Comment
Roger G. Noll: U.S. housing policy is a jumble of numerous programs, many of which appear to use similar means to accomplish the same objective. Professors Jaffee and Quigley do a comprehensive job of categorizing and assembling the financial costs of these programs. The only housing policy that is missing from the analysis is a group of programs that provide subsidies for housing the homeless, but collectively these programs are small compared to the programs analyzed in their essay.1 For other direct housing assistance programs and tax benefits, the authors provide estimates of the distribution of the financial gains from these policies by income class. For programs that promote homeownership through mortgage guarantees, the authors summarize other research that evaluates these programs and offer a policy reform proposal. The policy recommendation is that mortgage assistance programs be more targeted toward low-income households, that the government tax the debt of federal mortgage assistance agencies (to reflect the risk to the government of guaranteed mortgages), and that the government use the revenue to expand the housing voucher program.
These comments address two issues that could affect an assessment of housing policy, including the authors' proposed reforms, but are not addressed by them. The first is whether there is a rational basis for the complexity of federal housing policy and the implications of the relationships among programs. The second is the appropriate method for estimating the benefits and costs of housing programs. Whether it makes sense to tax the mortgage programs to [End Page 131] expand the voucher program and to orient mortgage programs more toward low-income households assumes certain answers to these questions.
The Mismatch between Policies and Instruments
A major puzzle about U.S. housing policy is why the portfolio of programs is so large. As Jaffee and Quigley note, U.S. housing policy apparently has two goals: to increase housing consumption among low-income households and to promote homeownership generally. As do Jaffee and Quigley, this section assumes that these goals are valid and focuses solely on whether they could be achieved more efficiently.
The mechanism for achieving housing goals is to lower the relative price of housing by offering subsidies, but the number of separate housing subsidy programs is over a dozen. Jaffee and Quigley document some of the complexity of housing programs. Whereas direct housing assistance to low-income households now focuses mostly on rental assistance vouchers, expenditures to support public housing remain almost as large as the voucher program. Other direct expenditure, tax expenditure, and mortgage guarantee programs seek to induce housing construction and renovation of rental housing for the poor, while other programs subsidize both supply and demand for homeownership among low-income households.2 In addition, most of the financial cost of tax expenditures and mortgage guarantee programs subsidizes families that are not poor.
Why the government needs numerous programs for each goal is a mystery. Jaffee and Quigley do not address why the federal government has so many housing programs and only implicitly address whether the efficiency of housing assistance could be significantly improved by simplifying the array of housing policies and programs or by reallocating effort among them. Implicit in their proposal to tax mortgage guarantee programs in order to expand voucher programs are three beliefs: the government's efforts in housing should be reallocated from promoting ownership to providing direct assistance; subsidized mortgage guarantees are an inefficient mechanism for promoting ownership (compared to direct and tax expenditures that do the [End Page 132] same thing); and as a means of reaching the poor rental vouchers are more effective than other housing programs, such as subsidized construction and renovation of rental housing or subsidized homeownership through down-payment and mortgage assistance. In context, their proposal arises from a discussion about how to internalize the fiscal risk of mortgage guarantees to the government, rather than more generally about how to improve the performance of the portfolio of housing assistance programs; however, the proposal implies two...
