«Moving to opportunity voluMe 14, nuMber 2 • 2012 U.S. Department of Housing and Urban Development | Office of Policy Development and Research ...»
Over 10 to 15 years, MTO’s treatment contrast shrank to 10 percent. We have no responsible way of knowing whether voucher holders in the Section 8 Program would achieve such a reduction across the same period. In Chicago, for instance, to achieve MTO’s long-term, 10-percent absolute contrast in poverty rates would entail families starting in neighborhoods that are about 29 percent poor and eventually living in neighborhoods that are 19 percent poor. This 10-percent decrease is possible in the national program but does not reflect that families in the Section 8 Program start off in neighborhoods less disadvantaged than MTO’s initial tract with 53 percent poor. It is almost impossible, therefore, for Section 8 Program families to experience a temporal pattern of neighborhood improvement as great as that in MTO. Because the treatment contrast in housing and neighborhood quality will be less in the Section 8 Program than in MTO, we must ask, “Would MTO’s health effects be replicated in the Section 8 Program, in which the neighborhood quality contrast is almost certainly smaller than in the MTO low-poverty treatment group and likely to be even smaller than in the MTO traditional voucher treatment group?
MTO Varied Housing and Neighborhood Quality, Whereas the Section 8 Program Also Varies Disposable Income For families living in public housing, obtaining a voucher replaces their public housing subsidy.
They can use their new voucher to purchase better housing and a better neighborhood in the private housing market, but that is all. On the other hand, families who are already in the private market can also use a new voucher to increase disposable income and pay for things such as clothes, car repairs, food, and phone service. The voucher works this way for them because families already in the private housing market can use their voucher to substitute for the rent they used to pay before getting the voucher. The size of this substitution depends on their income, rent, and local FMR values. In practice, most Section 8 Program families probably apportion their voucher’s monetary value between upgrading their housing and increasing their disposable income. Thus, Jacob and Ludwig (2012) estimated that a voucher enabled the average Chicago Section 8 Program household to spend about $3,840 more per year for housing and add $4,425 to its disposable income. Because public housing residents pay their new rent with a voucher and get nothing else, even if their rent is less than the voucher’s full value, it is highly likely that voucher holders coming from private-market housing experience smaller neighborhood (and housing unit) upgrades but larger cash transfers than those MTO produced.
How will the Section 8 Program affect health if most of its participants come from private housing, and so its neighborhood contrast is smaller than MTO’s but its disposable income supplement is larger? The additional income a family receives could reduce its members’ psychosocial stress, or it could purchase more health services. Either or both of these mechanisms could then improve disease-related biological processes and physical and mental health in both adults and children.
174 Moving to Opportunity Making MTO Health Results More Relevant to Current Housing Policy: Next Steps Numerous correlational studies imply a link among income, biology, and health, as do some laboratory analog studies described in Adler and Stewart (2010), as do well-identified causal analyses of the health effects of both food stamps (Almond, Hoynes, and Schanzenbach, 2011) and the Earned Income Tax Credit (Hoynes, Miller, and Simon, 2011). Still unknown, however, is how the total effect of combining the larger income supplement and the smaller neighborhood upgrade in the Section 8 Program writ large compares with MTO’s total health effect. Future research to examine this issue should also probe causal mechanisms. Is the average income supplement from vouchers substantial enough by itself to affect health to a degree that is meaningful for policy? Is the reduced neighborhood contrast relative to MTO nonetheless large enough to affect health to a meaningful extent? Perhaps especially important are questions about how income supplements and neighborhood improvements combine and interact to jointly influence health.
Housing Supply and Demand Would Probably Be Different in the Section 8 Program Than in MTO Imagine a policymaker who wants to use the MTO health results to justify redesigning the Section 8 Program so that its recipients can use vouchers only to move to neighborhoods with less than 10 percent poor households. Such a policy supposes two things that are very likely wrong. The first is that the supply of affordable rental units in these affluent settings can meet the increased demand from new voucher holders. Affluent communities tend to be characterized by a greater fraction of individually owned homes as opposed to rental units, and many rental units in these communities are more expensive than voucher-eligible families can afford, even with a voucher. Also, a national program restricting voucher use to affluent neighborhoods would surely bid up rents in those places.
Offering incentives to construct more rental units would, of course, offset such an increase. In affluent neighborhoods, however, we anticipate considerable reluctance to authorize the construction of more rental units at prices affordable for voucher-eligible families. Such resistance would probably be weaker for subsidized construction for elderly people and would probably be especially strong if the construction were for families with children, especially teenagers. In many affluent locations, it would be very difficult to achieve the number of affordable units needed to meet the increased demand that would follow from a Section 8 Program mandate to use vouchers only in affluent neighborhoods.
It is important to realize, however, that many voucher-eligible families might not want to live in affluent settings. Of families in the MTO experimental group, 53 percent did not use their voucher at all, one (of many) possible reason being that they did not want to live in places so different from the neighborhoods they were used to. Moreover, some of the families who moved initially did so again over the ensuing study years, after the requirement to live for 1 year in a neighborhood that was 10 percent poor lapsed. Most subsequent moves were to less affluent neighborhoods; that is, to settings more like those they initially left than like those into which they originally moved. Many reasons might explain this systematic mobility pattern, but one is surely that families from public housing preferred settings more sociologically like those they already knew. This predicament is most acute for racial or ethnic minority families who are fearful that affluent neighborhoods will tend to be majority White and replete with overt or covert racial prejudice. The MTO data suggest this
possibility, because almost all of MTO’s voucher holders from racial or ethnic minority groups moved into nearby affluent minority neighborhoods and not into nearby affluent White neighborhoods.
MTO families randomly offered a traditional Section 8 voucher, despite coming from public housing, probably provide the closest approximation to the behavior of the typical private-market housing family in the Section 8 Program. Families in the MTO Section 8 group were more likely to lease up (62 percent) than families in the experimental group (47 percent). Also, their spontaneous moves were to neighborhoods with fewer nonpoor families (71 percent nonpoor) than those of the lowpoverty voucher group (89 percent nonpoor). Again, we can invoke many reasons for such data, but one possibility is that the Section 8 group families voted with their feet in ways that reveal a preference for neighborhoods less affluent than those that MTO’s low-poverty voucher required.
What about families in the Section 8 Program who are relatively more affluent and more likely to be White? Will they be as inclined to avoid neighborhoods with poverty rates as little as 10 percent?
We do not know, but consider that they started in 2007 in Chicago from a base rate of 29 percent, not 53 percent, poor. Many of those families would therefore live in areas close to 10 percent poor, and for the others, the transition from 29 to 10 percent is less than from 53 to 10 percent. Our speculation is that fewer families in the Section 8 Program than in MTO would want to avoid neighborhoods that are 10 percent poor. This speculation means that scaling up the main MTO finding in the experimental group may be more difficult for most Section 8 Program voucher recipients, who are not as poor or as likely to be racial or ethnic minorities as are those in the MTO population.
They might be more likely to want to move into affluent neighborhoods, thus swelling the demand for units in areas where the supply is already limited. Scale-up would be less problematic, of course, if federal authorities issued many fewer new vouchers, or if they somehow managed to impose real constraints on local private housing markets to implement a policy with teeth that encouraged moves into affluent neighborhoods. Currently, neither policy seems likely. It is hard to see, therefore, how MTO’s main treatment arm could be scaled up within the Section 8 Program to capture MTO’s health results.
Beyond MTO’s Biological and Health Measures HUD did not originally design MTO with a central health focus. That focus emerged as primary halfway through MTO, when it became clear that the anticipated socioeconomic and educational effects were not occurring but that positive mental health effects were. Height and weight measures therefore gained new salience, and researchers added some health and biological measures to the final data collection wave. They obtained positive results for extreme obesity, diabetes, and HbA1c, suggesting a causal pathway between improved glucose regulation and reduced cardiovascular disease. The theoretical link among the three health outcomes, and from there to cardiovascular disease, makes the MTO health findings so credible, as does the fact that each is assessed in a quite different way—by the physical measurement of height and weight, self-report, and dried blood, respectively. Also adding credibility to MTO’s health findings is the consistency of the positive mental health findings obtained at both the study’s middle and end points, and for females in both their adult and youth years.
176 Moving to Opportunity Making MTO Health Results More Relevant to Current Housing Policy: Next Steps Many senior members of the medical research and policy establishments tend not to take social science findings seriously, especially if only self-reported or simple anthropometric assessments are available. They prefer biological measures that are part of well-established medical theories that manifestly predict subsequent serious diseases and are collected from, say, blood, sputum, or urine. They also prefer clinical assessments and cutoff values that are normative among health researchers and policy analysts. Like other scientists, they also seem to take more seriously findings that have a broad rather than a narrow reach. Justifying any housing policy because of its health consequences requires housing researchers to provide knowledge that the medical research and medical policy communities can freely embrace because the knowledge fits within their professional frames of reference. Thus, the concern in housing research on health is to use general clinical diagnoses, demonstrate biological mechanisms, assess clinical disease end points, and be applicable to large populations of individuals.
MTO went a considerable distance along this path, but probably not as far as it would have had it been initially framed as a study of housing and health. Its findings are from a smaller (but on average needier) population than the national population of Section 8 Program voucher holders.
Positive findings emerged for a category labeled “extremely obese” but did not statistically replicate for the larger and more commonly used “clinically obese” group, with its lower cutoff value. MTO examined asthma by self-report, but these reports did not vary by treatment group. Although MTO obtained a positive result for health outcomes and the HbA1c biomarker, indicative of improved glucose regulation, much past interest in how physical and social settings affect health has concentrated on immunological pathways that lessen resistance to pathogens and thus promote many kinds of disease, including cardiovascular disease (Adler and Stewart, 2010). Other biomarkers, such as Interleukin 6, C-reactive protein (CRP), and Epstein-Barr 18 Virus, therefore, also require careful examination. CRP was assessed in MTO, and it was marginally related to the low-poverty treatment, raising at least some hope that housing will affect pathways to disease based on regulating immunological and glucose functioning. We need a study of housing mobility that is initially and explicitly focused around causal links from housing to health.
What About Child Health?
The final measurement wave of MTO included not only adults but also youth. These youth were children when MTO began, but few health measures were taken from them as children, so MTO reports only youth results. There was a positive mental health effect for female (but not male) youth and no positive results for either gender for any physical health measure, so MTO’s preadult health story is a mixed bag.
There are good reasons, however, for expecting positive health consequences before adulthood.
Regular Section 8 Program voucher holders can move to better homes or neighborhoods, increase their discretionary income, or combine both. Each of these options should separately reduce psychosocial stress in the family, and reduced stress is a well-established mediator of improved biology and health in adults and even small children (Adler and Stewart, 2010). It is not, however, the only relevant causal mediator. Also relevant is that the discretionary income a voucher provides in the Section 8 Program can be used to access more and better health services. In addition, after moving, a family might also increase its members’ exposure to information and social models relevant to
leading a healthier lifestyle. All three mechanisms—reduced stress, more access to health services, and exposure to healthier lifestyles—should complement each other and promote better biology and health in general, including in children.