When 'Nothing' Beats 'Something'
In Washington’s policy battles, the time-tested adage says, “you can’t beat something with nothing.” The idea is that social ills require solutions and doing nothing is never the right thing to do — bad policy beats no policy every time. In the one-way federal ratchet, bad policies are met with less-bad policies and the growth of government continues its relentless march. But what if the facts on the ground say otherwise? What if “nothing,” on some issues, does beat “something”?
In October 2016, the U.S. Department of Housing and Urban Development released the long-term results from its Family Options Program (FOP). (Somehow in the carnival atmosphere of the 2016 election, I missed the report’s release.) FOP was a test of three different strategies for helping, or trying to help, homeless families. Families who were in emergency shelters were assigned to one of four approaches to developing secure, long-term housing:
- Housing Subsidy (SUB) — priority access to long-term housing subsidies
- Community-based Rapid Re-housing (CBRR) — priority access to a temporary housing subsidy for up to 18 months plus voluntary wrap-around social and human services support
- Project-based Transitional Housing (PBTH) — priority access to a temporary, services-intensive housing facility for up to 24 months with mandatory participation in wrap-around supportive services
- Usual Care (UC) — normal access to homeless and housing services but no priority access to housing subsidies or services
The purpose of FOP was to test whether subsidies alone were sufficient to improve outcomes or whether homeless families did better when subsidies were accompanied by services to reduce non-housing barriers. These services included case management, referrals for mental health and substance abuse services, workforce development/employment, and others. Outcomes were analyzed for a total of 2,282 families across 12 communities and 148 programs.
The chart below sketches the impact of the first three interventions (SUB, CBRR, and PBTH) verses UC. Most remarkably, the SUB model produced the best results on four of the five the measures studied (housing stability, family preservation, adult well-being, child well-being and economic self-sufficiency). And it wasn’t close. Over 37 months, the SUB families spent dramatically fewer nights in shelters or “doubled-up” with family or friends, had fewer family separations, lower levels of psychological distress among family heads, lower levels of domestic violence, less drug and alcohol abuse, fewer school changes for children and fewer absences, and less food insecurity.
On the negative side, fewer SUB family heads were employed compared to UC — a function of the housing subsidy without a work requirement — and there were more separated couples (the flip side of lower rates of domestic violence). Total costs for the SUB group were higher, almost $46,000 over 37 months, compared to $38,000 for CBRR and $40,000 for PBTH. The additional costs have to be weighed against the fact that the other two models found few if any positive results. Of greatest import, however, is the inverse relationship between intensity of service provision and positive outcomes: the fewer supportive services the families got, the better they did. In this project, nothing beat something.
At a recent AEI meeting on criminal justice, a senior researcher commented that a key takeaway from multiple evaluations of re-entry programs is that control groups who did not take part in programs often did as well or better than those who did, another instance of nothing beating, or at least fighting something to a draw. To me this suggests that people like Maurcio Miller, founder of the Family Independence Initiative, have developed an important insight: for people to improve their social and economic condition, the most important factor may be personal agency and strengths-based empowerment rather than services.
These data points are tantalizing. Without suggesting a wholesale revamping of homelessness, re-entry, or other social services programs, we ought to begin asking questions about where tests of alternative anti-poverty approaches might be useful. For instance, should we establish a principle saying that if the federal government wants to launch a new or expand an existing intervention-based approach, the sponsoring agency must analyze whether the same end could be served through an empowerment model? With a policy like this in place, experimental evaluations could be done on both approaches to see which approach has the greatest effect. We’ve tried “something” for decades. Let’s give “nothing” a chance, too.
Brent Orrell is a resident fellow at the American Enterprise Institute where he conducts research on workforce development, criminal justice reform, and social theory.