Serving the "hard-to-serve": the use of clinical knowledge in welfare reform.

This critical analysis of recent research and evaluations of welfare reform efforts describes how states have increasingly drawn on clinical knowledge in their efforts to move "hard-to-serve" recipients into the labor force. It argues that a clinical perspective is helpful as it brings attention to the mental health needs of low-income women. At the same time, however, this article suggests that states" use of a clinical framework is problematic in so far as it based on limited knowledge, dampens a broad discussion of the relationship between poverty and mental health, contributes to policy ambiguity, and increases recipient oversight.

**********

In 1996, the United States federal government passed the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA). PRWORA replaced Aid to Families with Dependent Children (AFDC), the federal entitlement program for single mothers and their children, with Temporary Aid to Needy Families (TANF), a time limited welfare-to-work benefits program. In its effort to reduce government "dependency" and promote economic "self-sufficiency," PRWORA gave states extensive flexibility in interpreting and implementing time limits and work participation requirements. Coupled with the unprecedented economic boom of the late 1990s, PRWORA measures contributed to a sweeping caseload decline. Although some states are currently experiencing a caseload rise, between 1996 and March, 2001 the TANF caseload fell 52% nationally (Savner, Strawn & Greenberg, 2002). In the face of this dramatic decline and upcoming TANF reauthorization in 2002, policy makers and researchers have devoted significant attention to recipients who remain on the rolls and have difficulty meeting work requirements. Currently referred to as the "hard-to-serve," many states have progressively looked to individual factors, including mental health problems, to account for recipients' persistent unemployment and welfare use.

This article focuses on the implications of welfare reform for the "hard-to-serve" population. In its critical review of recent research and initial evaluations of welfare reform efforts, this examination finds that two phenomena are occurring in several states. First, states increasingly frame difficulty with or failure to meet work requirements as potential clinical problems that mental health professionals must assess and treat. Secondly, "street level bureaucrats," or state and local administrators and line workers, as well as mental health professionals currently exercise growing discretion and oversight in determining clinical diagnosis and, ultimately, continued welfare eligibility. As we argue, these state-level developments hold mixed implications for women receiving welfare. A clinical lens is helpful as it brings attention to the previously ignored mental health needs of low-income women. Yet states' use of clinical knowledge is problematic in so far as it is based on incomplete information and dampens a broad discussion of the complex relationships among poverty, mental health, work, and welfare. Moreover, mental health diagnoses and interventions are not exact sciences and are subject to variation based on the characteristics of the clinician and the client as well as larger social and political influences. Taken together, worker discretion and a clinical framework reinforce one another and result in ambiguous policies that are inconsistently implemented and applied. Finally, we suggest that states' use of clinical diagnoses and interventions produces greater oversight of recipients. Again, this may benefit women and be welcomed by some, but it also adds an additional investigatory layer in which women must prove themselves worthy of financial assistance.

TANF and PRWORA

The PRWORA of 1996 established TANF, the work-based welfare program for low-income adults with children, that imposes a non-consecutive 60-month lifetime federal time limit on aid receipt. After two years of receiving assistance, recipients are required to participate in work or a work-related activity. The act set state-based work participation requirements at 30% in fiscal year (FY) 1997, 40% in FY 2000, and 50% in FY 2002 (Holcomb & Thompson, 2000). In the face of the 2002 TANF reauthorization, the Bush administration has proposed raising the work participation rate to 70% by 2007 ("Working Toward Independence," 2002). However, PRWORA allows states to exempt up to 20% of their caseloads from the time limits and grants states considerable flexibility in their implementation of work requirements. States can design their own welfare-to-work programs, decide who will be exempt from work requirements, and determine who will be exempt from time limits all together. They can also transfer block grant monies and maintenance of effort funds to pay for additional services to recipients, including mental health services (Derr, Douglas & Pavetti, 2001).

In light of time limits and work requirements, state policy makers became interested in the characteristics of those who remained on the rolls. Research findings increasingly suggest that women who remain on welfare despite a strong economy face multiple barriers to employment. While these barriers include traditional human capital variables, such as limited work histories and/or minimal employment skills, they also encompass factors such as poor health, substance abuse, histories of domestic violence, and mental health disabilities (Chandler & Meisel, 2000; Danziger et al., 2000; Danziger, Kalil & Anderson, 2000; Kalil et al. 1998; Polit, London & Martinez, 2001; Social Research Institute, 1999; Zedlewski & Anderson, 2001). With this seemingly new framework, states have classified many of the hard-to-serve as individuals with mental health disabilities and have paid growing attention to the relationship between mental health and welfare use.

Mental Health, Poverty, and Welfare Use

Psychiatric epidemiology has consistently shown an inverse relationship between psychiatric disorders and socioeconomic status (SES). Since the turn of the 20th Century, three generations of epidemiological research employing different definitions, sampling plans, and methods to measure mental illness have found persons with lower SES to have higher rates of psychiatric disorder (Bruce, Takeuchi & Leaf, 1991; Dohrenwend et al., 1992; Eaton & Muntaner, 1999; Regier et al., 1993). Most recently, the National Comorbidity Survey, a national survey, found that individuals in the lowest income group were almost twice as likely in a 12 month period and almost 1.5 times as likely within their lifetimes to develop a psychiatric disorder than individuals in the highest income group (Kessler et al., 1994). Importantly, high rates of psychiatric disorders among low-income women partially reflect gender differentials. Research indicates that women--due to a variety of psychological, biological, and environmental factors--are more likely than men to experience depression, Post Traumatic Stress Disorder (PTSD), and anxiety disorders (Belle, 1990; Bassuk, Buckner & Bassuk, 1998; Kessler, Foster & Saunders, 1995). Yet the inverse relationship between SES and psychiatric disorder also helps to explain these gender-based differences, as women are over-represented among the poor. In sum, this literature demonstrates that poverty is a definitive risk factor for psychiatric disorders. But it does not differentiate between low-income working women and women who receive welfare, nor the even more specific group of women on welfare who have difficulty meeting work requirements.

Research on the prevalence of mental health problems in the welfare population is a relatively new but popular development and has produced a wide array of estimates, ranging from 12% to 57% (Derr, Douglas & Pavetti, 2001; Lennon, Blome & English, 2001). These vast discrepancies among rates are in part attributable to differences in study instruments, samples, and diagnostic variables and also speak to the complexities and difficulties in measuring and defining mental health among low-income individuals. For example, in a national data set, researchers using the Composite International Diagnostic Interview (CIDI) found that 19% of welfare recipients met DSM-III-R diagnostic criteria for agoraphobia, generalized anxiety disorder (GAD), panic attacks, and/or major depression, which alone accounted for 12% of the disorders (Jayakody, Danziger & Pollack, forthcoming). However, in a study that used the same measures but sampled the TANF population in two California counties, researchers found rates of mental illness that ranged from 34% to 44% (Chandler & Meisel, 2000). In another national study that employed scales rather than standardized diagnostic criteria prevalence rates resembled the local study, as 35% of welfare recipients were found to have poor mental health including symptoms of anxiety, depression, and loss of emotional control (Zedlewski, 1999).

With welfare reform, emerging research has focused on comparative rates of psychiatric disorders between recipients who meet work requirements and those who do not. This body of literature is limited, but some initial reports do indicate that recipients who have a psychiatric disorder may be as much as 25% less likely to work than other recipients (Jayakody & Stauffer, 2000; Social Research Institute, 1999). One study found that women with major depression were significantly less likely to work than other recipients, yet psychiatric disabilities such as PTSD and GAD did not appear to affect employment chances (Danziger et al., 2000). In contrast, another study found no correlation between length of welfare receipt and mental health disorders, suggesting that long-term, hard-to-serve recipients' welfare use cannot be attributed to mental health issues (Chandler & Meisel, 2000). These preliminary findings begin to tease apart the complexities among women's poverty, welfare use, and mental health. Yet they raise more questions than they answer. While they point to a relationship among mental health, poverty, welfare use, and employment, the dynamics of these relationships remain unknown.

Some argue that a preexisting mental illness contributes to poverty, in so far as it limits employability. Others contend that poverty plays a causal role in the development of psychiatric conditions, most notably depression in women (Dohrenwend & Dohrenwend, 1969; Dohrenwend, 1990; Dohrenwend et al., 1992). Theorists seeking to explain this causal relationship often address: the role of poverty-related stress in the production of mental illness (Dohrenwend, 1990; Mechanic, 1975; Susser, Watson & Harper, 1985); the vulnerability of low-income individuals to physical and psychological disease given their lack of economic and social resources (Link & Phelan, 1995); and the impact of traumatic or cumulative negative life events on mental health (Turner & Lloyd, 1995). In spite of these causal hypotheses, many agree that the relationships between poverty and mental health have multi-directional components and that poverty and mental illness can form a mutually reinforcing cycle (Dohrenwend, 1990; Dohrenwend et al., 1992; Susser, Watson & Harper, 1985). For instance, in their discussion of depression and employment among low-income women, Lennon, Blome, and English (2001) suggest that depression may be a barrier to job attainment, but it may also affect job retention or constitute a response to job loss. Despite the gaps in knowledge and the complex nature of the relationships among mental health, poverty, and welfare use, the framing of these relationships is vulnerable to a reductionist clinical reasoning.

Clinical Reasoning in American Policymaking 

Scholars (Lyon-Callo, 2000; Navarro, 1985; Rosenheck, 1994; Stone, 1995) discuss the inappropriate application of clinical knowledge, or "clinical reasoning," that serves to skew, narrow, or incorrectly define social problems as individual mental health problems. According to Stone (1995), clinical reasoning has three primary attributes: (1) it is based on observations of individuals; (2) its techniques assess characteristics that supposedly only clinical--i.e., mental health or medical--professionals can evaluate; and (3) it promises objectivity through professional measurement and quantification. Clinical reasoning is politically attractive as it focuses on the individual, thus capitalizing on American ideology, and because it claims to be objective and therefore free of political interests (Stone, 1995). Its logic defines professionals as uniquely capable of providing objectivity and positions them as arbiters of clinical truth. Through its individualist orientation and claims of objectivity, clinical reasoning disguises the various political interests and conflicts that inform social problems and their solutions. It also deflects attention away from socio-structural inequalities and differential access to social and economic resources (Chambon & Irving, 1999; Foucault, 1999; Fraser, 1989; Lyon-Callo, 2000; Navarro, 1985; Rosenheck, 1994; Stone, 1995). Some strains in this line of thought tend to equate clinical knowledge and interventions with social control efforts and incorrectly negate the positive potential of clinical perspectives. Despite these limitations, theories of clinical reasoning are helpful in understanding the dangers of a reductionist clinical paradigm.

At a variety of historical junctures, issues of poverty and related social problems have been narrowed to the clinical mental health arena. Societal needs to manage and explain social problems, varying political and professional interests and practices, shifting race, class, gender, and generational relations, and perceived fiscal constraints have profoundly influenced who is deemed mentally ill and the type of treatment they receive (Foucault, 1973: Grob, 1972; Jones, 1999; Kunzel, 1993; Lunbeck, 1994; Magaro et. al, 1978; Rothman, 1971). In this context, individuals who are not mentally ill may be labeled so and the specific needs of mentally ill persons may be superceded by larger social, professional and/or political interests. A reading of mental health history provides several examples of such occurrences: the high rates of institutionalization of immigrants, the elderly, and the poor at different points in American history (Rothman, 1971); the explanation of mental illness as the root cause of homelessness during the 1980s (Rosenheck, 1994); state fiscal interests as a contributing factor in deinstitutionalization (Lerman, 1982); the use of psychiatric narratives to explain single motherhood and welfare use in the 1950s (Author; Solinger, 1992); and the popularization of child guidance in the early decades of the twentieth century (Jones, 1999). The contemporary response to welfare recipients who do not fulfill work requirements also contains elements of this phenomenon.

Clinical Knowledge and State Welfare-to-Work Policies

Initial reviews of TANF programs indicate that states have increasingly employed clinical knowledge in determining eligibility, work requirements, and benefits. To begin, many use clinical knowledge to account for and justify noncompliance with work activities. As of April 2000 32 states considered disability, generally including mental health disability, as "good cause" for noncompliance with work requirements, thereby protecting recipients from program sanctions (State Policy Documentation Project, 2001). In addition, as of 1998, 24 states exempted persons with disabilities from state program time limits altogether (Thompson, Holcomb, Loprest & Brennan, 1998). Notably, there is no standard definition of disability across states.

Reviews of selected state programs also show that some states rely on a variety of clinically-oriented strategies to define and identify mental health conditions among those who do not meet work requirements. The process of disability determination typically relies on clinical expertise. In their investigation of state welfare-to-work policies for people with disabilities, Thompson, Holcomb, Loprest and Brennan (1998) found that most states require some form of documentation from medical professionals to determine work readiness. Those with stricter participation requirements may subject this medical documentation to additional review by program staff, including internal and/or external medical review teams.

Identification of mental health disability typically occurs through self-report with supporting medical documentation (General Accounting Office, 2001). Interviewing and observation of client behavioral indicators by TANF staff, referrals to outside professional mental health assessment, and clinical inventories and questionnaires are also commonly employed assessment methods (Derr, Douglas & Pavetti, 2001; General Accounting Office, 2001; Johnson & Meckstroth, 1998; Pavetti et al., 1997; Thompson & Mikelson, 2001; Holcomb & Thompson, 2000). In Portland, Oregon, for example, all TANF applicants must participate in an "Addiction Awareness Class" where they complete the standardized Zung Depression Scale (Thompson & Mikelson, 2001) and in Florida outreach workers use a 32-item standardized measure that screen for substance abuse and mental health issues (Derr, Douglas & Pavetti, 2001). The multiple screening mechanisms further contribute to a variable definition of disability. States generally attempt to divert individuals with identified psychiatric disabilities to the Supplementary Security Income (SSI) program, which provides federally-funded assistance to persons with disabilities (Chandler & Meisel, 2000; Derr, Douglas & Pavetti, 2001; Holcomb & Thompson, 2000; Social Research Institute, 1999). But only the most disabled of these recipients will meet SSI eligibility criteria and many will remain on the TANF rolls for extended periods during the SSI application process. Given the concern with screening and assessment, states are considering devoting additional resources to the training of agency staff in order to better detect indicators of clinical mental health conditions (Derr, Douglas & Pavetti, 2001; Holcomb & Thompson, 2000; Thompson & Mikelson, 2001; Johnson & Meckstroth, 1998).

As advocates of mental health screenings suggest, enhanced assessment may be quite helpful to individuals who experience previously undetected disabilities that interfere with their ability to work or other areas of functioning (Holcomb & Thompson, 2000; American Psychological Association, 2001). Yet in their use of clinical knowledge and services, states largely equate non-work with psychiatric disability and construct continued employability as an indicator of psychological health. Here, clinical reasoning may be used as a primary means to define legitimate difficulty with or non-participation in TANF work programs, thus enabling government to partially shift the focus and responsibility of handling the larger social problem over to the clinical realm.

Complementing the popularity of clinical assessment tools, some states are drawing on clinical services to prepare hard-to-serve recipients for the labor market. Federal guidelines do not currently allow clinical activities to be counted as work and many states exempt persons with disabilities from work requirements altogether. Nevertheless some states permit certain types of clinical activities to partially fulfill state-mandated work requirements (General Accounting Office, 2001). Holcomb and Thompson's (2000) in-depth review of four state programs finds that TANF agencies have begun to develop service contracts with local mental health professionals and agencies. Although states report difficulties in service coordination, there also is indication of close coordination between TANF and mental health providers, and a few TANF agencies even co-locate with mental health providers (Derr, Douglass & Pavetti, 2001; Halcomb & Thompson, 2000; Johnson & Meckstroth, 1998). In other instances, TANF workers, whose traditional function primarily concerned eligibility determination, are being retrained to provide more intensive clinically oriented case management services. Some local TANF agencies have actually hired mental health professionals to conduct assessments and provide clinical services (Derr, Douglass & Pavetti, 2001). Other states continue to fund and/or refer recipients to mental health services as they near their time limits and even upon case closure. For example, Tennessee and Utah provide transitional mental health services for a year after case closing and other states automatically refer all clients nearing their time limits to mental health services (Derr, Douglas & Pavetti, 2001).

There is an underlying assumption that mental health services can cure mental health problems and, more specifically, move the hard-to-serve population into the labor force. Numerous studies have documented the effectiveness of a variety of psychotherapeutic and/or psychopharmological interventions for PTSD, GAD, and depression in particular (Beck, 1993; Charney et al, 1995; David et al., 1992; Elkin, 1994; Weissman, 1984). Psychotherapeutic interventions such as cognitive, cognitive-behavioral, and interpersonal therapies have been shown to reduce symptoms of depression, anxiety, and PTSD. Pharmacological interventions, or antidepressant medications such as tricyclic and selective serontonin reuptake inhibitors, have also proved to mitigate symptoms. Based on these positive outcomes, many advocate for the extension of clinical services to welfare recipients in the hopes that such intervention can mitigate mental health symptoms and help move women into the workforce (American Psychological Association, 2001; Holcomb & Thompson, 2000). These findings, however, may not be generalizable to the population of women receiving TANF for a variety of reasons. First, much of the research was not conducted on women or low-income persons or, more specifically, female recipients of cash assistance. Secondly, the study interventions were conducted in research settings that may not be easily replicated in the community. For example, psychotherapists in non-research settings may not follow strict intervention guidelines and prescription of certain medications may be restricted by type of insurance. Finally, although these studies have shown that certain therapeutic interventions decrease symptoms, very few have shown a positive relationship between therapeutic interventions and increased employability or ability to retain employment (Lennon, Blome & English, 2001). Although many agree that service should be extended to recipients with psychiatric disabilities who seek employment, it remains to be seen whether or not clinical interventions can truly support employment. In sum, states evidence a tendency to invoke clinical reasoning by equating non-work and poverty with psychiatric disability, even though the helpfulness of clinical services in reducing caseloads is open to debate.

Clinical Knowledge and Policy Outcomes

Worker Discretion

States' use of clinical knowledge also contributes to growing levels of worker discretion and ultimately produces ambiguous policies and policy outcomes. Lipsky (1980) provides a general framework detailing the use of discretion among public services workers. According to Lipsky (1980) "street-level" bureaucrats, or workers who determine eligibility to government programs and provide services within them, have considerable discretion due to three main factors: their work context is too complex to be completely defined and monitored; they must respond to human situations which require inherently subjective observation and judgment; and because the exercise of discretion reinforces the power of the worker and thus legitimacy of the services. Lipsky's framework reveals the potential for a significant amount of worker discretion, and thus variation, in the implementation of policies targeting the hard-to-serve. In light of the lack of a standard definition of disability, the use of multiple assessment tools, and limited knowledge concerning barriers to employability among hard-to-serve welfare recipients, there is significant room for welfare workers' decisions to be influenced by a host of individual, social, and institutional factors. These may include personal dispositions, understandings of psychiatric disability, values concerning work and government aid, feelings about recipients, interpretations of job duties, and levels of supervisory oversight. While welfare workers possess a significant amount of discretion in determining TANF eligibility, clinicians may possess even more.

As discussed, mental health professionals are increasingly involved in disability determination and may significantly influence program eligibility and the distribution of resources under welfare reform. In making recommendations regarding recipients' service plans and ongoing eligibility status, clinicians offer a form of "expert testimony" to TANF staff. For instance, in Utah, licensed social workers participate in review hearings for sanctioned clients and those nearing their time limits (Derr, Douglas, & Pavetti, 2001). The intrinsic nature of mental health assessment and diagnosis further enhances the considerable discretion of mental health clinicians.

Mental health assessment and diagnosis are not exact sciences and are influenced by a variety of forces. Kirk and Kutchins (1991) describe how the complexity of human experience--or what they refer to as fundamental factors--undermines a simplistic classification process. They argue that people possess a "bewildering array of personal and interpersonal troubles, and possess all manners of idiosyncratic personal histories ... a life-time of personal experiences, environmental stresses, and psychological confusion" that do not fit neatly into clearly defined categories (Kirk & Kutchins, 1991). In addition, those who are diagnosed are active participants in the process and thus influence its outcome. Likewise, characteristics of the clinician and the client along with external influences--or what Kirk and Kutchins refer to as extrapsychiatric factors--also shape the assessment process (Killian & Killian., 1990; Phelan and Link, 1996; Kirk & Kutchins, 1991). These factors include the discipline and training of the clinician; race, gender, ethnicity, and socioeconomic status of clinician and/or client; the insurance status, voluntary or involuntary status, and physical presentation of the client; the service setting; and the cultural distance between the clinician and client. The purposes of evaluation, institutional needs to regulate client flow and rationalize decision-making, clinicians' desire to protect patients from harm, and the larger political context may further shape psychiatric diagnoses (Killian & Killian., 1990; Kirk & Kutchins., 1991).

Both fundamental and extrapsychiatric factors are likely to play large roles in the conferring of diagnoses and thus program eligibility for the hard-to-serve TANF population. In addition to the varying personal characteristics of workers and clients, welfare workers' and clinician's judgments are subject to time limit pressures and the states' 20% exemption rate. Welfare workers and clinicians often may not have sufficient information about the complexity of factors in the lives of TANF recipients in order to fully inform clinical judgment. Workers may over-diagnose mental health disabilities as it protects clients by maintaining eligibility; yet they may also under-diagnose in order to control client flow and move as many recipients off the rolls.

Policy Ambiguity and Oversight

Importantly, worker discretion interacts with a variety of additional forces to produce high levels of ambiguity in TANF's clinically-oriented policies and greater oversight of recipients. Theorists (Lipsky, 1978; Matland, 1995; Pressman & Wildavsky, 1984) argue that a policy possesses high ambiguity when the technology needed to meet the policy's goals is not known and/or when the various actors in the implementation process are unclear about their roles. Given the lack of standard definition of disability, inconsistent prevalence rates of mental health disabilities, and the dearth of knowledge about effective interventions to assist in attaining and maintaining employment, effective technologies for moving hard-to-serve TANF recipients into the workforce are largely unknown. Moreover, welfare workers' roles are in the process of redefinition from basic eligibility worker to clinical case manager and assessment specialist, and clinicians must also shift the focus of their interventions to emphasize the goal of employment.

This high level of ambiguity directly affects policy implementation by adversely influencing the degree to which the policies are reliably understood and monitored (Lipsky, 1978; Matland, 1995). In other words, the ambiguous nature of these clinically-oriented policies produce "scientific" assessments and interventions that vary from state to state, program to program, and worker to worker. Ambiguity may also increase the power of different actors and importance of the environmental context in implementing policy (Berman & McLaughlin, 1978; Matland, 1995; Van Meter & Van Horn, 1975). The vagueness of these clinically-oriented policies creates opportunity for various stakeholders, such as state and local welfare agencies, clinicians, welfare entrepreneurs, and welfare advocacy groups, to fashion implementation to reflect their own interests. Contextual factors, including time limits, sanctions, local, state and national economies, may further influence implementation and contribute to inconsistencies in clinically-orientated interventions and diagnoses. For example, the current recession may pressure states to expand their definition of mental health disability in order to qualify recipients for continued eligibility. In protecting recipient eligibility, these clinically-oriented policies could mitigate some of the worst effects of welfare reform time limits. Yet these ambiguous policies may simultaneously result in unreliable diagnoses and untested interventions. They also create a new class of recipients who are defined as ill and in need of greater oversight.

The adoption of clinical reasoning and ambiguous state policies together contribute to greater state oversight of TANF recipients. Under AFDC, financial need and family structure were the primary determinates of eligibility and benefits. With TANF, however, individuals may now be subject to psychological assessments and examinations in their effort to secure and maintain financial assistance. States do not refer recipients to ongoing mental health treatment without their consent, but participation in treatment can become mandatory once it is included in an employment plan and those who fail to participate may be subject to sanctions (Derr, Douglass & Pavetti, 2001). More intensive case management and individualized service planning delve deeper into recipients' "private" lives than traditional eligibility determination. Mental health assessments, clinical services, and increased communication between service providers affect not only recipients' eligibility status, but other significant areas of recipients' lives as well. For instance, some (Lennon, Blome & English, 2001; Thompson & Mikelson, 2001) suggest that clinical assessment and intervention may lead to greater detection of potential problems in parenting and thereby produce more referrals to Child Protective Services. Relatedly, referrals by TANF staff to mental health assessment and services raise confidentiality issues. Mental health providers cannot speak with TANF staff without recipients' consent and release of information forms often detail the nature of the information to be disclosed (Derr, Douglass & Pavetti, 2001; Thompson & Mikelson, 2001). Nevertheless, clinicians possess significant discretion in interpreting the boundaries of the confidentiality agreement and may divulge privileged information to TANF staff. Greater state involvement may assist recipients by promoting attention and perhaps sensitivity to individual circumstance and mental health issues and expanding access to needed services. But given its tendency to produce greater recipient oversight and state involvement, TANF programs' use of clinical reasoning perpetuates what feminist theorists refer to as the "two-tiered" welfare state in which female beneficiaries of state assistance experience greater levels of state scrutiny and supervision than male recipients (Abramovitz, 1988; Fraser, 1989; Gordon, 1990; Mink, 1998).

Conclusion

The use of clinical knowledge in TANF programs holds multiple implications for the development of state policies and for TANF recipients themselves. This growing attention to mental health concerns may be helpful to low-income women who experience psychiatric disabilities. Yet states' adoption of a clinical framework also provides reason for pause. As this paper has argued, states have increasingly attributed ongoing welfare use to individual barriers such as mental illness. This narrowing perspective, however, may ignore the complex relationships between poverty, mental health and employment and fails to address the fundamental social problem of unequal access to and distribution of economic resources that informs welfare use and poverty. Moreover, little is known about the hard-to-serve population and the effectiveness of clinical interventions in moving recipients from welfare to work. The meager and contested state of our knowledge base suggests that it is naive and perhaps negligent to assume that clinical interventions can solve the problems of the hard-to-serve. Futhermore, the ambiguity accompanying these clinically-oriented policies makes them highly susceptible to political manipulation and subject to significant amounts of worker discretion. There seems to be no reliable method to predict how an individual will be "processed" as she goes through the system regardless of whether or not she truly has a mental health disability and individuals may be subject to a variety of treatments and classifications. Taken together, these forces culminate in a new class of poor women who states define as mentally ill yet potentially amenable to rehabilitation through higher levels of oversight and intervention.

Under the Bush administration's recent proposal for TANF reauthorization, states would be allowed to count rehabilitative services towards federal work requirements for a limited length of time ("Working Toward Independence," 2002). Advocates are calling for an even greater emphasis on mental health needs and have proposed federal requirements that would mandate states to provide mental health screening and psychological services (Polk, 2001). Before support can be lent to such policy recommendations or alternative recommendations can be fully developed, our current knowledge base about the effectiveness of mental health services on the sustained employment of long term welfare recipients and the effects of clinical knowledge at the implementation level needs expansion. If service provision does increase based on our current state of knowledge, it is questionable as to whether or not it will consistently serve the best interests of recipients or sufficiently tackle the issues of welfare use, unemployment, and poverty.

References

Abramovitz, M. (1988). Regulating the lives of women: Social welfare policy from colonial times to the present. Boston: South End Press.

American Psychological Association. (2001, November). Mental health issues in TANF reauthorization. Available: www.apa.org/pi/urban/tanfreg.html.

Bane, M. J. and Enwood, D. T: (1994). Welfare realities: From rhetoric to reform. Cambridge, MA: Harvard University Press.

Bassuk, E., Buckner, J. C., Perloff, J. N., & Bassuk, S. S. (1998). Prevalence of mental health and substance use disorders among homeless and low-income housed mothers. American Journal of Psychiatry, 155(11), 1561-1564.

Beck, A. T. (1993). Cognitive therapy: Past, present and future. Journal of Consulting and Clinical Psychology, 61, 194-198.

Bell, W. (1965). Aid to dependent children. NY: Columbia University Press.

Belle, D. (1990). Poverty and women's mental health. American Psychologist, 45, 385-389.

Bruce, M. L., Takeuchi, D. T., and Leaf, P. J. (1991). Poverty and psychiatric status: Longitudinal evidence from the New Haven epidemiological catchment area study. Archives of General Psychiatry, 48, 470-474.

Chandler, D. & Meisel, J. (2000). The CalWORKS project prevalence report: The prevalence of mental health, alcohol and other drug and domestic violence issues among CalWORKs participants in Kern and Stanislaus counties. Sacramento, CA: California Institute for Mental Health.

Charney, D. S., Miller, H. L., Licino, J., and Salomon, R. (1995). Treatment of depression. In A.F. Schatzberg and C.B. Nemeroff (Eds.), The American psychiatric press textbook of psychopharmocology. Washington DC: American Psychiatric Press.

Committee on Ways and Means, U.S. House of Representatives. (2000). Green book. Washington, DC: U.S. Government Printing Office.

Danziger, S., Corcoran, M., Danziger, S., Heflin, C., Kalial, A., Levine, J., Rosen, D., Seefeldt, K., Siefert, K., and Tolman, R. (2000). Barriers to the employment of welfare recipients. In In R. Cherry and W.M. Rodgers, III (Eds.), Prosperity for all? The economic boom and African Americans. NY: Russell Sage Foundation.

Danziger, S., Kalil, A. & Anderson, N. (2000). Human capital, physical health, and mental health of welfare recipients: Co-occurrence and correlates. Journal of Social Issues, 56(4), 635-654.

Derr, M., Douglas S. & Pavetti, L. (2001). Providing mental health services to TANF recipients: Program design choices and implementation challenges in four states. (Mathematica Policy Research Report No. 8736-403). Washington, DC: Mathematica Policy Research.

Dohrenwend, B.P. and Levav, I, Shrout, P.E., Schwartz, S., Naveh, G., Link, B. G., Skodol, A. E., and Stueve, A. (1992). Socioeconomic status and psychiatric disorders: The causation-selection issue. Science, 255(5047), 946-952.

Elkin, I. (1994). The NIMH treatment of depression collaborative research program: Where we began and where we are. In A.E. Bergin and S.L. Garfield (Eds.), Handbook of Psychotherapy and Behavior Change (pp. 114-136). NY: John Wiley and Sons.

Foucault, M. (1973). Madness and civilization: A history of insanity in the Age of Reason. NY: Vintage Books.

Foucault, M. (1999). Social work, social control, and normalization. In A. Chambon, A. Irving & L. Epstein, L. (Eds.), Reading Foucault for social work (pp. 83-97). NY: Columbia University Press.

Fraser, N. (1989). Women, welfare, and the politics of need interpretation. In N. Fraser (Ed.), Unruly practices: Power, discourse, and gender in contemporary social theory (pp. 144-160). Minneapolis: University of Minnesota Press.

Goldman, H. H. and Morrisey, J. P. (1985). The alchemy of mental health policy: Homelessness and the fourth cycle. American Journal of Public Health, 75(7), 727-731.

Goodman, L. (1991). The prevalence of abuse among homeless and housed poor mothers: A comparison study. American Journal of Orthopsychiatry, 61(4), 489-496.

Gordon, L. (Ed.). (1990). Women, the state, and welfare. Madison: The University of Wisconsin Press.

Grob, G. (1972). Mental Institutions in America: Social Policy to 1875. NY: Free Press.

Harris, K. M. (1996). Life after welfare: Women, work, and repeat dependency. American Sociological Review, 61,407-426.

Hien, D. & Bukszpan, C. (1999). Interpersonal violence in a "normal" low-income control group. Women and Health, 29(4), 14.

Holcomb, P., and Thompson, T. (2000). State welfare-to-work policies for people with disabilities: Implementation challenges and considerations. Washington D.C.: US DHHS.

Holzer, H. (1996). What employers want: Job prospects for less-educated workers. NY: Russell Sage Foundation.

Ingram, H. (1990). Implementation: A review and suggested framework. In N. & A. Wildavsky (Eds.). Public Administration: The State of the Discipline (pp. 462-480). Chatham, NJ: Chatham House Publishers.

Jayakody, R., Danziger, S., & Pollack, H. (forthcoming). Welfare reform, substance abuse, and mental health. Journal of Health Politics, Policy, and Law.

Jaykody, R. & Stauffer, D. (2000). Mental health problems among single mothers: Implications for work and welfare reform. Journal of Social Issues, 56(4), 617-634.

Johnson, A. & Meckstroth, A. (1998). Ancillary services to support welfare to work. Princeton, NJ: Mathematica Policy Research Inc.

Kalil, A., Corcoran, M., Danziger, S., Tolman, R., Seefeldt, K., Rosen, D. & Nam, Y. (1998). Getting jobs, keeping jobs, and earning a living wage: Can welfare reform work (Institute for Research on Poverty Discussion Paper No. 1170-98). Ann Arbor, MI: University of Michigan.

Kessler, R. C., McGonagle, K. A., Zhao, S., Nelson, C. B., Hughes, M., Eshleman, S., Wittchen, H.U., & Kendler, K.S. (1994). Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the United States: Results from the national comorbidity survey. Archives of General Psychiatry, 51, 8-19.

Kessler, R. C., Foster, C. L., and Saunders, W.B. (1995). Social consequences of psychiatric disorders. American Journal of Psychiatry, 152, 1026-1032.

Killian, T. M. & Killian, L. T. (1990). Sociological investigations: A review. Hospital and Community Psychiatry, 41(8), 902-911.

Kirk, S. A. & Kutchins, H. (1991). The selling of the DSM: The rhetoric of science in psychiatry. In J. Best (Ed.), Social problems and social issues: NY: Aldine De Gruyter.

Kingdon, J.W . (1995). Agendas, alternatives, and public policies. NY: Addison-Wesley Publishers.

Lennon, M. C., Blome, J., & English, K. (2001). Depression and low-income women: Challenges for TANF and welfare-to-work policies and programs. Research Forum on Children, Families and the New Federalism. New York: National Center for Children in Poverty, Mailman School of Public Health, Columbia University. Available: www.researchforum.org

Lerman, P. (1982). Deinstitutionalization and the welfare state. New Brunswick, NJ: Rutgers University Press.

Link, B. G. and Phelan, J. (1995). Social conditions as fundamental causes of disease. Journal of Health and Social Behavior, extra issue, 80-94.

Lipsky, M. (1978). Standing the study of public policy implementation on its head. In W. D. Burnham & M. Weinberg (Eds.) American politics and public policy. Cambridge, MA: MIT Press

Lipsky, M. (1980). Street-level bureaucracy: Dilemmas of the individual in public services. NY: Russell Sage Foundation.

Loring, M. & Powell, B. (1988). Gender, race, and DSM-III: A study of the objectivity of psychiatric diagnostic behavior. Journal of Health and Social Behavior, 29, 1-22.

Lunbeck, E. (1994). The psychiatric persuasion : Knowledge, gender, and power in modern America. Princeton, NJ: Princeton University Press.

Lyon-Callo, V. (2000). Medicalizing homelessness: The production of self-blame and self-governing within homeless shelters. Medical Anthropology Quarterly, 14(3), 338-45.

Magaro, P., Gripp, R., McDowell, D. & Miller, I. (1978). The mental health industry: A cultural phenomenon. NY: John Wiley & Sons.

Matland, R. E. (1995). Synthesizing the implementation literature: The ambiguity-conflict model of policy implementation. Journal of Public Administration Research and Theory, 5(2), 145-174.

Mechanic, D. (1975). Sociocultural and social-psychological factors affecting personal responses to psychological disorder. Journal of Health and Social Behavior, 16, 393-404.

Olson, K & Pavetti, L. (1996). Personal and family challenges to the successful transition from welfare to work. Washington, DC: The Urban Institute.

Pavetti, L., Olson, K., Nightingale, D., Duke, A. & Isaacs, J. (1997). Welfare-to-work options for families facing personal and family challenges: Rationale and program strategies. Washington, DC: The Urban Institute.

Phelan, J. C. & Link, B. G. (1996). The labeling theory of mental disorder (I): The role of social contingencies in the application of psychiatric labels. In A. V. Horwtiz and T.L. Scheid (Eds.), A handbook for the study of mental health: Social contexts, theories and systems (pp. 139-149). Cambridge, UK: Cambridge University Press.

Polit, D., London, A. & Martinez, J. (2001). The health of poor urban women: Findings from the project on devolution and urban change. NY: Manpower Demonstration Research Corporation.

Polk, C. (2001). Rep. Mink introduces fixes to 1996 welfare law. [National Organization of Women enews]. Available: www.now.org.

Pressman, J.L. & Wildavsky, A. (1984). Implementation. Berkeley: University of California Press 

Regier, D. A., Narrows, W. E., Rae, D. S., Manderscheid, R. W., Locke, B. Z. & Goodwin, F. K. (1993). The de facto U.S. mental and addictive disorders service system: Epidemiological catchment area prospective 1-year prevalence rates of disorders and services. Archives of General Psychiatry, 50, 85-94.

Rosenheck, R. (1994). Editorial: Homelessness in America. American Journal of Public Health, 84(12), 24-25.

Savner, S., Strawn, J., & Greenberg, M. (2002). TANF reauthorization: Opportunities to reduce poverty by improving employment outcomes. Washington DC: Center For Law and Social Policy.

State Policy Documentation Project (2001). Sanctions: Good cause for noncompliance [On-line report]. Available at www.spdp.org.

Stone, D. (1995). Clinical authority in the construction of citizenship. In H. Ingram & S. R. Smith (Eds.) Public policy for democracy. Washington D.C.: The Brookings Institution.

Stone, D. (1997). Policy paradox: The art of political decision making. NY: W.W. Norton and Company.

Thompson, S., Holcomb, P., Loprest, P., and Brennan, K. (1998). State welfare-to-work policies for people with disabilities: Changes since welfare reform (US DHHS Report No HHS-100-95-0021). Washington D.C.: US DHHS.

Thompson, S. & Mikelson, K. (2001). Screening and assessment in TANF/welfare-to-work: Ten important questions TANF agencies and their partners should consider (US DHHS Report No. HHS-100-99-0003). Washington, DC: US DHHS.

Turner, R. J. & Lloyd, D. (1995). Lifetimes traumas and mental health: The significance of cumulative adversity. Journal of Health and Social Behavior, 36, (December) 360-376.

Working toward independence. (2002). White House Press Office. Available: www.whitehouse.gov/news/releases.

Zedlewski, D. & Alderson, D. (2001). Do families on welfare in the post-TANF era differ from the pre-TANF counterparts?. (The Urban Institute Report No. 01-03). Washington, DC: The Urban Institute.

Rufina Lee

Columbia University

School of Social Work

838 President Street

Brooklyn, NY 11215

[e-mail].

Laura Curran

Rutgers University 

School of Social Work

536 George Street 

New Brunswick, NJ 08901

[e-mail]