<?xml version="1.0" ?> <tei> <teiHeader> <fileDesc xml:id="0"/> </teiHeader> <text xml:lang="en"> <head>INTROD UCTION<lb/></head> <p>By definition, homeless men and women have one very important and<lb/> central common denominator: they need a home. Homeless women,<lb/> however, also have many other problems that differentiate them from<lb/> homeless men. One of the major differences between homeless men and<lb/> women is that most homeless women are single heads of families with<lb/> the direct responsibility of children in their custody, which is almost<lb/> never the case for homeless men.<lb/></p> <p>There is a consensus that the concerns of women have been under-<lb/>represented in the literature on homeless populations <ref type="biblio">(Milburn & D'Er-<lb/>cole 1991; Hagen 1987)</ref>. Five recent noteworthy studies <ref type="biblio" >(Breakey,<lb/> Fischer, Kramer, et al. 1989; Maurin, Russell & Memmott, 1989;<lb/> Hagen, 1987; Burt & Cohen, 1989; Crystal, 1984)</ref> have directly com-<lb/>pared homeless men and women. These studies have found that com-<lb/>pared to homeless men, homeless women are younger, better educated,<lb/> more likely to be caring for dependent children, more likely to be on<lb/> welfare and less often below the poverty line, and less likely to have a<lb/> history of arrest or jail. Studies disagree on whether homeless women<lb/> have higher <ref type="biblio">(Breakey & Fischer, 1989; Maurin, Russell & Memmott,<lb/> 1989; Crystal, 1984; Crystal, Ladner & Towber, 1986; Rosnow, Shaw &<lb/> Concord, 1986; Wright & Weber 1987)</ref> or lower <ref type="biblio">(Hagen, 1987)</ref> rates of<lb/> psychiatric problems than their male counterparts.<lb/></p> <p><ref type="biblio">Burt and Cohen (1989)</ref> subdivided women into those with children in<lb/> their custody and those without. They found that solitary women,<lb/> compared to women with children, had higher proportions with histo-<lb/>ries of psychiatric hospitalization and suicide attempts, had been home-<lb/>less longer, were more likely to be working, and were less often depen-<lb/>dent on welfare.<lb/></p> <p>The current paper reporting from a study of 900 randomly selected<lb/> and systematically interviewed homeless individuals describes a set of<lb/> analyses dedicated to furthering our understanding of the differences<lb/> between male and female homeless populations. More complete appre-<lb/>ciation of the unique circumstances and problems of homeless men and<lb/> women will permit development of more effective intervention pro-<lb/>grams that attend to the separate needs of each subgroup. Given the<lb/> continuing influx of women and children into the ranks of the homeless<lb/> <ref type="biblio">(Burt & Cohen, 1989)</ref>, the special needs of homeless women and their<lb/> families will become increasingly urgent in the near future.<lb/></p> <head>METHODS<lb/></head> <head>Sampling<lb/></head> <p>The sample of 600 homeless men and 300 homeless women was drawn from all<lb/> overnight and daytime shelters in St. Louis, as well as locations on the street or other<lb/> public areas where the homeless are known to congregate. Greater detail is provided<lb/> elsewhere on this study's working definition of homelessness, inclusion criteria, the<lb/> kinds of shelters and day programs sampled, and the street sampling methods <ref type="biblio">(Smith,<lb/> North & Spitznagel, 1992; 1993)</ref>. All 14 of the night shelters for men and all but two of<lb/> the 13 night shelters for women, all 4 day centers, and both of the city's rehabilitation<lb/> programs cooperated with the project.<lb/></p> <p>From overnight shelters, 195 men and 251 women were interviewed; 150 men and 29<lb/> women were interviewed from day centers; 76 men were sampled from specialized<lb/> rehabilitation programs; and 20 women were sampled from 24-hour emergency shel-<lb/>ters. Random sampling was conducted proportionally to the numbers of persons in the<lb/> various overnight or day shelter programs. Male subjects sampled from unsheltered<lb/> locations (N = 179) were systematically drawn from the streets, parks, and other public<lb/> areas. We were unable to find homeless women to interview from these locations.<lb/></p> <p>Interviewing proceeded to completion in 12 months, with sampling of study subjects<lb/> throughout all parts of months in all four seasons. Interviews were conducted by<lb/> professional trained interviewers, and lasted two hours on average. Subjects received<lb/> $10.00 for their participation. The completion rate was 91% for men and 96% for<lb/> women.<lb/></p> <head>Instruments<lb/></head> <p>The Diagnostic Interview Schedule/Homeless Supplement (DIS/HS) was used in this<lb/> study. The DIS <ref type="biblio">(Robins, Helzer, Croughan, Williams & Spitzer, 1981)</ref> permitted us to<lb/> make DSM-III-R diagnoses including schizophrenia, bipolar affective disorder, major<lb/> depression, panic disorder, generalized anxiety disorder, alcohol abuse/dependence,<lb/> drug abuse/dependence, and antisocial personality disorder. The Homeless Supplement<lb/> was designed for use in this study and it included questions about the homeless<lb/> experience and subjective impressions of the origins of their problems. Subjects were<lb/> also asked a series of questions about their educational, occupational, and financial<lb/> history, and family and childhood background.<lb/></p> <head>Data Analysis<lb/></head> <p>Chi square tests were performed in analyses comparing dichotomous variables. If<lb/> expected counts in cells were 5 or less, then Fisher's exact tests were performed. Means<lb/> are presented with standard deviations. T-tests were performed for prediction of contin-<lb/>uous variables. Level of significance was set at p _< .05.<lb/></p> <p>For certain analyses, women were subdivided into two groups: '%olitary" women who<lb/> had no children under age 16 in their physical custody (N = 98), and mothers who had<lb/> children under age 16 with them (N = 202). (In comparison, 594 men were "solitary"<lb/> and 3 had children with them).<lb/></p> <head>RESULTS<lb/></head> <head>Demographics<lb/></head> <p>The overwhelming majority of women (89.7%) were mothers; a majority<lb/> of the men (60.5%) also were fathers (x 2 = 81.37, dfl, p < .001). Most<lb/> women with children under age 16 had some or all of them in their<lb/> physical custody (80.2%), but men rarely did (1.2%) (x 2 = 331.82, dfl, p<lb/> < .001). Overall, 67.3% of women and 0.5% of men had children under<lb/> age 16 with them.<lb/></p> <p>The majority of both men and women were nonwhite, the women's<lb/> sample having a significantly higher proportion of African-Americans<lb/> (84.0% vs. 69.7%; x 2 = 21.60, dfl, p < .001); 4.0% of women and 2.8% of<lb/> men were of other minorities. The women were significantly younger<lb/> than the men (29.0 • 8.6 vs. 35.9 • 10.8 years; t = 9.68, p = .0001).<lb/> Most subjects had never married (60.3% of women, 54.5% of men; x 2 =<lb/> 2.77, dfl, p = .10). More men than women had graduated from high<lb/> school (51.4% vs. 42.3%; x 2 = 5.87, dfl, p = .015).<lb/></p> <p>Solitary women were more likely to be white than were women with<lb/> children (26.5% compared to 4.9%; x 2 = 29.10, dfl, p _ .001), and the<lb/> solitary women were racially like solitary men. Solitary women were<lb/> significantly older than other women (mean age 33.1 • 11.8 vs. 27.0 •<lb/> 5.6 years; t = 6.04, p = .0001), but younger than solitary men (35.9 •<lb/> 10.8 years; t = 2.42, p = .016). Solitary women did not differ from other<lb/> women in the proportion with a high school education.<lb/></p> <head>Employment and Financial Support<lb/></head> <p>Few subjects were employed; men were more likely to be working than<lb/> women (21.5% vs. 10.7%; x 2 = 14.11, dfl, p < .001). More solitary<lb/> women (16.3%) than mothers with children (8.4%) were working, but<lb/> the difference was not statistically different, nor were they different<lb/> from solitary men (21.5%). In spite of most women having children, only<lb/> 6.2% indicated that lack of affordable day care was an obstacle to<lb/> employment (which was never a problem for men).<lb/></p> <p>Women reported higher annual incomes for the last year than did<lb/> men. More than half (53.0%) of women said they earned at least $4000,<lb/> compared to 28.6% of men (x 2 = 38.19, dfl, p < .001). Solitary women<lb/> were statistically indistinguishable from solitary men on income. Most<lb/> men depended primarily on their earnings (60.0% compared to 32.4% of<lb/> women; x 2 = 53.48, dfl, p < .001), and women more often depended on<lb/> welfare (43.8% compared to 2.5% of men; x 2 = 209.58, dfl, p < .001).<lb/></p> <p>Solitary women were less likely to be supporting themselves primarily<lb/> on welfare than were mothers with children (10.0% compared to 57.8%;<lb/> x 2 = 5.33, dfl, p < .001) and more likely to be supporting themselves on<lb/> their earnings (42.5% compared with 28.1%; x 2 = 5.33, dfl, p = .021).<lb/></p> <head>History of Homelessness<lb/></head> <p>Men were more likely than women to have been homeless prior to the<lb/> current episode (58.2% vs. 50.2%; x 2 = 5.22, dfl, p = .022), and they also<lb/> reported more total lifetime years of homelessness (mean 2.75 :e 3.79<lb/> for men, 1.15 • 2.33 for women; t = 6.70, p --.0001). Solitary women<lb/> were more likely than mothers with children to report more than a year<lb/> of total lifetime homelessness (42.9% compared to 16.4%; x 2 = 24.51,<lb/> dfl, p < .001), but they were less likely than solitary men to have<lb/> experienced more than a year of total lifetime homelessness (42.9%<lb/> compared to 54.7%; x 2 = 4.75, dfl, p = .029).<lb/></p> <p>A sizable portion of women had spent most of the last year in their<lb/> own home or apartment (39.2%, compared to 20.2% of the men; x 2 =<lb/> 35.86, dfl, p < .001), or with their families (33.8% vs. 13.3%; x 2 = 50.45,<lb/> dfl, p < .001). Very few women (none with children) had spent most of<lb/> the last year living on the streets or other unsheltered locations (2.0%),<lb/> although many men gave such a history (17.2%) (x 2 = 42.28, dfl, p <<lb/> .001).<lb/></p> <p>Subjects were asked to identify reasons for their having first become<lb/> homeless. For men, the most frequent reason given was unemployment<lb/> (23.5%), and for women it was familyconflict (6.0%). More men than<lb/> women cited divorce or separation (7.7% of men, 3.0% of women;<lb/> Fisher's exact p = .005) as a precipitant of their homelessness. Only a<lb/> small fraction identified substance abuse as the main reason; men more<lb/> often than women mentioned alcohol (7.2% vs. 0.7%; Fisher's exact p =<lb/> .000003) or drug (8.8% vs. 4.4%; x 2 = 5.90, di], p = .015) problems.<lb/> Mental illness other than substance abuse was a very infrequent an-<lb/>swer (1.2% of men, 0.7% of women).<lb/></p> <head>Legal History and Social Networks<lb/></head> <p>Men reported more legal difficulties than women. Two-thirds (65.8%) of<lb/> men and nearly one-fourth (23.2%) of women said they had ever been<lb/> incarcerated. One-fourth (24.2%) of the men and 4.6% of the women<lb/> were convicted felons (x 2 = 12.58, dfl, p < .001). Nearly half (43.6%) of<lb/> the women reported that they had suffered physical abuse by a spouse<lb/> or partner, and 39.2% of those abused reported more than one abusive<lb/> relationship.<lb/></p> <p>Neither men nor women were as disaffiliated as one might expect<lb/> from other published reports <ref type="biblio">(Bassuk, Rubin, & Lauriat 1984; Fischer,<lb/> Shapiro, Breakey, Anthony & Kramer 1986; Bassuk & Rosenberg 1988;<lb/> Fischer & Breakey 1986; Milburn & D'Ercole 1991)</ref>. More than half of<lb/> the subjects (52.3% of men, 55.1% of women; p > 0.5) had relatives. The<lb/> vast majority of both men (92.5%) and women (81.8%) with relatives<lb/> indicated that they themselves refused to have contact with their rela-<lb/>tives (x 2 = 5.68, dfl, p = .017), and many subjects also said that their<lb/> relatives refused contact with them (36.2% for men, 52.7% for women;<lb/> x 2 = 5.02, dfl, p = .025). Most (76.2%) of the women reported that they<lb/> spent their time with their children, whereas men rarely did (6.0%) (x 2<lb/> = 263.56, dfl, p < .001), saying that they spent their time with friends<lb/> (30.4%), other people at the shelters (23.2%), or, most often, alone<lb/> (49.4%).<lb/></p> <head>Family and Childhood<lb/></head> <p>Family and childhood backgrounds of the subjects revealed few differ-<lb/>ences between men and women. The majority (62.4%) were raised in<lb/> broken homes, with fathers absent, and men and women did not differ<lb/> in this regard. Nearly one-fourth (23.1%) of the women reported having<lb/> been sexually abused as a child, but few men (4.4%) acknowledged this<lb/> experience (x 2 = 63.99, dfl, p < .001).<lb/></p> <head>Psychiatric Disorders<lb/></head> <p>The personal psychiatric histories of the subjects were remarkable for<lb/> high rates of psychiatric disorders-nearly half (47.7%) of the women<lb/> and more than three-quarters (76.7%) of the men qualified for a lifetime<lb/> diagnosis. A history of substance abuse was reported by 74.7% of the<lb/> men, leaving virtually no psychiatric disorders in men uncomplicated<lb/> by a substance abuse history. Almost one-third (30.8%) of the women<lb/> reported a history of substance abuse, accounting for the majority of<lb/> women with any lifetime psychiatric disorder. The only non-substance<lb/> Axis I diagnosis with rates significantly different in men and women<lb/> was major depression, which was more prevalent among women (24.7%<lb/> vs. 18.4%; x 2 = 4.99, dfl, p = .025).<lb/></p> <p>Compared to women with children in their physical custody, solitary<lb/> women were more likely to meet lifetime criteria for schizophrenia<lb/> (9.3% compared with 1.0%, Fisher's exact p < .001) or alcohol abuse/<lb/> dependence (25.3% compared with 12.7%; x 2 = 7.26, dfl, p = .007).<lb/> Details of specific rates for other psychiatric disorders are described in<lb/> previous work <ref type="biblio">(Smith, North & Spitznagel, 1992; 1993)</ref>.<lb/></p> <head>DISCUSSION<lb/></head> <p>The main sociodemographic attributes that homeless men and women<lb/> share are the high proportions of minorities and the prevalence of<lb/> unemployment and poverty. Even these variables, however, are not<lb/> uniform between men and women. Homeless women were younger than<lb/> homeless men, more likely to be members of a minority group, and<lb/> likely to have young children with them. Although they reported<lb/> higher incomes than men, more were unemployed and dependent on<lb/> welfare. Women had more often than men had emerged into homeless-<lb/>ness from the homes of their families, and the majority reported that<lb/> their families refused contact with them. The family as a safety net has<lb/> apparently failed them, and the burden of many children on limited<lb/> family income and space may be a source of conflict that has strained<lb/> family relations beyond the breaking point.<lb/></p> <p>Solitary women were different in many respects from homeless<lb/> mothers with children in their custody. Compared to mothers with<lb/> children, solitary women were more often white, more often relied jobs<lb/> rather than welfare, had a longer history of homelessness, and had<lb/> higher lifetime rates of schizophrenia and alcoholism. The population of<lb/> homeless women is therefore heterogeneous, with at least two sub-<lb/>groups.<lb/></p> <p>Homeless women overall, compared to men, were found to have lower<lb/> rates of substance abuse, incarceration, and felony conviction. They had<lb/> not been homeless as long or as frequently men, had spent less time on<lb/> the streets and in other unsheltered locations, and spent less time<lb/> alone. Many women had been sexually molested as children and had<lb/> endured physically abusive partners.<lb/></p> <p>While major non-substance mental illness was over-represented<lb/> among the homeless, most subjects, male or female, did not qualify for a<lb/> diagnosis. Aside from the higher rates of major depression in women,<lb/> these data do not support findings from other studies that reported<lb/> higher rates of psychiatric disorders among homeless women than men,<lb/> a difference that may reflect this study's systematic assessment. Dilu-<lb/>tion of the homeless male population by large numbers of primary<lb/> alcoholics <ref type="biblio">(Fischer & Breakey, 1991)</ref> may yield higher apparent rates of<lb/> other psychiatric illness in homeless women.<lb/></p> <p>Homeless women with children comprise the fastest growing sub-<lb/>group of the homeless <ref type="biblio">(Burt & Cohen 1989)</ref>. Our data indicate that the<lb/> needs of homeless women and their families are very different from<lb/> those of most homeless men. Unlike homeless men, most homeless<lb/> women have children in their custody to think about and deal with.<lb/> This factor alone makes the experience of homelessness a qualitatively<lb/> different experience for women, as well as suggesting different origins<lb/> of homelessness and potential solutions to it. Children may paradox-<lb/>ically be both a source of support and a source of stress to homeless<lb/> mothers <ref type="biblio">(Milburn & D'Ercole 1991)</ref>. Homeless mothers may perceive<lb/> themselves as in perpetual risk of losing their children, and they must<lb/> constantly struggle to provide the financial and emotional support,<lb/> physical care and safety, and developmental stimulation that domiciled<lb/> children receive.<lb/></p> <p>Homeless men, on the other hand, do not have the access to welfare<lb/> eligibility that many homeless mothers do, yet homeless mothers have<lb/> the problem of making the welfare money spread far enough to support<lb/> them and their dependent children throughout each month. While the<lb/> majority of homeless mothers seem to depend on welfare, it is clearly<lb/> inadequate to support their needs. Their incomes, although higher than<lb/> those of homeless men, have been outstripped by the demands of their<lb/> dependent families upon them.<lb/></p> <p>The problems that have made these women vulnerable to homeless-<lb/>ness are problems that also may adversely affect their children: drug<lb/> and alcohol abuse, mental illness and behavioral problems, abject pov-<lb/>erty, and lack of intact families. Every problem a homeless mother<lb/> faces is fraught with complications related to responsibility toward her<lb/> dependent children.<lb/></p> <p>Not only should research investigations of the problems related to<lb/> homelessness examine and report on men's and women's experiences<lb/> separately, but interventions designed to help homeless persons over-<lb/>come homelessness and achieve stability must be tailor made to the<lb/> specific needs of men and women <ref type="biblio">(Crystal 1984)</ref>. To be effective, all<lb/> programs to aid homeless women, whether aimed at problems of sub-<lb/>stance abuse or other psychosocial problems, must take into account<lb/> their special needs in association with their dependent families, and the<lb/> many important socioeconomic and interpersonal elements of the cir-<lb/>cumstances of homeless women.</p> </text> </tei>