In April 1986, Joyce Brown, a former New Jersey secretary (who also calls herself Billie Boggs), had a fight with her sisters with whom she was then living—and hopped on a bus to New York City. Something happened, and she ended up living near the hot-air vent of Swensen’s restaurant at Second Avenue and 65th Street. She stayed there for a year, during which time her hair became tangled and matted; she insulted passers-by (especially black men, at whom she hurled racial epithets, although she herself is black), she burned the money she was given by sympathetic observers, and she relieved herself on the streets.
Eventually she came to the attention of Project Help, a mobile psychiatric unit that monitors mentally-ill homeless people in lower Manhattan. Until recently, Project Help would not involuntarily transport a “street person” to a hospital unless he was an immediate danger to himself or others, and it interpreted that criterion strictly. When the Koch administration decided to apply a less strict interpretation—on the ground that living on the street is dangerous for the mentally ill even if not immediately so—Joyce Brown became the first person removed from the streets to Bellevue Hospital for psychiatric evaluation. She was diagnosed as a chronic schizophrenic, and the city held her in the hospital for twelve weeks, during which time her doctors attempted to obtain permission to have her medicated. But with the aid of Robert Levey, a lawyer from the New York Civil Liberties Union (NYCLU), Miss Brown not only avoided medication but successfully litigated for her release.
At this point, under the guidance of Levey and NYCLU president Norman Siegel, Miss Brown declared that she had been “appointed the homeless spokesperson.” After shopping trips to Saks Fifth Avenue, Lord & Taylor, and Bloomingdale’s, and dinner at Windows on the World, Miss Brown and her lawyers hit the lecture circuit. She spoke at New York University Law School and at the Cardozo Law School, she was interviewed on 60 Minutes and Donahue, and she received half-a-dozen book and film proposals. Then, on February 18, 1988, Miss Brown, Levey, and Siegel all participated in the Harvard Law School Forum on “The Homeless Crisis: A Street View.”
Levey (describing himself as a kind of “warm-up act . . at a rock concert”) spoke first. He wanted, he said, only to raise some questions, of which the key one was why our society had decided to make Joyce Brown into a celebrity. Was it because we wanted to sweep the problem under the rug by focusing on the fate of a single individual who had successfully challenged the city and gotten off the streets? This seemed an odd question and a still odder answer since it was precisely Levey and his colleagues at the NYCLU who had made Joyce Brown a celebrity; and the last thing in the world they had in mind was to sweep the problem of homelessness under the rug.
At length, Miss Brown herself spoke. Her speech was slurred, and she dropped a few lines from her prepared statement, but she certainly gave a creditable performance. The first part of her talk sounded very much as if it had been stitched together out of slogans made familiar by advocates for the homeless: homelessness is caused by policies that help the rich and not the poor; it will be solved only by building low-income public housing; etc. Of her stay in the hospital she said, “I was a political prisoner.”
Much more interesting was what Miss Brown called “my street view” of homeless life. She said nothing about her history either of heroin and cocaine abuse or of mental illness. According to her, she had had only two problems in being homeless. The first was police harassment. She claimed that at some point she had been beaten with night sticks and kicked by several police officers. Project Help also had degraded and humiliated her and had denied her what she called “my right to live on the street.” Never, she said, had Project Help been of any use, except to offer her a sandwich.
Her second problem was that, obviously, she had been unable to find an apartment, which was why she had ended up on the streets. With no place of her own and since there are no public toilets, she had to use the streets as a bathroom. She did not explain why she did not return to live with her sisters in New Jersey, who had been looking for her during the year she was living on the streets. Moreover, if she had been eager all along to come indoors, what was the point of fighting tooth and nail for the right to live on the streets?
Joyce Brown’s claim that she had been beaten by the police also presented difficulties. She provided no details of the event. When she was asked during the question period if she intended to press charges against the police for their abuse, her response was, “Everyone knows New York City cops are killers.”
Another difficulty was her complaint of being degraded by Project Help. It is true that for most of the time she was on the streets, outreach workers only offered her sandwiches. The reason for this was that Miss Brown spurned all other help. Project Help workers kept regular tabs on her, they coaxed her to accept further services, such as transportation to washroom facilities and a women’s shelter. Miss Brown refused every time.
In other words, until the city decided to bring her inside against her will, Project Help was following exactly the policy that her lawyers told the Harvard audience ought to be followed in these cases: the city kept an eye on her, offered whatever services she would accept, and tried to win her confidence. It was only after months of such attempts that this approach was abandoned.
Joyce Brown ended her talk, received applause, and sat down. Norman Siegel then rose to speak. He too claimed that homelessness was essentially a housing problem, or rather that it was an issue of economic justice and equality and in no sense a mental-health issue, still less a matter of public order or law enforcement. The main cause of homelessness was the construction of high-rise developments for the rich and the destruction of the single-room-occupancy hotels (SROs) that many poor people once lived in. Allowing landlords to warehouse apartments, failing to require developers to build low-income housing, and the unwillingness of the city to redevelop the abandoned apartments it had seized—these were the real problems.
As Siegel saw it, the only way to deal with these problems was by radical political action. Just as law students from Harvard had gone to the South during the 1960’s to organize and register blacks as part of a progressive political movement, so in the late 1980’s lawyers should go to the bus terminals and park benches of America’s cities and organize for economic justice. Homeless people needed “guerrilla legal tactics” in order to win, though “non-legal solutions,” pressed forward “in a harassing way,” were also important.
Whatever else one may say of the approach that Siegel and Levey put forward at the Harvard Law School Forum, it is very close to the position taken by most advocates and researchers on the homeless. Writers like Jonathan Kozol in his new book, Rachel and Her Children: Homeless Families in America;1 activists like Mitch Snyder of Communities for Creative Non-Violence (CCNV) and Robert Hayes of Coalition for the Homeless; and research centers like New York’s Community Service Society all agree with Joyce Brown’s lawyers in their general analysis of the problem.
The thrust of this analysis is as follows:
- Homelessness is a huge problem and it is getting worse. The size of the homeless population is at least two to three million (a figure originally advanced by CCNV in 1982), perhaps as large as four million, and growing.
- Homelessness is simply or primarily a housing issue. As Kozol puts it, “The cause of homelessness is lack of housing” due to federal cutbacks and urban redevelopment.
- Mental illness and other disabilities, such as alcoholism, while frequent among the homeless, have been greatly exaggerated. About one-third of the homeless are members of homeless families, who are neither mentally ill nor otherwise disabled.
- Radical tactics and objectives are necessary if the problem is ever to be solved. Homelessness is a systemic problem, caused by the structure of the economy and society in general. As such it cannot be effectively addressed either by the charity of the welfare state or by benefits conferred at the whim of legislatures. What is needed is the enactment of a constitutional “right to shelter” that would be enforceable through the courts.
Let us take up these claims one by one, especially as they relate to the Joyce Brown case.
First, as to the size of the homeless population. The CCNV estimate of two to three million as of 1982 was based on an unsystematic telephone survey of shelter providers and advocates. It was never clear just how CCNV went from these local to its national estimates, or even exactly how CCNV defined homelessness.
To try to get things straight, the Department of Housing and Urban Development (HUD) conducted its own survey, released in 1984, which concluded that, as of 1983, there were about 250,000 to 350,000 homeless people in America.
The HUD report was widely attacked. Mitch Snyder of CCNV declared that the officials who had inspired it reminded him “of nothing so much as a school of piranha, circling, waiting to tear the last ounce of flesh.” Various methodological critiques were also brought against the report, but their validity remained uncertain in the absence of an independent cross-check on its methods.
Such an independent cross-check was finally completed in 1986 by Richard B. Freeman and Brian Hall of Harvard University in their Report of the National Bureau of Economic Research, “Permanent Homelessness in America?” The bottom line of the Freeman-Hall study was that “the much-maligned” HUD figure of 250,000-350,000 was “roughly correct.” Freeman and Hall’s exact estimate for 1983 was 279,000 (and Freeman estimates that as of 1988 the number has jumped to about 400,000). Further, the key data on which they based their numbers were confirmed by surveys of homeless people done by other researchers using several different methods in Boston, Chicago, Nashville, Washington, D.C., Phoenix, Pittsburgh, and Los Angeles. I myself have also confirmed these findings in a survey of homeless people in San Diego.
To cite only the example of Chicago: CCNV had reported one estimate of 250,000; later, Coalition for the Homeless cut this by a factor of ten, to 25,000. Neither of these estimates was based on systematic scientific research, nor was either organization willing to say exactly how they were arrived at. Then after two surveys of his own, Peter Rossi of the University of Massachusetts released his conclusions: in the winter of 1986 there were about 2,020 homeless people in Chicago (give or take about 275).
These studies, all independent of one another, using various methodologies, and all arriving at approximately the same conclusions, reinforce one basic point: the estimate circulated by advocacy groups of between two and three (and even up to four) million homeless people is about ten times too high.
When confronted with this evidence, Jonathan Kozol and the others frequently argue (although without ever withdrawing their own claims) that, in Kozol’s words, “Whether the number is one million or four million or the administration’s estimate of less than a million, there are too many homeless people in America.” Or as Chester Hartman, a housing analyst at the Institute for Policy Studies, testified before Congress during its hearings on the HUD report, “. . . the real issue is that in a society with the wealth of the United States, there should not be a single involuntary homeless person.”
Now, it is undoubtedly true that homelessness is a tragedy no matter how few or how many people it touches, and it is also true that 400,000 is a large number. But such statements as Kozol’s and Hartman’s will not do. It is as though someone were to claim that the unemployment rate is 60 percent and then, upon being informed that the real rate is closer to 6 percent, were to respond: “No matter whether the rate is 60 percent or 6 percent, too many people are unemployed. The real issue is that in a society this wealthy not a single person should be involuntarily unemployed.”
One final point on the size of the homeless population. Surveys done in Nashville and Boston, and shelter counts in New York, suggest that the growth of the homeless population has leveled off. One should therefore be very skeptical about recent claims by the National Coalition for the Homeless and the U.S. Conference of Mayors that the number of people on the streets has grown by 25 percent over the last year. Homelessness undoubtedly did increase throughout the early 1980’s, but it may by now have reached its peak.
What about the cause of homelessness? Is it, as Kozol says, lack of housing? Before we can evaluate this claim, we have to re-acquaint ourselves with a few basic facts.
First, it is indeed true that New York (to take a city with an especially large homeless population, and which is the subject of Kozol’s recent book) faces serious housing problems. Between 1981 and 1984 the number of apartments renting for under $200 a month dropped from 437,000 to 256,000. As single-room-occupancy hotels have been torn down for development, thousands of dwelling units for the poor have vanished.
Yet the fact is that, as in most other cities, the housing stock in New York is so large (about 1.8 million rental units), and the number of homeless families is so “small” (about 5,000 at any given point and about 12,000 in the course of a year) relative to that stock, that a simple “lack” of housing cannot be the trouble. The real trouble is that the current housing market is prevented from making a rational allocation of such housing as exists. For the housing market could easily meet the demand that a few thousand homeless families impose on it if it were allowed to, and it could do so in a manner consistent with current standards of decency and fairness.
Again, the Freeman and Hall study throws light on the subject. They report that for the U.S. as a whole there was no dramatic decline in the number of “affordable” units (i.e., those renting for under $200 in real terms) during the recent increase in homelessness between 1979 and 1983. In central cities, as the case of New York shows, the number of such units did decline during this period (by about 5.4 percent according to Freeman and Hall), but this decline is in itself too small to have caused homelessness. Nor is it correct that the number of public-housing units declined during this same period. In fact, public housing units actually increased from 1,178,000 in 1979 to 1,250,000 in 1983.
What did happen, however, was a sharp rise in the number of people looking for such units. Between 1979 and 1983 the number of poor unattached individuals increased by about 21 percent and the number of poor families by 45 percent.
Yet this increase need not, in itself, have led to homelessness. In an open market, landlords and perhaps some developers would have responded to the new demand by providing more cheap housing. Through some combination of dividing up old units, renovating abandoned buildings, renting out space formerly used as garages and basements and the like, these newly-poor renters could have had their demand met. Admittedly such accommodations would have been of inferior quality, but they would have prevented homelessness.
This demand was never met because housing regulations in New York (and some other cities) made it difficult for the market to adjust. For example, New York offered a bonus of $6,000 to landlords who would put up homeless families. There were few takers, even when the bonus was raised to $9,700. One reason seems to be that participating landlords would have had to spend more than the bonus to bring their buildings up to the required standard.
That homelessness is not due to a lack of housing is also shown by the fact that most homeless families in New York do manage to find a place to stay fairly soon after they enter the shelter system. Half leave the system within two to five months, and two-thirds leave within a year. Writers like Kozol who think that most families in welfare hotels are in effect permanently homeless are focusing on the long-term stayers and missing the majority who do leave after several months. In other words, homeless families simply need more assistance in finding housing more quickly in the stock that already exists.
Since advocates for the homeless claim that homelessness is entirely or primarily caused by a housing shortage, they typically deemphasize the role that disabilities like mental illness, alcoholism, and drug abuse play in the plight of the homeless. This is the theme sounded by Peter Marcuse, a professor of urban affairs at Columbia University. Marcuse cautions against “blaming the victim,” a tendency which (according to him) holds that “The homeless are not like you and me. There is something wrong with them or they wouldn’t be homeless.” We should, as Louisa Stark, president of Coalition for the Homeless puts it, “Blame the System, Not Its Victims.”
To get a sense of how valid such claims are, we have to distinguish for the moment between homeless individuals and homeless families. This approach is controversial. Indeed, one reason that disability among the homeless is sometimes thought to be less of a factor than is popularly supposed is that many of the homeless are members of homeless families. Since these families—especially the children—have much lower rates of mental illness and other disabilities than homeless individuals do, amalgamating the two groups brings the disability rate down. At the same time, however, it conceals the true dimension of disability among homeless individuals. The best way to proceed, therefore, is first to consider individuals and families separately and then to reaggregate them for the total picture.
In some journalistic accounts of homelessness during the mid- to late-1970’s, all homeless individuals were assumed to be mentally ill. It was thought that most such mentally-ill homeless individuals had once been patients in mental hospitals before they were “deinstitutionalized.”
It is clear, however, that deinstitutionalization—if by this is meant the policy of reducing the number of patients in mental hospitals with an eye to having them cared for by community mental-health centers—cannot explain the plight of today’s homeless, because this policy was implemented mostly during the 1960’s. Very few homeless people today came to the streets or shelters directly from mental hospitals. (In the Freeman and Hall study the number of such people was only 1 percent of the sample.)
On the other hand, while few of the homeless are direct victims of deinstitutionalization, we learn from Freeman and Hall that more than 90,000 of the people on the streets today would have been in mental hospitals in the days before the policy of deinstitutionalization came into being. Almost all surveys indicate that between a quarter and a third of the homeless are indeed mentally ill. For example, the Freeman and Hall study found 33 percent of its sample to be mentally ill. Rossi, in a survey of eighteen studies of homeless individuals, found that the average rate of chronic mental illness was 36.5 percent.
According to some writers, these findings demonstrate that the perception of the homeless as being mentally ill is merely a stereotype. Thus Louisa Stark tells us that “Although only one-third of homeless people nationwide are mentally disabled, since the early 1980’s homelessness has become synonymous in the public mind with mental illness. What we have done, then, is taken one stigmatized illness, alcoholism, and replaced it with another, mental illness, as a stereotype for the homeless.” Peter Marcuse presents the other side of this observation when he remarks that “most of the mentally ill are not homeless. More mentally ill are housed than homeless.”
But neither of these superficially correct observations can break the link between mental illness and homelessness. For a rate of mental illness of 33 percent is very high. In the general population, according to Freeman and Hall, the rate is less than 2 percent, and it is not much more than that even among the dependent poor: Rossi has found that among individuals receiving General Relief in Chicago the rate of previous hospitalization for mental illness is between 3 and 4 percent.
Further, to say that the majority of the mentally ill are housed and that therefore mental illness is not a key factor in the current homeless problem is like saying that during the 1930’s the majority of Americans were employed and that therefore the Depression had nothing to do with unemployment.
Yet those researchers who stress that the majority of the homeless are not mentally ill are making an important point. Just as it is misleading to assume that homelessness is simply a housing problem (although it is partly that), so too is it misleading to assume that homelessness is entirely a problem of mental illness (although it is partly that, too). There are several other groups among homeless individuals with different problems of their own.
The most obvious are alcoholics and hard-drug abusers. Freeman and Hall report that 29 percent of their sample suffered from alcohol abuse (the rate for the general population is 13 percent), and they found a rate of hard-drug abuse of 14 percent (the rate for the general population is less than 1 percent). My own guess is that something like a fifth to a third of homeless individuals suffer from some combination of these disabilities.
Mental illness, alcoholism, and drug abuse are all regarded as quasi-medical problems. But of course straightforward medical or physical problems can be disabling, too. Indeed, such disabilities turn out to be characteristic of significant percentages of homeless individuals. For example, 36 percent of Rossi’s Chicago sample reported “fair” or “poor” health (a level of self-reported ill health roughly twice that of the general population); 28 percent reported a hospital stay of more than twenty-four hours in the last year; and 28 percent were unable to work for health reasons. Also, a survey of over 8,000 clients in New York City’s shelters for individuals (conducted in 1984 by Stephen Crystal and Mervyn Goldstein, then of New York City’s Human Resources Administration) found almost two-fifths had a current medical problem.
It is difficult to extrapolate from these numbers to a hard estimate of physical disability among the homeless. But a safe guess is that about 25 to 30 percent of homeless individuals suffer from such problems.
In addition to such strictly or quasi-medical difficulties, there are problems of a qualitatively different type that can be contributing factors to a spell of homelessness. Thus 41 percent of Rossi’s sample had been in jail for periods of longer than forty-eight hours, 28 percent had at some point been convicted by a court and given probation, and 17 percent had felony convictions behind them. And Freeman and Hall found 39 percent of their sample had at some point been in jail.
One objection to regarding a criminal record as a disability characteristic of a subpopulation of the homeless is that such a history might be considered a result rather than a precipitating or contributing cause of homelessness. My experience in San Diego suggests that many homeless people are frequently convicted of misdemeanors like jay-walking or littering as a form of harassment by police. Yet Freeman and Hall found that 61 percent of the time spent in jail by their sample was before the subjects became homeless. This indicates that being convicted of a crime can indeed be a contributing cause of homelessness. Moreover, it is unlikely that minor police harassment can account for a felony conviction rate of 17 percent.
What, then, is the overall picture of disability among homeless individuals? First, these data—dramatic as they are—do not support the idea that homelessness is some “special” problem quite unlike any other social problem. The disabilities involved are not unique to the homeless; they are found among the extremely poor and the underclass in general. In this sense, those who caution against “stereotyping” the homeless have a point.
However, a point which these data highlight with at least equal force is that homeless individuals are much more disabled than the general population or even than the poor in general. Somewhere between 70 and 80 percent of homeless individuals suffer from one or more major disability. Thus Rossi found that 82 percent of his survey had at least one of the following disabilities: poor or fair health; previous mental hospitalization; previous stay in a detoxification unit; clinically high scores on psychological tests for depression or psychotic thinking; sentence(s) by a court.
Similarly, the Crystal and Goldstein study of New York’s shelters for singles found that 74.9 percent of the men and 70.4 percent of the women suffered from at least one of the following disabilities: hard-drug abuse; alcoholism; jail record less than an 8th-grade education; never employed; physical/medical problems; psychiatric problems; over sixty-five years old.
The homeless, then, are not “just like you and ine” and most of them do have “something wrong with them” which contributes to their being homeless. (We should also keep in mind, however, the approximately 20 to 30 percent of the homeless who are relatively able and who have been described in other studies by Crystal as “economic-only” clients.)
When advocates for the homeless are confronted with figures like the above, they often choose to shift attention away from homeless individuals to homeless families. For example, a recent publication of the National Coalition for the Homeless argues:
. . . not only is homelessness increasing in numbers, it is also broadening in reach. The old stereotype of the single, male alcoholic—the so-called “skid-row derelict”—no longer applies. Increasingly, the ranks of the homeless poor are comprised of families, children, ethnic and racial minorities, the elderly, and the disabled. The face of America’s homeless now mirrors the face of America’s poor: skid row has become more democratic. Perhaps the starkest indication of this diversity is the fact that, today, the fastest growing segment of the homeless populaion consists of families with children. In some areas, families with children comprise the majority of the homeless . . . families with children now account for about 30 percent of the homeless population.
Kozol and other advocates also claim that the rate of disabilities like drug abuse and mental illness among such families is very low. Although “Many homeless individuals may have been residents of such [mental] institutions,” writes Kozol, “in cities like New York, . . . where nearly half the homeless are small children, with an average age of six, such suppositions [of former mental hospitalization] obviously make little sense.”
There are two questions here: (1) Just how many homeless family members are there and what percentage of the total homeless population do they represent? (2) What are the rates of disability among homeless family members?
As to the question of numbers, the Freeman and Hall study again provides the best source. They estimate that in 1983 there were 32,000 homeless family members, and that in 1985 there were 46,000 such people.
Right away, we can see from the Freeman-Hall data that Kozol is exaggerating wildly when he tells us that 500,000 children are currently homeless. Since only 46,000 members of families were homeless in 1985 (this includes both adults and children), homeless families would have had to grow by more than ten times in three years for Kozol’s figure even to approach reality—not to mention that there are in fact no more than a total of 400,000 homeless people in 1988. The truth is that the widely circulated estimate that families with children represent 30 percent of the total homeless population is more than twice too high.
So far as the question of disability goes, we know roughly what the rates are among homeless individuals, but at this point we do not know enough to conclude either that homeless families suffer from no more disabilities than do other poor families, or that in fact they do have special problems which contribute to their being homeless. Such studies as we do have suggest that the rate of disabilities among such families may be significant. But until better information is available on this subject, we have to regard the confident pronouncements of advocates that homeless families are simply victims of a tight housing market as being largely conjectural.
What are we to make of the advocates’ interpretation of the politics of homelessness? It is important to understand that they see homelessness not as an aberration or a failing but as a natural outcome of the ordinary workings of social policy and the economy in general. As Mitch Snyder of CCNV writes (in collaboration with Mary Ellen Hombs):
We live in a disposable society, a throwaway culture. The homeless are our human refuse, remnants of a culture that assigns a pathologically high value to independence and productivity. America is a land where you are what you consume and produce. The homeless are simply surplus souls in a system firmly rooted in competition and self-interest, in which only the “strongest” (i.e., those who fit most snugly within the confines of a purely arbitrary norm) will survive.
Similar sentiments are expressed by Robert Hayes of Coalition for the Homeless (“the homeless are indeed the most egregious symbol of a cruel economy, an unresponsive government, a festering value system”) and by Peter Marcuse of Columbia (“homelessness in the midst of plenty may shock people into the realization that homelessness exists not because the system is failing to work as it should, but because the system is working as it must”).
Yet it ought to be obvious that homelessness cannot be the result of “a festering value system,” or “free-market capitalism,” or “a system firmly rooted in competition and self-interest,” or any other long-term systemic feature of American society, for the simple reason that all these have remained more or less what they always have been, and so cannot explain the rise of homelessness now.
Advocates seem to recognize this at some level, since, after having made sweeping pronouncements against American society in general, they usually concentrate their fire on the Reagan administration’s social and economic policies. But while poverty did increase during the Reagan years, the Reagan administration can hardly be blamed either for deinstitutionalization or for the housing policies—such as rent control and byzantine regulations—which in cities like New York have undoubtedly exacerbated the situation.
To sum up: homelessness is a much smaller problem, in terms of the number of people affected by it, than is commonly thought, but it is also much more intractable than advocates understand. This intractability stems from the fact that the great majority of homeless individuals, and possibly some significant proportion of homeless families, are afflicted with behavioral or medical disabilities or both. Dealing with such problems requires a willingness to assert authority—for example, in refusing to allow people like Joyce Brown to live on the streets—as much as it requires an expenditure of resources. Nonetheless, an important part of the problem is economic, and can best be addressed not by building more public housing but by raising the income of the extremely poor and by removing regulations which block the allocation of housing to them.
In short, a reformist agenda—one aimed at enabling our mental-health system to treat people who need treatment, at reducing extreme poverty through income supports, and at allowing housing markets to function—can go a long way toward the elimination of homelessness.
This agenda is a far cry from the radical systemic measures such as a constitutional “right to shelter,” or the “guerrilla legal tactics” and the “harassing” techniques recommended by Joyce Brown’s lawyers and other activists in the field. But it has the virtue of being based on a truthful diagnosis rather than on wild and tendentious analyses which, by simultaneously exaggerating the dimensions of the problem and underplaying its difficulties, make it harder rather than easier to help the homeless at all. In taking up the case of Joyce Brown, and in securing her release, the NYCLU lawyers may have furthered the “delegitimation” (as Peter Marcuse calls it) of a social system they consider evil. But as for Joyce Brown herself, only a short time after her appearance at Harvard, she was found begging on the streets and hurling abuse at passers-by who refused to give her money.
1 Crown, 261 pp., $16.95.