«reNtal HousiNg Policy iN tHe uNited states Volume 13, Number 2 • 2011 U.S. Department of Housing and Urban Development | Office of Policy ...»
The second market failure is the inability to write enforceable contracts promising the payment of rent. Partly this inability arises from state landlord-tenant codes that make instantaneous eviction impossible; but even without these codes, few landlords would evict all tenants who were a few days late on the rent and promised to pay next week. Because such contracts are not available, some Pareto-improving tenancies never occur. (An equivalent insurance contract would compensate landlords for tenant delinquencies. This contract also appears to be unavailable.) Some might argue that tenancies can be long-term relationships with flexible prices, and so all mutually beneficial insurance arrangements can be worked out over time. But tenancies can become long-term only if they are allowed to become so, and all tenancies start short. Hubert (1995) and Miceli and Sirmans (1999) showed that serious adverse selection problems prevent first-best contracts from being negotiated in tenancies that have a chance of becoming long-term, essentially because incentive-compatibility constraints need to be satisfied. Benjamin, Lusht, and Shilling (1998) estimated that tenants who are liquidity constrained and cannot make security deposits borrow effectively from landlords at an annual interest rate of around 30 percent.
130 Rental Housing Policy in the United States Rental Housing Assistance for the 21st Century More deeply, the people who are going to lose their jobs or get sick are going to do these things no matter where they live (to a first approximation). Some landlord is going to have to bear this cost. The time and energy that landlords devote to screening for this possibility is pure waste from a social perspective, since the only goal is to make some other landlord bear this cost instead of themselves. It is just a game of hot potato. Similarly, the distorting contract menus designed to combat adverse selection create obvious deadweight social losses, but are useful only for playing hot potato. Whether they combat adverse selection well or not is irrelevant, as long as they are costly.
In this sense, by making landlords less concerned with where people are living when adverse events occur, rental insurance can promote efficiency. A formal analysis of this issue has not been done yet, and would be helpful.
On the other hand, rental insurance might work against efficiency by decreasing geographic mobility—“locking in” people to the wrong locations when, for instance, jobs have shifted elsewhere. These efficiency costs, however, may not be large, since rental insurance payments would probably last less than 6 months, and migration decisions often take longer than that.
Why Can’t Private Firms Provide This Insurance?
Maybe they can, but because of the external costs of homelessness and instability, they would have to be subsidized.
Rental insurance markets, however, are likely to suffer from adverse selection. Prospective tenants are likely to have private information about their health and employment situation, and so may small landlords. Landlords also have private information about the channels through which they recruit tenants.
Subsidies offset adverse selection, and a private subsidized insurance scheme might be viable. But it might be better for HUD to start a program with a large, mainly involuntary base built in. For instance, apartment buildings and multifamily homes with FHA or GSE backing might be required to purchase such insurance. The history of mortgage insurance shows that sometimes the government needs to get a market started.
What About Moral Hazard?
Like every other kind of insurance, rental insurance will cause some moral hazard problems.
Designing intelligently can reduce the size of these problems when they occur.
On the landlord side, the obvious behavioral response to rental insurance is to loosen screening standards. This loosening may be desirable, in part to reduce homelessness. Nor is it obvious that markets produce optimal screening standards. Screening by landlord A affects the pool of prospective tenants that landlord B sees, and hence the standards that landlord B finds it optimal to adopt.
So the social costs of moral hazard on the landlord side may not be particularly big.
On the tenant side, several behavioral responses may occur. Tenants may seek higher quality and higher rent apartments because they have less to fear if something goes wrong; they may also seek riskier jobs because they have less to fear if the job falls through. The social costs of these behavioral changes may not be large. On the other hand, tenants may work less diligently at their jobs
or take less good care of their health. This behavioral response is a real concern, but the amounts involved are modest and the clawback provisions can be adjusted to offset this behavior.
Fraud may also occur, of course. Landlords, for instance, could invent fictitious tenants. A lot needs to be learned about how to run these programs right.
Who Is This For?
Rental insurance is a complement to existing HUD low-income assistance programs, not a substitute for them. These programs have a variety of goals, and this article cannot comment on all of those goals. Unlike these existing programs, however, rental insurance is meant to be an entitlement, and so should be one way of mitigating the unfairness of denying assistance to some households while giving large amounts of assistance to other households that are in no obvious way more deserving.
Because rental insurance is an entitlement, HUD-assisted tenants should be able to purchase it.
This new activity would make HUD’s current programs work a little better. Currently, the risk of tenant nonpayment is borne by public housing authorities and a multitude of investors and landlords. Many of these are unsophisticated, risk averse, and geographically undiversified; many of them also worry about cash constraints at times of recession. Rental insurance for HUD-assisted tenants shifts some of this risk to the national level, where it can be handled much better. A more efficient allocation of risk should reduce the per-household cost of these programs (including any losses from rental insurance), and allow them to serve a few more households.
Should rich people be eligible for rental insurance, too? Probably they should be: healthy people are eligible to buy medical insurance. Whether tenants need the help depends on their current circumstances, not their past circumstances, and eligibility rules have to be based on past circumstances. But if federally sponsored insurance is losing money and rich people are heavily involved, the subsidy should be reduced so that private companies can enter the market. The danger in operating a program without explicit income guidelines is that it might be relatively more attractive to rich people than poor. If this were to happen, the program parameters could be adjusted in cost-neutral ways to make it more attractive to poor people and relatively less attractive to rich— for instance, by reducing upfront premiums and raising clawback percentages.
Of course, rental insurance is not ready for wide-scale adoption yet. To learn about how it works, starting with specific groups defined by hard-to-change characteristics would be informative and probably productive. For instance, young adults aging out of foster care and returning veterans are two groups that often have trouble finding and keeping apartments. Rental insurance could open doors for them with landlords who might otherwise fear rent delinquencies, and provide social agencies with early warning signs of trouble.
Summary Designing a practical rental insurance program is tough. Many parameters and details must be chosen carefully and consistently. For many questions, experience will have to be the only guide.
Some relevant experience is available to draw on: Insurent, the VA’s mortgage insurance program, and the emergency payment systems that some social service agencies run all have some similarities to rental insurance. But real experience can be acquired only through real experiments.
132 Rental Housing Policy in the United States Rental Housing Assistance for the 21st Century Better Neighbors The basic low-income housing subsidy programs were designed many years ago to address the problems that bothered people then. Structural conditions seem to have been the major concern.
If people lived in homes that were too small, or too drafty, or without enough sunlight, or without proper water and sewer connections, they and their children would not be healthy, physically or morally. Their ill health would spread by contagion and crime. So the entire society would gain by improving the structural conditions of the housing of the poor. Thus, public housing in the 1930s and even Housing Choice Vouchers (HCVs) in the 1970s contained strict rules about structural characteristics but almost no rules about location.
Times have changed. Structural quality of unsubsidized housing stock has improved tremendously, especially at the bottom of the distribution. The rest of the nation has become more concerned about how well low-income children read, and less about whether they have tuberculosis. The general public worries about diabetes and fast food outlets, not polio and poor ventilation. Crime remains a great concern, but the causes are seen less as inadequate sunlight and more as inadequate role models.
The externalities, in other words, now come more from location and less from structure. A complete study of subsidized housing location externalities in this sense has never been done, and may be impossible to do. However, something is known about some of the relevant parameters.
This section looks at how housing location affects what kind of citizens residents are or become— the traditional focus of housing policy (especially ownership policy). Health, jobs, and education will be the focus in this section. The third section briefly considers two other possible externalities of subsidized housing: how it affects the value of neighboring properties, and how tenant selection affects the cost of other public programs. The final section is about crime, which brings all of these issues together.
This article concentrates on very specific, measurable externalities. Broad questions like whether subsidized housing should be in wealthy neighborhoods or minority neighborhoods will not be addressed. An enormous amount has been written on these questions, yet no consensus on the answers exists. Consensus exists, however, on more direct issues like carbon monoxide, particulate matter, bad schools, and long commutes; and we can put rough numbers on these costs. This article will concentrate on these areas because the science is pretty well known. It is easier to think about trees than to think about forests.
Health Basic Empirical Results Recent empirical work in health economics establishes a few fairly strong links from location to health. Currie, Neidell, and Schmieder (2009); Currie et al. (forthcoming a); and Currie and Walker (forthcoming) found that higher carbon monoxide levels induce greater school absenteeism and poor infant health outcomes, even at low levels. Living close to traffic congestion is bad for kids. There is also substantial evidence that particulate matter is harmful. (These dangers seem to
be quite localized, with impacts measured in hundreds of feet; census tracts are too coarse.) Currie et al. (forthcoming b) found that children with more access to fast-food restaurants were more obese.
Exogenous Nuisances First, suppose that the location and intensity of noxious sites are fixed and exogenous. Then the role of housing policy is to discourage people from living near them. (The next section examines the more realistic case of endogenous nuisances.) In a perfectly functioning private housing market, fixed noxious sites would be no problem. Housing consumers, realizing that living near these sites was harmful, would lower their bid-rent by the full amount of the damage. No houses would be built near the noxious sites if the lowered bid-rent were not enough to pay for structure and to bid the land away from nonresidential uses. If houses were built, the residents would be compensated for the harm by lower rents, and the landowner would bear the damage cost.
Subsidized housing as it is now run short-circuits this adjustment. To a large extent, the rent that a landlord receives is independent of the attractiveness of the location.2 This holds whether the landlord is a public housing authority (PHA) or a private landlord with HCV tenants. Everything else being equal, the absence of location in the subsidy formula is an incentive to place subsidized tenants in the worst possible locations. For supply-side projects, this is because land is cheapest (hence forgone property taxes are the least) in those locations. For HCV tenants, this is because the opportunity cost of renting to a subsidized tenant is least there. Thus, we expect current assistance programs to do worse than the market would in promoting tenant health, not better.
But even doing as well as the market does is not good enough. Tenants do not pay the full cost of poor health, and parents do not fully internalize the health of their children (especially in the long run). Tenants may also not be aware of the best current research. The argument for subsidized housing can only be that it does a better job than the market.
(That is why the many exceptions to the picture of complete independence of rent from location amenities do not change the basic picture. Landlords of HCV tenants with rents above Fair Market Rent [FMR] receive the full value of any change in bid rent at the margin, and so will locate away from nuisances, at least those that are smaller than the difference between actual rent and FMR.
Developers planning developments that include both market-rate units and subsidized units like HOPE VI will bear the cost of nuisances to the extent that part of the project is unsubsidized at the margin, but they will bear only a fraction of the true cost.) Can current housing assistance programs be revamped so that they do the job they should be doing?