Human Service Integrative Course

Homeless Women With Children The Role of Alcohol and Other Drug Abuse Marjorie J. Robertson I I For some women with children, alcohol and other drug use may be an important risk factor for homelessness be- cause it may interfere with a woman's capacity to compete for scarce resources such as housing, employment, or ser- vices. The impact of various policy decisions on homeless women, their dependent children, and the family unit is considered, including women's right to privacy, crim- inalization of drug use, and scarcity of appropriate treat- ment programs for homeless women.

Homeless families are a heterogeneous population, and the causes of family homelessness are multiple and in- teractive (McChesney, in press; Weitzman, Knickman, & Shinn, 1990; Wood, Schlossman, Hayashi, & Valdez, 1989). The most accurate and comprehensive depiction of homeless women with children and the risk factors for their homelessness is essential for the design of adequate programs and services to prevent or alleviate homelessness in this population. Often, descriptions of the individual characteristics of homeless persons, such as alcohol- or other drug-related problems, have been inappropriately interpreted as ex- planations of their homelessness. On one hand, such sim- plistic explanations tend to ignore or minimize structural or environmental factors that influence the extent and course ofhomelessness (Hopper, 1990). Principal among these are the crisis in affordable housing, reduced social welfare benefits, and increasing poverty among single women with children in the United States. On the other hand, the experience of homelessness is not a random event that is independent of individual characteristics or experiences. In the context of poverty and high housing costs (Dolbeare, 1989), serious personal and family problems such as alcohol and other drug abuse can make an individual less able to compete for scarce resources such as low-cost housing or employment (Hop- per, 1990; Wood et al., 1989). In the words of one re- searcher, serious alcohol or other drug abuse represents an additional "burden of vulnerability" for homeless and other poor people (Fischer, 1989). Thus, although alcohol or other drug abuse should not be construed as the ex- planation of family homelessness (Weitzman et al., 1990), it may serve as an important contributing risk factor for some women with children. Alcohol and other drug use among contemporary homeless adults is increasingly identified as a serious Alcohol Research Group, Institute of Epidemiology and Behavioral Medicine, Medical Research Institute of San Francisco, Berkeley, CA II problem that is more prevalent among the homeless than among the general population (Fischer, 1989; Garrett, 1989; Milburn, 1988). However, almost nothing is known about the direct contribution of alcohol and other drug abuse to homelessness among women with dependent children. In this article, I explore critical ways in which alcohol and other drug use can aggravate, precipitate, or prolong homeless episodes among homeless women with dependent children. Epidemiologic data on the impact of alcohol and other drug use on women and their chil- dren, and on the stability of the family unit, are discussed briefly. Policy considerations are introduced, including available treatment and policy options. Because the lit- erature is limited, substantive findings will be integrated into a conceptual inquiry into the ways in which alcohol and other drug abuse affects homeless women with chil- dren.

Alcohol and Other Drug Use Among Homeless Women With Children Alcohol-related problems have been prominently re- ported in studies of homeless persons in past decades, and the contemporary generation of homeless adults is distinguished by reports of high rates of illicit drug use, as well (Garrett, 1989). A review of the empirical literature suggests that alcohol and other drug problems are more prevalent among the homeless population than among the general population, with estimates of alcohol problems ranging from 12% to 68%, and estimates of other drug problems ranging from 1% to 48% (Fischer, 1989). In general, alcohol and other drug problems are reportedly less prevalent among homeless women than among homeless men (Fischer). Homeless women with dependent children seem to be a distinctive subpopulation among homeless adults, with lower rates of substance use and related problems than have been reported for homeless men or for homeless women without children (Burr & Cohen, 1989; Johnson Preparation of this article was supported by National Institute of Mental Health Grant RO1 MH46104. The author is grateful to Elizabeth A. Smith, Herb Westerfelt, Scott O. Swain, and anonymous reviewers for comments on earlier drafts of this article. Correspondence concerning this article should be addressed to Marjorie J. Robertson, Alcohol Research Group, Institute for Epide- miology and Behavioral Medicine, 2000 Hearst Avenue, Suite 300, Berkeley, CA 94709. 1198 November 1991 • American Psychologist Copyright 1991 by the American Psychological Association, Inc. 0003-066X/91/$2.00 Vol. 46, No. 11, 1198-1204 & Krueger, 1989; Robertson, in press-a). Estimating the prevalence of alcohol and other drug abuse in this special population is difficult, however. The task of estimation is hampered by the paucity of empirical work on homeless families. Also, because of major methodological limita- tions common to studies of homeless populations, most current findings on homeless women with children are neither comparable with one another nor representative of the population as a whole. Studies of homeless families vary significantly by focus, design, measurement, and def- initions of homelessness and family unit. These variations prohibit meaningful generaliTation or comparisons across studies. For example, studies of homeless families tend to be based exclusively on families in shelters, leaving uncertain how these homeless families differ from those that are not in shelters (Solarz, in press). Virtually all of the studies are cross-sectional, overrepresenting persons who have been homeless longer and likely overrepresent- ing the prevalence of alcohol- and other drug-related problems, inasmuch as these problems are more likely to be associated with longer term homelessness. Moreover, the lack of rigorous sampling methods often prevents generalization beyond the homeless family members in- terviewed. With regard to instrumentation, most studies of homeless families are not designed as epidemiologic studies of substance abuse. Because studies of homeless families tend not to focus on alcohol and drug use, sparse findings must be gleaned from studies of homeless families that have been designed with some other focus, such as physical or mental health. This is a common strategy for examining alcohol and other drug use among homeless persons generally (Fischer, 1989). Furthermore, if alcohol and other drug use are assessed at all, standardized in- strumentation or specific criteria for alcohol or drug abuse are often lacking. Prevalence rates cannot be compared across studies because rates vary as a function of instru- mentation and whether one uses lifetime or current rates, clinical or descriptive definitions of abuse, or treatment histories (Fischer). Also, information about use of a spe- cific drug is seldom reported, and alcohol and other drug indicators are often aggregated into a single rate of sub- stance use, abyse, or treatment (Robertson, in press-b). Although these methodological issues constrain our ability to generalize across samples to the larger popu- lation, the available literature suggests a high prevalence of alcohol and other drug use and related problems among homeless women with children compared with other poor women with children. For example, in a case-control study of poor, female-headed families in Boston, higher alcohol and other drug problems were reported for homeless mothers compared with poor-but-housed mothers (16% compared with 6%, respectively, on the basis of structured psychiatric interviews and applying Diagnostic and Statistical Manual of Mental Disorders, 3rd edition criteria (Bassuk & Rosenberg, 1988). Simi- larly, in New York City, more homeless parents (i.e., those requesting access to city shelters) reported recent, personal substance-abuse problems than did housed parents re- ceiving public assistance (4.8% vs. 1.1%, respectively; Knickman & Weitzman, 1989; Weitzman, Shinn, & Knickman, 1989). In the same study, detoxification treatment for alcohol or drugs was a significant predictor of homelessness in a multivariate model. In Los Angeles, homeless mothers reported more alcohol or other drug use than did poor, housed mothers (43% vs. 30%; Wood, Valdez, Hayashi, & Shen, 1990). The difference in drug use stemmed mainly from a higher prevalence and more frequent use of cocaine, including crack, among homeless mothers (25.3% of homeless mothers used cocaine com- pared with 16.8% of housed mothers, and cocaine was frequently used by 8.7% of homeless mothers compared with 3.1% of housed mothers; personal communication, Hayashi, January 18, 1990; Wood et al., 1989). In sum, there is insufficient empirical evidence on which to base a reliable prevalence estimate for alcohol and other drug abuse among homeless women with chil- dren. More germane than specific prevalence rates, how- ever, are the patterned findings that suggest that homeless women with children consume less alcohol and other drugs than do either homeless men or homeless women without children (Robertson, in press-b). However, homeless women with children have higher rates of al- cohol and other drug use than do other poor women with children. These findings are descriptive and not explanatory, however, and a causal relationship between substance use and homelessness cannot be assumed. However, recent literature suggests that alcohol and other drug abuse can be a contributing factor in family homelessness (Shedlin, 1989; Weitzman et al., 1989), with apparent consequences for the woman, her children, and the family as a unitJ Consequences for the Woman, Her Children, and the Family Unit Alcohol and other drug abuse among homeless women represents a personal health risk for the women, beyond the effects of drug use itself. Use of alcohol or other drugs l Discussion here will focus on alcohol or other drug use by homeless women with dependent children. However, substance abuse by the wom- an's partner may also contribute to family homelessness, especially when alcohol or other drug abuse is combined with domestic violence toward the mother or her children (Somers, in press). Histories of domestic violence are more common among homeless mothers than among other poor but housed mothers with children (Bassuk & Rosenberg, 1988; Weitzman, Shinn, & Knickman, 1989). Some homeless women report that alcohol or drug abuse by their partner may be coupled with domestic violence that precipitated their homelessness (Dail, 1990). For example, in a Massachusetts study, more than 40% of homeless mothers reported that their most recent boyfriend or spouse was a substance abuser, and two thirds reported that battering episodes were most commonly alcohol related. Although 40% of the mothers reported that they had been in an abusive relationship at some time in the past, they seldom identified domestic violence as a precipitant of their current homeless episode (Bassuk, in press). Similarly, compared with housed tingle mothers, more homeless single mothers in Los Angeles reported their most recent male partners to have problems with drug and alcohol use and domestic vi- olence (Wood et al., 1989). Howevea; a connection between the substance use and violence is not specified and in many instances is unclear (Somers, in press). November 1991 • American Psychologist 1199 may result in disinhibition, which may in turn lead to sexual risk taking and increased chance of pregnancy or exposure to sexually transmitted diseases, including hu- man immunodeficiency virus (HIV). Cocaine use in par- ticular has been reported to be associated with syphilis among women, due in part to related sexual behaviors such as trading sex for drugs (Rolfs, Goldberg, & Sharrar, 1990; "Sex-for-Drngs Trade," 1990; "Study Points," 1990). Intravenous use of heroin or cocaine puts one at additional risk for exposure to HIV. In order to procure drugs, women who use crack or other illicit drugs may participate in criminal behaviors, such as drug dealing (McChesney, in press) or prostitution (Inciardi, 1989; Rolfs et al., 1990), which have attendant risks of incar- ceration and consequent family and employment disrup- tion (Williams, 1991). The impact of a woman's alcohol and other drug use represents multiple health risks for her children. For ex- ample, any substance ingested by a pregnant woman may affect the developing fetus, and the specific detrimental effects of alcohol and other drugs have been documented (McKay, 1983). 2 Intravenous drug use increases the risk of prenatal HIV exposure and infection ("Study Points," 1990). Furthermore, drug use combined with homelessness may represent a barrier to prenatal health care, inasmuch as substance-abusing women in general (McKay, 1983; U.S.

House of Representatives, 1989) and homeless women in particular (Robertson & Cousineau, 1986; Shedlin, 1989) are less likely to obtain prenatal or other health care than are other women. Under the rubric of child neglect, abuse, and endangerment laws, in many counties, infants with positive toxicology screens of their urine at birth are re- moved from their mother's custody or placed under su- pervision of the court or department of children's services ("Pregnant Addicts," 1990; Shedlin, 1989; U.S. House of Representatives, 1989; Wilcox, 1990). Other research indicates that the consequences of a homeless mother's drug use may include the neglect, vi- olent treatment, or abandonment of her children. Some neglect may stem from intoxication that interferes with a woman's ability to monitor and meet the needs of de- pendent children. It may also stem from use of limited family resources for illicit drug purchase. In addition, neglect and violence are also attributed to the specific emotional and behavioral alterations caused by crack or other drug use (Inciardi, 1989; Shedlin, 1989). For ex- ample, Shedlin reported that drugs that caused violent behavior, extreme mood swings, or withdrawal were more likely to be associated with violence and neglect of chil- dren. In that study, more violence and neglect of children was reported for homeless crack-addicted mothers than for homeless heroin-addicted mothers. Although poor parenting skills, stress, and a history of family violence also may contribute to neglect and abuse of children by homeless mothers, homeless women in one study believed drug use per se was the main reason for the neglect and abuse of their children and for similar behavior of other women (Shedlin). Several studies of homeless children (Bassuk & Ro- senberg, 1988; Molnar, Rath, & Klein, 1990; Wood et al., 1989) reported higher rates of child abuse or neglect or involvement of child protective agencies, compared with other poor families. However, it is worth noting that some reported incidents of child abuse and neglect of homeless children may to some degree be a product of their parents' homelessness and exaggerated poverty. In addition, by virtue of their homelessness, homeless par- ents are subject to more intense scrutiny by the public, shelter and other service providers, and public welfare staff. Observers report the same behaviors as more abusive for low-income parents than for middle-income parents (Molnar et al.). Also, homelessness itself is interpreted to be child neglect under some child protection statutes. Women with alcohol- or other drug-related problems may experience more difficulty than do other poor women in keeping their families together and housed. Such women face increased difficulty competing in a tight housing market, in which low-income women with chil- dren, welfare recipients, and women of color are already at a disadvantage (City of New York, 1986; Sullivan & Damrosch, 1987). Alcohol or other drug abuse poses an additional obstacle to finding and keeping employment for women who are already at a disadvantage in the job market because of educational deficits, lack of vocational skills, or limited employment histories (Bassuk & Rosen- berg, 1988; Milburn & Booth, 1989-1990; Mills & Ota, 1989; Shedlin, 1989; Sullivan & Damrosch, 1987). Fur- thermore, a homeless mother who uses alcohol or drugs risks exclusion or expulsion of both herself and her chil- dren from emergency shelters or transitional housing.

Those settings are often centers for access to welfare ad- vocacy, health care, child care, and other essential services.

Consequently, women with alcohol- or drug-use problems may be less able to take advantage of rehabilitation and other services (Sullivan & Damrosch), both because of the woman's impaired ability to negotiate bureaucratic barriers (such as keeping appointments) and because of policies of exclusion from programs on the basis of her substance use (Robertson, in press-a). Clearly, alcohol and other drug use may impair a woman's ability to compete for scarce resources, and thereby threaten family integrity. Also, drug or alcohol abuse is likely to place a homeless woman at higher risk of losing custody of her children) Separation may occur 2 Specific outcomes of prenatal alcohol and other drug exposure include prematurity, low birth weight, hypertonicity, low Apgar scores, and longer hospital stays, often in the intensive care nursery (Petitti & Coleman, 1990; U.S. House of Representatives, 1989). Other risks of prenatal drug exposure include fetal alcohol syndrome or the infant's dependence on cocaine, heroin, or methadone at birth (Bullard, 1983; McKay, 1983). Congenital syphilis is a special risk among infants born to crack users who trade sex for drugs ("Sex-for-Drugs Trade," 1990). Crack cocaine and alcohol are reported to be significant parts of the poly-drug pattern of substance abuse among pregnant women (U.S. House of Representatives, 1989). 3 Many homeless women have already experienced separations from their children. For example, two thirds of women in New York City single-adult shelters had children who were not present (Weitzman, Shinn, & Knickman, 1989). 1200 November 1991 • American Psychologist when the mother informally places a child with a friend or relative--often the child's grandparent (Wilcox, 1990).

Other separations occur when children are placed in foster care, either voluntarily by the mother or as a result of intervention by authorities, such as county child protec- tive services (Wood et at., 1990). The removal of children from their mother's custody may have several conse- quences for the woman, beyond the psychological impact of being involuntarily separated from her children. Loss of children may affect support from the extended family.

According to professional staffin one New York City wel- fare hotel, children were seen to enhance a woman's kin- ship ties and consequent emotional and economic support with the children's father's family (Shedlin, 1989). On a more material level, loss of custody may also affect eli- gibility for public housing or for family benefits through Aid to Families with Dependent Children (AFDC). Loss of benefits further reduces a woman's ability to secure or maintain stable housing for herself and her children. Once the child has been placed in foster care, it may be ex- tremely difficult to regain custody, even in cases in which there was no prior evidence of parental abuse or neglect (Solarz, in press). Many agencies will not restore custody until the mother demonstrates that she has adequate and stable housing and income; ironically, she is not eligible for AFDC or housing benefits until the children reside with her (Williams, 1991). In sum, a woman's alcohol and other drug use may result in increased residential instability for herself and her children, including prolonged or episodic homeless- ness, foster placement or other relocation of the children, reduced contact with extended family members, reduced access to health care and other social services, and risk of neglect or abuse for the children.

Treatment Barriers There are multiple barriers for homeless women with children who seek treatment for their alcohol or other drug use. One barrier common to all women is the in- adequate number of alcohol- and other drug-treatment programs in the United States that are geared to the spe- cial needs of women (Comfort, Shipley, White, Griffith, & Shandler, 1990; Ridlen, Asamoah, Edwards, & Zimmer, 1990). Also, there are long waiting periods for drug treat- ment, even for pregnant women (Robin-Vergeer, 1990; U.S. House of Representatives, 1989). For example, of an estimated 4 million women in the United States who currently need alcohol or other drug treatment and who cannot find placements, 250,000 are pregnant ("Survey Finds," 1990). Moreover, a recent survey showed that more than one half of New York City treatment programs refused to enroll pregnant women, and nearly 90% refused pregnant women on Medicaid who were addicted to crack ("Pregnant Addicts," 1990). Lack of access to treatment programs creates an ad- ditional burden for homeless women who have minor children. Residential treatment programs tend to be de- signed for adults only, and they only rarely make provi- sions for women with children. Some innovative programs are designed to work with pregnant women who are drug users, as an alternative to the state's taking custody of the newborn (Wilcox, 1990), but even these facilities are not designed to accommodate a woman's other children. And although a few counties allow grandparents to become foster parents to their own grandchildren (Fagan, 1990), the extended family more often lacks resources to support the children of poor or homeless women in treatment.

To enter a residential treatment program, a homeless mother may have to place her children in foster care (Weinreb & Bassuk, 1990), knowing that she may not be able to retrieve them even if she successfully completes the treatment program. Few treatment and recovery programs are sensitive to service needs of indigent clients (Thomas, Kelly, & Cousineau, 1990). Also, treatment programs usually are not prepared to deal with clients who have a history of poly-substance use or comorbid mental health and sub- stance-use problems, both of which appear to be more common among homeless women than among women in the general treatment population. Homeless women have an additional obvious barrier to treatment: the lack of stable housing that would permit use of outpatient treatment and aftercare services (Wittman, 1989).

Policy Considerations Because there is a paucity of empirical literature on homeless women with children, it is difficult to make comprehensive policy recommendations. Nevertheless, available literature is sufficient to identify three related issues that require choices on the part of policymakers.

Although these policies are played out here in the lives of homeless women, they also affect other poor women and women in the community at large.

Women's Versus Children's and Fetal Rights In the previous discussion, one theme that emerges is the conflict between an individual woman's right to privacy and a child's right to protection by the state. Child pro- tection statutes are invoked, presumably, in the interest of the child's welfare. However, legal safeguards for the mother may be seriously curtailed, including her right to confidential treatment as well as her right to privacy as guaranteed under the Fourteenth Amendment to the U.S.

Constitution ("Pregnant Addicts," 1990; Robin-Vergeer, 1990). 4 This conflict is especially apparent in the prenatal 4 In the Roe v. Wade decision in 1973, the Court recognized that one aspect of the "liberty" protected by the due process clause of the Fourteenth Amendment is "a right of personal privacy, or a guarantee of certain areas or zones of privacy." This right of personal privacy includes, "the interest in independence in making certain kinds of im- portant decisions .... (1)t is clear that among the decisions that an individual may make without unjustified government interference are personal decisions relating to marriage; procreation; contraception; family relationships; and child rearing and education" (Robin-Vergeer, 1990, p. 785). Also, the Supreme Court has historically recognized that freedom of personal choice in matters of family life is a fundamental liberty interest protected by the Fourteenth Amendment (Robin-Vergeer). November 1991 • American Psychologist 1201 screening of pregnant women for illicit drug use and the potential removal of newborns from their mothers by the state. Besides constitutional issues, there are practical consequences of the threat of state intervention in the name of child protection. For example, because their drug use may be detected during prenatal examinations, homeless pregnant women who use street drugs are thought to avoid or delay prenatal care, potentially risking the health of the woman and the developing fetus (McKay, 1983; "Pregnant Addicts," 1990). Also, removal of the newborn child from the woman's custody may affect the family welfare income and further destabilize the family unit and weaken the mother's ability to take care of other children in the family. As a consequence, the mother may lose her welfare support and even her other children. In addition to constitutional and practical concerns, removing newborns or other dependent children from their mothers has broader social implications. States that have laws placing drug-exposed newborns under state custody often overwhelm the foster care system ("Preg- nant Addicts," 1990; "Public Health Problems," 1990; Robin-Vergeer, 1990; U.S. House of Representatives, 1989). Drug-exposed newborns who have been medically cleared for discharge may remain in the hospital for var- ious reasons, including eventual abandonment by their mothers, the lack of available and appropriate foster care placement, or delayed protective services evaluation, producing "boarder" babies in many hospitals (Robin- Vergeer, 1990; U.S. House of Representatives, 1989). In addition to disrupting the family, this is an inefficient use of state resources. If counties intervene on behalf of all infants with positive toxicological screens, resources will be spread so thinly that, inevitably, those infants whose lives truly need thorough investigation will receive only cursory attention. In short, removing infants from mothers who use drugs results in the "warehousing" of infants to the detriment of both mother and child (Robin- Vergeer).

Criminalization Versus Treatment of Drug Users There are two competing policy responses to women who use drugs during pregnancy. One is criminalization, in which drug use is seen as willful criminal behavior that prompts criminal prosecution or intervention by a child protection agency. The second response is treatment, in which drug use is seen as an illness or as a problematic hcalth-risk behavior that prompts health intervention ("Pregnant Addicts," 1990). In its popular "war on drugs," the Bush Administration has continued to place massive political and budgetary emphasis on law enforce- ment control of illicit drugs. In contrast, the administra- tion's policy has placed inadequate emphasis on preven- tion of and treatment for drug use. It has also failed to target dangerous legal drugs, including alcohol and to- bacco. Furthermore, the administration has not addressed the social context that helps engender drug problems ("Groups Decry Lack," 1991). Although the current administration emphasizes le- gal over medical solutions, legal intervention is not uni- form and appears to be applied primarily to poor women and women of color. For example, legal intervention often begins in medical institutions in which prenatal and ob- stetrical care can include tests for drug use. In this context, the threat of criminal prosecution of patients for drug use raises several legal, ethical, and practical questions.

Responses to a drug-exposed infant vary within states, and positive toxicological screens are not subject to man- dated reporting. Furthermore, protocols for administering toxicological screens on infants vary among hospitals:

Many hospitals do not use them at all, whereas public hospitals that serve large numbers of poor and minority women do tend to screen (Robin-Vergeer, 1990). In hos- pitals that do screen, the screening criteria themselves may contribute to disproportionate screening of poorer women and their children. For example, infants whose mothers have had no prenatal care are typically screened, overrepresenting homeless and other poor women, who are less likely to have received prenatal care (Robin-Ver- geer). Homeless women are at further risk, in that children with positive screens may be immediately removed if child welfare agencies expect difficulty in locating the mother after discharge (Robin-Vergeer). Within hospitals, treating physicians or nurses exercise discretion, potentially re- sulting in discriminatory decisions as to who will be screened (Robin-Vergeer). Legal reaction to a woman's drug use during preg- nancy is inconsistent within states and includes coercive interventions such as removal of newborns from their mothers; court-ordered detention of pregnant, drug-using women; and in several states, prosecution for injury or death of infants when the mother's drug use during preg- nancy was a contributing factor ("Pregnant Addicts," 1990; Robin-Vergeer, 1990). There is some evidence that women of color and low-income women may be dispro- portionately affected by punitive measures ("Policy Statement 9020," 1991; "Pregnant Addicts," 1990). As mentioned earlier, one consequence of the use of medical care in the service of drug criminalization is that women who might want medical care for themselves and their babies may not feel free to seek treatment because of the threat of criminal prosecution related to illicit drug use (Robin-Vergeer, 1990). Furthermore, the interven- tionist approach ignores more pervasive problems preg- nant women face, including inadequate access to prenatal care, insufficient and inadequate treatment programs for substance abuse, and the host of socioeconomic condi- tions that make it difficult for poor women to care for themselves and their fetuses during pregnancy (Robin- Vergeer). When considering criminalization of drug use by pregnant women, policymakers should consider the source of the appeal for this approach. For example, is a woman subject to prosecution because she puts a devel- oping fetus at risk or because she is using drugs, a highly stigmatized behavior? Protecting the fetus seems to be pretext for punishing illicit drug use. For example, preg- 1202 November 1991 • American Psychologist nant women may be prosecuted for use of illicit drugs but not for alcohol or cigarette use, both of which have demonstrated detrimental health effects on a developing fetus. Similarly, a woman is not prosecuted if she know- ingly puts the fetus at risk because of some other behavior, such as not eating well or driving a car without using a seat belt. Finally, second parties are not prosecuted for providing illicit drugs, such as cocaine, to pregnant women. Several groups have recommended decriminalizing drug use by pregnant women, in favor of treating the drug problem. In 1990, the governing council of the American Public Health Association (APHA) recom- mended viewing illicit drug use by pregnant women as a public health problem rather than a criminal justice problem ("Policy Statement 9020," 1991). They rec- ommended that no punitive measures be taken against pregnant women when no other illegal acts have been committed. They further encouraged development of outreach programs and services to meet specialized needs of women and their children, including drug treatment facilities and halfway houses to serve women who use illicit drugs, their newborns, and their other dependent children. In 1991, the APHA called for programs to es- tablish drug treatment as an alternative to incarceration.

Traditional Versus Tailored Treatment Programs for Homeless Women A third policy concern is the inadequacy of alcohol or other drug treatment programs to meet the specific needs of homeless women with children. Those interested in providing such treatment may need to tailor creative treatment interventions to these women and to intervene at multiple levels, creating in effect, family-centered drug treatment ("Policy Statement 9020," 1991; Weinreb & Bassuk, 1990). Types of support needed for homeless women with children include long-term supported hous- ing, financial support, health care, job skills training, compensatory education, and child care (McChesney, in press; Weinreb & Bassuk, 1990; Wittman, 1989). For women with alcohol or other drug problems, Weinreb and Bassuk (1990) recommended a treatment model that addresses both their homelessness and their alcohol or other drug use. The model includes long-term, structured, and supported drug-free residential programs that maintain family integrity. This model provides the stability and predictability of a drug-free environment and the comprehensive services necessary to address the multiple health, psychosocial, and financial difficulties common to some homeless families. Other models for delivery of tailored treatment to homeless women have been developed, including a recent research demonstra- tion designed to evaluate the effectiveness of a residential treatment program specifically for poly-drug-addicted homeless women with preschool children (Comfort et al., 1990). An alternative model might provide the option for women to have a voluntary but temporary separation from their children during detoxification or at other times in their treatment program. Summary To date, only a few studies of homeless women with chil- dren have been completed that report any findings on alcohol or other drug use. Some researchers have reported to me that they had been encouraged to exclude alcohol or other drug assessment from their studies, or that at the least they were encouraged to not report "damning" in- formation about alcohol or drug use by homeless mothers.

This lack of reliable empirical information is unfortunate, inasmuch as the most accurate and comprehensive de- piction of homeless women with children is essential to the design of adequate policies, programs, and services for them. Despite its limitations, the existing literature suggests that in many respects, homeless women with children are much like other poor women with children. However, alcohol and other drug use may be an important risk factor for homelessness, one that interferes with some women's ability to compete for scarce resources such as housing, employment, or services. Studies further suggest that homeless women with children who have alcohol or other drug problems require highly tailored and integrated services that address their homelessness, treatment needs, and family issues. Access to treatment is an essential ele- ment in service interventions for some women because alcohol or other drug use may jeopardize other health, financial, employment, or social interventions (Robert- son, in press-a; Weinreb & Bassuk, 1990). At the epidemiological level, research on homeless women with children is needed that specifically assesses the prevalence of alcohol and other drug abuse and its relation to mental health problems, the health of the chil- dren, treatment and other service utilization patterns, and the family's course of homelessness. On the treatment level, research demonstration projects are needed that examine innovative strategies for providing drug treat- ment to this population (Weinreb & Bassuk, 1990). At the policy level, choices have to be made with a clear understanding of the consequences of such choices, in- cluding whose needs get prioritized and at what costs to individual women, families, and the larger community.

REFERENCES Bassuk, E. L. (in press). Women and children without shelter: The char- acteristics of homeless families. In M. J. Robertson & M. Greenblatt (Eds.), Homelessness: The national perspective.

New York: Plenum. Bassuk, E. L., & Rosenberg, L. (1988). Why does family homelessness occur? A case control study.

American Journal of Public Health, 78, 783-788. Bullard, 1. D. (1983). Maternal/child nursing. In G. Bennett, C. Vourakis, & D. S. Wolff (Eds.), Substance abuse: Pharmacologic, developmental, and clinical perspectives (pp. 240-251). New York: Wiley. Burt, M. R., & Cohen, B. E. (1989, March 28).

Feeding the homeless: Does the prepared meals provision help?

(Report to Congress on the Prepared Meal Provision, Vol. 1, rev.). Washington, DC: Urban In- stitute. City of New York. (1986).

A one day "snapshot'" of homeless families at the Forbell Street Shelter and Martinique Hotel.

New York: Human Resources Administration, Office of Program Evaluation. Comfort, M., Shipley, T. E., White, K., Griffith, E. M., & Shandler, I. W. (1990). Family treatment for homeless alcohol/drug-addicted November 1991 • American Psychologist 1203 women and their preschool children. Alcoholism Treatment Quarterly, 7, 129-147. Dail, P. W. (1990). The psychosocial context of homeless mothers with young children: Program and policy implications. Child Welfare, 64, 291-307. Dolbeare, C. N. (1989). Out of reach: Why everyday people can "t find affordable housing. Washington, DC: Low Income Housing Infor- mation Service. Fagan, K. (1990, August 25). New community center to benefit parents, youth. Oakland Tribune, p. A-9. Fischer, P. J. (1989). Estimating the prevalence of alcohol, drug and mental health problems in the contemporary homeless population: A review of the literature. Contemporary Drug Problems, 16, 415- 452. Garrett, G. R. (1989). Alcohol problems and homelessness: History and research. Contemporary Drug Problems, 16, 301-332. Groups decry lack of emphasis on treatment in drug control. (1991, April). The Nation's Health, 21(4), 1, 23. Hopper, K. (1990). Deviance and dwelling space: Notes on the resettle- ment of homeless persons with drug and alcohol problems. Contem- porary Drug Problems. 16, 391-414. Inciardi, J. A. (1989). Trading sex for crack among juvenile drug users: A research note. Contemporary Drug Problems, 16, 689-700. Johnson, A. K., & Krueger, L. W. (1989). Toward a better understanding of homeless women. Social Work, 34, 537-540. Knickman, J. R., & Weitzman, B. C. (1989). A study of homeless families in New York City: Risk assessment models and strategies for prevention (Final report, Vol. 1). New York: Health Research Program of New York University. McChesney, K. Y. (in press). Homeless families: Four patterns of poverty. In M. J. Robertson & M. Greenblatt (Eds.), Homelessness: The na- tional perspective. New York: Plenum. McKay, S. R. (1983). Substance abuse during the childbearing years. In G. Bennett, C. Vourakis, & D. S. Wolff(Eds.), Substance abuse: Phar- macologic, developmental, and clinical perspectives (pp. 135-155). New York: Wiley. Milburn, N. G. (1988, July). Drug abuse: Is it a serious problem among homeless people? Paper presented at the meeting of the National As- sociation for Welfare Research and Statistics, Baltimore. Milburn, N. G., & Booth, J. A. (1989-1990). Sociodemographic, home- less state, and mental health characteristics ofwomen in shelters: Pre- liminary findings. Urban Research Review,, 12(2), 1--4. Mills, C., & Ota, H. (1989). Homeless women with minor children in the Detroit metropolitan area. Social Work, 34, 485--489. Molnar, J. M., Rath, W. R., & Klein, T. P. (1990). Constantly compro- mised: The impact of homelessness on children. Journal of Social Issues, 46(4), 109-124. Petitti, D. B., & Coleman, C. (1990). Cocaine and the risk of low birth weight. American Journal of Public Health, 80, 25-28. Policy Statement 9020: Illicit drug use by pregnant women. (1991). American Journal of Public Health, 81, 253. Pregnant addicts provoke legal, ethical issues. (1990, July-August). ADAMHA News, 16(4), 11-12. (Office of Substance Abuse Prevention Suppl.) Public health problems fuel surge in children in foster care. (1990, Jan- uary). The Nation's Health, pp. 1, 11. Ridien, S., Asamoah, Y., Edwards, H. G., & Zimmer, R. (1990). Outreach and engagement for homeless women at risk of alcoholism: Family treatment for homeless alcohol/drug-addicted women and their pre- school children. Alcoholism Treatment Quarterly, 7, 99-109. Robertson, M. J. (in press a). Homelessness: The national perspective. In M. J. Robertson & M. Greenblatt (Eds.), Homelessness: the national perspective. New York: Plenum. Robertson, M. J. (in press b). The prevalence of mental disorder among homeless persons. In R. Jahiel (Ed.), Homelessness: A prevention-ori- ented approach. Baltimore: Johns Hopkins Press. Robertson, M. J., & Cousinean, M. (1986). Health status and access to health services among the urban homeless. American Journal of Public Health, 76, 561-563. Robin-Vergeer, B. I. (1990). The problem of the drug-exposed newborn: A return to principled intervention. Stanford Law Review, 42, 745- 809. Rolfs, R. T., Goldberg, M., & Sharrar, R. G. (1990). Risk factors for syphilis: Cocaine use and prostitution. American Journal of Public Health, 80, 853-857. Sex-for-drugs trade pushes U.S. syphilis rates up. (1990, January). The Nation's Health, 20(1), 17. Shedlin, M. G. (1989). The health care of homeless mothers and children: Impact of a welfare hotel. New York City: Medical and Health Research Association of New York City. Solarz, A. L. (in press). To be young and homeless: Implications of homelessness for children. In M. J. Robertson & M. Greenblatt (Eds.), Homelessness: The national perspective. New York: Plenum. Somers, A. (in press). Domestic violence survivors. In M. J. Robertson & M. Greenblatt (Eds.), Homelessness: The nationalperspective. New York: Plenum. Study points up AIDS risks of crack, alcohol and other drugs. (1990). Alcoholism and Drug Abuse Week, 2(25), 1-3. Sullivan, P. A., & Damrosch, S. P. (1987). Homeless women and children. In R. D. Bingham, R. E. Green, & S. B. White (Eds.), The homeless in contemporary society (pp. 82-98). Newbury Park, CA: Sage. Survey finds 250,000 pregnant women need treatment. (1990). Alco- holism and Drug Abuse Week, 2(25), 4-5. Thomas, L., Kelly, M., & Cousineau, M. (1990). Alcoholism and sub- stance abuse. In P. W. Brickner, L. K. Scharer, B. A. Conanan, M. Savarese, & B. C. Scanlan (Eds.), Under the safety net: The health and social welfare of the homeless in the United States (pp. 204-214). New York: Norton. United States House of Representatives. (1989). Appendix VI: Survey. Addicted infants and their mothers. In No place to call home: Discarded children in America (A report together with additional and dissenting views of the Select Committee on Children, Youth, and Families; pp. 205-214). Washington, DC: U.S. Government Printing Office. Weinreb, L. E, & Bassuk, E. L. (1990). Substance abuse: A growing problem among homeless families. Family and Community Health, 13, 55-64. Weitzman, B. C., Knickman, J. R., & Shinn, M. (1990). Pathways to homelessness among New York City families. Journal of Social Issues, 46(4), 125-140. Weitzrnan, B. C., Shinn, M., & Knickman, J. R. (1989, October). Mental health problems as risk factors for homelessness. Paper presented at the meeting of the American Public Health Association, Chicago.

Wilcox, J. (Speaker). (1990, July 23). Children of crack \[Cassette re- cording\]. San Francisco: KCBS Radio. Williams, C. W. (1991). Child welfare services and homelessness: Issues in policy, philosophy, and programs. In J. H. Kryder-Coe, L. M. Sal- amon, & J. M. Molnar (Eds.), Homeless children and youth: A new American dilemma (pp. 285-299). New Brunswick, NJ: Transaction.

Wittman, E D. (1989). Housing models for alcohol programs serving homeless people. Contemporary Drug Problems, 16, 483-504. Wood, D., Schlossman, S., Hayashi, T., & Valdez, R. B. (1989). Over the brink: Homeless families in Los Angeles. Sacramento, CA: As- sembly Office of Research.

Wood, D., Valdez, R. B., Hayashi, T., & Shen, A. (1990). Homeless and housed families in Los Angeles: A study comparing demographic, economic, and family function characteristics. American Journal of Public Health, 80, 1049-1052. 1204 November 1991 • American Psychologist