(Updated) It sounds like common sense: if you provide free food and shelter for homeless alcoholics and allow them their booze, they’ll probably drink themselves to death. But conventional wisdom isn’t always correct. Now a new study confirms earlier research on so-called “wet houses,” showing that residents who are allowed to drink actually drink less.
The homeless residents in the study cut the number of drinks they consumed daily by 40% over the course of two years in a home that did not require abstinence. Moreover, for every three months of their stay, participants consumed 8% fewer drinks on average on their heaviest drinking days. The occurrence of delirium tremens, or DTs — potentially life-threatening withdrawal symptoms — also declined by more than half, with 65% of residents reporting suffering DTs in the month before being housed, compared with just 23% in the month afterward.
The study followed 95 participants, mainly white and American Indian or Native Alaskan men in their 40s, who lived in Seattle’s 1811 Eastlake residence. Unlike other shelters, the Seattle program does not require participants to observe curfews or abstain from drinking — rules that in other cases tend to keep the heaviest drinkers on the streets and in and out of jail or the emergency room. At least 40% of chronically homeless people suffer from alcoholism.
“These individuals have multiple medical, psychiatric and substance abuse problems, and housing that requires them to give up their belongings, adhere to curfews, stop drinking and commit to treatment all at once is setting them up to fail. The result is that we are relegating some of the most vulnerable people in our community to a life on the streets,” Susan Collins, lead author and assistant professor of psychiatry at the University of Washington, said in a statement.
MORE: Happy Hour? Wet Housing Allows Alcoholics to Drink with Surprising Results
The Seattle program is intended to be a permanent residence. And researchers found that it wasn’t just that those with the least severe alcohol problems who were able to stay and cut back; they didn’t see differences in demographics or drinking levels in those who stayed versus those who left.
Participants did continue to drink heavily, however. Still, while they were averaging 20 drinks a day at the start of the study, two years later, that number was down to 12. This is still far from a healthy level of alcohol intake, of course, but residents did show improvements in overall health.
So why might housing itself help reduce drinking? “Participants in the study told us that they’re happy to have a home, and happy that they no longer have to drink to stay warm or to put themselves to sleep or to forget that they’re on the streets,” Collins said.
The research follows an earlier study of the same participants, which was published in the Journal of the American Medical Association. That study found that in one year, this single group of homeless people had cost the city $8.2 million in medical bills, policing, detox and incarceration, before the study began. Per person, the cost to taxpayers was $4,066 per month on average. After the first year in wet housing, however, that cost dropped to $958 per person per month. Spending was cut by 53%, compared with the costs to the city for homeless people on a waiting list for the program.
MORE: Alcohol in Pregnancy: It’s Never Safe, Especially Not in the First Trimester
The new study was done in part to address the main critique of such “harm reduction” programs — that failing to require total abstinence not only “enables” addiction, but actually makes it worse. “A lot of people believe in the ‘enabling hypothesis’ — that allowing homeless, alcohol-dependent individuals to drink in their homes will enable them to drink more, and their drinking will spiral out of control,” Collins said. “But instead what we found are across-the-board decreases in alcohol consumption and problems.”
Indeed, the study found that rather than worsening alcoholism over time, the program was associated with better health in participants, who were able to remain housed. They did not get sober — but that hadn’t been achieved in repeated episodes of detox and abstinence-based treatment, either.
Since the study was not randomized, it is not possible to conclude that the provision of housing without abstinence directly led to the reduced drinking. It would be quite interesting to conduct a trial with a similar population that randomizes participants to either wet or dry housing, with appropriate supports for harm reduction and abstinence, respectively, and follow the outcomes of both dropouts and those who remain housed long-term.
Until then — and probably long afterward, if the protracted debate about needle exchange is anything to go by — people will continue to argue about what’s best for severe addiction based on conventional wisdom about “enabling,” rather than research evidence.
MORE: Appreciation: G. Alan Marlatt Brought Compassion to Addiction Treatment
The study was published in the American Journal of Public Health and co-authored by the late Alan Marlatt, a pioneer in addiction research focused on harm reduction.
Correction [Jan. 20, 2012]: The original version of this story misstated that for every three months of their stay in a “wet house,” homeless residents in the study consume 8 fewer drinks on average on their heaviest drinking days. That figure should be 8% fewer drinks on average.
Maia Szalavitz is a health writer for TIME.com. Find her on Twitter at @maiasz. You can also continue the discussion on TIME Healthland’s Facebook page and on Twitter at @TIMEHealthland.