Scapegoating Rent Control: Masking the Causes of Homelessness

Richard P. Applebaum, Michael Dolny, Peter Dreier and John I. Gilderbloom

[Reprinted from: Richard P Appelbaum, Michael Dolny, Peter Dreier, and John I. Gilderbloom, "Scapegoating Rent Control: Masking the Causes of Homelessness," Journal of the American Planning Association, vol. 57, no. 2 (Spring 1991). Copyright © 1991 by The American Planning Association. Notes and references omitted.]

The U.S. Congress has recently considered legislation that would withhold federal housing funds from the numerous locales that have adopted rent control. Such legislation is supported by HUD [Department of Housing and Urban Development] Secretary Jack Kemp, who strongly believes that rent control is partly responsible for discouraging badly needed investment in rental housing. Sixteen states currently have laws that restrict the ability of localities to enact rent control, while another 29 have been targeted for such laws by the National Apartment Owners' Association and the National Multi-Housing Council. While the belief that rent control has adverse consequences for housing markets has long been advanced by housing economists, a new claim has recently emerged in support of anti-rent control legislation: the assertion that rent control should be dismantled because it is the chief underlying cause of homelessness. The evidence for this claim can be traced to a single study by journalist William Tucker (1987a, 1987b, 1989a, 1989b). Since homelessness is such a visible national issue, local rent regulations -affecting millions of tenants nationwide-are more vulnerable to federal anti-rent-control legislation than at any time in the recent past. ...

Despite the widespread attention it has received. Tucker's research is seriously flawed. The link between rent control and homelessness it purports to demonstrate does not withstand serious scrutiny. Given the political context in which the research appears, the following critique of Tucker's thesis is doubly important. Unchallenged, Tucker's work represents a serious threat to local rent control by linking it with a national problem of high concern. Pointing the finger at rent control can only divert attention from a serious effort to uncover and address the actual causes of homelessness.

The growth of homelessness during the 1980s is not linked with the efforts by a handful of local governments to regulate skyrocketing rents. Homelessness is directly related to the overall level of poverty, to the availability of affordable housing, and to the accessibility of support services for people suffering from mental illness or alcoholism. It is no accident that the number of homeless Americans increased dramatically during the 1980s. The past decade has witnessed growing poverty, especially among the "working poor"; a decline in low-rent housing, including sharp cuts in federal low-income housing assistance; and a failure to adequately serve the deinstitutionalized mentally ill. As a result, since the early 1980s the homeless population has increased between 20 and 25 percent a year, according to the U.S. Conference of Mayors annual surveys (1989, 2). Moreover, the profile of the homeless population includes a growing number of families with young children, as well as individuals with jobs (US. Conference of Mayors 1989).

This assessment of the underlying causes of America's homeless problem would seem to suggest fairly straightforward remedies directed at increasing the wages of America's working-poor, expanding the supply of affordable housing, and providing residential and social support programs for the nation's mentally ill. A comprehensive examination of the evidence gives no support to the claim that rent control is the root cause of homelessness in the United States. ...


Two hundred cities and counties currently have some form of rent regulation. This group includes over 100 communities in New Jersey, as well as cities and counties in Massachusetts, New York, Virginia, Maryland, Alaska, Connecticut, and California. Most of these ordinances were first enacted in the early 1970s. Approximately 10 percent of the nation's rental housing stock is estimated to be covered by some form of rent control (Baar 1983). Current rent control measures can be categorized as moderate, in comparison with the more restrictive rent control that was in effect in New York City during the immediate postwar period.

Moderate rent controls permit rent increases sufficient for the landlord to maintain an adequate return on investment, while protecting tenants against rent gouging. All ordinances currently in effect are moderate in nature. Such controls typically peg annual rent increases to increases in the landlords' costs, and exempt newly constructed rental units from controls altogether. They also often require adequate maintenance as a condition for annual rent adjustments: tenants in buildings that are inadequately maintained can appeal their rent increases. Some rent control laws permit vacated units to be temporarily decontrolled so that rents can be raised to market levels for incoming tenants, after which they are recontrolled. Moderate rent controls thus contain a number of provisions explicitly designed to encourage both construction of new rental housing and maintenance of existing units.

In a few highly inflationary California housing markets, some controls are coupled with an additional provision: they exclude increased mortgage costs from the formulas relating landlords' costs and allowable rent increases. This provision is designed to discourage speculation in rental housing. Under such an exclusion, a landlord who has incurred increased capital costs (either through recent purchase or through refinancing to obtain equity capital) cannot pass the higher financing costs through to tenants in the form of rent increases.

In sum, current rent controls contain provisions that are intended to guarantee the landlord a fair and reasonable rate of return on investment, while protecting the interests of tenants by preserving affordable housing. Maintenance is strongly encouraged; newly built units are not controlled at all.

Nonetheless, critics continue to argue that rent control discourages investment in rental housing. According to Tucker (1987a, 1987b, 1989a), for example, localities that enact rent control rob landlords of their rightful returns. So deprived, landlords cut costs. Maintenance suffers; buildings are abandoned. Badly needed new units are never constructed. Although rents may be lowered in the short run, the argument goes, housing scarcity eventually results. Scarcity, in turn, causes homelessness. In posh areas like Santa Monica, Cambridge, or the Upper West Side of Manhattan, yuppies squeeze out low income tenants in the fight for scarce apartments. In blighted areas like the South Bronx, buildings are abandoned, and eventually razed by arsonists or government bulldozers. Either way, says Tucker, the poor are relegated to the streets and shelters.

This analysis is not original to Tucker; on the contrary, it is shared by a number of housing economists as well as many people in the real estate community. For example, ten years ago a national survey of economists found virtually unanimous agreement that "a ceiling on rents reduces the quantity and quality of housing available" (Kearl et al. 1979). These conclusions are not based on empirical studies, but on theoretical assumptions about how housing markets are supposed to operate. The real estate lobby has been highly effective in communicating this analysis to its members and the media. Major news organizations, including the Wall Street Journal and Forbes magazine, have editorialized against rent controls (Gilderblooml983).

Numerous empirical studies have been conducted on the effects of moderate rent control on rental housing investment; none support the views just described. A comprehensive review... finds that such controls have not caused a decline in construction, capital improvements, maintenance, abandonment, or demolition of controlled units relative to noncontrolled ones. This is because of the nonrestrictive nature of moderate controls, which, as we have seen, guarantee landlords a fair and reasonable rate of return. Rent controls eliminate extreme rent increases, particularly in highly inflationary markets, but they do not eliminate the profits necessary to encourage investment in private rental housing (Gilderbloom 1984, 1986; Heffley and Santerre 1985; Mollenkopf and Pynoos 1973; Daugh-erbaugh 1975; Vitaliano 1983). In particular, the vacancy decontrol-recontrol provision in some localities results in significantly higher average rents than those that would occur in the absence of such a provision (Gilderbloom and Keating 1982; Hartman 1984; Clark and Heskin 1982; Rydell 1981; Los Angeles Rent Stabilization Division 1985). While moderate rent control is successful in eliminating exorbitant rent increases, its impact on redistributing income from landlords to tenants clearly depends on the degree to which market conditions would otherwise have led to rent increases that greatly exceed the allowable, rent levels.


Tucker's study is the first to look at the impact of rent control on homelessness. In order to support his argument that rent control produces homelessness by discouraging investment and thereby creating housing scarcity, Tucker sought to show that cities with rent control had lower vacancy rates and greater homelessness than cities without rent control.

For his primary data set, Tucker relied on the single comparative study of homelessness that had been done at the time of his study - the HUD survey of homelessness in 60 metropolitan areas (1984). HUD had conducted a random sample of 20 cities in each of three size strata (50,000- 250,000; 250,000-1,000,000; and over, 1,000,000). For each city, HUD telephoned people they labeled "knowledgeable informants" and asked for their estimates of the homeless street population in their areas. (Shelter estimates were more accurately obtained from information provided by shelter operators.) The various estimates for each locale were then combined into an average figure that was weighted to reflect the presumed reliability of the different sources. Tucker took the HUD estimates for the 40 metropolitan areas in the two largest strata. He then computed a homeless rate for each city by dividing HUD's estimate of the total number of homeless by the population of the core city for each metropolitan area.

Tucker did not rely exclusively on HUD's random sample of places; rather, he modified the HUD sample in several ways. First, he dropped six cities from among HUD's 40 metropolitan areas over 250,000 in population: Dayton, Davenport, Colorado Springs, Scranton, Raleigh, and Baton Rouge. These six places were reportedly eliminated because of "the great difficulty in determining local vacancy rates" (Tucker 1989a, 5, n. 4). For unexplained reasons, Tucker then added to his list one of HUD's smallest (under 250,000) metropolitan areas - Lincoln, Nebraska. He also mistakenly classified Hartford as a city with rent control. Finally, he added 15 additional cities "to include some notable HUD omissions" (1987a, 1); he does not explain how these cities were selected out of thousands of possible places across the United States. Since these cities were not a part of HUD's original study, Tucker developed his own homeless estimates by making telephone calls to unspecified informants in each city. This misguided sampling methodology yielded a final list of 50 places for his analysis.

Once he had obtained his list of places, Tucker identified factors that might be important determinants of homelessness. He originally chose rates of poverty, unemployment, public housing availability, and rental housing vacancy; total population; mean annual temperature; and the presence (or absence) of rent control. Two additional variables - population growth rate and mean annual rainfall - are employed in a recent study (1989a) but apparently not in the original studies (1987a, 1987b); nonetheless, the appendix in the recent study reports only the seven original variables. High rates of poverty and unemployment are indicative of an economically marginal population, and therefore should be associated with greater homelessness. Public housing availability, on the other hand, provides one form of protection against homelessness, and so should be associated with lower rates. Low vacancy rates indicate scarcity in the private rental housing market, and - according to Tucker - should be associated with both rent control and homelessness. Larger, faster-growing places may well attract the unemployed with the lure of jobs, thereby contributing to homelessness in such cities. Finally, locales with warm temperatures and low rainfalls have an obvious appeal to the homeless.

Having selected these key variables, Tucker employed them in two-and three-variable regression equations predicting homelessness. While his results vary somewhat among his various reports, he generally found that the only variables that made any substantial difference in the rate of homelessness were the local vacancy rate and rent control - and that the latter statistically accounts for much of the impact of the former. In fact, Tucker found that rent control by itself explains fully 27 percent of the difference in homelessness among cities; when combined with mean temperature, it accounts for 31 percent. According to these findings, homeless people are attracted to cities with hospitable climates; when such places have rent control, increased housing scarcity is assumed to result, and - with it - greater homelessness.

In evaluating Tucker's findings, it is important to bear in mind that he classified only 9 of the 50 cities as having any form of rent control at all. Since all of the cities had homeless problems to varying degrees, it is obvious that rent control cannot be the principal cause of homelessness, as Tucker contends. Miami, with the highest rate of homelessness in the cities under study, does not currently have rent control. Nor does St. Louis, which ranks second. Nor does Worcester, which ranks fourth. The fact that three out of four places with the most severe homeless problems lack rent control would seem to provide a prima facie case for rejecting Tucker's claim out of hand.

The first major difficulty in Tucker's study lies with his use of HUD's measure of homelessness (1984) as his key variable. According to two congressional hearings that examined HUD's methods in detail, that measure was highly unreliable. HUD relied on what it called "knowledgeable informants" - police departments, social service agencies, shelter staffs - who simply guessed at the numbers of homeless people in the 60 areas HUD reviewed. There was no actual count of the number of homeless in the streets, park benches, abandoned cars, and elsewhere - and certainly no estimate of the "invisible" homeless temporarily living in overcrowded apartments with friends or relatives. Although the guesses were mainly for downtown neighborhoods, HUD acted as if they applied to much larger metropolitan areas - Rand McNally marketing areas (RMAs), areas with four or five times as many people. This method, not surprisingly, produced very low rates of homelessness for the metropolitan areas HUD studied, since they guaranteed that homeless people outside the downtown areas would be excluded from the study. Tucker's principal variable, therefore, substantially undercounts the homeless.

The second major problem results from the questionable procedures by which Tucker arrived at his 50 cities, which - as will be demonstrated in the next section of this article - skew his results towards his foregone conclusions. As noted above, he began with HUD's random sample of 40 medium and large metropolitan areas, added one smaller HUD metropolitan area, selectively eliminated six places, and then added 15 others of his own choosing. Since only five of HUD's cities were among the more than 200 places with rent control, Tucker made certain that three rent-controlled cities were included among those he added. But sampling problems are compounded by the fact that the three rent-controlled cities he added are already presumably included in HUD's homeless estimates: Newark and Yonkers are part of the New York City metropolitan area, while Santa Monica is part of Los Angeles.

Tucker's third major error is his failure to consider the possibility that high rents might themselves be a chief cause of homelessness, while at the same time causing tenants to demand rent control. In other words, his reported correlation between rent control and homelessness might be an artifact of the association of both with high rents. He nowhere looks at the possible causal effect of rent on homelessness. ...


The United States now faces the worst housing crisis since the Great Depression. The underlying problem is a widening gap between what Americans can afford to pay and what it costs to build and operate housing. In this situation, the poor are the most vulnerable to joining the ranks of those without a home.

The number of poor Americans, now about 33 million people, is growing, and the poor are getting poorer (Center on Budget and Policy Priorities 1988, 1; Children's Defense Fund 1989, 16-26, 100-106, 115; U.S. Joint Economic Committee of Congress 1988, ch. VII). The largest increase is among the "working poor" - people who earn their poverty on the job because of low wages. Among the "welfare poor" - primarily single mothers and their children - Aid to Families with Dependent Children (AFDC) and other benefits have declined far below the poverty level. These are people who are only one rent increase, hospital stay, or layoff from becoming homeless. In fact, a recent report by the U.S. Conference of Mayors (1989, 2) found that almost one-quarter of the homeless work, but simply have wages too low to afford permanent housing.

The plight of the poor is worsened by the steadily rising housing costs that have plagued the economy throughout the past decade (see U.S. Comptroller General 1979 for an early announcement of the housing crisis). On one hand, rising homeownership costs have forced many would-be first-time buyers into the status of reluctant long-term renters, greatly increasing pressures on the rental housing market. Homeownership rates have been declining steadily since 1980, particularly among first-time homebuyers. Among households where the head was under 25, for example, ownership has declined from 23.4 percent to 15.1 percent of all households, a drop of 36 percent; for those headed by someone aged 25 to 34, the decline has been from 51.4 percent to 45.1 percent, or 12 percent (Apgar 1988, 24). In 1973, it took 23 percent of the median income of a young family with children to carry a new mortgage on an average-priced house. Today, it takes over half of a young family's income (Children's Defense Fund 1988,57).

On the other hand, renters confront chronic production shortages and rising rents. Between 1970 and 1983 rents tripled, while renters' income only doubled. As a result the average rent-income ratio grew from roughly one-quarter to one-third; the proportion of tenants paying 25 percent or more income into rent increased from one-third to one-half. By 1985, close to one out of every four renters paid over half of their income for housing costs. Eleven million families now pay over one-third of income on rent; 5 million pay over half.

The problem is especially acute for the poor, who are now competing with the middle class for scarce apartments. It is estimated that by 1985 there was a national shortage of some 3.3 million affordable units for households earning under $5,000 - an increase of more than 80 percent since 1978 (Leonard et al. 1989, 9). Among the nation's nearly 7 million poor renter households, 45 percent spent more than 70 percent of their income on housing in 1985; 65 percent paid more than half; while 85 percent - some 5.8 million households - paid more than the 30 percent officially regarded as "affordable" under current federal standards. The median tenant household paid almost two-thirds of its income on rent (Leonard et al. 1989,1-2). The typical young single parent pays 81 percent of her meager income just to keep a roof over her children's heads (Children's Defense Fund 1988,59).

Despite the severity of these problems, less than one-third of poor households receive any kind of housing subsidy (Leonard et al. 1989, 27; U.S. Congressional Budget Office 1988,3). This housing subsidy level is the lowest of any industrial nation in the world. Some 6 to 7 million low-income renter families receive no housing assistance whatsoever, and are therefore completely at the mercy of housing markets that place them immediately at risk of being homeless. And, while the number of poor families has risen during the 1980s, the number of low-rent private apartments has plummeted as a result of rising rents, urban redevelopment activities, condo conversions, and arson. Between 1974 and 1985, the number of privately owned, unsubsidized apartments renting for less than $300 (measured in 1988 dollars) fell by one-third, a loss of nearly 3 million units (Apgar et al. 1989,4). The swelling waiting lists of even the most deteriorated subsidized housing projects are telling evidence of the desperation of the poor looking for affordable homes.

The already existing shortages of affordable private housing were worsened considerably by the short-sighted actions of the Reagan administration. The 1986 Tax Reform Act, for example, removed many of the tax benefits that previously made it profitable for the private sector to rent housing to poorer families. It is estimated that the loss of tax shelters for housing will eventually reduce the value of income property by 20 percent, forcing compensating rent increases of 25 percent by the early 1990s. The National Association of Home Builders predicted that rental housing construction would decline by half as a direct result (Furlong 1986,16); an MIT market simulation predicted an eventual loss of 1.4 million units (Apgar etal. 1985,1).

The Reagan administration's budget cutbacks virtually eviscerated publicly owned and subsidized housing, all but eliminating the already small federal commitment to providing housing for the poor. Not only were safety net programs cut in general, but housing was selected to bear the brunt of budgetary retrenchment. Between 1981 and 1989 federal expenditures for subsidized housing declined by four-fifths, from $32 billion to $6 billion. Total federal housing starts declined from 183,000 in 1980 to 20,000 in 1989 (Low Income Housing Information Service 1989). The administration even proposed to sell off 100,000 units of public housing, an effort that was stymied largely because public housing tenants were too poor to afford their units. A number of specific programs, including several directed at the needs of the homeless, were "zeroed out" in the 1989 budget. It should be pointed out that, as severe as these measures may appear, President Reagan's proposed cuts were still deeper: philosophically committed to ending federal involvement in housing altogether, he was prevented from doing so only by the lobbying efforts of low-income housing advocates before a Democrat-controlled Congress. A single statistic tells the story in unambiguous terms. When President Reagan came to office in 1981, the federal government spent seven dollars on defense for every dollar on housing. When he left office in 1989, the ratio of dollars spent was 46 to one.

In sum, declining incomes at the bottom have converged with rising housing costs to produce a potentially explosive situation, which unwise short-term federal policies have served to worsen. Rent control plays no role in this unfolding tragedy. According to one estimate (Clay 1987, i), by 2003 "the gap between the total low-rent housing supply (subsidized and unsubsidized) and households needing such housing is expected to grow to 7.8 million units," representing an affordable housing loss for nearly 19 million people. This figure represents the probable constituency of the homeless, as the United States moves into the twenty-first century.

On its own, rent control can't solve the housing crisis. It is merely one tool available to local governments for confronting skyrocketing rents and a shortage of affordable housing. Tucker's study does not demonstrate what it sets out to do, and so cannot be used to justify a scapegoating of rent control for the mounting tragedy of homelessness.