In late July, researchers outlined their proposal in a working paper published by the Community Development Investment Center at the Federal Reserve Bank of San Francisco. The authors found that participants would see such a significant reduction in rates of extreme obesity and type two diabetes that the intervention would save more than it costs in 10 years.
The authors explored the idea because prior evaluations of housing mobility programs have found lower rates of extreme obesity, diabetes and major depression among the households who moved to low-poverty neighborhoods compared to those who remained in high-poverty neighborhoods. The paper assumes that if a housing authority provided housing vouchers, along with counseling and assistance searching for better places to live, the administrative cost for 400 initial movers would be about $2.2 million over 10 years. The reduced medical costs could yield anywhere from $100,000 to $1.6 million in net benefits.
One of the authors, Mary Cunningham of the Urban Institute, explained in a blog post that state and local governments have to explore elaborate financial arrangements involving foundations, the private sector and nonprofits because of waning federal support for housing vouchers and the relocation services that appear to make vouchers most effective.
Governing spoke with the paper's lead author, Dan Rinzler, a manager at the Low Income Investment Fund, to better understand the proposal. The following transcript has been edited for clarity and length.
Do you think it's a problem that the most optimistic break-even point, where savings exceed costs, takes place after seven years?
We recognize that this is a longer timeline than has been seen in pay-for-success transactions to date. We wanted to do the modeling in an honest and conservative way, but we wanted to show, with the evidence that we had, that this was at least possible. Hopefully our analysis has enough integrity and is thorough enough to get people to take a closer look.
To clarify, when you say "this was at least possible," you mean that it's possible that purely from savings gleaned from health-care spending on extreme obesity and diabetes, these housing mobility services are worth the upfront investment?
Right. That the services could pay for themselves in a reasonable timeframe. This is just a basic dollars and cents benefits-cost analysis. And there are a variety of other social outcomes that could be included and accounted for in a housing-mobility pay-for-success transaction, such as health and educational attainment for children. We just happened to focus on one area, adult metabolic health, where we could run a simple cost analysis.
The most conservative scenario assumes that the break-even point where benefits exceed costs occurs almost at the end of the 10-year period.
Right. In that scenario, the benefits do not even start until five years down the line. We wanted to show increasingly conservative scenarios, so that's what "linear year three" and "linear year five" are intended to represent. The reason we ran multiple scenarios with various levels of aggression is that the data [from prior evaluations] do not tell us when these [health] benefits happen. So, yes, the most conservative scenario doesn't come out so rosy.
Both your paper and the recent GAO report of pay-for-success financing talked about the "wrong pocket problem." Housing mobility seems like an area that would be particularly vulnerable to the wrong pocket problem because the health benefits you would expect to see aren't enjoyed by the housing authority investing in housing mobility. Do you think that the logistics of having to coordinate across a Medicaid agency, a housing authority and potentially other agencies makes this type of intervention harder to do?
On the surface, yes, the cross-sector issues are daunting. Housing mobility is a classic example of the wrong pocket problem because it's an investment in one area that generates impact in other areas like medical cost savings and increased tax revenue from higher incomes. But I think pay-for-success could provide a safe testing ground for an initiative that involves collaboration between sectors and between different levels of government. This could be the beginning of a fruitful relationship between these agencies that normally don't get together. In other words, it's a great way to address the wrong pocket problem.
How come you focused on diabetes and extreme obesity but not major depression, which the research says is another health benefit associated with housing mobility services?
Diabetes and extreme obesity are areas where medical cost savings are the economic benefit that occurs and that accrues to government budgets -- because this is a very low-income population that largely relies on the government to cover medical costs. So this is an area where you could potentially have a single payer, like a Medicaid agency, where you're getting all of that benefit. In the case of major depression, societal costs are primarily around lower economic output from lost worker productivity. I guess you could file that under a non-cashable benefit, but you wouldn't necessarily incorporate it into the pay-for-success transaction.
Does the federal government have a role to play in a project like this?
Medicaid costs are shared across state and federal government and it would be ideal to be able to capture the savings to both state and federal coffers because the savings would generate to both. A pay-for-success housing mobility initiative could involve multiple sectors of government -- health, housing, potentially education -- as well as different levels of government.
The government could just pay for this itself, right? There's no reason that the private sector has to get involved, except for lack of political will. Or am I wrong about that?
That's right, but we talked earlier about housing mobility being an example of the wrong pocket problem. It's probably worth it to a Medicaid agency to support housing mobility. However, the federal government needs to see it that way, too. The health sector might be understandably skeptical that a housing program without any medical services will actually achieve these considerable results in diabetes and extreme obesity, despite gold-standard evidence. So, pay-for-success could be a good way to pilot it, and demonstrate impact in a safe, controlled environment where there's a good evaluation, and then that will hopefully, down the line, change government spending behavior in the way that you're talking about.
