Although a District program designed to help homeless families find private apartments has significantly grown over the past five years, a blistering report out this week says the program is “deeply flawed,” that its success is “illusory,” and that it does not always prevent families from returning to the shelter system.

For the report, titled “Set Up to Fail,” the Washington Legal Clinic for the Homeless examined data on D.C.’s rapid rehousing program run by the local Department of Human Services. Rapid rehousing provides homeless families a minimum of one year of rental subsidies coupled with case-management services. Families pay between 40 and 60 percent of their income on rent, and the program covers the rest. It’s a step removed from D.C.’s family shelters, which are so full that officials have had to rely on hotels.

The District invests over $31 million a year in rapid rehousing, three-quarters of which is funded by D.C. tax dollars and the other quarter from the federal government. The program’s goal is to keep families out of shelter and get them on the path to independent living. But based on extensive client work, WLCH says rapid rehousing is often just a temporary solution. It has found that after families’ subsidies expire, few of them can hold onto their housing independently.

To make matters worse, the group says, rapid rehousing comes with a host of problems for the families D.C. intends to serve. More than 4,000 people—or about 1,350 households, including about 2,300 children—currently participate in rapid rehousing. Eighty percent of those WLCH worked with for its analysis reported poor housing conditions like rats, broken appliances, and leaks. In addition, WLCH notes that, on average, families enrolled in the program can only cover 40 percent of a unit’s market rent.

In an interview, DHS director Laura Zeilinger defends the program and says the report is based on “a misguided understanding of the purpose and intent of rapid rehousing.” Zeilinger says the vast majority of rapid rehousing clients—85 percent—do not fall back into homelessness after the program, which is meant to offer people stable housing while they look for employment and access services. It is not intended to function as long-term affordable housing as other vouchers do, she says. She notes that homelessness is down in D.C. in the wake of recent reforms and that WLCH’s clients are a “small subset” of the program’s participants.

“Rapid rehousing is part of our homeless-services continuum that helps people exit shelter as soon as possible,” Zeilinger says. “There’s a philosophical disagreement that everybody has a right to long-term affordability—which is a beautiful idea, but it’s not practical when you’re looking at the role of the homeless-services system.” That role is “narrow and specific,” she adds. It’s a “crisis-response system.”

In one of five client cases cited in the report, a woman and her children have cycled in and out of rapid rehousing three different times. (The report does not identify any of these clients by name, and Zeilinger says DHS can’t comment on individual circumstances.) In her second stint in rapid rehousing, the woman moved from a shelter into a four-unit building where her apartment had “severe rodent and mold infestation” and charged $1,450 in monthly rent. After a TV news report on the building, the woman returned to shelter, then proceeded to another rapid rehousing unit. She had to leave the apartment a few months later due to domestic violence, and again entered shelter, where she remains with her kids.

In a second client case, a man with children exited from the deteriorating D.C. General family homeless shelter (which the District plans to replace) into a rapid rehousing unit. His case management abruptly ended a year into the program, and his cash-assistance benefits through the Temporary Assistance for Needy Families program declined to $277. But he was still on the hook for $208 of his rent each month, meaning he only had $69 to pay for utilities—which aren’t covered by rapid rehousing—and other basic needs. He became two months behind on his rent when his rapid rehousing subsidy ended, prompting his landlord to file for eviction. An administrative judge dismissed that case, ordering D.C. to restore the man’s subsidy and furnish him with a new case manager. But he still spends most of his income on rent.

“Unfortunately, the problems highlighted by these stories are recurring features of rapid rehousing rather than one-off events,” report author and WLCH attorney Max Tipping writes. “After being terminated from rapid rehousing, many return to homelessness, now with an eviction or rental debt on their record.” That means it becomes harder for affected families to find stable housing in the future—lowering their morale.

Tipping has assisted more than 100 households receiving rapid rehousing since 2015. Participants earn less than $500 a month in average income, whether through TANF, Social Security and disability, or work. At the same time, WLCH explains, the average rent for a rapid rehousing unit is nearly $1,200 a month, and there’s evidence that some landlords who accept the program’s clients charge artificially high rents.

Criticism of the program is not new, but WLCH’s analysis distills it into a few dozen pages. It also comes as elected officials shape D.C.’s budget for next year, including funds for homeless services. The group calls for the District to cease the rapid rehousing “cliff,” or strict time limits on clients being able to receive benefits, arguing that homelessness is “the direct and logical outcome of terminating a family’s housing subsidy,” and that “arbitrary time limits fail to account for individual circumstances.” The report also recommends creating stricter policies for dealing with slumlords who take advantage of families in the program, and financing “long-term, deeply affordable housing programs,” given D.C.’s high costs.

One in eight families in the shelter system had been in rapid rehousing at least once, as of last October. But DHS says the rate at which families are returning to shelter described in the analysis is not accurate. “A lot more people would be in system right now, but we have fewer people in it overall,” Zeilinger says.

Still, proponents of the program, Tipping writes, “have ignored or downplayed the obvious structural flaw of using a rapid-rehousing model in a high-rent jurisdiction like D.C.—most homeless families cannot increase their income quickly enough to afford market rent, and as a result they risk cycling right back into homelessness when the program ends… [Hence], the program effectively sets families up to fail, offering a façade of stable housing but sending families over a financial cliff when the subsidy expires.”

“It has been operating as an extension of the District’s shelter system,” he contends, “a holding place for homeless families that are technically no longer considered homeless but are always right on the edge.”

Tipping credits Mayor Muriel Bowser‘s administration with initially establishing a “more humane and reasonable policy” than in the past, of allowing families to stay in rapid rehousing for longer than a year if they can’t afford rent on their own. But then, in May 2016, DHS allegedly began “mass terminations” of subsidies—citing time limits—and over the last 18 months, 55 percent of program terminations have resulted from time limits. Almost 575 families exited from the program during that period, WLCH reports.

Zeilinger says the “mass terminations” claim is “totally wrong.” Under previous mayors, she says, families in rapid rehousing would often be cut off within four months of enrolling based on non-compliance with the program, and some families did not last even a month. While DHS rewrote the regulations for rapid rehousing and retrained case managers to look at “the totality” of a family’s situation, it paused the program. “What we couldn’t do is say we’re going to extend everybody on the program,” Zeilinger says.

Through a Freedom of Information Act request, WLCH found that roughly 45 percent of the families who have received rapid rehousing for a year or more have been sued for eviction by their landlords. (The average time a family spends in the program is now two years.) Though not all of these lawsuits led to physical evictions, Tipping says “this astounding figure shows that rapid rehousing in D.C. is so poorly designed and managed that these families are unable to ‘achieve stability’ even while they are receiving a subsidy and case management.” He adds that eviction risk may incentivize families to “abandon” units.

(As for the presumed eviction rate, Zeilinger says WLCH’s “numbers don’t seem consistent at all with” those at DHS, but she adds that it’s difficult to comment without seeing their calculations.)

Sometimes clients even live in slum properties such as buildings owned by Sanford Capital, which has benefitted from more than $100,000 a month in guaranteed rapid rehousing payouts. Tipping says “there is no clear mechanism for cutting off the program portion of the rent if there are housing code violations,” and it can take months for clients to be relocated. Some families have reported having multiple case managers in the program.

And that’s only after moving out of shelter through rapid rehousing, which Tipping says has become the District’s “one-size-fits-all” approach to addressing family homelessness. (The Bowser administration is currently seeking landlords to lease large buildings for homeless families under a new initiative and is also aiming to tighten the requirements to access family shelter.) WLCH says most families spend six months in shelter before finding a rapid rehousing unit, in large part because of landlords who discriminate on the basis of income or housing status. This has produced overt segregation: Some 90 percent of program participants live east of the Anacostia River, while only 1.5 percent live west of North or South Capitol streets. Only one rapid rehousing family lives west of 16th Street NW, per WLCH’s FOIA results.

“It’s unfortunate that some of the landlords who have been the worst offenders are the landlords who have been most willing to rent in the program,” Zeilinger admits, adding that the D.C. Housing Authority conducts housing-quality inspections before move-ins. Of the conditions at Sanford Capital buildings specifically, she says “no one believes those types of housing situations are anywhere near acceptable.”

Zeilinger says the results of the January 2017 “point-in-time” count of D.C.’s homeless will be released shortly.