Direct cash transfer programs have been popular tools to alleviate poverty in developing countries over the last two decades. Because of this, programs in North America have rapidly emerged in recent years, with some aimed at people experiencing homelessness. As the research base quickly grows, it is critical to define these programs, evaluate their effectiveness, weigh key decisions, explore challenges, and identify future areas of study. With a focus on the United States but attention to global programs, this brief draws on cutting edge research as well as past and ongoing programs to achieve these aims.
At a fundamental level, direct cash transfer programs provide monetary assistance to recipients directly, rather than distributing goods or services through intermediary organizations (e.g., food or housing). Proponents of cash transfer programs emphasize their simplicity, flexibility, and effectiveness — by providing cash directly, recipients make their own decisions about how best to use financial resources to meet their varying needs.
This research brief summarizes five key distinctions related to cash transfer programs: whether they are conditional (vs. unconditional), universal (vs. targeted), contributory (vs. non-contributory), cash (vs. cash plus), and immediate (vs. delayed).
The brief also describes two pilot cash transfer programs aimed at individuals experiencing homelessness, one in Canada and one in the United States, as well as a related program using “Centralized Diversion Funds” in the United States.
What is a direct cash transfer?
At a fundamental level, direct cash transfer programs provide monetary assistance to recipients directly, rather than distributing goods or services through intermediary organizations (e.g., food or housing). Proponents of cash transfer programs emphasize their simplicity, flexibility, and effectiveness — by providing cash directly, recipients make their own decisions about how best to use financial resources to meet their varying needs.
Many different types of cash transfer programs have undergone evaluation. Outlining key definitions can help to better identify and test specific characteristics of direct cash transfer programs.
Conditional vs. Unconditional
A core distinction between cash transfer programs is whether the receipt of financial assistance is conditional or unconditional. Conditional cash transfer programs provide financial assistance only if certain requirements are met or behaviors are performed by the recipient. These programs have often been used to target the health, nutrition, or education of children, for example by mandating school enrollment or ensuring regular health check-ups (Millan et al., 2019). One of the most prominent conditional cash transfer programs was the “Prospera” program (formerly Oportunidades and Progresa) in Mexico, which ran from 1997 to 2019 (Martinez-Martinez et al., 2019; World Bank, 2014). This taxpayer-funded program used government resources to provide a monthly payment to eligible families for each child currently enrolled in school. Prospera served as a model for many conditional transfer programs around the world, most of which have been implemented in low- or middle-income countries. The United Nations maintains a global database of these programs, and many scholars have reviewed their effectiveness (Kabeer & Waddington, 2015; Lagarde et al., 2009; Millan et al., 2019; Owusu-Addo & Cross, 2014). One effort to implement a conditional cash transfer program in the United States took place in New York City from 2007 to 2010, but achieved mixed results (Riccio & Miller, 2016).
In contrast, unconditional cash transfers provide financial assistance to recipients with “no strings attached” (Marinescu, 2018). These programs may provide locations, resources, or advice on how to access supplemental services (e.g., money management training), but financial benefits are not contingent on enrollment. Historically, large-scale unconditional cash transfer programs have taken the form of governmental tax breaks or dividend programs (Fuller et al., 2022; Marinescu, 2018). However, in recent years, relatively smaller pilot studies of unconditional cash transfer programs have emerged. Funded publicly or privately, these programs aim to establish their effectiveness in alleviating poverty, to generate support for larger programs. Dwyer and colleagues (2023) and GiveDirectly have created separate, publicly available databases of unconditional programs and traction is growing in the United States. For example, many U.S. mayors and counties advocate for unconditional cash transfers through Mayors for a Guaranteed Income (MGI). According to their 2023 year-end report, MGI has funded over 60 pilot programs in its first three years, delivering “at least $250 million in direct, unconditional relief to everyday Americans.”
Universal vs. Targeted
Another important characteristic of cash transfer programs is whether they are universal or targeted. Universal cash transfer programs, oftentimes called a “universal basic income,” involve monetary payments to everyone in a geographic area, regardless of need (Marinescu, 2018). Many countries implemented universal cash transfer programs during the COVID-19 pandemic, although some countries like the United States placed limitations on high-income earners (making the payments not technically universal).
Some scholars further distinguish between a “universal basic income” and a “universal base income.” A basic income is one that is sufficient to live on, generally amounts that are comparable to social security benefits. Theoretically, such a model could replace means-tested welfare programs, like social security and Medicare (Thigpen, 2016). Conversely, a base income model provides more modest distributions, around several hundred dollars per person, per month. These models provide an income floor but not enough money to live on, meaning recipients still need to find work or other forms of support (Thigpen, 2016). Universal income programs have entered and exited US political and popular discourse since the 1960s. Due to the size of investment needed, large scale universal programs must rely upon a shared resource, like the Alaska Permanent Fund or the Eastern Cherokee Nation’s casino profit sharing program (Akee et al., 2010). Alternatively, they must be funded by raising taxes or reducing existing programs, both of which can be politically treacherous.
To overcome the sticker-shock of universal programs, targeted cash transfers disburse funds only to individuals that meet certain criteria. Eligibility criteria may be as simple as being below a certain income level, such as those used by the negative income tax experiments in the U.S. during the 1960s and 1970s (Levine et al., 2005). Alternatively, criteria may be as specific as a recipient’s housing situation (e.g., those experiencing homelessness; Dwyer et al., 2023) or a recipient’s familial role (e.g., low-income mothers; Aizer et al., 2016). One drawback of targeted programs is that screening for eligibility is costly and time-consuming to both the funding organization and the recipient; even so, these programs overcome a primary weakness of universal programs, which is leakage to the non-poor (Besley, 1990).
Contributory vs. Non-contributory
Most direct cash transfer programs are non-contributory, meaning that recipients do not need to provide financial resources themselves to receive a benefit. This sets direct cash transfer programs apart from other forms of contributory assistance, like employment pensions (Zhao et al., 2023 [Ch. 1]).
Cash vs. “Cash Plus”
Although many cash transfer programs provide cash alone as their intervention, it is becoming increasingly common to implement “cash plus” models. Such approaches combine cash transfers with other interventions or services, like mental health support, behavior change programming, or instrumental resources (e.g., linking recipients to health care, food services, or educational opportunities; Roelen et al., 2017). These “plus” features target factors that are theorized to be necessary for cash to be most effective — by improving recipients’ mental well-being or helping them connect with a medical provider, cash can better influence development outcomes and more permanently lift individuals out of poverty (Little et al., 2021). Cash plus programs are increasingly used to support children in low- and middle-income countries, where meta-analytic evidence indicates combining cash with food transfers or primary health care improves child health and well-being to a greater extent than cash alone (Little et al., 2021). Whether this helps in Western contexts is less clear, however. One pilot program in Vancouver, B.C., targeting individuals experiencing homelessness, found that cash alone was equally as effective as cash paired with a money management workshop and financial coaching (Dwyer et al., 2023).
Immediate vs. Delayed
Lastly, some cash transfer programs prioritize flexibility by giving a larger, one-time payment up front to recipients. Alternatively, other programs disburse smaller monthly or annual payments to foster reliability and promote income stability over time. Some global evidence suggests that a larger, single payment increases spending on durables, and improves psychological well-being and empowerment to a greater extent than smaller monthly transfers, whereas smaller and regular transfers improve food security to a greater extent (Haushofer & Shapiro, 2016). Reflecting these mixed effects, some researchers conclude that “consistent, long-term cash transfers have been shown to have more beneficial effects than one-time payments,” (p. 25, Neighly et al., 2022) while others have stressed that lump sums enable “maximum purchasing freedom and choice (e.g., rent, durable goods), whereas smaller repeated transfers [do] not” (p.2, Dwyer et al., 2023).
As an extreme example of delayed payments, Harlem Children’s Zone is piloting a savings program in which K-12 youth are given $10,000, which cannot be accessed until they are 25 years old. The organization will use professional investors to manage the money, require youth to hit certain milestones to receive it (e.g., graduating high school and college), enroll recipients in financial literacy courses, and set up a board to ensure the funds are used for “wealth building purposes.” Using the definitions laid out in this research brief, we can classify this program as conditional, targeted, non-contributory, “cash-plus,” and delayed.
What are some known benefits of direct cash transfers?
Now that this brief has addressed core characteristics of cash transfer programs, what does the literature say about their effectiveness? This is a challenging question, because cash transfer programs are necessarily context dependent (Zhao et al., 2023 [Ch. 2]; Glassman et al., 2013). Each program unfolds in a specific social context with unique eligibility criteria and disbursements, all of which influence generalizability. Moreover, the specific objectives of a program (e.g., reducing the number of nights spent homeless vs. improving child nutrition) inform its methods (e.g., cash plus housing vouchers vs. cash plus child health care access).
Nonetheless, researchers have reported on general impacts of cash transfers by looking across many programs and summarizing findings using meta-analysis or systematic review. These studies have largely supported the efficacy of conditional and unconditional cash transfers in lifting individuals out of poverty and improving a range of positive outcomes.
Conditional Cash Transfer Evidence
Conditional cash transfer programs have mostly been implemented in low- and middle-income countries targeting families with young children. The short-term effectiveness of these programs has been well established, with reviews showing that conditional cash transfers raise savings and investments by around 10-20% of the size of the transfer without reducing labor force participation (Kabeer & Waddington, 2015), positively influence household consumption by better balancing spending and saving (Kabeer & Waddington, 2015), and increase the use of preventative health services while improving objective and subjective health (Lagarde et al., 2009). Regarding young children, conditional cash transfers lower rates of child labor, particularly for boys (Kabeer & Waddington, 2015), improve school enrollment and attendance (Baird et al., 2013), increase the uptake of maternal and newborn health services (Glassman et al., 2013), and generally benefit child health by improving access to care, increasing child and maternal nutrition and immunizations, and reducing morbidity risk and household poverty (Owusu-Addo & Cross, 2014). Conditional cash transfers may be more effective in middle-income (vs. low-income) countries, as these countries more often have functional health systems that recipients can use their cash to access (Lagarde et al., 2009; Owusu-Addo & Cross, 2014).
Some reviews have tried to estimate longer-term effects of conditional cash transfers, asking whether these programs break intergenerational cycles of poverty. Evidence for this is considerably more mixed: conditional cash transfers appear to provide some important long-term impacts on schooling, but evidence for other outcomes is limited due to methodological challenges or weak/non-existent long-term impacts (Millan et al., 2019).
Unconditional Cash Transfer Evidence
Because unconditional programs place no requirements on how cash is used, their effectiveness is relatively more comparable across contexts. There has also been increased funding and research of unconditional transfer programs in western countries, allowing for reviews in the United States and Canada (Fuller et al., 2022; Marinescu, 2018; Neighly et al., 2022), low- and middle-income countries (Pega et al., 2022), and global comparisons (Zhao et al., 2023 [Ch. 1]; Siddiqi et al., 2018). Taken together, positive short-term benefits of unconditional cash transfers appear across many contexts, especially for poorer and more vulnerable recipients (Neighly et al., 2022).
In the United States and Canada, unconditional cash transfers have been shown to reduce poverty (Berman & Reamey, 2016) and contribute to financial stability (Neighly et al., 2022). They also appear to positively impact nutrition and food security (Marinescu, 2018; Neighly et al., 2022), improve education outcomes (e.g., school attendance, grades, test scores, and educational attainment; Marinescu, 2018; Neighly et al., 2022), and benefit the health, development, and cognitive skills of children (Fuller et al., 2022). Although less frequently studied, evidence from the U.S. and Canada also suggest that unconditional cash transfers may be associated with reduced hospitalizations (Marinescu, 2018), improved mental health (Akee et al., 2015; Neighly et al., 2022), decreased self-reported criminalized activities (Marinescu, 2018) and reduced alcohol and cannabis dependence (Akee et al., 2015). Importantly, unconditional cash transfers do not appear to significantly discourage recipients from working (Marinescu, 2018). In fact, one pilot program in Stockton, California, found the opposite, reporting that the full-time employment rate among recipients of a monthly, $500 unconditional cash transfer rose from 28% to 40% after one year of payments, outpacing the control group which rose from 32% to 37% (West et al., 2021).
Evidence regarding unconditional cash transfers in low- and middle-income countries largely parallels emerging findings in the U.S. and Canada. In these contexts (where absolute poverty levels are higher on average) unconditional cash transfers have been shown to reduce the risk of living in extreme poverty and the likelihood of having any illness in the last few months, while increasing food security, dietary variety, health care spending, and school attendance (Pega et al., 2022). A review of global programs that target pregnant women or mothers of children under one year old finds they improve infant health as measured by birth weight and infant mortality (Siddiqi et al., 2018).
Evidence Synthesis and Conclusions
Some reviews have directly pitted conditional and unconditional cash transfers against one another to test if one method is superior. In low- and middle-income countries, high-quality reviews conclude that there are no meaningful differences between the effectiveness of conditional and unconditional cash transfers (Pega et al., 2022; Siddiqi et al., 2018). In high-income countries however, reviews have suggested that work conditionalities (e.g., making cash payments contingent on employment) are detrimental to the health status of recipients (Siddiqi et al., 2018). One review focused on the U.S. and Canada also reported that beneficial results for educational outcomes are greater for unconditional as compared to conditional programs (Neighly et al., 2022).
In sum, both conditional and unconditional cash transfers appear effective at improving a range of financial, health, and education outcomes. Conditional programs have a longer research history and are mostly conducted in low- and middle-income countries, whereas unconditional programs are relatively newer approaches favored by Western nations like the U.S. and Canada. Although most evidence is on individuals in poverty (rather than individuals experiencing homelessness), recent reviews suggest that programs are most beneficial for the poorest and most vulnerable recipients (Neighly et al., 2022). This means these programs may be highly effective at helping people experiencing homelessness, and two pioneering programs in the U.S. and Canada are starting to test these questions directly.
What are some examples of direct cash transfer programs targeting people experiencing homelessness?
As referenced earlier, GiveDirectly, Dwyer and colleagues (2023), and the United Nations maintain databases of active and past cash transfer programs. Honing in on Community Solutions’ interests, this research brief describes two pilot cash transfer programs aimed at individuals experiencing homelessness, one in Canada and one in the United States, as well as a related program using “Centralized Diversion Funds” in the United States.
New Leaf Project, Vancouver, B.C.
Beginning in 2018, researchers at the University of British Columbia partnered with Foundations for Social Change to pilot the first documented North American direct cash transfer program for individuals experiencing homelessness. Through the New Leaf Project, recently homeless individuals were randomly assigned to either receive a one-time, unconditional cash transfer of $7,500 CAD (n = 50) or to a control condition that did not receive the cash transfer but were paid a small honorarium for completing surveys (n = 65). The cash transfer — provided electronically to a bank account set up for all participants — was roughly equivalent to the 2016 annual income assistance in British Columbia or approximately 60% of participants’ average annual income ($12,580). To ensure the cash did not impact other benefits, the program secured an agreement with the local government. Researchers were also interested in whether a money management workshop and financial coaching were beneficial, so they randomly assigned participants to receive this training or not. To mitigate any potential risks from providing cash transfers to some participants but not others within the same shelter, researchers used a clustered design in which shelters (n = 22) were randomly assigned to experimental conditions.
Eligible participants had to be Canadian citizens between 19 and 65 years old who were recently homeless (less than two years) and did not have severe levels of substance use, alcohol use, or mental health problems. These screening criteria were implemented to minimize the potential harm of a cash transfer (e.g., due to overdose) and to give the pilot program the best opportunity to succeed. However, this resulted in only 31% of all screened participants (229/732) being eligible. This means that the sample represents a “high-functioning subset of the total homeless population in Vancouver” and that findings “may not extend to people who are chronically homeless, or experience higher severity of substance use, alcohol use, or psychiatric symptoms” (pp. 6-7, Dwyer et al., 2023).
Researchers followed-up with participants at one, three, six, nine, and 12 months after disbursement to measure a range of outcomes, including housing stability, employment, education, income, spending, executive function, subjective well-being, food security, substance use severity, and social service use. Their primary finding was that over the course of the year, cash recipients spent 99 fewer days homeless and 55 more days in stable housing on average than control participants. Researchers estimated that reductions in shelter nights as a function of these cash transfers led to societal net savings of $777 per person. Recipients of the cash transfer also reported that they retained more savings ($1,160) and increased their monthly spending more ($429) than control participants. Most of this increased spending was on durable goods (e.g., car, furniture), rent, food, and transit, while spending on temptation goods (e.g., alcohol, drugs, cigarettes) did not differ between groups.
Effects of the cash transfer were most pronounced in the first three months, possibly due to the high cost of living in Vancouver (e.g., participants spent most of the money quickly), the fact that control participants eventually “caught up” over time (Segaert et al., 2017), or because attrition decreased statistical power to detect effects at later time points. There were limited effects on cognitive and well-being outcomes, save for cash recipients showing higher positive affect at 1 month and higher executive functioning at three months. The money management workshop and financial coaching were also not effective, perhaps because they were focused on aspirational goals and money planning. Debriefing interviews suggested that instrumental supports like getting IDs replaced, completing resumes, or finding affordable housing, may have been more helpful.
In summary, this cash transfer program — the first-of-its-kind in North America — reveals that a one-time unconditional cash transfer can reduce homelessness and increase housing stability while generating net savings for society. Even so, these findings are limited by their small sample (n = 115) and strict eligibility criteria (recently homeless without severe alcohol, drug, or mental health problems). For additional details, see their published paper, the University of British Columbia’s press release, or Foundations for Social Change’s website.
Denver Basic Income Project
Initiated in 2022, the Denver Basic Income Project (DBIP) explores whether moderate monthly payments or a larger up-front payment followed by smaller monthly transfers provide differential benefits to those experiencing homelessness. Participants from the Denver area received $1,000 a month for 12 months ($12,000 total), $6,500 upon enrollment and $500 a month for the next 11 months ($12,00 total), or $50 a month for 12 months ($600 total; the active control group). Person-level randomization was used to assign participants to one of the three conditions. In total, 807 participants were enrolled from November 2022 to February 2023. All participants were given a cell phone to complete surveys and could receive cash through direct deposit or a reloadable debit card. The cash provided was unconditional (no workshops or training were offered), but DBIP did not arrange for recipients to have their existing public assistance unaffected.
Eligible participants had to be 18 years old or older, access services from at least one of DBIP’s partner agencies, not have severe and unaddressed mental health or substance use needs, and be experiencing homelessness. DBIP defined homelessness as “individuals without fixed, regular, and adequate nighttime residence” (p. 4, Brisson et al., 2023). Similar to the New Leaf Project, findings must be interpreted cautiously, as participants with severe mental health or substance use needs were ineligible.
To evaluate the effectiveness of this project, DBIP commissioned the Center for Housing and Homelessness Research at the University of Denver to produce a 6-month interim report and a one-year evaluation. These reports were based on biweekly text-based surveys and long-form surveys completed by participants at enrollment, six months, and ten months into the study. The center also conducted interviews with approximately eight participants in each experimental condition, documented in an interim and final qualitative report. Using these mixed methods, researchers captured a variety of housing, financial, physical health, mental health, family, social network, and public service use outcomes from participants.
Quantitative findings
Of the 807 participants who were enrolled in the study, 631 completed a baseline survey, 457 completed a six-month survey, and 396 completed a 10-month survey. Participants who completed all surveys were roughly evenly distributed across the three experimental conditions. The main takeaway from this study is promising but counter-intuitive: all payment groups similarly improved on key outcomes, including participants in the control group that received only $50 per month. Notably, the proportion of all participants who reported renting or owning homes increased from between 6-12% at baseline to around 43-48% ten months into the study. Consistent with this finding, the proportion of participants living in housing they considered to be stable increased from around 20% at enrollment to around 50% at ten months. All experimental groups experienced roughly equivalent improvements in perceived financial well-being and equivalent decreases in food insecurity, time spent accessing resources, and parenting distress as well. One of the only outcomes that differentially improved for the higher payment groups as compared to the control group was participants’ ability to pay bills: the percentage of participants who were able to pay their bills in the two higher payment groups doubled from enrollment to 10 months, whereas participants in the control group reported only a slight increase. Outcomes that remained largely unchanged over the course of the study for all three groups were participants’ transportation security, as well as their health, energy, stress, anxiety, hope, and agency.
How did participants spend their money? Based on the 461 participants who accepted a DBIP issued debit card and consented to having their spending data tracked, most spent payments on manual cash disbursements (42.3%), retail outlets (24.3%), or utility services (5.2%). Regarding overall cost effectiveness, the researchers posit that the program resulted in public service cost savings of approximately $589,214 for the 342 participants that provided these data. This figure was based on reductions in self-reported hospital, jail, and shelter utilizations from enrollment to 10 months. Importantly, this cost saving figure does not take into account the amount DBIP spent on the program, which for these 342 participants would have been approximately $2.8 million in payments alone (114 participants receiving $600 each [control group], plus 228 receiving $12,000 each [experimental groups]). While the report prepared by DBIP indicates substantial public service cost savings, this should not necessarily be interpreted as evidence of cost-effectiveness.
Qualitative findings
Interviews with participants of DBIP gave some insight into how exactly cash payments may improve well-being for individuals experiencing homelessness. Participants in all groups expressed gratitude for the DBIP, commenting specifically on how the flexibility in how they could spend the funds set this support apart from other services like SNAP. Once participants felt they were able to trust that DBIP payments would come in monthly, some remarked on how these payments helped fill gaps between other services and gave them room to breathe. Some commented on being able to spend money on products they normally would not purchase for others (e.g., presents for friends or family) or themselves (e.g., new clothes), but needing to dismantle a sense that these purchases were frivolous, after having to go without them for so long. Most participants were parents, and these individuals talked extensively about the positive impact the cash transfer had on their finances and relationships with children or grandchildren.
Participants in the higher payment groups suggested that receiving these transfers largely decreased their anxiety and stress and instilled a sense of hope and relief. Many reported they had greater confidence and agency in managing their money at the 10-month interview. Although these findings did not emerge in the quantitative data, it may indicate that greater sums of money do improve participant mental health whereas smaller amounts have no impact on mental health. Some unexpected challenges emerged for participants receiving the larger payments, however. Specifically, almost all of these recipients noted a decline in other benefits (e.g., SNAP), with some reporting that they fully lost access to SNAP. Others also reflected on minor stressors and anxiety that these payments engendered, including having to navigate new financial responsibilities, family dynamics, or service systems (e.g., housing). In all experimental conditions, participants noted how structural barriers remained that cash could not solve, including prior evictions and criminal records. When asked to reflect on the end of the program, many participants expressed feelings of gratitude mixed with sadness, fear, and anxiety.
Summary and Limitations
On the whole, participating in the DBIP improved participants’ housing and well-being similarly, whether one received $1,000, $500, or $50 per month. Why would a $50 per month payment result in equivalent improvements relative to $1,000 per month or a $6,500 lump sum and $500 per month after? This is a difficult question to answer, but some possibilities are worth considering. First, the DBIP offered a suite of services to all treatment groups, including a phone with 12 months of service, a debit card, other DBIP program services, and bi-weekly opportunities to complete additional surveys for more financial compensation. They used this active control group approach to better respect participants’ time and energy, but this may be an explanation for why the active control group looked so similar to experimental groups. Second, the authors of the report point to other factors outside the study that may serve as confounding variables, including the availability of temporary housing vouchers during the COVID pandemic. Factors such as these, that all participants had access to, may be why all participants’ outcomes improved similarly, not the cash transfers themselves.
Like the New Leaf Project, participants could not have severe substance use or mental health conditions, which limits generalizability. A further limitation of the DBIP study is that all participants knew which group they were assigned to when they completed the baseline survey. This may have contaminated the baseline data, as a participant might be jealous, frustrated, or elated when completing their initial survey depending on what experimental condition they were assigned to. Relatedly, researchers noted that participants at the 10-month time point were stressed and concerned about the end of the program, which may have influenced these responses. Lastly, like the New Leaf Project, substantial participant attrition may have biased results. For example, participants that stopped responding to surveys may have differed in important ways (e.g., were experiencing worse outcomes) from those who completed all surveys. In their final report, researchers noted that they will continue to examine patterns of missingness to explore this possibility.
In summary, the DBIP was an ambitious, pioneering cash transfer program in the United States, accompanied by robust quantitative and qualitative evaluations. This project shows that even small amounts of cash ($50 per month) may help individuals transition out of the homelessness system. Qualitative findings hint at the many possible mechanisms whereby cash helps, which is a notable strength of this contribution. Nonetheless, findings are limited by the selective nature of their recruitment process and lack of a true control group. Future work should consider building off DBIP’s mixed-methods evaluation approach to test the effects of cash transfers using greater experimental control.
Centralized Diversion Funds, Washington State
Although not identical to direct cash transfers, some funders disburse cash with limited strings attached using centralized diversion funds. As a primary example, the “A Way Home Washington Centralized Diversion Fund” provides funds to youth and young adults experiencing homelessness or at imminent risk of entering the homelessness system. In this model, local service providers or case managers make a request for one-time financial assistance on behalf of a young person. This request is then processed by a fiscal administrator, and if approved, funds are disbursed via checks, gift cards, bill payments, or other mechanisms no more than 72 hours after the request was received. Fiscal administrators are trained to evaluate whether providing a payment would result in the young person being housed outside of the homelessness system within 30 days. This approach removes administrative burdens, referrals, and wait times for recipients to provide cash support flexibly and efficiently. However, they do not fit a strict definition of direct cash transfers because assistance is requested by a case manager on a per-need basis, rather than via access to “no strings attached” funds.
In their first year, A Way Home Washington supported over 220 youth and young adult households through their centralized diversion fund across four counties in Washington State (Chao et al., 2021). Eligibility criteria were kept as broad as possible, and included any young person between 12 and 24 experiencing a housing crisis or who were unaccompanied or at imminent risk of becoming unaccompanied. Ninety-six percent of recipients did not return to homelessness within three months, and 92% of households were successfully diverted from homelessness. The top three requests were housing deposits, rental assistance or arrears, and rental application fees. The average amount disbursed was $1,356. The report emphasizes the cost-effectiveness of this intervention, claiming that this average cost is substantially lower than a comparable emergency shelter ($2,423) or rapid re-housing approach ($8,649). Additionally, the median processing time for a centralized diversion fund (two days) resulted in less time experiencing homelessness when compared to emergency shelters (13 days) and rapid re-housing (144 days). Finally, the report demonstrated that the flexibility of these funds made them ideal for reaching young people in historically marginalized communities, such as those from minority racial, ethnic, sexual orientation, or gender identity groups, as well as those with disabilities. Critically, this evaluation did not randomly assign recipients to receive a fund or not — a major limitation of this evidence as compared to the evaluations provided by New Leaf and DBIP. Conducting randomized controlled trials is a necessary next step to rigorously test these funds’ effectiveness.
Summary and Emerging Evidence
At the time of writing, the New Leaf Project and DBIP represent two primary, pioneering cash transfer programs for individuals experiencing homelessness in North America. These programs demonstrate the initial effectiveness of these transfers, but findings are limited by the eligibility criteria employed. The A Way Home Washington program suggests that centralized diversion funds may also be a way to provide flexible and efficient financial support to young people experiencing homelessness. Emerging from this initial work in Washington State with centralized diversion funds, Point Source Youth describes additional programs across the country that provide direct cash transfers to youth experiencing homelessness. In their 2022 annual report, they suggest that preventative direct cash transfers (DCT-P) may help stop vulnerable young people from entering homelessness. Their work and others indicate that combining direct cash transfers, centralized diversion funds, and other community approaches can most effectively reduce youth homelessness (Chao et al., 2021; Morton et al., 2020). Hybrid approaches taken by this organization and others are worth monitoring as the research space expands.
What are key program design decisions?
Any group planning to implement a cash transfer program must make numerous design decisions. The distinctions in the definitions section of this brief should be weighed against the budget, scope, and goals of an organization. In addition to these definitional decisions, more specific choices around eligibility, characteristics of the disbursement, study design, and data collection will follow. The two programs described above are useful templates for homelessness-specific programs, and a recent review of 43 unconditional programs in the Global North (39 in North America) and 44 in the Global South can illuminate the approach taken by programs in general (Zhao et al., 2023 [Ch. 1]). The following section breaks down these decision points, with a focus on standard approaches taken in the Global North.
Eligibility Criteria
Who gets into a cash transfer program is a highly defining feature, as it largely informs what conclusions are drawn. Programs in the Global North have targeted no group in particular (n = 15), or specifically enrolled children (n = 11), parents (n = 13), women (n = 9), ethnic minorities (n = 8), or vulnerable groups (n = 6). If vulnerable groups are targeted, having a clear definition of homelessness is critical for evaluating eligibility and communicating findings. Screening out individuals with severe mental health problems, alcohol use, or drug use, appears to be necessary to ensure a pilot program has a strong chance of succeeding and to limit harm (e.g., by reducing the risk of overdose). Both programs described above used screeners to achieve these goals (Cameron et al., 2007; Saunders et al., 1993; Staley & el-Guebaly, 1990). The New Leaf Project also screened out chronically homeless individuals, and some programs like Point Source Youth target highly vulnerable groups (e.g., homeless youth), both of which may be worth considering.
Characteristics of the disbursement
The total amount and frequency of payments are also critical. The amounts given must be enough to impact recipients’ financial situation, but not so great that it restricts sample size. Most programs in the Global North provide an annual cash amount between $6,000 and $13,000 per person (mean = $7,753; median = $6,403) in 2021 US dollars. This is greater than the average amount in the Global South (mean = $1,414; median = $799). Yet, due to the weaker purchasing power of cash in the North, these programs confer an average increase in relative wealth of only 12% (median = 9%) whereas that figure in the South is 37% (median = 17%). Regarding frequency, most programs in the North provide cash transfers monthly (n = 39) while a minority provide weekly (n = 3), quarterly (n = 3), annual (n = 1) or one-time (n = 2) payments. The most common total payment window is one year (n = 19), but some programs last for less than a year (n = 5), for 2 years (n = 12), for 3 to 5 years (n = 9), for more than 5 years (n = 1), or are ongoing (n = 2).
These average figures and the two pilot programs are useful ballparks to begin with when deciding on the amount and frequency of transfers. Importantly, local service providers, community organizations, and potential recipients must be consulted, as on-the-ground knowledge will most accurately characterize the needs of a specific community.
Study Design and Data Collection
To generate actionable and replicable findings, high-quality study design and data collection procedures are necessary. Both the New Leaf Project and DBIP partnered with academic teams at local universities to evaluate the impact of their cash transfer programs. This approach is highly recommended to increase the rigor and impartiality of an assessment. Evaluation organizations should be approached early in the process to help with design, implementation, and evaluation. Randomization, sample size, disbursement method, and outcomes are key decisions to consider.
Randomization: Randomization is almost always needed to test cash transfer programs. This involves comparing recipients’ outcomes to those of a control condition, or a group of randomly selected participants who did not receive a cash transfer. To facilitate retention, it is common to have moderate payments for control groups (e.g., $50 per month or $30 per survey completion). Some programs use cluster randomization, in which shelters (rather than individuals) are randomly assigned to one treatment condition or another. This avoids giving some service users at a shelter a transfer while withholding it from others, which likely increases safety. Some programs in the Global North (n = 5) have included multiple treatment groups (i.e., transfers of different sizes, frequencies, or components) to test the effects of disbursement characteristics. Both DBIP and the New Leaf Project followed this model, though adding conditions increases the required sample size.
Sample Size: The sample sizes of cash transfer studies vary dramatically. On average, programs are smaller in the Global North as compared to the Global South due to implementation costs. Most programs in the North involve between 100 and 1,000 recipients (n = 21), while some recruit fewer (n = 10) and some recruit more (n = 10). The median number of recipients in the North is 130. When landing on a target sample size, it is necessary to factor in participant attrition, which can be high in populations experiencing homelessness. In the New Leaf Project, 79 of the 115 enrolled participants completed responses at the 12-month mark (30% attrition). For DBIP, 457 of the 807 participants completed six-month surveys (43% attrition).
Disbursement Method: When deciding how to disburse funds, it is critical to prioritize ease of access. Research suggests that extreme and chronic scarcity — characteristic of poverty and homelessness — can negatively impact cognitive capacity, planning, and decision making (Shah et al., 2012). Cash transfer programs must be deployed in ways that accommodate these challenges (Zhao et al., 2023 [Ch. 2]). Although most programs in the Global North do not specify how cash is delivered (n = 29), some provide transfers directly into a bank account (n = 8) or a prepaid bank card (n = 6) set up by the organization. Further, the New Leaf Project and DBIP gave recipients phones to be used for surveys. These techniques and others can reduce cognitive load on recipients, increase safety, and facilitate data collection (Dwyer et al., 2023; Zhao et al., 2023 [Ch. 2]).
Outcomes: Lastly, an organization must consider what outcomes to measure, how frequently, and for how long after disbursement. These decisions should be made in close collaboration with the evaluation team, who can identify appropriate statistical methods to analyze data. The most common “primary outcome domains” measured by cash transfer programs in the Global North have been health (n = 16), assets (n = 11), well-being (n = 11), spending (n = 5), community (n = 4) labor (n = 4), housing (n = 4), and education (n = 3). The primary follow-up period for data collection in the Global North usually occurs between 0 and 1 year after initial payment (n = 42), but others have surveyed between 1 and 2 years later (n = 30) or between 3 and 5 years later (n = 23). As mentioned earlier, organizations should consider following participants further into the future to contribute much needed knowledge on the long-term benefits of cash transfers.
What are some main challenges to implementation?
One must be clear eyed about challenges to implementing cash transfer programs in the United States. This research brief touches on four here: benefits cliff, high costs, disbursement challenges, and public perceptions.
Benefits cliff
Cash transfers can sometimes interfere with a recipient’s eligibility for ongoing or existing means-tested benefits (e.g., welfare). This is not desirable, as most transfer programs are designed to be supplemental to existing services (Neighly et al., 2022). To avoid this, it is common to partner with local governments or pass legislation to dismantle the “benefits cliff” for a program. For example, the New Leaf Project secured an agreement with the provincial government in Canada to ensure that their cash transfer did not impact participants’ existing or future benefits (Dwyer et al., 2023). The state of California also recently passed legislation to make their pilot guaranteed income payments not impact other benefits. These steps appear necessary to maximize the impact of cash payments while minimizing service disruptions.
High cost in the U.S.
Another well-known limitation of cash transfer programs in the United States is the relatively weak purchasing power of cash as compared to low- and middle-income countries (Zhao et al., 2023 [Ch. 1]). High costs of living — particularly in urban areas — means that cash payments disbursed must be large to meaningfully improve a recipients’ financial situation. To illustrate this, the average cash transfer program in the Global North provides $7,753 per year in 2021 U.S. dollars, which only represents an increase in relative wealth of 12%. In the Global South, an average program of $1,414 per year translates into a relative wealth increase of 37% (Zhao et al., 2023 [Ch. 1]). High costs also means recipients typically need to spend their cash transfer quickly (Dwyer et al., 2023), and pilot programs can only recruit around 100 to 200 recipients due to funding constraints.
Disbursement barriers
Cash transfers are highly appealing because of their ease, flexibility, and immediacy, but some disbursement barriers exist. Enrolling individuals experiencing homelessness and contacting them for follow-up surveys has been difficult for the New Leaf Project and DBIP program, both of which experienced significant attrition. Another disbursement challenge is the blinding of recipients and researchers to which experimental condition they are assigned to (cash transfer vs. control group). Some have suggested that if a recipient knows what experimental group they are in, this can bias results or contribute to frustration among those assigned to the control group. Adequate blinding and recruiting participants before they are informed about their eligibility for a cash transfer may help to retain an unbiased sample. Electronic payments and cell phones can also improve communication and make transfers easier to provide and data easier to collect.
Public perceptions
Finally, the United States may be a uniquely hostile social environment in which to implement cash transfer programs. Most Americans hold negative attitudes towards unconditional cash support: the Pew Research center in 2020 found that 54% of Americans oppose the federal government providing a guaranteed income of about $1,000 per month for all adult citizens (Gilberstadt, 2020). Americans also hold negative attitudes about individuals experiencing homelessness, with these groups being rated as less warm and competent (Cuddy et al., 2007), and having less important psychological needs than those who are not homeless (Schroeder and Epley, 2020).
Therefore, it is not surprising that emerging research on attitudes towards cash transfer programs for individuals experiencing homelessness reveals high levels of mistrust and negative beliefs. The same team that evaluated the New Leaf Project conducted a large survey of the U.S. public (N = 1,114) to explore how Americans thought recipients of a cash transfer would spend their money. On average, respondents said that recipients experiencing homelessness would be more likely to spend their money on temptation goods (e.g., cigarettes, alcohol) as compared to recipients who were not homeless. The authors wrote that this reveals “a public mistrust of individuals experiencing homelessness in their ability to manage money,” which may act as a barrier to cash transfer programs (p. 7; Dwyer et al., 2023). Optimistically, a second study (N = 1,373) showed that support for a cash transfer program could be increased by counteracting negative stereotypes (i.e., showing that individuals do not disproportionately spend on temptation goods) or by emphasizing the social benefit of cash transfer programs (i.e., their cost-effectiveness).
What still needs to be known?
The provision of direct cash transfers is a new idea that is rapidly expanding in North America. Many programs are underway, meaning the knowledge about best practices and program effectiveness is changing quickly. In this time of expansion, researchers have identified some future directions that are especially valuable to explore.
How long do benefits last?
Because conditional cash transfer programs have existed for decades, some research has been able to examine long-term benefits of such programs, finding mixed results (Millan et al., 2019). As unconditional cash transfer programs expand for the first time in Western nations, long-term evidence of effectiveness is badly needed. Although challenging, researchers must consider implementing extensive follow-up windows to contact recipients long after they have stopped receiving payments. Population-level data collected across time could also be leveraged to evaluate long-term outcomes. A promise of cash transfers is their ability to break cycles of poverty, but more data is needed to demonstrate if this claim is true.
Who exactly benefits from cash transfers?
The two pioneering cash transfer programs for individuals experiencing homelessness in North America used screening criteria to administer cash transfers to individuals who are homeless but do not experience severe mental health, substance use, or alcohol use problems. Future work might consider loosening these criteria in controlled, monitored ways, to determine if cash transfers may be more effective for some populations and not others. The length of time spent homeless may be a key moderating variable to consider, as cash could differentially impact those who are chronically homeless vs. those experiencing homelessness for the first time.
Are there any backfire effects?
Almost all the research synthesized in this brief concludes that cash transfer programs offer some benefits to recipients, or at least do not inflict harm. However, one recent study in the U.S. found that a one-time, unconditional cash transfer of a modest sum ($500 or $2,000) given to impoverished individuals was related to worse self-reported financial, psychological, and health indices as compared to a control group that did not receive any cash (Jaroszewicz et al., 2022). This unexpected finding was stronger for subjective measures than for objective ones, leading the authors to conclude that modest, one-time cash disbursements may make recipients feel worse by rendering their unmet needs more salient. This study was pre-registered and well-powered (N = 5243) but is inconsistent with most research on unconditional cash transfers that suggest they are beneficial. Even so, due to the potential of harm to recipients, exploring backfire effects of cash transfers is a critical future research direction.
Is this cost-effective?
Review papers on cash transfers have long lamented that cost-effectiveness estimates are not readily available (Glassman et al., 2013; Lagarde et al., 2009). These estimates are hard to obtain, as a valid assessment must compare the effectiveness of a cash transfer program to the effectiveness of a means-tested program of similar cost. The New Leaf Project did find that their program led to societal net savings of $777 per person, per year, due to reductions in the number of nights recipients spent homeless (Dwyer et al., 2023). Other programs should conduct similar cost-benefit analyses to add to the growing knowledge of whether these programs effectively spend resources.
How do these transfers work?
Finally, some have noted it is difficult to determine what aspects of conditional cash transfer programs increase well-being and improve health (Lagarde et al., 2009; Glassman et al., 2013). For example, it can be hard to tell if a program improved health outcomes because of the money provided, or because of the requirement to attend regular health check-ups. For unconditional programs, it is a little easier to tell if cash directly leads to improvements, but much remains to be known. Does cash lead to a reduction in life stressors, thereby improving well-being and cognitive capacity? Does cash provide instrumental support directly, allowing recipients to simply purchase housing? Does cash reduce the urgency of daily expenses, freeing recipients up to search for employment? These types of “pathway and mechanism questions” are just beginning to be explored, and qualitative research like focus groups or debriefing interviews can help to determine how cash helps people in their own words.
Conclusions
This research brief has provided a broad overview of direct cash transfer programs across the globe, with a focus on those in North America targeting homelessness. Many different types of cash transfer programs exist, and while the impact of a program is highly context-dependent, reviews of the evidence have largely supported their effectiveness in lifting individuals out of poverty. Research is newer regarding individuals experiencing homelessness in North America, but two pilot programs demonstrate promising initial findings. Organizations considering implementing a cash transfer program can use this brief as an introduction to the evidence, critical program decisions, barriers, and future directions. As the literature base expands, it is necessary to design, administer, and evaluate new pilot programs in ways that are replicable and rigorous.