«reNtal HousiNg Policy iN tHe uNited states Volume 13, Number 2 • 2011 U.S. Department of Housing and Urban Development | Office of Policy ...»
A fraction of eligible households receive a large subsidy. A larger fraction of eligible households receive no assistance. The distribution is capricious.
For example, Currie (2006) reported that under current rental subsidy policies, more than 70 percent of households below the poverty line are not served, and more than 40 percent of the households who are served are not in poverty. A couple of years ago, it was reported (Quigley,
2008) that less than one-third of renters with incomes below 30 percent of the local median received housing assistance, and less than one-fifth of renters earning between 31 and 50 percent of local median received housing assistance.
Could Anything Really Be Done?
Maybe nothing can be done to correct this imbalance, but something closer to equitable assistance for poor renters could be achieved if the eligibility rules for housing subsidy were made more realistic and if the program were financed using the successful models employed in other U.S.
housing subsidy programs.
First, eligibility rules for rental housing assistance would need to be tightened. Under current law, households with incomes below 80 percent of the Area Median Income (AMI), adjusted for
household composition, are eligible for rental housing subsidies. In 2006, this average cutoff income was about $52,000 for a family of four (Quigley, 2008). At that time, eligibility for food stamps for four-person households was confined to those with incomes less than one-half as large ($25,000). Eligibility under the EITC program was limited to households (with one or more children) earning one-third less per year ($37,000). Eligibility for rental assistance would have to be tightened considerably to replace a national lottery program with a broader and more equitable program for housing assistance for very low-income renters.
Objections exist to the labor supply implications of adding another income-based entitlement program to the array of federal policies. Much of this concern seems misguided, or else simply rationalizes opposition to a more transparent and universal housing assistance program.1 Second, a broader program would require additional support outside the community of housing advocates and professionals, and the continuity of the program would be problematic. One way to increase political support, and to reduce administrative costs as well, would be to follow the politically successful programs of subsidies to homeowners and subsidies to builders for the construction of low-income housing. (New construction of low-income housing is discussed further in this article.) These programs use the relatively efficient systems of the Internal Revenue Service (IRS) to determine eligibility and to distribute the benefits.
Currently in the United States, the multibillion-dollar home ownership subsidies are distributed largely by the IRS. Individual taxpayers need not report the dividend (that is, the imputed rent) on owner-occupied housing. Interest and property tax payments are deductible as personal expenses, and capital gains on sale are accorded special treatment in the computation of tax liability. (Jaffee and Quigley [2007 and 2010] provide estimates of the costs of these provisions. The costs are large.) The distribution of these subsidies to qualifying households is relatively painless. The subsidy is paid as a credit against the other tax liability of a homeowner. But subsidies provided under these provisions of the tax law for owner occupants are not refundable to the taxpayer.
In contrast, EITC is fully refundable to the taxpayer. Eligibility for the credit can be established on line (using the “EITC assistant,” for example). Alternatively, the IRS will establish eligibility and will compute the credit due—and they will also send along a check—to any qualifying taxpayer.
A refundable credit is not hard to administer.
In fact, a housing program that the IRS already administers could serve as the template for a lowincome rental housing subsidy program of this kind. The Mortgage Credit Certificate Program authorized by the Deficit Reduction Act of 1984 entitles selected homeowners to claim a tax credit for some portion of the mortgage interest paid in any year, rather than the tax deduction afforded other homeowners. (See Greulich and Quigley  for a detailed discussion.) A taxpayer in Evidence indicates weak effects of the EITC on the labor supply of recipients (Eissa and Hoynes, 2006), and stronger effects of welfare receipt upon labor market outcomes. (See Moffitt  for a review.) Less is known about the work incentives of the food stamp program (Hoynes and Schanzenbach, 2010). Fischer (2000) found small effects of housing assistance on work incentives, and Shroder (2002) found essentially no effects of housing assistance upon the employment of recipient households. Clearly, it is to be expected that any labor market distortions under a universal housing assistance program would be smaller than these estimates, not larger (because an entitlement program would undoubtedly provide a lower level of benefit, but one enjoyed by many more households).
possession of a Mortgage Credit Certificate issued by a unit of state or local government merely checks a box on his or her tax return (on line 54 of Form 1040) and submits an additional form (Form 8396, only 11 lines long) to claim the nonrefundable credit.
To claim the low-income housing subsidy under this more equitable low-income renter assistance program, the taxpayer would need to submit a form issued by a local housing authority and check a box added to the current IRS form. The additional form would certify that the household was renting a dwelling that meets the minimum habitation standards imposed by HUD. That form, together with the income reported by the household, the number of dependents in the household, and the postal address of the household, would be sufficient to compute the low-income renter subsidy payable to any household.2 The computation could be made by any taxpayer (perhaps on line) or by the IRS, as is the case with the EITC. Of course, H&R Block or any other commercial tax preparer could also make the computations. The private sector would have an incentive to help administer the program.3 The appropriate credit could be mailed in monthly installments to the low-income household, to the local housing authority for distribution to the household, or to the landlord directly.
Financing the low-income renter assistance program through the IRS and administering the program through HUD may also facilitate long-overdue reforms to the tax code in the treatment of housing.
For example, a revenue-neutral way to finance a low-income renter assistance program of roughly the same size as all current HUD low-income subsidies would be to eliminate the capital gains exclusion currently afforded to owner occupants when they sell their dwellings (Jaffee and Quigley, 2007, 2010). Reducing the limits on the deductibility of interest payments for high-income homeowners could easily finance a more equitable universal program. Using the tax code to support low-income renters may thus further national goals of equity in the tax treatment of housing by the federal government.
What About Unit-Based Housing Subsidies?
Viewing rental housing subsidies as part of the modern welfare system is very different from conceptualizing these subsidies as part of an infrastructure investment program—the rationale for the program 70 years ago. Ensuring equal treatment of eligible households as a part of a national welfare program is quite different from a policy of using rental subsidy funds to design and build new dwellings to be rented at below-market rents—at any conceivable budget. And the reason is obvious.
The cost of providing decent-quality housing through new construction is much greater than the cost of providing it by using the existing depreciated stock of housing. This fact is well known to It would be simple to differentiate the payment by postal code (and hence metropolitan area) if it was considered desirable to vary subsidies by the local cost of living. Most economists would probably not support such differentiation (for example, Glaeser, 1998), but many politicians might.
The computations would be a bit more complicated than those under the EITC because incomes of potential recipients may vary within the year. Currently, no administrative mechanism exists in the IRS for recertifying incomes within the year (perhaps on a quarterly basis). But this duty could be assigned to local housing authorities, whose other burdens would be reduced under the proposed housing subsidy program.
builders and developers, who almost never target new construction of rental units to the bottom one-half of the income distribution. (This fact is also quite well known to slumlords, who offer small quantities of housing services to the poor, using the oldest and most obsolete portion of the housing stock.) These cost differences in shelter provision for low-income households were thoroughly documented in conjunction with the Experimental Housing Allowance Program a quarter of a century ago and, more recently, by the General Accounting Office (GAO, 2001, 2002), now the Government Accountability Office. The latter study concluded that the present-value lifecycle costs of new construction subsidies were from 19 to 38 percent more than were the costs of voucher programs for comparable housing.4 No conceivable budget that sought to cover all renters below some income cutoff level could make provisions for the expenditures that are required to provide newly constructed housing for assisted households.
Should programs that provide unit-based housing assistance (that is, housing subsidies tied to particular dwelling units) be discouraged? Currently, a little more than two-thirds of the low-income renter households subsidized through the Section 8 program (2.2 million recipients) receive vouchers that can be used locally and which often are portable regionally. These vouchers make up the difference between 30 percent of household income and the estimated costs of “just standard” housing. In contrast, about one-third of Section 8 recipients receive equivalent financial subsidies in designated dwelling units under long-term contract to HUD. The benefits provided to recipients under the latter arrangements (unit-based assistance) are not as valuable as those provided by unconstrained vouchers. The recipient of a voucher could always choose the designated dwelling unit if it most closely matched his tastes and preferences, but the recipient of an unconstrained voucher could also choose another unit. A voucher recipient can also change dwellings in response to job changes, schooling needs, or neighborhood conditions.
Beyond these advantages to the individual recipient, the system of unconstrained vouchers provides an additional incentive for the deconcentration of low-income households in urban areas. Those who are willing to incur search costs (and perhaps other nonmonetary costs associated with integrating neighborhoods by income and race) make it easier for those searching subsequently in the housing market. None of these spatial incentives are present in unit-based Section 8 housing.
Indeed, some concern exists that unit-based assistance concentrates disadvantaged renters.5 The imposition of geographically determined unit-based housing assistance under Section 8, instead of portable and flexible vouchers available under the same program, might be understandable— This additional cost is apparently well known, if not extensively documented. For example, the analysis of the Moving to Opportunity (MTO) experiments by Kling et al. (2007) concluded that the MTO treatments pass the cost-benefit criterion because “the MTO intervention[s] produced large mental health improvements and because other research suggests that it is cheaper to provide a unit of subsidized housing with vouchers than in a public housing project” (2007: 108).
Little evidence supports the theory that supply-side housing assistance helps to disperse low-income households across neighborhoods. For example, Freeman (2004) reported that newly subsidized units created under the tax credit program are located in neighborhoods that contain a disproportionate share of Black residents. These neighborhoods also have considerably higher poverty rates, lower median incomes, and lower house values than typical metropolitan neighborhoods.
It should also be noted, however, that little evidence indicates that voucher recipients use vouchers to purchase housing in better neighborhoods. Under the MTO experiment, for example, recipients of MTO vouchers were required to move to neighborhoods with much lower concentrations of low-income households. After about a year, however, a large fraction of MTO recipients moved again––to neighborhoods with higher concentrations of poor households. (See Quigley and Raphael  for a discussion.) 152 Rental Housing Policy in the United States Rental Housing Assistance if the former was cheap enough relative to unrestricted vouchers. But no evidence suggests that this is the case.6 Absent a compelling rationale based on comparative costs, it would be administratively simple to unwind unit-based housing assistance over time, as contracts with landlords expire, and replace them with more flexible vouchers.
Of course, there may be circumstances in which just standard, but depreciated, units are, in fact, cheap enough to make federal leases of these units competitive with demand-side assistance— especially if these leases also reduce negative external effects. Under current market conditions, in regions with high vacancy rates and abandoned properties, federal assistance to private or nonprofit owners in return for shelter provided to low-income renters may well pass a strict welfare test. If acquisitions can be made cheaply enough, if external effects (for example, from abandoned or foreclosed properties) are large enough, a place-based subsidy program might be quite sensible, at least in the short term.
What About Newly Constructed Low-Income Housing?