The Impact of Foster Care and Temporary Assistance for Needy Families (TANF) on Women's Drug Treatment Outcomes

Abstract

Text

The purpose of this study was to assess how having a child in foster care, and/or receiving cash benefits through Temporary Assistance for Needy Families (TANF) impacted women's residential drug treatment completion rates. The study took place at a women's residential drug treatment center in the Midwest where drug treatment, child welfare, and TANF are provided through three different agencies who work independently of each other, and where there is little, if any, interagency collaboration.

Examining the impact of foster care and welfare on women's drug treatment completion rates seems especially relevant in light of the passage of the Adoption and Safe Families Act of 1997 (ASFA) and the Personal Work and Responsibility Act of 1996 (PRWORA). As service delivery systems, child welfare, TANF, and substance abuse services have separate goals, philosophies, legal mandates, and timelines. With the new ASFA timelines, low-income women in recovery are expected to completely overcome their addiction in a relatively short period of time (Azzi-Lessing & Olsen, 1996). While not empirically tested, these various timelines make it more difficult for mothers to work toward recovery. For example, accounts from pilot tests of women's residential drug treatment programs suggest that women be allowed to progress at their own rate and be offered services they could use. Personnel from some of the pilot projects suggested that they felt pressured to emphasize a rapid reentry into the workforce. Additionally, child welfare would often dictate the level and timing of parenting classes when women are in drug treatment. Requirements from both of these agencies put tremendous pressure on these pilot sites to rush women through drug treatment (CSAT, 2001).

Previous research has examined how women's substance abuse and drug treatment involvement has impacted outcomes that are desirable to child welfare and welfare agencies, but there has been less attention on how these services impact outcomes that are desirable to drug treatment programs; namely, ongoing recovery. Further, though the agencies often serve the same clients, there is a paucity of studies that examine the impact of child welfare, welfare, and drug treatment program within one study. To illustrate, a study in the child welfare arena examined how mothers' participation in drug treatment impacted the well-being of their children who were in foster care (Haight, 2005). Similarly, in the welfare arena, one study examined the relationship between TANF recipients' history of substance abuse and their employment functioning (Gorske, Larkby, Daley, Yenerall, & Morrow, 2006). While the well-being of children and having employment are certainly important, less is known about how having a child in foster care and TANF work requirements affect women's recovery process.

Several studies indicate that better drug recovery outcomes are achieved when women receive more services, in addition to drug treatment services (Ashley, Marsden, & Brady 2003; Smith, 2002; Sun, 2000). However, voluntary and involuntary services may differ in their overall benefit. Based on current policies, foster care services and receiving TANF cash benefits have consequences associated with noncompliance, and these consequences may mitigate the potential benefit of additional services (Beckman, 1994).

The study's hypothesis was that residential drug treatment completion rates for women who had a child in foster care and/or who received TANF cash benefits would differ from completion rates of women who did not receive these services. The study used a nondirectional hypothesis since this was an exploratory study. While most studies indicate that women who have their children with them in treatment were likely to complete treatment (Hughes, Coletti, Neri, Urmann, 1995; Knight, Logan, & Simpson, 2001; Nishimoto & Roberts, 2001; Stahl, Sicillian, & Anthony, 1995; Szuster, Rich, Chung, & Bisconer; Strantz, & Welch, 1995), at least one study suggests that women who had custody of their children and who had significant child care responsibilities were less likely to complete treatment (Knight, Logan, & Simpson, 1999). Further, much of the relevant TANF research is descriptive, and there is little research that assesses the impact of receiving TANF cash benefits on women's drug treatment outcomes (Gutman, Ketterling & McLellan 2003; Keesee & Williams, 1997; Montoya, Atkinson & Struse, 2001).

In child welfare, drug abuse is often viewed as the most serious problem (Bush & Sainz, 1997), and is a factor in three-fourths of all foster care placements (Day, Robison, & Sheikh, 1998). Substance abuse places women at risk for poor parenting and losing their children (Brady & Randall, 1999; Kettinger, 2000). Losing custody of their children due to drug use is a major threat for women in treatment (Nelson-Zlupko, Dore, Kauffman, & Kaltenbach, 1995). Women were willing to receive drug treatment when referred by child welfare agencies, even though they may perceive child welfare to be intrusive and unfair (Carten, 1996). Once referred, women must decide whether or not they will complete drug treatment, and the factors that impact women's drug treatment decisions are complex. Women's responsibilities for children and the related services they receive are often factors that are part of this decision-making process and may contribute to whether they complete treatment. One study found that women who had two or more children with them in treatment, or who reported a child welfare case open tended not to complete treatment (Knight, et al, 2001). On the other hand, another study indicated that women whose children were in foster care and who completed drug treatment were more likely to be reunified with their children (Smith, 2003).

Overall, there is a lack of current data to examine post-1996 TANF reform and the literature generally relies on self-report data. Based on nationally representative datasets, less than 20% of recipients use illegal drugs in a given year. Because of the lack of data, the impact of welfare reform on the subgroup of poor women who also may use drugs are largely unknown (Pollack, Danziger, Jayakody, & Seefeld, 2001). Two recent studies of African-American women who received TANF services found that they had a high level of childhood abuse and psychiatric impairment (Gorske, et al., 2006), and were more likely to engage in unsafe and illegal activities than women who continued to receive benefits (Lam, Wechsberg, & Zule, 2002).

One TANF study found that drug treatment completion was associated with maintaining employment (Metsch, McCoy, Miller, McAnany, & Pereyra, 1999). Since this study was conducted with out-of-treatment women, it provides little insight into how receiving TANF cash benefits affects women while they are in drug treatment. TANF recipients who have substance abuse concerns often have other physical and mental health disorders and face many complex barriers, such as domestic violence. A study of one state's TANF participants found that chemical dependency of the payee or payee's partner, mental health concerns, and child welfare involvement were the major challenges facing their TANF recipients (Keesee & Williams, 1997). In contrast, a more recent study found that there were no significant differences in psychosocial barriers between drug users and nonusers in a welfare-to-work program (Montoya, et al., 2001).

Additional research is needed on the effectiveness of drug treatment for women (Hanson, 2002), especially since many women are referred to drug treatment from child welfare or welfare agencies (Grella, 2006). Further, the inflexible timelines established by ASFA and PRWORA seem to be in juxtaposition to current research on the recovery process. Recovery for most individuals is a highly individualized process, and treatment is most effective when treatment providers respond to the uniqueness of each individual (CSAT, 2001; DiClemente, 2006). The strict timelines imposed by ASFA and PRWORA may work against treatment providers' ability to address each woman's individual situation. While this exploratory study does not address all the complexities found when women are served by more than one agency, it does seek to shed some light on how such a treatment environment might affect women's recovery. Specifically, this study seeks to contribute to our understanding of whether having a child in foster care or receiving cash benefits impacts women's drug treatment completion rates.

Methods

This study is part of a larger research project examining the impact of this multiple agency service environment on women's drug recovery outcomes over time. Project data were gathered using both retrospective data collected when women first entered residential treatment and prospective data collected on significant service events occurring during the study period. Data on women's drug treatment history, foster care, and TANF cash assistance, were collected using a Services Life History Calendar developed for this study. Secondary data from women's case records were used to gather information on whether or not women completed treatment, demographic data and scores on the Addiction Severity Index (ASI) (McLellan, Luborsky, O'Brien, & Woody, 1980). Both bivariate and multivariate statistics were used. Approval for the study was obtained from the Institutional Review Board for the Protection of Human Subjects (IRB) prior to any data being collected, including pilot study data.

Results

Most of the women had children, as approximately 90% or 105, were mothers. Of these 105 women, only 27, or 23% of the women brought their children with them. For a few of these women, the state had legal custody of their children, while they retained physical custody. Only 24 women (21%) had children in foster care. Some women were receiving the supportive child welfare and TANF services designed to preserve families and increase financial independence at the time they were admitted into residential treatment. Nineteen women (17%) were receiving either family preservation or informal supervision through the child welfare agency. Twenty-five women (21%) were accepted for TANF cash benefits, and another 8 had applied for benefits after entering treatment.

Overall, 92 of the 117 women (73%) had completed residential drug treatment, 21 (18%) had not completed treatment, and information for 4 women (3%) was missing, bringing the sample size to 113. It is difficult to compare this completion rate with rates reported in the literature, as these treatment completion rates were measured after a longer interval. For example, one residential treatment program for women reported that 27% of women left treatment prematurely, but was measured within the first 90 days (Knight, Hood, Logan, & Chatham, 1999). The 11 original grantees that piloted CSAT women's residential treatment programs reported an overall dropout rate of 38%, but this was also measured between 30 and 60 days of treatment or after 6 to 9 months in the residence (CSAT, 2001). Thus, given the shorter time period of 30 days to measure residential treatment completion, an overall rate of 18% of women who did not complete treatment may be comparable to that of other residential treatment programs reported in the literature.

The bivariate analysis indicated that women who received cash benefits or who had a child in foster care were less likely to complete treatment than women who did not receive these services. As indicated in Table 1, 38% of women who received cash benefits did not complete treatment and 62% completed treatment and these differences were statistically significant (χ² = 6.5, p < .05). Of women who did not receive cash benefits, 14% did not complete treatment and 86% completed treatment. However, women who had not received TANF at all during the year prior to entering treatment were more likely to complete treatment than women who were dropped from TANF, having received it for part of the year prior to entering treatment (91% and 65% respectively).

Although the results were not statistically significant, women with children in foster care were somewhat less likely to complete treatment than women without children in foster care placement. Of women with children in foster care placement, 25% did not complete treatment and 75% completed treatment. A larger percent of women with children who were not in foster care completed treatment, as only 17% did not complete treatment, compared to the 25% rate of women with children in foster care. It should be noted that mothers who brought their children with them into treatment had a similar completion rate to mothers whose children were in foster care (69% and 75% respectively). Most women without children or with children over 18 completed treatment (82%).

T-tests were used to compare women who completed treatment with those who did not complete treatment on their overall alcohol use/abuse, drug use/abuse, employment problems and psychological severity scores as measured by the ASI (Table 2). Again the results were not statistically significant, probably due to the small sample size. The results do suggest that women with higher psychiatric problems scores were less likely to complete treatment than women with lower scores. Alcohol use, drug use, and employment severity scores seemed not to affect likelihood of treatment completion. T-tests were also used to compare women with a child in foster care or who received cash benefits with those who didn't receive these services. Our findings indicated that there were no significant differences on the alcohol, drug, employment, or psychiatric severity scores between women receiving these services and those who did not receive these services.

For the Cox regression analyses, all records were used because receiving cash benefits and having a child or children in foster care, the variables of theoretical importance, are time-variant. While completion rates for all women were examined in the bivariate analysis, women without minor children were not included in the multivariate analysis shown in Table 3. The exclusion of women without minor children brings the sample to 100 women for that analysis only. Records for all 113 women were used for the multivariate analysis presented in Table 4.

As shown in Tables ​Tables33 and ​and4,4, Model 1 measures the effects of the control variables on the dependent variable. Model 2 combines the control variables with the independent variables of foster care and TANF cash benefits. Model 3 combines the control variables with the independent variables of problem severity scores. Model 4 is the most parsimonious model, retaining only the control variables significant in the prior models and the theoretically important independent variables.

In terms of the independent variables in Model 4, women with a child or children in foster care were less likely to complete residential treatment than other women with children under the age of 18. The risk of not completing treatment was 28% higher for women with a child or children in foster care than for other women with minor children. Model 4 also indicates that younger women were slightly less likely to complete treatment. For each previous drug treatment episode, the risk of not completing treatment increased by 10%. Women with higher drug and alcohol usage scores were more likely to complete treatment, however, than women with lower drug and alcohol usage scores. Women's employment problems score did not have significant effects. Having psychiatric problems greatly increased the risk that women would not complete treatment. Marital status, education, and race/ethnicity were not significant factors.

To separately determine the effects of TANF cash benefits, the analyses were also run for all 113 women and without the foster care variable, using 2713 records (Table 4). The findings indicated that women receiving cash benefits were less likely to complete treatment than women not receiving cash benefits. The risk of not completing treatment was 32% higher for women who received cash benefits than for women who did not. For women with minor children, having a child in foster care seems to be a greater deterrent to completing treatment than receiving benefits, as the TANF variable was not a significant predictor in the analysis including only mothers. The analysis including all women, however, suggests that receiving cash benefits also negatively impacts women's likelihood of completing treatment.

Discussion

These results support our hypothesis that residential drug treatment completion rates for women who have children in foster care will differ from those of women who received residential drug treatment only. As shown in model 4 of the first Cox regression analysis (Table 3), women with a child or children in foster care were less likely to complete treatment than other women with children under age 18 years of age. This finding is consistent with earlier research suggesting that women with fewer children remained in treatment longer (Ingersoll, Lu, & Haller, 1995). Even when taking alcohol, drug, employment, and psychological problems into account, women with children in foster care were less likely to complete treatment than other women in treatment. Women receiving TANF cash benefits were also less likely to complete treatment than women not receiving cash benefits (Table 4). These results also support our hypothesis that residential drug treatment completion rates for women who receive TANF cash benefits will differ from women who receive drug treatment only.

Women who have children must confront a plethora of complex factors that weigh upon their decision of whether or not to complete treatment. Because of the complexity of the issues, and the exploratory nature of this study, the reasons for these differences in treatment completion rates cannot be ascertained. For example, it is not known whether these differences were due to the individual services and their respective mandates, the way service providers interacted with each other and with the women and families they served, or whether characteristics of the women themselves could account for these different completion rates. While it is critical to study factors that predict treatment retention, ultimately there are too many factors that contribute to an individual's decision to remain in treatment, and the overall goal of being able to predict treatment compliance for any population is elusive (Kolden, Howard, Bankoff, Maling, & Martinovich, 1997).

Women whose children are in foster care may feel “pushed out” of treatment, when facing the hurdles and short time frame they have to complete child welfare requirements for reunification with their children. Similarly, women who receive cash benefits may feel an urgency to meet TANF requirements that may affect their decision to leave treatment. It is also possible that women who receive TANF may experience a “pull,” or a “check effect,” that precipitates their departure. Swarz, Hsieh, and Baumohl (2003) describe such a phenomenon with disability payments. In their study, participants tended to purchase illegal drugs close to the time they received their disability payment. In our study, it is not known whether TANF recipients experienced consequences beyond the three-month follow-up period as a result of not completing treatment.

Women with the highest completion rates were those who had not received any TANF benefits during the past year, as 91% of these women completed treatment (Table 1). Women who had been dropped from TANF during the twelve months prior to entering treatment had similar treatment completion rates as women who still received TANF (65% and 67% respectively), suggesting that these women may have more in common with each other than women who are not receiving TANF. Instead of being in foster care, children of women who had been dropped from TANF were primarily in the custody of their biological fathers, grandparents, or other relatives. Though the exact reasons that these children were placed in the custody of their father or relatives were beyond the scope of this paper, it is possible these women relinquished custody to focus on their own recovery.

Young and Gardner (1998) described the impact of ASFA and PRWORA on women in drug recovery as the dilemma of the “four clocks.” As a first clock, ASFA requires that children achieve permanency within twelve months of being placed in foster care. For the second clock, PRWORA mandates that women can only receive a total of five years of benefits. Third, drug recovery programs emphasize a one-day-at-a-time philosophy. As a fourth clock, women in treatment may feel the pressure of their children's developmental timetable, especially younger children, while they are in treatment.

Though each agency has different priorities, they all share a common goal of women's recovery. For child welfare, recovery means mothers can provide a safe and stable home for their children. For welfare agencies, recovery lays the groundwork for self-sufficiency and ultimate independence from receiving cash benefits. Given their shared purpose and the reality of the “four clocks,” it is imperative that child welfare, welfare, and drug treatment agencies work together toward their common goal. Findings from preliminary pilot studies suggest that such collaborative arrangements can benefit clients (Young & Gardner, 2002). Since first identifying the “four clocks,” Young and Gardner (2002) have added the concept of the “fifth clock,” to signify that time is running out on child welfare, welfare, and drug treatment agencies which fail to develop cooperative working agreements. In the absence of agency leadership to develop such collaborative programs, policy makers may step in and mandate change.

Although illicit drug use is rare among TANF recipients, most women who do use drugs have spent some time receiving cash assistance. Identifying TANF recipients with drug use disorders would have major implications for public health, whether or not women achieved economic self sufficiency. While drug use among TANF recipients is a concern, other barriers to self-sufficiency, such as depression, post-traumatic disorder, physical health problems, education, and transportation are even larger concerns (Pollack, et al. 2001). Again, the lower completion rates of TANF recipients suggest that these women face more barriers to self-sufficiency than women who do not receive TANF.

While these results suggest that receiving cash benefits under current legislative mandates do not necessarily foster drug treatment completion, sanctioning, or further restricting access to benefits may not be the answer. Additionally, as suggested by these preliminary findings, foster care placement does not appear to encourage women's recovery through treatment completion. Sanctioning and loss of custody of children can be viewed as punishment, which also has mixed results in fostering recovery. Rather than extinguishing behaviors, punishment may simply suppress them (Craighead, Craighead, & Illardi, 1995). As discussed previously, women who lose benefits may engage in risky illegal activities, at risk to themselves, their families, and society (Wechsberg, et al., 2002).

Though we can point to some trends and tendencies, research does suggest that recovery is highly individualized (DiClemente, 2006). In light of this, perhaps decisions on whether women should retain custody of their children and/or receive cash benefits should be based on their individual situation, instead of mandating a “one size fits all” approach through legislation. Building flexibility into child welfare, welfare, and drug treatment policies could encourage staff in all three agencies to fully engage women in their recovery process. Being fully engaged with your client, establishing rapport, and taking an individualized approach to recovery has been shown to be effective in fostering recovery (DiClemente, 2006). In contrast, the current system, with its legislative mandates may inadvertently encourage providers to be more passive and less engaged with these most vulnerable women because many of the important decisions have already been made.

In terms of limitations, the study could benefit from a longer follow-up period, and a larger sample size. Future studies should follow women for at least twelve months, which corresponds to child welfare's 12-month timeframe. These findings may not be generalizable to all women's drug treatment programs, especially when the program employs a different treatment approach. Other drug treatment programs may use more structured women-specific interventions, may have longer periods of residential treatment, or may operate within a state that uses a waiver for the Gramm amendment. While the study's methodology may limit generalizability, these findings do lend support to the benefits of interventions that apply women-specific interventions (Wechsberg, Sule, & Lann, 2004), and those that seek to increase coordination between these three service delivery systems (McLellan, et al., 2003).

Implications

These findings suggest several implications for social work practice and policy. Having mothers who are drug-free and actively working on their recovery program is in the interest of all three service delivery systems, and not just drug treatment agencies. Since women served by more than one agency do not seem to differ in the severity of their concerns, it is critical that we understand how having a child in foster care and receiving cash benefits impacts women who are in drug treatment.

Previous research suggests that substance-abusing women on welfare may be more likely to complete drug treatment when there are additional services working together to address their multiple concerns (McLellan, et al, 2003). Treatment protocols should seek to build on the positive relationships of parents and their children as a way to foster recovery (Collins, Grella, & Hser, 2003). While much progress has been made, more needs to be done to develop treatment programs for women. For example, it is possible that women with children may have different treatment needs than women who do not have children. Both groups of women could benefit from more individualized approaches to treatment.

Residential treatment for women should include a holistic approach, and include a broad range of psychosocial assessments (Comfort & Kaltenbach, 2000). A family-centered model and the woman-focused intervention are two approaches that offer a broad perspective and that research suggests may benefit women served by these agencies. The family-centered approach can improve reunification outcomes for children by addressing the treatment needs of their families, including substance abuse (Lewandowski & Pierce, 2004). A culturally-specific woman-focused intervention has been effective in increasing women's self-efficacy and improving drug treatment outcomes (Wechsberg, et al., 2004).

Policymakers should explore options for introducing flexibility into existing child welfare, welfare, and drug treatment policies. In building flexibility, these agencies should be encouraged to work together, and develop collaborative partnerships. Leadership is needed within substance abuse treatment, child welfare, and welfare agencies to develop systematic strategies to promote interagency collaboration. To be successful, agency leaders should promote role clarity, mutual respect, and having a unified purpose among the three agencies (Lewandowski & GlenMaye, 2002).

Finally, policy makers should increase their efforts to foster women's drug recovery, with a special emphasis on how ASFA and PRWORA may be impacting women's drug recovery process. Though ASFA may be successful in reducing time children spend in foster care, the policy does little to decrease the number of children entering foster care because of parental substance abuse. Policy makers are divided on how welfare reform impacts women and their children. Proponents of reform argue that requiring mothers to get jobs provides the most reliable pathway out of poverty, while opponents argue that reforms have stressed parents in ways that result in poorer child outcomes (Chase-Lansdale, Coley, Lohman, & Pittman, 2006). As noted by Metsch and Pollack (2005), PRWORA was enacted during a period of prosperity, and it is not known whether the new system will be as effective during more difficult economic times.

Directions for Future Research

Future research should focus on how the established legislative timelines affect women in recovery. Studies with a larger sample could shed more light on the relationship between the range of services women may receive and their drug recovery outcomes. With a larger sample, future research could also evaluate the interaction effects between the services women receive and individual women's characteristics, such as having employment and psychiatric problems.

In light of current mandates, research should especially focus on how losing custody of her children and/or her cash benefits affects women and their families as they progress through their recovery process. These dual losses, potentially a consequence of the “four clocks,” may have a profound effect on vulnerable families. Some states do require TANF participants to undergo drug testing, and sanction those who test positive (Howard, 2000). Though there were very few women in this sample who were actually sanctioned by TANF, further research is needed to assess the impact of sanctioning and pressures to find employment on women's drug recovery outcomes. Implementation studies should also be conducted to evaluate how the current state of the art in addiction research can continue to be incorporated into women's drug treatment programs.

Though these findings are preliminary, the study was conducted in a context where there was little collaboration between the drug treatment, child welfare, and welfare agencies. Given that ASFA and PRWORA remain in effect, further research is needed to assess how collaborative agency partnerships affect women's drug treatment completion rates. Such studies should also assess how these partnerships affect foster care placement decisions and the impact of cash benefits on women's recovery and their drug treatment outcomes.

In closing, it has been several years since Young and Gardner (2002) challenged agencies through the metaphor of the fifth clock, to develop collaborative interagency relationships. Meanwhile, mothers who are in drug treatment must conduct an elaborate juggling act to continue in recovery, keep their family together, and maintain some form of economic viability. For them, the clocks are still ticking, and time is, in fact, running out.

Acknowledgements

Footnotes

References