Manhattan Institute Report Urges States to Expand Recovery Housing

The Manhattan Institute published a report Thursday arguing that recovery housing — the peer-run, substance-free residences that sit between emergency shelter and permanent apartments — should carry more of the weight in state efforts to reduce homelessness among people with addiction, and that the sector needs more certification rather than less oversight.

The report, by senior fellow Stephen Eide, concludes that states should expand certification for recovery homes, write exceptions for those homes into eviction law, and keep the sector under state rather than federal control. A federal push to tie homelessness programs to behavioral health goals and to favor nonmedical approaches to addiction has made the question timely, the report argues.

Four recommendations, aimed at state capitals

Eide's recommendations are narrow and procedural rather than programmatic. State government is and should remain the main theater of recovery housing policymaking, the report says, because the sector is highly decentralized. States should expand certification for recovery houses to establish quality and legitimacy. They should also create exceptions in eviction law for recovery housing, on the reasoning that a house cannot maintain a substance-free environment if a resident who relapses can only be removed through a full eviction proceeding.

The fourth recommendation is a caution rather than an expansion: government should generally be wary of integrating recovery housing into homelessness services systems, because the funding streams and eligibility rules attached to those systems can reshape what the homes do.

Where the homes sit in the housing system

The report's argument rests on a description of the current system that Eide's institute summarizes as a two-ended framework. Permanent supportive housing, the Housing First model that became standard over the past decade, and short-term outreach and shelter cover the extremes; the institute's announcement cites surveys suggesting as many as 70 to 80 percent of homeless people have a substance use disorder, a group it says the middle of the system often does not serve well.

Recovery housing is an intermediate-length residential program for people in recovery from drug or alcohol addiction. Residents are prohibited from using drugs or alcohol, and the program provides no medical treatment at all. It relies on peer support and the social setting of the house itself, with the goal of extending sobriety long enough that it holds. The report puts the average stay in a long-term residential treatment program at 66 days and in a short-term program at 25 days, and notes that recovery housing is generally a longer commitment than either.

Scale is not trivial. The institute estimates that 250,000 to 300,000 people already use recovery housing each year.

Counting an under-regulated sector

The report's central finding is that the problem is not overregulation but its opposite. About one-third of states have no official definition of recovery housing, and most states do not maintain a database of the homes, according to the report. The gap has produced cases of fraud and low-quality programs, and it also makes the sector difficult to count.

That counting problem shows up in the report's own figures. The federal government has estimated that about 8,000 certified recovery housing programs operate nationwide; a separate analysis put the total number of programs, certified or not, near 18,000; and a 2026 peer-reviewed study estimated a range of 10,000 to 18,000. The best-documented model, Oxford House, has nearly tripled since 2010, growing from 1,450 houses to 4,029. The National Alliance for Recovery Residences, founded in 2011, certifies programs at four levels of support, with its first level corresponding to the peer-run Oxford House model.

What the report does not settle

The report is explicit that recovery housing is not treatment. Residents typically arrive after detox or inpatient care, in a non-acute phase of their condition, and the report frames the homes as "the setting is the service" rather than a clinical intervention. That distinction matters for how any of the recommendations would work in practice: certification can set standards for a house, but it cannot make a house a treatment provider, and the report does not propose that it should.

The release also does not attach a cost to expanding certification or estimate how many homeless people would move into recovery housing if more states adopted the recommendations. It argues the conditions are favorable and leaves the arithmetic to state legislatures.

For readers weighing options, recovery housing is one rung on a longer ladder of care; the clinical side of that ladder includes residential treatment and outpatient programs, and the underlying diagnosis is covered in more detail under substance use disorder. Michigan's own state-funded expansion, which added 205 recovery housing beds through a recovery housing investment program, is covered in earlier reporting.