Atrómitos Consulting Done Fearlessly
CAREERS

Menu

A Roof Is Not Enough. Someone Has to Pay the Rent

Michealle Gady

·

Paying for the Last Link: A Six-Part Series on Housing Finance in Washington, Part 5 of 6

The previous article looked at two ways Washington could make its money for buildings go further. Even the best-financed building, though, cannot house people who cannot pay the rent. As Part 2 described, households with the lowest incomes need ongoing help with rent on top of the capital that builds or preserves their homes. That help is part of the last link in the chain of care I described in July. Three proposals could help those households afford and keep their homes: a statewide rental voucher, outreach that finds people before they lose their homes, and direct cash assistance. The evidence for them ranges from strong to thin, and each carries risks the state would need to manage.

1. Create a Statewide Rental Voucher on Stable Funding

Washington has no broad, ongoing statewide rental voucher like those in Massachusetts and Minnesota. In state fiscal year 2025, only 14% of households leaving shelters that the state’s Consolidated Homeless Grant funded moved into stable housing, far below the target of half that the Washington State Department of Commerce (Commerce) sets for shelter exits to permanent housing. For many, there was likely nowhere affordable to go next.

The state’s ongoing rent assistance reaches only specific groups. The document recording fee described in Part 3 pays for some rent assistance and project-based vouchers, but its revenue rises and falls with the housing market. The state-funded Housing and Essential Needs program provides rent assistance to adults who cannot work. As Part 3 described, Commerce’s report finds that this help, meant as a temporary bridge, has effectively become a permanent subsidy. Most clients now qualify through the state’s Aged, Blind, and Disabled program, and many have permanent disabilities. Apple Health and Homes pays rent only for people with a medical or behavioral health condition who qualify for its supportive housing services.

Massachusetts and Minnesota show how a state voucher works. Massachusetts has run its current program since 1992, serving nearly 11,000 households a year. Minnesota created Bring It Home in 2023, and its program guide sets out how the program works. State appropriations fund it, together with a share of the Twin Cities metro-area sales tax for housing. Local housing agencies and Tribal governments administer it, and nonprofits can step in where no local agency has the capacity. Renters who earn up to half the area median income and spend more than 30% of their income on rent pay 30% of their income toward rent, and the program covers the rest, up to a limit. Households with children and incomes at or below 30% of area median income get priority. In August 2025, Minnesota Housing approved more than $123 million in two-year grants to help approximately 5,000 families.

This proposal has the strongest evidence of the three. As Part 2 described, the Family Options Study, run by the U.S. Department of Housing and Urban Development (HUD), randomly offered vouchers to some families with children staying in shelters. About three years later, families offered a voucher were less than half as likely to have spent a night homeless or doubled up in the previous six months as families who received only the usual help available in their communities. I have found no newer randomized study of how vouchers affect homelessness, though HUD has designed a 12-year follow-up of the same families. A 2026 California Policy Lab paper, described in Part 2, again cites the study’s finding that long-term rent subsidies reduce homelessness. I have not found rigorous evaluations of the Massachusetts or Minnesota programs, but they follow the same basic design as federal vouchers.

The risks are cost, fragility, and fit. A voucher is a recurring obligation that grows with rents. As Part 2 noted, rising costs have led Massachusetts to pause new mobile vouchers, which move with the tenant. Starting July 1, 2026, it also capped rent increases at 3% for units rented with state vouchers. The federal Emergency Housing Voucher shortfall that Parts 2 and 3 described showed how quickly a funding gap can put households’ homes at risk. Part 2 also described how households with vouchers can struggle to find a unit in a tight market.

Washington can reduce those risks, though not eliminate them. Stable funding, the subject of the final article in this series, would leave households less exposed to the next shortfall. Attaching some vouchers to specific buildings, as project-based assistance does, would help in markets where households struggle to find a unit. The tenant’s share also matters. Federal Housing Choice Vouchers generally set it at about 30% of income for rent and utilities, which Part 2 showed is more than many households with the lowest incomes can bear. A Washington program could instead set the tenant’s share on a sliding scale based on the University of Washington’s Self-Sufficiency Standard, so that the households with the least income pay the smallest share. The Standard estimates how much a household needs to cover basic costs in each county without public help. A sliding scale would cost more per household, so the same budget would reach fewer households.

Rent assistance alone has limits, however. Harvard’s Joint Center for Housing Studies found that capping rent at 30% of income for every burdened working-age renter household would reduce the share who cannot cover basic needs by only 1.3 percentage points, though lower-income households would receive larger amounts than higher-income households. As the researchers concluded, “costs are too high and incomes are too low.”

Given the strength of the evidence and the size of the shelter problem, the benefits outweigh the risks. Washington should create a statewide voucher, starting with the households least able to pay rent, and fund it from a source that will hold up in a downturn.

2. Find People Before They Become Homeless

Washington’s homelessness system mostly responds after people lose their homes. Los Angeles County’s Homelessness Prevention Unit tries to reach them first. The California Policy Lab and the county built a predictive model from records that 11 county agencies share. The model ranks residents who recently used county health or mental health services by their risk of becoming homeless. People the model flags become homeless at nearly 3.5 times the rate of the eligible population as a whole. Case managers then contact them, offer about six months of help, and provide an average of about $6,500 in flexible financial assistance. Over 18 months, participants were 71% less likely than comparable high-risk residents who did not enroll to go into a shelter or to have contact with street outreach teams. That finding comes from an early pilot, and people who chose to enroll may differ from those who did not. A randomized trial will test that question, and researchers expect its results in 2027.

The main benefit is reaching people before a crisis. As Part 1 showed, Washington’s prevention programs cost a fraction of what shelter costs for each household that ends up in stable housing, though the two serve different households, so the comparison is not direct. Until January 2026, the unit operated out of the county’s Department of Health Services, which made it an example of health and social care integration at work. That month, the county moved it to a new Department of Homeless Services and Housing.

Washington is well positioned to try the approach. The Department of Social and Health Services’ Integrated Client Databases combine information from more than 30 of the state’s data systems and are already used for predictive modeling, and the state’s Snapshot of Homelessness brings together client records from three state agency data systems. Using the Integrated Client Databases for outreach, however, would likely require new data-sharing agreements.

The risks involve privacy, fairness, and reach. The approach depends on sharing sensitive health and service records. A model can also miss people or treat groups differently. Reach is limited as well: the Los Angeles model flags about 10,000 people a year, far more than the 400 to 600 people the unit can serve. Washington can manage these risks with clear privacy protections, an equity review of the model, and a pilot sized to the help the state can actually provide. With those safeguards, the benefits outweigh the risks, and Washington should test the approach in a pilot.

3. Test Direct Cash Assistance

Washington has already asked whether cash could help. According to Commerce’s 2025 annual report, the department asked in 2025 to allow cash assistance in the Housing and Essential Needs program, and lawmakers limited it to gift cards and other cash equivalents. A separate proposal for a direct cash pilot serving young people did not advance that year, and Commerce is exploring whether existing resources could fund one for young people leaving foster care and other public systems of care.

The Denver Basic Income Project shows how a cash program works. The randomized study enrolled 807 people experiencing homelessness and gave them money with no conditions on how to spend it. One group received $1,000 a month for a year, a second received $6,500 up front and then $500 a month, and a comparison group received $50 a month. At 10 months, between 43% and 48% of participants in each group, including the comparison group, had their own house or apartment, up from 6% to 12% at enrollment. When Mayor Mike Johnston’s proposed 2025 budget left the project out, a spokesperson for the mayor said the year-one report “did not show a statistically significant difference in homelessness resolution between the groups that received large cash transfers and those who did not”. A smaller analysis of 221 participants aged 50 and older, which researchers released as a conference abstract, found that both groups receiving larger payments were significantly more likely than the comparison group to obtain housing.

A small randomized study in Vancouver, British Columbia, found that people experiencing homelessness who received a one-time, unconditional payment of CAD$7,500 spent fewer days homeless over the following year than those who did not. The study included only people who had been homeless for less than two years and whose substance use and mental health symptoms were not severe, so its results may not apply to people with greater needs.

The appeal of cash is flexibility: people can spend it on whatever stands between them and a home. The risk is that it may not work better than other help, and the evidence for it is the least consistent of the three proposals in this article. A small, well-evaluated pilot for a defined group, such as young adults leaving foster care or older adults experiencing homelessness, would limit that risk. At that scale, the benefits of finally answering a question the state has already asked outweigh the cost.

Measure What Matters

Most of the five proposals in this article and the previous one (two in Part 4 and three here) rest on thin evidence, and Washington has not funded any of them or put any of them fully into practice statewide. The state should design an evaluation for each one before the first dollar goes out and build that evaluation with the people it affects, including tenants, people who have experienced homelessness, providers, developers, local governments, and Tribes. An evaluation that the state imposes on them risks measuring what is easy to report rather than what matters.

Each evaluation also needs to do more than count units financed or households served. As Part 1 described, Washington already tracks homelessness and its own activity closely. The more useful questions are about outcomes and quality: whether people stay housed, whether their health and stability improve, whether the housing is safe and well maintained, and whether the help reaches the people who need it most.

Whether any of these proposals lasts depends on how the state pays for it.

Next in this series: how Washington can pay for these proposals, and why housing belongs among Washington’s protected priorities.

Sources