«reNtal HousiNg Policy iN tHe uNited states Volume 13, Number 2 • 2011 U.S. Department of Housing and Urban Development | Office of Policy ...»
Put aside taxes for a moment, and assume that rents increase each year. In equilibrium, the marginal owner’s cost should be indifferent between the cost of owning and renting.
Gross rent is the cash flow expense to the tenant; the sum of OCC and operating expenses is the cash flow expense to the landlord. If rents are expected to grow (as they should in a developing economy), the cash flow expense of owning at the beginning of an ownership period is greater than the cash flow expense of renting. As rents rise, the relationship can reverse, but, for those facing consumption constraints, the cost of owning in early years can be greater than the cost of renting. For owners with large mortgages, the cash flow expense takes the form of interest payments;
for owners with large amounts of home equity, the cash flow expense is the foregone income from an investment other than housing.
For a poor person—and in particular for a poor young person who would need to take on a large mortgage to buy a house––who is consumption constrained, the ability to consume housing with less cash flow in a year could be very desirable. For an elderly, low-income person, converting home equity into an alternative investment can make the ability to consume rental housing quite desirable. There can be no doubt that such households are better off (or at least not worse off) when the option is available.
Policies Favor Owner Housing and Discourage the Development of Rental Housing Despite the fact that rental housing can advance desirable policy goals, the United States has a pervasive policy bias toward owning. One possible consequence of such policy is that the market alone will not provide adequate levels of rental units at sufficiently low prices. Although rental subsidies might create distortions in an otherwise first-best world, they may improve welfare in a secondbest world (it is also unlikely that the distortions that cause capital to flow to owner-occupied housing will be removed).
Policies that are beneficial to owner-occupied housing (or hostile to rental housing) may be divided into three types: (1) zoning policy (which is implemented by local governments), (2) tax policy (some of which is implemented at the local level, some of which is implemented at the state level, and some of which is implemented at the federal level), and (3) financial policy, which is implemented at the federal level.
Zoning Policy Zoning came into existence for a very sensible policy purpose: to eliminate (or at least reduce) the negative effects—or externalities—that one type of land use might have on another. The iconic example is the oil refinery that wishes to locate next to a neighborhood of houses: the private benefit to the refinery might be large, but the cost to the adjoining neighborhood would also be high— likely higher. Such a use is sometimes referred to as a “nuisance use.” The benefit of zoning is not only that it prevents a nuisance use, but also that it reassures people that they are avoiding the risk of a nuisance use. Hence, in principle, zoning is sensible policy.
Two metrics measure in theory whether zoning is optimal. First, if zoning exists, such that land values are maximized (that is, that no change in zoning would increase total property values), then zoning is optimal. Alas, knowing whether values are at their maximum possible value at a given level of zoning requires the knowledge of the full set of values that would be produced given a full set of zoning options, which is not possible.
A short note by Colwell and Dehring (1999) implies a somewhat better metric. Although the model contains many strong assumptions, it offers some insight. The model has a market with two types of land use, one of which produces negative externalities (commercial) and one of which does not (residential). In the absence of zoning, when the two types of land uses abut each other, their values will be equal—the value of the residential property will be encumbered by the commercial property, and equilibrium requires them to be equal at their boundary. As residential uses become more distant from commercial uses, the negative effect of the commercial uses on residential uses becomes smaller. On the other hand, because commercial property is not affected by residential property, its value remains the same everywhere (this conclusion assumes that views, topography, and the distance from the central business district, etc., were considered). Consequently, residential land that is not next to commercial land is more valuable than commercial land. Because commercial land is relatively inexpensive, commercial uses tend to get too much investment, and residential uses too little. Optimal zoning would restrict commercial development until the value of 42 Rental Housing Policy in the United States Thoughts on Rental Housing and Rental Housing Assistance commercial land is just equal to the value of residential property that is sufficiently far from commercial activity that its value is not affected by it. A test of optimal zoning, then, is whether property values away from boundaries between uses are the same or different. If commercial uses are restricted to the point that commercial values become higher than residential values, zoning has gone too far; it prevents the optimal development of commercial uses.
Communities seem to have a presumption that apartments have a negative effect on single-family houses—this presumption would explain why apartments are almost always subject to more restrictive zoning than single-family houses. Yet no compelling evidence indicates that apartments produce negative externalities.3 Certainly, apartments create a smaller fiscal burden for cities, because only about 15 percent of apartment dwellers in buildings with more than 50 units have children, whereas about 37 percent of dwellers in detached single-family homes have children.4 The question of whether communities restrict apartments too much5 is not a settled question, although a previous HUD report essentially argues that it is (Ashey and Kean, 1991). It is a question worth revisiting, and HUD could survey whether apartment land values are more or less equal to single-family property values across a number of U.S. cities.
If communities are restricting apartment construction beyond what is socially justified, the effect of zoning policy is to levy a tax on apartment dwellers. According to the 2007 AHS, the median income of those living in detached houses is more than twice the median income of those living in large apartment buildings. Equity suggests that low-income apartment dwellers should receive some sort of subsidy. Too much zoning also leads to a shortage of rental units, which is economically inefficient.
The alternative is to induce municipalities to loosen up zoning codes; the politics of such actions are problematic. Although economists are suspicious of subsidies in general, in a second-best world, undoing the effects of one distortion by introducing another may improve social welfare.
Tax Policy Nearly every country in the world uses tax policy to encourage homeownership. The method is passive: equity owners of housing earn imputed rent, and such rent is rarely taxed. The Congressional Budget Office (CBO) includes nontaxation of net imputed rent as a tax expenditure.
One could argue that such treatment (nontaxation) makes sense from the standpoints of transparency and simplicity. Taxing imputed rent would be difficult to explain to taxpayers, and so it is not transparent. As for simplicity, taxes are best levied on things that are easily measured. The difficulty that the Bureau of Labor Statistics has with measuring the owner-housing component of Consumer Price Index suggests that imputed rent is NOT easily measured. Indeed, the fact that before 2007 Although apartments are not synonymous with rental housing, multifamily properties have a higher share of the rental market than do single-family homes. According to the 2008 AHS, 80 percent of units in buildings with more than 50 units had renters and 86 percent of detached single-family units had owners.
Children are not negative externalities, but the short-term cost of educating them is.
Communities also use ordinances such as occupancy codes to prevent nonrelated people from renting together, which is effectively a restriction on rental housing.
the CBO estimate of the tax expenditure on imputed rent was smaller than the tax expenditure on the home mortgage interest tax deduction (additional information follows) shows how difficult such a measurement is. Until 2007, the value of home equity in the United States was higher than the value of mortgage debt outstanding (Federal Reserve, 2011), and returns to equity should be higher than returns to debt. The implication is that imputed rent should have been larger than mortgage interest.6 It is easy to understand why imputed rent is not taxed, but the fact that it is not taxed makes the mortgage interest deduction something of a puzzle.7 Many English-speaking countries (Canada, Australia, and the United Kingdom) do not have policies that allow homeowners to deduct mortgage interest. Moreover, the idea that mortgage interest deduction was designed to spur homeownership is something of a myth: it is actually a residual of the 1913 Income Tax Code, which allowed all consumer interest to be deductible. The predecessor proposals to the Tax Reform Act of 1986 would have eliminated the deductibility of all consumer interest. Brilliant lobbying on the parts of the National Association of REALTORS®, the National Association of Home Builders, and the Mortgage Bankers Association convinced Congress that the mortgage interest deduction was crucial to maintaining and increasing the homeownership rate in the United States (Birnbaum and Murray, 1987).
Little, if any, evidence suggests that this claim is correct. The countries listed previously that lack a mortgage interest deduction policy have homeownership rates similar to that of the United States, which is 66.4 percent: in Australia it is 71 percent, in Canada it is 65 percent, and in the United Kingdom it is 69 percent (Proxenos, 2002). Some simulations (Green and Reschovsky, 1999) imply that the ownership rate would fall by a miniscule amount if the mortgage interest deduction were eliminated. Another simulation (Green and Vandell, 1999) shows that a targeted tax credit would be a far more effective method for encouraging homeownership than the current mortgage interest deduction.8 The largest problem with the mortgage interest deduction is that it provides virtually no benefits to below-median-income households. Such households, even when they own, may not take the mortgage interest deduction at all because, for them, the standard deduction will be more valuable than itemized deductions, including one for mortgage interest.9 And even if they do itemize, the marginal benefit of the deduction, relative to the standard deduction, will be quite small, meaning that it will provide little incentive for ownership. At the same time, the deduction provides disproportionately large tax relief to those at the upper end of the income distribution (Follain and Ling, 1991). Finally, the mortgage interest deduction encourages households to take on debt. Canadians One could calculate imputed rent by calculating the total value of home equity and multiplying that total by some interest rate. But it is difficult to know the “correct” rate; as such, a rate would be a function of both expected duration and risk.
Both of these factors could vary substantially across homeowners.
Landlords can deduct mortgage interest, but they also pay tax on the rental income they receive.
Classic papers on the effect of the mortgage interest deduction include Poterba (1992), Rosen (1979), and Rosen and Rosen (1980). More recent papers include Glaeser and Shapiro (2003), Green and Reschovsky (1999), and Green and Vandell (1999).
Rosen and Rosen (1980) wrote their paper at a time when the standard deduction was much smaller, so that the mortgage interest deduction was more valuable to lower income households.
and Australians are just as likely to be homeowners as Americans, and they use mortgages, but they pay their mortgage balances more quickly, meaning their households are less leveraged and, therefore, less vulnerable. Only a limited number of economists (Woodward and Weicher, 1989) have kind things to say about the mortgage interest deduction. They argue that an elimination of the deduction gives an advantage to those who can buy houses with cash (presumably higher income people) relative to those who need mortgage financing.
The mortgage interest deduction may also make rents more expensive than they otherwise would be. Bid rent theory implies that the bidder who most values the land will set the land prices. Consider two bidders: one has the benefit of the mortgage interest deduction, and another does not (that is, a renter). Because the bidder who uses the deduction faces a lower after-tax cost of capital, she will outbid the renter, thus leaving the renter with the options of paying the high price set by the subsidized bidder, or settling for an inferior location (which might have higher commuting costs). Consequently, the total cost of renting per unit of housing quality is driven upward by the mortgage interest deduction.
Despite the fact that many policy analysts have issues with the mortgage interest deduction, the last time it was threatened was during the debate about the Tax Reform Act of 1986. Gyourko and Sinai (2003) estimated the net benefits of the mortgage interest deduction by congressional district.
On the one hand, they found that if the proceeds from the mortgage interest deduction were returned as lump sum payments to taxpayers, most districts would benefit from the elimination of the mortgage interest deduction, but by a small amount. On the other hand, the districts that benefit from the mortgage interest deduction would lose large amounts if it were eliminated. When a policy’s benefits are concentrated and costs are diffused, it will usually have substantial lobbying support.
As with zoning, the mortgage interest deduction leads to a suboptimal allocation of rental housing, and makes rental housing more expensive. It may be reasonable for the federal government to offset these negative effects with subsidies.