly bring Housing First to scale for all populations” (p. 42). “All In” also doubled down on claims of the effectiveness of Housing First: “When implemented to fidelity, the model is a proven solution that leads to housing stability as well as improvements in health and well-being” (p. 45). As such, the new qualifier when implemented to fidelity effectively asserted that providers were not implementing Housing First “correctly”—a point to which we will return below. While similar statements were made earlier in the period, the evidence for such conclusions proved far weaker by this time. Most dubious are the claims about health and well-being, as summarized in a 2018 consensus study, “Overall, except for some evidence that [permanent supportive housing] improves health outcomes among individuals with HIV/AIDS, the committee finds that there is no substantial published evidence as yet to demonstrate that PSH improves health outcomes or reduces health care costs” (National Academies of Sciences, Engineering, and Medicine 2018, p. 4). Nevertheless, the plan reiterated the intention to leverage federal funding to pressure the adoption of Housing First principles by providers (pp. 46, 49, 60). As of spring 2023, USICH’s website stated, “USICH believes in evidence-based practices, particularly Housing First. Compared to other interventions, Housing First has been proven to quickly rehouse people and to reduce the likelihood of experiencing homelessness again.”Footnote 2

Along with this recommitment to Housing First, the plan also reintroduced a single, measurable goal. Far shy of the ambitious benchmarks of the prior decade, the plan’s goal was to “reduce overall homelessness by 25 percent” from 2022 to 2025 (p. 70). Estimates from 2022 are the most recent ones available as of this writing. Only time will tell if doing the same thing over again will produce different results this time around.


The United States’ federal mission to prevent and end homelessness through the twenty-first century offers profound lessons that call into question the prospects of an expanded role for the State to mobilize and drive societal responses to grand challenges. The case illustrates how most or all of the necessary ingredients for a successful moonshot mission were present: bipartisan political support, increased government expenditures, a laser-like focus on evidence, clear indicators of success, strong federal leadership, broad and cross-sector collaboration, and a stated emphasis on local engagement and solutions. In fact, few initiatives in history can boast such a constellation of promising factors. And yet, even if such conditions are necessary for success, it is unambiguously clear that these factors were not sufficient to achieve the desired results.

A number of scholars have carefully assessed federal homelessness policy to offer specific explanations for the shortcomings observed (Corinth 2017; Eide 2022; Lucas 2017). Rather than delving into this literature, we conclude by considering how the case of homelessness offers general insights into the prospects of state-led missions directed at grand challenges.

The Mirage of “What Works”

One of the most striking aspects of the federal homelessness mission was the relentless emphasis on “evidence.” That Housing First was an “evidence-based” policy became a quasi-religious axiom in homelessness discourse across public agencies, homeless service providers, and advocates alike. It is safe to say that it would have been difficult to mobilize a comparable level of bipartisan, cross-agency political support for massive increases in homelessness spending and programs without the rallying cry behind “the evidence.” In turn, calls to focus on “what works” were repeated so frequently that Housing First’s ability at ending homelessness became something of a taken-for-granted fact among stakeholders in the homelessness context. By letting evidence “guide” policy, proponents argue, the only obstacle becomes something like “political will” needed to fund that policy to fruition.

The federal homelessness mission reveals several flaws in this premise. The first is that evidence cannot “guide” anything. Evidence comes from experts, and experts are human beings subject to imperfect knowledge and error (Koppl 2018). Evidence is not applied automatically to social problems, but is interpreted, negotiated, and extrapolated by policy actors with limited information and a variety of interests (both well-meaning and self-serving). As such, evidentiary information must be considered in the context through which the evidence is produced and utilized. In this context, Lucas (2018) suggests that evidence-based policy can be understood through the analytical lens of “public entrepreneurship,” wherein actors leverage and apply evidence subjectively to achieve their private interests for policy change. He documents how the cultivation of “evidence” was pivotal to the policy dominance of Housing First and the broader emergence of a federal homeless mission. However, the problem was that Housing First quickly came to be prescribed for many purposes beyond what the academic research actually showed (Eide 2020).

Most of the studies about Housing First’s efficacy at housing retention showed results (i) for a specific homeless subpopulation, the chronically homeless, and (ii) at the individual level. It was clear that individuals who had repeated or long-term bouts of homelessness with a cooccurring physical and/or psychiatric disability were, on average, more likely to become and stay housed through the Housing First programs that followed the Pathways Housing First approach as introduced in New York City. Notably, however, these studies said nothing about the actual goals of homelessness policy: the substantive reduction of homelessness across many groups at the population level. In other words, the question of whether a Housing First program improves housing retention for an individual is not the same question as whether “retooling” the homeless shelter system to a Housing First approach will reduce the amount or rate of homelessness in a community. The federal homelessness mission inappropriately extrapolated the evidence beyond this unique group of chronically homeless individuals to many different homeless subpopulations, who experience homelessness very differently in terms of duration, repetition, and services utilized. It also inappropriately extrapolated the insights about individual-level housing outcomes to community level results—an archetypal example of the “atomistic fallacy.” In fact, the research that has emerged since that time casts doubt on whether increases in permanent supportive housing bed provision correspond to any reductions in community-level homelessness (Corinth 2017). Ultimately, federal funding’s main function has been to provide more beds in the shelter system and thereby increase the total number of people who are counted as homeless (Lucas 2017).

The two epitaph quotes reveal a core problem of “what works,” i.e., the limits of evidence as a tool for state-led social and environmental missions. The evidence does not speak for itself but the political process uses it in a manner that oversimplifies, overpromises, and underperforms. Whereas the federal strategic plan from 2010 emphasized that “homelessness is solvable,” careful academic researchers a decade later were quick to conclude that we have yet to find clear evidence on how to achieve this goal. It appears that “what works” remains an open question.

An increasingly common response by Housing First advocates to the above analysis is that Housing First was not actually tried at scale throughout the period. To be fair, researchers have developed careful scales of fidelity to the programmatic tenets of Housing First. Using these, they have found considerable variance across programs that purport to utilize the approach (Gilmer et al. 2013). As such, it is unclear how much of the growth of permanent supportive housing and rapid rehousing bed types conforms to a Housing First approach. Like modern-day socialists, the claim is something like, “actual Housing First has never really been tried.”

For our purposes, this argument only weakens the federal homelessness policy’s premises. For one, consider the claim of the relative cost-effectiveness of Housing First, which is questionable considering the massive funding increases over the period and limited results. To evaluate and monitor each local provider’s programming for fidelity to Housing First is both a major oversight challenge and practical cost that would dramatically increase the requirements to implement. In addition, the word “fidelity” does not appear in any of the strategic plans until 2022—indicating that this predictable concern was not a consideration in program success even as funding was doubling to nearly USD 9 billion annually.

While one might say it is hard to predict such a setback in advance, that is precisely the point. Whether fidelity concerns “could” have been foreseen by policy makers, they evidently were not. The fidelity counterargument thus illustrates a pragmatic reality endemic to many efforts to engage with all complex social challenges: many obstacles to success are only observable ex post. Because problems like hunger, homelessness, health, and crime are complex and interdependent, promising solutions often run into unforeseen obstacles and failures. What is needed for such problems is not “prescription” of “the solution,” but support for the emergence of new solutions and the development of feedback mechanisms to reward those that may show promise in a particular context. By contrast, top-down efforts to mobilize a specific “solution” to such a problem in ways not backed by research and for purposes beyond the extant research basis mistakenly prioritize “outputs” over “outcomes”—and are almost certain to fall short in the process.

Whither State Leadership?

Finally, the case of the federal homelessness mission illustrates the profound tensions associated with state “leadership” that is also supposedly “collaborative” and open to “bottom-up” solutions. In the development of its strategic plans, notices of funding availability, and implementation of programs, the federal government relentlessly worked to emphasize “local” context, knowledge, and collaboration in the fight against homelessness. The word “local” is mentioned 74 times in the 2010 federal strategic plan, which even indicates that each community’s homelessness strategies “should be locally driven, reflecting local conditions, since a one-size-fits-all plan does not exist.” The plan goes on: “interdisciplinary, interagency, and intergovernmental action is required to effectively create comprehensive responses to the complex problem of homelessness” (USICH 2010, p. 30).

On the surface, the federal government’s leadership was thus structured to facilitate the use of local knowledge and context-specific, bottom-up solutions—an approach that some scholars explicitly call for (Lucas 2020). However, the reality was far removed from the stated commitment to bottom-up, emergent organizing. Communities were invited to develop their own strategies and introduce innovative programs…so long as these accorded with “best practices”—aka Housing First. Federal funding decisions, in turn, were not based on the success of programs, but on organizations’ and communities’ commitment to the programs that the federal government prioritized. The demise of transitional housing during this period is clear evidence of this; such programs were essentially penalized simply for their programmatic features—regardless of how successful they may (or may not) have been. Little to no funding was reserved for alternative but promising programs that some organizations were implementing. For instance, the New York City Doe Fund’s Ready, Willing and Able program was a work empowerment program purporting to offer highly effective programming for previously homeless persons. Austin Texas’ Community First was a tiny home community that offered a variety of social and economic activities and programs alongside housing. Unfortunately, there is not yet robust evidence about the efficacy of these alternative programs. But this does not mean they are inferior; it only means they are understudied. True federal leadership would facilitate such locally driven experimentation and support the evaluation of promising alternatives so that effective programs can be identified and replicated as appropriate. Overall, the “collaborative leadership” of the federal government in fact moved significantly toward centralizing the approach to an already top-heavy homeless response system, functionally eliminating any substantive collaboration that would enable the emergence of entrepreneurial, flexible, and bottom-up solutions to this complex problem.


The philosopher George Santayana famously wrote, “Those who cannot remember the past are condemned to repeat it.” While the United States’ misadventures in its mission to prevent and end homelessness in the early twenty-first century are but recent history as of this writing, its lessons are notable for those who would seek to promulgate a future where the state is the lead player in developing and driving missions to face grand social challenges. Homelessness is an important problem for societies to grapple with, especially those societies that have generated immense wealth and technological advance. However, if the federal government persists with a monolithic focus on failed solutions of the past—as the most recent federal strategic plan indicates—then we must be prepared for the letdowns of the past to reemerge as well.