n.V1-7.20 | HOUSING VOUCHERS AND ASSISTED MOVING

01| If where you live determines what your children earn, one solution is to change where you live. Moving to Opportunity is the only randomised test of it: families in high-poverty American public housing were randomly offered vouchers to move to lower-poverty neighbourhoods, with a control group that stayed.

02| Children who moved before age thirteen earned $3,477 more per year in their mid-twenties — 31 percent above a control mean of $11,270 — and were more likely to attend college. Children who moved as adolescents did slightly worse, most plausibly from the disruption of the move itself. Adults showed no earnings effect at all.

03| The structure of the result is what makes it causal rather than correlational:

"The gains from moving to a better neighborhood are proportioned to the exposure time during childhood. Every year of exposure to a higher-opportunity environment during childhood improves long-term outcomes, demonstrating that neighborhood environment operates as a cumulative, causal determinant of upward mobility."— Raj Chetty & Nathaniel Hendren (The Effects of Exposure to Better Neighborhoods on Children: New Evidence from the Moving to Opportunity Experiment, American Economic Review, 2018, Vol. 108, No. 6, p. 1104)

04| Selection cannot produce that shape. If better families simply chose better places, the effect would not scale with years of exposure. It scales, and the scaling is difficult to explain any other way.

05| Adults gained something too, and not what was expected:

"The long-term evaluation of Moving to Opportunity demonstrates that moving from a high-poverty to a lower-poverty neighborhood leads to sweeping, long-term improvements in physical and mental health, including lower rates of severe obesity and depression, even in the absence of significant impacts on adult economic self-sufficiency."— Jens Ludwig et al. (Neighborhoods, Obesity, and Diabetes — A Randomized Social Experiment, New England Journal of Medicine, 2011, Vol. 365, p. 1512)

06| Same income, different place, measurably better physical and psychological state. The place acted on the person directly, without going through the wage.

07| Here is the part that anyone arguing for improving neighbourhoods has to absorb. The decisive experimental evidence that place is causal was produced by an intervention that did not improve any place. Nobody fixed anything. The mechanism was removal, it worked on young children, and it worked in proportion to how many years of childhood were spent elsewhere. The strongest evidence for the importance of place is simultaneously the strongest evidence for getting people out of it.

08| The follow-up finding is harder still. Voucher holders were always free to move to high-opportunity areas and only about 15 percent did, which was generally read as preference — people are attached to where they live, have family there, and choose to stay. Bergman, Chetty, DeLuca, Hendren, Katz and Palmer added information, help with deposits, search assistance and introductions to landlords. Take-up went from 15 percent to 53 percent.

09| The barrier was friction, not preference. When leaving was made genuinely easy, most people left. Any argument that residents want to stay and rebuild has to be made against that number, and made honestly: much of what looks like attachment is the cost of the paperwork, the deposit and the phone calls.

10| The limits are severe and cut in the other direction. All of this is American, inside metropolitan areas with functioning rental markets, a voucher system, enforceable tenancies and somewhere better within commuting distance. Nothing about an informal settlement resembles that, and in many of the cities this book part is concerned with there is no obviously better place to move to that the household could afford or would be admitted to. Whether the friction result would replicate where the destination does not exist is not merely untested — it may not be a coherent question.

11| And one claim that gets made constantly on the other side has, as far as this review found, no evidence behind it at all: that a neighbourhood losing its most capable residents declines faster than a comparable neighbourhood that keeps them. It is the load-bearing argument for asking people to stay, it is stated as though established, and no study testing it was located. It also carries a cost that is rarely priced: if capable people are asked to stay, they are being asked to forgo something, and MTO puts a number on part of what — roughly 31 percent of their children's adult earnings.

12| The comparison that would resolve the entire argument has never been run: the same budget spent moving people versus improving the place, outcomes measured the same way. That is the question this whole literature turns on, and it does not appear to have been asked experimentally.