Is the Low-Income Housing Tax Credit “Keeping Cities Segregated”?

The front page of the New York Times this morning features a full-frontal assault on the low-income housing tax credit, the largest federal subsidy for the development of affordable housing. The charge against the credit is that the housing units it subsidizes are “disproportionately built in majority nonwhite communities,” which “means . . . that the federal government is essentially helping to maintain entrenched racial divides.” The first part of that claim is indisputably true: developments receiving low-income housing tax credits are, indeed, disproportionately located in communities with large nonwhite populations. But it does not therefore follow that the federal government, through the credit, is perpetuating residential segregation.

The low-income housing tax credit (abbreviated “LIHTC” and pronounced “lie-tech”) is an enormously complicated program, but to summarize (and somewhat oversimplify), it functions as follows: The federal government, through section 42 of the Internal Revenue Code, allocates credits among states based on population, and state agencies then decide which developments the credits should go to. Qualifying projects must satisfy either a “20–50 test” (20% of the units are occupied by renters whose incomes are 50% or less of the area median) or a “40–60 test” (40% of the units are occupied by renters whose incomes are at or below 60% of the area median). These units must be rent-restricted, with the aim that rents should not exceed 30% of the occupants’ income. The credit calculation is even more complicated, but the basic idea is that for every $100 spent building affordable units, the investors in the project will receive — over the course of 10 years — tax credits worth up to $70 of construction costs. There are many more rules that apply to credit-financed projects; the important point for now is that the program provides substantial subsidies for the development of affordable housing — totaling around $8.9 billion in fiscal year 2018 and rising to $10.0 billion by 2020, according to the Joint Committee on Taxation. (For full disclosure: My spouse represents nonprofit and for-profit developers in LIHTC-financed transactions.)

Under certain circumstances, developments can receive credits covering even more than 70% of construction costs. One way to do so is by building in a census tract where at least half of households have incomes less than 60% of the area median or at least a quarter of households fall below the federal poverty line. Unsurprisingly, this means that a lot of LIHTC-financed investment is concentrated in relatively low-income neighborhoods, and since race and income are correlated, that in turn means a lot of credits go to developments in majority nonwhite areas.

Now, one might ask whether it’s a bad thing that the federal government is encouraging investment in communities that have experienced decades of disinvestment. But the Times article does not grapple with that question. Instead, it assumes (with qualifiers like “fair-housing advocates say”) that if the federal government is investing in affordable housing in majority nonwhite neighborhoods, then it’s “keeping cities segregated” — and that this scandal should be page one news.

The argument against federal financing for affordable housing in majority nonwhite neighborhoods — the argument that animates the Times article — goes something like the following: Housing desegregation will only happen if nonwhites move to majority white communities, which means (given the race-income correlation) that we need to build housing for low- and moderate-income families in neighborhoods with white majorities. Moreover, moving low-income families to higher-income (often whiter) neighborhoods will be good for kids. As the Times says, “Research suggests that when children from low-income households grow up in affluent communities, they tend to get a better education and earn more money as adults.”

Now, the research on the long-term effects of moving low-income families to more affluent neighborhoods is much more ambiguous than the Times article acknowledges. The best evidence comes from the much-studied “Moving to Opportunity” experiment in the 1990s, which randomly selected families for vouchers that enabled them to move to more affluent areas. A recent analysis by Raj Chetty, Nathaniel Hendren, and Lawrence Katz finds that moving to a more affluent area significantly improves educational and economic outcomes for children who are younger than 13 when their families relocate, but “the same moves have, if anything, negative long-term impacts on children who are more than 13 years old when their families move.” In other words, “moving to opportunity” is not an unalloyed good.

My own view, for what it’s worth (which is not much), is that federal housing policy should seek to facilitate the relocation of low-income households to more affluent communities, and that programs such as Moving to Opportunity should be scaled up. But at the same time, Moving to Opportunity is not infinitely scalable, and it cannot be our entire strategy for residential desegregation. No one (or at least, hopefully no one) thinks that housing policy should aim to empty out Chicago’s predominantly African-American, lower-income South Side and move everyone to Chicago’s predominantly white, higher-income North Side. We can’t just depopulate communities that have suffered from decades of disinvestment; we also need to invest there. And if we really want to achieve residential desegregation, we’ll have to encourage white families to move to majority nonwhite communities, as well as encouraging nonwhite families to move in the other direction. Residential desegregation is a two-way street.

A comprehensive housing policy that improves conditions in historically low-income neighborhoods while encouraging racial and economic desegregation on a metro-wide scale will need to be a three-legged stool (to repurpose a metaphor from the health care debates). The three legs are:

— (1) Encouraging and facilitating the movement of nonwhite and lower-income families to whiter and more affluent neighborhoods;

— (2) Encouraging and facilitating the movement of white and higher-income families to nonwhite and less affluent neighborhoods;

— (3) Promoting investment in nonwhite and less affluent neighborhoods that have experienced decades of disinvestment.

When credits flow to developments in nonwhite and lower-income communities, LIHTC functions as a leg-three (and to some extent, leg-two) policy. A new paper by economists Rebecca Diamond and Tim McQuade, both at Stanford’s Graduate School of Business, finds that LIHTC-financed investments in low-income communities lead to increased property values and lower crime rates. Moreover, Diamond and McQuade find that LIHTC construction in high minority areas leads to a statistically significant increase in the share of non-black homebuyers moving into those areas. In the authors’ words, “it appears that building affordable housing in high minority areas may lead to lower racial segregation.” That’s exactly the opposite of what the Times headline claims. And yet the Diamond and McQuade study is only briefly mentioned in a single sentence at the end of paragraph 40 of the 51-paragraph story.

To put the point more bluntly: The headline is that “Program to Spur Low-Income Housing Is Keeping Cities Segregated,” but what’s probably the most comprehensive study of the subject suggests that the program probably makes cities more integrated.

None of this is to dispute the fact that the LIHTC statute could be revised so that more credits go to affordable housing developments in whiter and more affluent neighborhoods. (A bipartisan bill introduced in the Senate this past March, the Affordable Housing Credit Improvement Act of 2017, would advance that goal by making it harder for local opposition to stop projects from being funded. The bill also would scale up the program by 50%.) Nor is this to dispute the claim that states should change their credit allocation criteria so that more credits go to projects in whiter and more affluent communities (as Texas did in response to a Fair Housing Act lawsuit in 2013). Leg one of the stool is still critically important.

But it’s also important not to mischaracterize a program that provides billions of dollars a year in federal subsidies to communities that have historically received less than their fair share of federal housing dollars. As the Trump administration proposes to cut federal housing funding in next year’s budget by $6 billion, it’s a surprising time for the Times to attack one of the few affordable housing programs that’s not on the chopping block. At the very least, we should acknowledge (in more than a deeply buried sentence) that on this extraordinarily complex issue, there is more than just one side.