Robert Frost once described home as "the place where, when you have to go there, /They have to take you in." This is precisely what the homeless lack: not only a home of their own but also a claim on anyone else -- kin, community, or government -- to take them in and give them a home. Chronic illness, physical as well as mental, commonly accompanies homelessness and may have contributed to it in the first place. Often, however, it is only when the homeless become acutely ill that they are able to make a large claim on resources from a system that must take them in. And of course that claim falls immediately on hospitals and ultimately on the public purse, or as political economists sometimes aptly refer to it, the "public household."
Elsewhere in this issue of the Journal, Salit et al. report that hospital stays for the homeless in New York City cost far more than stays for other patients and suggest that this added burden be taken into account when policies for housing and supportive services are developed. The phenomenon the authors identify follows a familiar pattern. Failure to deal with a social problem "upstream" (lack of housing, education, health insurance, substance-abuse prevention) leads to added costs for resources "downstream" (police, prisons, hospital care). The downstream institutions are not only expensive, but also poorly equipped to deal with the underlying social problems. Many people conclude, therefore, that preemptively attacking the problems upstream would be both more efficient and more effective, but the pattern stubbornly persists. In the case at hand, we continue paying to put the homeless in hospital beds while not providing them with ordinary beds of their own.
This result seems irrational, even tragic, but the failure to do more earlier is not due simply to short-sightedness. In tracing the genesis of any complex social problem, there are usually several factors to consider, and each one presents its own difficulties. The causes of homelessness involve changes in labor and housing markets, mental health policy, and social behavior. Since the 1970s, housing costs have risen more rapidly than the incomes of the least skilled workers, and although the number of subsidized housing units has increased, it falls far short of demand. The deinstitutionalization of the mentally ill (without adequate provision of community care) and the use of addictive drugs, notably "crack" cocaine, have also contributed to the increase in the homeless population. The decline in marriage rates has exposed more poor people to the risk of homelessness when they are unable to meet the costs of housing individually. (3)
None of these sources of homelessness has a cheap and obvious solution. Upstream interventions might be more affordable if they could be directed exclusively to people who end up in trouble downstream, but often the number of people at risk is much larger than the number who are affected. (This is where the metaphor breaks down; streams are usually narrower the further up one goes.) For example, the number of people who are eligible for subsidized housing because their incomes are low far exceeds the number who actually become homeless. Many of the poor live doubled up with relatives and roommates, although they are only an argument or an acute illness away from living on the streets. Subsidized housing for all these people would sharply reduce homelessness, but it would be very expensive.
It seems hard to believe, however, that the problem is intractable. After all, there was a time within living memory when homelessness hardly existed in the United States. As recently as the 1960s, the homeless population was extremely small, although we do not know exactly what its size was. But by 1987 about 500,000 to 600,000 Americans were homeless on any given night, according to a widely cited estimate (4) (which one critic revised downward to 350,000). (3) The National Law Center on Homelessness and Poverty estimates that there has been a 5 percent annual increase since 1987, but there is no consensus on the current figure. Most of the homeless are without homes for only short periods, and thus a far larger number of people experience a spell of homelessness over time. From 1988 to 1992, about 3 percent of the population of New York City used public shelters at some time. (5)
Although homelessness has increased greatly in recent decades, overall housing conditions in the United States have not declined. On the contrary, by any of the usual measures, such as the number of square feet per residence, housing has substantially improved. Higher housing standards may, in fact, have indirectly contributed to homelessness. The low-grade, crowded dwellings that once provided homes to the very poor can no longer be built under current housing codes. Many communities do not want cheap housing in their backyards, and local zoning regulations often restrict the supply of cheap rental units by preventing houses from being subdivided. Given the current standards for living space and amenities, public subsidies support relatively fewer low-income housing units than they would under the standards that prevailed in the past.
High-minded, low-budget public policy thus seems to have had a role in creating homelessness. Should we be surprised that more people are unable to find an affordable place to live after governmental policies raise housing standards and restrict the private supply of cheap apartments while real earnings are stagnant and housing subsidies reach only a minority of the poor? Or that more people in psychological distress are on the streets after states have emptied mental hospitals without funding alternative community services? In regard to both housing and mental health, the comfortable have not been willing to pay the costs of policy, and the homeless have ended up paying the price.
If people with very low incomes are to afford homes, either their incomes must be higher or housing must be cheaper. Increases in the minimum wage and higher wage subsidies, such as earned-income tax credits, help raise more of the poor to self-sufficiency. Increased subsidies for low-income housing or reduced regulation can make housing cheaper or cheap housing more available. The remedy does not have to be a single magic bullet. And the government itself need not be in the business of building or providing the homeless a place to live; private nonprofit organizations have been some of the most effective suppliers of low-income housing. But governmental action to raise incomes or reduce costs is necessary to narrow the housing-affordability gap. Unfortunately, some policies that have been adopted recently, notably time limits on welfare benefits, may contribute to future increases in homelessness even if only a minority of the former recipients of public aid are unable to find work.
Failing the necessary upstream policies, the health care system will be left with the extra burden of the homeless. No one approves of this result, but whether the public and its representatives disapprove strongly enough to prevent it is another matter. Health care professionals can perform a great service by helping to mobilize that disapproval to put the public household in better order.
Paul Starr, Ph.D.
Princeton, NJ 08544-1010