The social and economic impact of sanctions and time limits on recipients of temporary assistance to needy families.

A central feature of the reforms enacted through the Personal Responsibility and Work Opportunity Reconciliation Act (welfare reform) has been the adoption of strategies to involuntarily remove Temporary Assistance to Needy Families (TANF) recipients from the welfare rolls, including increased use of sanctions and time limits on welfare receipt. Drawing on data from a three year panel study of women who had been receiving welfare in a state which adopted stringent sanctioning and time limit policies, we investigate predictors of recipients' TANF status after implementation of welfare reform, and identify differences in post-reform material resources, hardships and quality of life based on TANF status. Almost half of all welfare case closures during the first time period after reforms were implemented through involuntary strategies. Relatively few baseline characteristics predicted different outcomes once welfare time limits and sanctions were implemented. Those who were timed off welfare had substantially lower incomes in the year following their removal. One third of all respondents, regardless of reason for leaving TANF reported having insufficient food, housing problems and lack of access to needed medical care.

The passage of the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA, 1996) revamped federal welfare efforts to emphasize participation in the labor force as a primary strategy for reducing the dependence of single mothers and their children on public assistance. To amplify the consequences for failing to comply with new program requirements, Congress passed a mandatory time limit of 60 months for receipt of Temporary Assistance to Needy Families (TANF), and allowed states the option of imposing stricter sanctions on families that were not following through on mandated activities. As a result, states now have greater latitude to involuntarily remove TANF families from the welfare rolls, without regard to their social or economic circumstances.

Research on welfare caseloads has largely focused on identifying differences between welfare "leavers" and "stayers," but these studies have not differentiated between those who leave TANF voluntarily because they have obtained other income, and those who are removed from TANF involuntarily, either through time limits or sanctions. Descriptive information about the characteristics of sanctioned families suggests that they may possess certain demographic or human capital characteristics that may make them more vulnerable to involuntary removal. However, longitudinal research on this topic is limited. Even fewer studies have examined the longitudinal impact of involuntary leaving for TANF families' material resources, hardships, and quality of life.

Drawing on data from a three-year panel study of women who had been receiving welfare in Louisiana, a state that has adopted stringent sanctioning and time limit policies, we address the following questions:

1. What is the TANF status of study participants once welfare reform rules, including time limits and increased sanctions, have been implemented?

2. What baseline characteristics are predictive of later TANF status?

3. Are there differences in subsequent financial resources, hardships, and quality of life measures based on earlier TANF status?

Background

PRWORA replaced Aid to Families with Dependent Children (AFDC), a means-tested public assistance program created through the Social Security Act of 1935 to provide financial assistance to impoverished single mothers with children (Gordon, 1994). A central feature of the reforms enacted through PRWORA has been the adoption of strategies to involuntarily remove TANF recipients from the welfare rolls. These strategies were designed to serve both as an anticipatory incentive to engage in work efforts and as a punishment for non-compliance with welfare regulations (Corcoran, Danziger, Kalil & Seefeldt, 2000; Ferber & Storch, 1998). The most prominent of these efforts has been the creation of time limits and the imposition of stricter sanctions for women deemed to be non-compliant with various welfare rules.

Sanctions and Time Limits Policies

The federal government has devolved responsibility for developing welfare policy to the states, and as a result states vary widely in their applications of time limits and sanctions. Sanctions impose financial penalties on clients for failing to comply with administrative rules such as participating in mandatory work activities, pursuing child support enforcement, obtaining immunizations for children and providing required paperwork. Although time limits have received increased public attention as a new element introduced by PRWORA, more families are affected by sanctions than time limits; by one estimate, almost four times as many families will experience sanctions as time limits (Bloom & Winstead, 2002). Estimates of sanction rates range from 5 percent to 52 percent depending on the sampling methodology used (Pavetti, 2003).

Prior to 1996, many states were experimenting with reforms that were incorporated into PRWORA, including the use of a full family sanction, which allows the welfare agency to terminate benefits to an entire family for non-compliance. Under AFDC, when a parent was non-compliant with work activities, their portion of the public assistance grant was withheld, but the children's portion continued to be paid (Ferber & Storch, 1998). As of 2001, 36 states impose full family sanctions at some point in the process of deeming a client to be non-compliant, and eighteen of these states terminate all benefits immediately upon any instance of non-compliance (Center for Law and Social Policy, 2001). In seven states, noncompliance with work activities can lead to lifetime ineligibility for TANF benefits (Bloom & Winstead, 2002).

Under PRWORA rules, states are prohibited from using TANF dollars to provide cash assistance to families for longer than 60 months. States are free to continue to provide assistance to clients using their own resources after the 60 months have been surpassed, and ten states do so, including New York and California which comprise a significant portion of the nation's welfare caseload (Bloom, Farrell, Fink & Adams-Ciardullo, 2002). States also have the option of imposing shorter time limits than the federal maximum, which 17 states have opted to do (Bloom, Farrell, Fink & Adams-Ciardullo, 2002), the shortest time limit being found in Tennessee, which ends welfare payments after one year (Kim, 2000). As of 2002, 93,000 families have had their case closed because of time limits (Bloom, Farrell, Fink & Adams-Ciardullo, 2002).

Correlates and Outcomes for Involuntary Leavers

A handful of studies have identified correlates of sanctioning, but fewer have looked at which families are involuntarily removed from welfare because of time limits. Most studies note that mothers are more likely to be sanctioned if they have lower educational levels (Edeloch, Liu & Martin, 2000; Goldberg & Schott, 2000; Hasenfeld, Ghose & Larson, 2004; Kalil, Seefeldt & Wang, 2002; Westra, 2000). Contrasting findings are noted regarding age, with some finding that younger women are more likely to be sanctioned (Hasenfeld, Ghose & Larson, 2004; Kalil, Seefeldt & Wang, 2002) and others finding that sanctioned leavers are older (Lindhorst, Mancoske & Kemp, 2000). In some studies, being sanctioned is also associated with race and disability with African Americans more likely to be sanctioned (Hasenfeld, Ghose & Larson, 2004; Kalil, Seefeldt & Wang, 2002), and those who are disabled having a greater likelihood of sanctioning (Hasenfeld, Ghose & Larson, 2004).

Federal law does not require that states assess the impact of sanctions or time limits on families. As a result, only a handful of studies are available that investigate outcomes, and these tend to focus on reported income and employment. The purpose of sanctioning is to increase women's compliance with work mandates, yet, whether sanctioning acts as an incentive to employment is unclear. Survey research with welfare recipients indicates that sanctioned leavers have lower employment rates than other welfare leavers (Lee, Slack & Lewis, 2004; Moffitt & Roff, 2000); however, when using administrative data to compare state policies and rates of work among recipients, Kim (2000) found that the probability of working is higher for recipients who live in states that employ a full family sanction for non-compliance. It is also unclear whether material hardship differs among sanctioned and voluntary leavers. In some studies, being sanctioned is associated with greater risk of having utilities turned off (Kalil, Seefeldt & Wang, 2002; Lee, Slack & Lewis, 2004; Lindhorst, Mancoske & Kemp, 2000), experiencing food insecurity (Cherlin et al, 2001; Lee, Slack & Lewis, 2004; Lindhorst, Mancoske & Kemp, 2000) and having unmet medical needs (Lindhorst, Mancoske & Kemp, 2000). Other studies, though, have found that these hardships are common for all welfare leavers regardless of the reason they left (Bloom & Winstead, 2002).

Similarly, those who left welfare because of time limits also do not appear to have significantly higher levels of material hardship relative to voluntary leavers (Bloom, Farrell, Fink & Adams-Ciardullo, 2002). Time limits do not appear to have succeeded in encouraging work among recipients post-PRWORA (Bloom, Farrell, Fink & Adams-Ciardullo, 2002; Kim, 2000), likely because those who are more employable leave before time limits are imposed.

Background on Welfare Reform in Louisiana

Louisiana has a historically high level of poverty and welfare use. In 2000, Louisiana had the highest percentage of children in single mother families living below the poverty line (50 percent), and was 48th out of 50 states for the percentage of children living in poverty (30 percent). Louisiana is second to last among the states in the percentage of children whose parents do not have full-time, year round employment. Given these facts, it is not surprising that Louisiana has one of the highest percentages of children living in high risk environments in the country (Annie E. Casey Foundation, 2000).

Welfare reform was instituted in Louisiana through the creation of the Family Independence Temporary Assistance Program (FITAP). Recipients are required to spend 20 hours per week in approved work activities, except for women with children under one year of age. Louisiana elected to implement the Family Violence Option, which allows states to grant a temporary waiver of these program requirements to any person who is a verified victim of domestic violence. Louisiana is one of a handful of states that has opted to impose both full family sanctions, as well as a shorter time limit, timing recipients off welfare after 24 months of benefits (as compared to the five-year maximum required by Congress). Families must wait two years before reapplying after they have exceeded the limit (Louisiana Department of Social Services [LaDSS], 2003). While Louisiana had already imposed sanctions of the loss of individual benefits for any household head who failed to obtain work or work training within three months, beginning March 1, 1998, the whole family could be terminated from benefits (LaDSS, 1998b).

Since 1993, the welfare caseload has decreased 72.7 percent in Louisiana, placing it well above the national mean of 56 percent in its rate of reduction (U.S. Department of Health and Human Services [US DHHS], 2004). In 1999, when the first wave of recipients reached the 24-month time limit in the state, approximately 4,200 people stopped receiving welfare benefits (DeParle, 1999). Temporary exemptions from termination were given to another 2,000 people, mainly because of physical health problems of the mother, or her care for a disabled child (Finch, 1999). The state did not record any exemptions for reasons of domestic violence in the first wave of time limits. In the year after time limits were enacted (the second year of the present study), the welfare caseload in the state declined by 48 percent (US DHHS, 2004). Louisiana is one of five states which account for the largest number of families timed off welfare (Bloom, Farrell, Fink & Adams-Ciardullo, 2002).

Methods

From 1998 through 2001, a panel study of welfare recipients was conducted to evaluate outcomes related to implementation of PRWORA in Louisiana (McElveen, Mancoske & Lindhorst, 2000). To create the panel, a random sample, stratified to represent the rural and urban distribution of the state's welfare caseload, was created from a recipient listing provided by the state Department of Social Services (LaDSS). Subjects were eligible for participation if they were 18 years or older and receiving welfare payments as the guardian of a dependent child during the first year of the study in 1998. Child-only cases in which welfare payments did not include the adult recipient were excluded from the panel, as these cases were exempted from many of the new PRWORA regulations. Respondents were enrolled in the panel study prior to the implementation of time limits and other increased sanctions which were initiated in Louisiana in January 1999. In year one, a response rate of 72 percent was attained. At each wave of data collection, in-person interviews were completed by either a social worker or a Master's in Social Work student, and respondents were given small financial incentives for participation.

In the second year, 348 respondents were resurveyed, representing 61.1 percent of the original sample. In reports from federally funded studies of welfare leavers, re-interview response rates vary from 51 to 75 percent (Isaacs & Lyon, 2000), indicating that the response rate for this panel study is consistent with other longitudinal studies of welfare recipients. Higher attrition rates are not unusual in longitudinal research within low-income communities where mobility is high and access to telephones can be sporadic (Katz, E1-Mohandes, Johnson, Jarrett, Rose, & Cober, 2001). For the first set of analyses, we use data for the 348 women who were interviewed in year one and year two, using first year responses to construct predictors of welfare status in year two. For the second analyses, we use data for 277 women for whom responses from all three years are available to investigate consequences of involuntary welfare leaving.

Measurement

The dependent variable of TANF Status was determined by the client's self-report, first of whether they were currently receiving TANF payments, and if not, the primary reason that they were no longer receiving TANF Based on a list of case closure statuses provided by the LaDSS and interviewee responses, 3 additional categories were created. Voluntary leavers consisted of those persons who were dropped from the welfare rolls because they had received other income through work or marriage, who were no longer eligible because the child turned 18 or no longer lived with the respondent, or who voluntarily chose not to reapply for benefits. Timed off leavers were those who reported reaching the 24 month time limit and having their benefits ended. Sanctioned leavers included respondents who were involuntarily removed from the rolls because they did not meet the work requirements, did not cooperate with child support, missed an appointment, didn't know why their benefits ended, or believed the welfare office had made a mistake.

Human capital characteristics: To assess potential barriers to work and reasons for continued welfare use, we measured eight areas associated with these outcomes in previous studies. Education was a dichotomous measure of whether the respondent graduated from high school/obtained a Graduate Equivalency Diploma (GED) or not (0 = no; 1 = yes). Respondents were asked whether they had ever married (0 = never married; 1 = ever married); and whether they were currently employed, defined as self-report of any paid work outside the home, either full-time or part-time of at least 20 hours per week (0 = not employed; 1 = employed). Mothers were asked whether they or any child they cared for were currently disabled (0 = no disability; 1 = disabled). Measurement of domestic violence used the Epidemiological Survey of Intimate Partner Violence designed by the Louisiana Office of Public Health (Kohn, Flood, Chase & McMahon, 2000). Two screening questions asked respondents if they experienced physical violence (defined as having been hit, slapped, kicked, punched or beaten) or harassment (defined as being stalked or threatened with violence by someone known to the victim) in the past year. Two measures of longer-term poverty spells were also used. Recent poverty measures the proportion of time that a respondent reported she had received TANF in the previous five years (range = 0-5). Childhood poverty was measured by a proxy variable asking the respondent if, during her childhood, either parent had ever received welfare payments.

Results of leaving: To investigate possible outcomes of differing welfare statuses, we assess associations between TANF status and three areas in the third year: available resources, material hardships, and quality-of-life experiences. Available financial resource is a proxy for income that measures the total monthly amount the respondent reported receiving from any of the following sources: employment, TANF payments, Food Stamps, child support, or other financial resources such as Supplemental Security Income, Social Security Disability payments, etc. We also assess current employment as noted above, and the percentage of people that receive Food Stamps. We also measured resource limitations in the areas of food, housing and health in the past 3 months. Insufficient food is measured as the percentage of respondents who answered "yes" to any of three questions about food insecurity: went without food for a day or more because there wasn't any money; had to go to a food bank or a soup kitchen; or had to skip meals or eat less because there wasn't enough money (Carlson, Andrews, and Bickel, 1999). Housing problems were measured as the percentage of respondents who answered "yes" to any of six questions that asked whether in the past 3 months, the respondent had been unable to pay her rent, had been without any shelter, had the electricity turned off, had to move in with others, have others move in to help cover expenses, or had the phone turned off. Health hardships were defined as the percentage of people needing, but not receiving Medicaid, and being unable to obtain needed medical care for themselves or their children.

We also measured areas related to quality of life in the third year. A measure of health from the Health-Related Quality-of-Life Measure (Newschaffer, 1998) was used to assess general health status, where respondents noted the number of days in the past 30 days in which they were in poor health. Mental health was measured using the Center for Epidemiological Studies Depression scale (CES-D, Radloff, 1977). The CES-D is a self-report scale of depressive symptomotology widely used to identify possible depression in research samples. We categorized each respondent as one who would likely meet clinical criteria for a diagnosis of major depression if their score exceeded 22 on the scale (Measurement Excellence and Training Resource Information Center, 2004). In addition, we measured the lifetime incidence of domestic violence and sexual assault (0 = no violence; 1 = violence) in the third year using a modified version of the Louisiana epidemiological survey mentioned above. Since these questions assess lifetime prevalence rather than incidence post-sanctioning, they should be interpreted as associations with TANF status rather than as consequences.

Data Analysis

The first research question regarding TANF status was analyzed using descriptive statistics. In order to evaluate the second research question related to predictors of TANF status, a multinomial logistic regression equation was constructed since TANF status is composed of four non-orderable categories (Demaris, 1992). In this analysis, the status of remaining on TANF is compared to each of the three leaver statuses: leaving voluntarily, being timed off welfare, or being sanctioned. We present the odds ratio for each predictor variable which can be interpreted as the change in odds (greater than 1 = increased odds; less than 1 = decreased odds) of being in one of the leaver categories relative to those remaining on TANE Overall model fit is evaluated using the -2 log likelihood test (Pedhazur, 1997). The pseudo [R.sup.2] describes the proportion of variance explained by the independent variables, and its interpretation is similar to that of the [R.sup.2] in OLS regression. Both figures are reported at the end of the tables. The second research question regarding outcomes associated with TANF status is assessed using one way Analysis of Variance, with the post-hoc Bonferroni test (Castaneda, Levin & Dunham, 1993).

Results

Description of sample

Approximately 90 percent of the sample was African American, consistent with the demographics of the LaDSS caseload (LaDSS, 1998a). Just over half (53 percent) had received a high school diploma or GED, and 36.5 percent had been married at some point in their lives. Average age of respondents was 34.5 (SD = 12.28) at the start of the study, and they had 2.5 (SD=1.48) children on average, with 48.4 percent having a child under the age of five. One fifth of the respondents were either disabled themselves (19.4 percent) or caring for a disabled child (19.8 percent). Eight percent reported experiencing serious physical violence or harassment in the previous year. Almost eighty percent had received TANF payments for more than a year in the past five years (recent poverty), but only 33.2 percent had parents who received welfare payments when the respondent was a child. Thirty percent of recipients were working in the first year.

TANF status in the year after implementation of welfare reform

After enactment of sanctioning and time limit policies, 38.3 percent of respondents continued to receive TANF, 33.0 percent had left welfare voluntarily, 12.5 percent were timed off, and 16.2 percent were sanctioned off welfare for other reasons. Using the 62 percent (n = 213) whose TANF ended in the second year as the whole, 54 percent left TANF for voluntary reasons, usually because they obtained employment (29.7 percent) and their work income made them ineligible for benefits, or because they received other income (14.4 percent), most frequently SSI or Social Security Disability payments. Over 45 percent of the respondents had their TANF benefits discontinued involuntarily, with the majority of these (21.8%) being because the mother had reached the twenty-four month time limit imposed in 1999. Families experienced full family sanctions for not meeting work requirements (7.9 percent), missing appointments (4.5 percent) or not assisting with child support enforcement (1.9 percent). Although the welfare office noted that cases could be closed if children were not immunized, didn't meet school attendance requirements, or if the parent was no longer eligible because of citizenship status, no respondent reported these as primary reasons for their involuntary removal.

Factors predicting TANF status

The next analysis assesses the role of demographic and human capital variables in predicting the likelihood of achieving a particular TANF status. Odds ratios are presented for the statuses of "Voluntary leaver," "Timed off," and "Sanctioned" relative to those still on TANF (see Table 1).

Overall, relatively few characteristics differentiated any of the leaver statuses from those who remained on TANF. Voluntary leavers differed from those who remained on TANF in that they were less likely to have children under 5 years of age (OR = .55) and were more likely to have been working at baseline (OR = 3.24). Those who were timed off TANF were also less likely to have young children (OR = .52), were more likely to be working at baseline (OR = 3.15) and were less likely to have a disabled child (OR =.23). Respondents who were sanctioned off TANF were not significantly different from those remaining on TANF on any of the characteristics. The overall model is significant on the goodness-of-fit [chi square], with the full model explaining approximately eight percent of the variance in TANF status.

Effects of TANF Status on Later Outcomes

Differences between TANF statuses exist in fewer than half of the categories related to resources, material hardships and overall health and mental health, indicating that the groups are more alike than different. Significant differences appear in financial resources, receipt of Food Stamps, and medical care hardships. Monthly financial resources differ significantly across the four groups. Timed off leavers have the lowest monthly financial resources of the four groups, even though a sizeable proportion is working. Recipients were working in all groups, with voluntary leavers reporting the highest work level of 50 percent. Declining Food Stamps use among voluntary leavers who are working might be anticipated since increased income limits Food Stamps eligibility. However, given that the incomes of voluntary leavers are lower than TANF recipients the likelihood is that many of these families remain eligible for Food Stamp participation, whether they receive Food Stamps or not.

Material hardships differ among the three groups in two categories related to medical resources. Twenty-nine percent of timed off leavers and almost one quarter of the voluntary leavers were unable to obtain medical care that they needed, a significant difference when compared to those continuing to receive TANF. Almost three times as many of the voluntary, timed off and sanctioned leavers reported that they needed Medicaid, but were unable to obtain it. Markedly smaller percentages reported that they were unable to access needed medical care for a child, likely reflecting the availability of special child insurance programs. Almost one-third of the families across all TANF statuses reported food insecurity, but the differences between the groups were not significant. Thirty-one to 44.4 percent of respondents also reported serious housing problems, such as not being able to pay rent, or having the electricity or phone disconnected, but again, TANF status was not associated with these hardships.

TANF status approaches significance in its association with having ever experienced domestic violence. Almost twice as many timed off and sanctioned leavers reporting having ever experienced domestic violence in their lifetime as compared to voluntary leavers. When these two groups are collapsed into "involuntary leavers," they report significantly higher levels of domestic violence than do voluntary leavers (F=3.74, p<.05). While no significant differences were found between the TANF status groups in regards to their report of health or their level of depression, between 31 and 42 percent of the women endorse feelings consistent with a diagnosis of major depression. A relatively high number of women across each status also reported having experienced sexual assault in their lifetime, particularly among those still receiving TANF benefits.

Discussion

In this study, almost half of respondents noted that their cases were closed for involuntary reasons, contradicting the general public's view that women leave welfare because they find work (DeParle, 2004). Despite the emphasis in welfare reform policies on helping families to achieve economic self-sufficiency through work, less than one third of case closures in this study were as a result of achieving employment, and the majority of involuntary leavers were not working. Louisiana has been successful in implementing time limits, as this accounts for the highest percentage of involuntary leavers. These findings underscore econometric data indicating that time limit policies in particular are responsible for a significant portion of the decrease in welfare caseloads since their implementation (Grogger 2002; Fang & Keane, 2004).

A minority of the welfare leavers in this survey (29.7 percent) actually exited welfare because they had obtained employment. During the last year of the survey, 50 percent of the voluntary leavers, 45 percent of the sanctioned leavers, 35 percent of TANF recipients, and 33 percent of those timed off welfare were working. If increased sanctions and time limits were acting as an incentive to work, one would expect to see higher rates of employment among involuntary leavers, given their lack of access to public financial assistance. Since a minority of women who have been sanctioned or timed off welfare were working, the policies do not appear to be successful in significantly increasing work force participation. This is consistent with earlier findings that sanctions and time limits do not act as significant incentives to improve employment outcomes (Moffitt & Roff, 2000; Pavetti, 2003).

Perhaps the most striking in these analyses are the similarities among TANF recipients, voluntary leavers and those timed or sanctioned off welfare. Very few demographic or human capital characteristics differentiated these groups. Voluntary leavers and those who were timed off welfare had similar baseline characteristics, as did respondents who remained on TANF or who reported being sanctioned off. These findings suggest that other structural rather than individual factors may be more relevant to understanding why some people leave via sanctions and time limits and some are able to obtain employment, gain other income or leave for other voluntary reasons.

Regardless of TANF status, a significant number of respondents reported serious material hardships such as having insufficient food, serious housing problems or an inability to obtain necessary medical care for the recipient (but not the child). It appears that even families leaving voluntarily left for situations that did not substantially improve their family's social or economic circumstances. Although differences between leavers and TANF recipients were not seen in their reports of food and housing hardships, one third of respondents reported these problems, with timed off and sanctioned leavers reporting the highest levels. Other studies indicate that between 25 and 33 percent of all leavers experience food insecurity (Isaacs & Lyon, 2000; Loprest & Zedlewski, 1999), so the families in Louisiana do not differ markedly in this regard from other TANF families.

Time limits are associated with decreased monthly financial resources. As noted in other research (Pavetti, 2003), the monthly financial resources of women involuntarily removed from TANF are significantly less than either voluntary leavers or TANF recipients. Using the monthly financial resources figure as a rough proxy for actual yearly income, timed off leavers report receiving $6,608 yearly, an amount that is 50.3 percent of the poverty threshold of $13,133 for a parent with two children set by the U.S. Census Bureau for this time period (U.S. Census Bureau, 1998). Since women who are timed off TANF report the lowest monthly income and the lowest employment rates, it appears that being involuntarily removed from welfare leads to increased economic stress, which is not resolved through greater labor force participation.

Sanctioned leavers differed significantly from TANF recipients in terms of their ability to access medical resources. In this regard, they shared with voluntary leavers an inability to obtain Medicaid, and, likely as a consequence, had difficulty in obtaining medical care they needed for themselves. Although Medicaid services were administratively delinked from TANF receipt (Health Resources Services Administration, 2003), these results suggest that when TANF benefits are ended, a sizeable minority of recipients who might be income eligible for Medicaid no longer receive it.

Although having been a recent victim of domestic violence did not predict TANF status, differences in the lifetime prevalence of abuse indicate that it may be associated with sanctioning and time limits. Almost twice as many timed off and sanctioned leavers report a history of domestic violence compared to voluntary leavers, and almost one third of those still on TANF report having experienced domestic violence in their lifetimes. These data suggest that abuse may have a cumulative effect making women more vulnerable to sanctioning, rather than a strictly proximal one.

Conclusions

This study provides a glimpse into the experiences of families receiving TANF in a single state, during the beginning period of implementation of welfare reform regulations, within a more stringent policy regime than is the case in many states. Although these results cannot be generalized to the national level because of the differences in policy and economic environments among the states, these results do raise certain cautions. First, while it is possible to achieve successful caseload reduction through the implementation of time limits and sanctions, this does not necessarily translate into increased economic self-sufficiency on the part of families who are the targets of these policies. In fact, families that are timed off the welfare rolls may instead be experiencing even deeper poverty and deprivation.

Second, time limits and sanctions do not appear to be significant motivators for employment. This may in part be related to the fact that the labor market environment for most welfare recipients consists of unstable, low-paying, geographically inaccessible positions without benefits that do not significantly improve social or economic circumstances of these families. Effective services need to build on the strengths of families and to address the many structural as well as human capital barriers that hinder them from achieving financial independence through employment rather than individualistic attempts to shape complex family outcomes. TANF policymakers can either view sanctions and time limits as guides pointing them to the families that are most in need of supportive interventions, or these strategies can be used as ways to easily exit challenging clients. If welfare reform is to be truly labeled a success, it cannot be at the expense of those who are the most vulnerable and least able to access necessary resources.

References

Annie E. Casey Foundation. (2004). Kids count data book: State profiles of child well-being. [On-line]. Available: http://www.aecf.org/kidscount/kc2004/.

Bloom, D., Farrell, M., Fink, B., & Adams-Ciardullo, D. (2002). Welfare time limits: State policies, implementation and effects on families. [Online.] Available at: http://aspe.hhs.gov/hsp/welf-time-limits02/htm.

Bloom, D. & Winstead, D. (2002). Sanctions and welfare reform. Brookings Institute Policy Brief No. 12. Washington, DC: Brookings Institute.

Carlson, S. J., Andrews, M. S., & Bickel, G. W. (1999). Measuring food insecurity and hunger in the United States: Development of a national benchmark measure and prevalence estimates. Journal of Nutrition, 129 (2S Supplement), 510-516.

Castaneda, M. B., Levin, J. R., & Dunham, R. B. (1993). Using planned comparisons in management research: A case for the Bonferroni procedure. Journal of Management, 19 (3), 707-724.

Center for Law and Social Policy. (2001). Sanctions for noncompliance with work activities. [Online.] Available at: www.spdp.org/tan/sanctions/sanctions_ findings.htm.

Cherlin, A., Burton, L., Francis, J., Henrici, J., Lein, L., Quane, J., & Bogen, K. (2001). Sanctions and case closings for noncompliance: Who is affected and why. Policy Brief 01-1. [Online.] Available at: www.jhu.edu/~welfare/ 18058_Welfare_Policy_Brief.pdf

Corcoran, M., Danziger, S. K., Kalil, A. & Seefeldt, K. S. (2000). How welfare reform is affecting women's work. Annual Review of Sociology, 26, 241-269.

Demaris, A. (1992). Logit modeling: Practical applications. Newbury Park, CA: Sage Publications.

DeParle, J. (1999). As benefits expire, the experts worry." New York Times, October 10, Section 1, p. 1.

Deparle, J. (2004). American dream: Three women, ten kids, and a nation's drive to end welfare. New York: Viking.

Fang, H. & Keane, M. P. (2004). Assessing the impact of welfare reform on single mothers. Brookings Papers on Economic Activity, 1, 1-115.

Ferber, J. & Storch, R. (1998). Full-family sanctions: Not work the risk. [On-line]. Available: http://www.welfarelaw.org/family_sanctions.html.

Finch, S. (1999). Welfare reform works for many in LA." Times-Picayune/States-Item, January 10, Al.

Goldberg, H. & Schott, L. (2000). A compliance-oriented approach to sanctions in state and county TANF programs. Washington, DC: Center on Budget and Policy Priorities.

Gordon, L. 1994. Pitied but not entitled: Single mothers and the history of welfare. Cambridge, MA: Harvard University Press.

Grogger, J. (2002). The behavioral effects of welfare time limits. American Economic Review 92(2), 385-89.

Hasenfeld, Y., Ghose, T., & Larson, K. (2004). The logic of sanctioning welfare recipients: An empirical assessment. Social Service Review, 78 (2): 304-319.

Health Resources Services Administration. (2003). Louisiana Medicaid & S-Chip eligibility. [Online.] Available at: http://www.hrsa.gov/tpr/states/Loui siana-Eligibility.htm.

Isaacs, J. B., & Lyon, M. R. (2000). A cross-state examination of families leaving welfare: Findings from the ASPE-funded leavers studies. Scottsdale, AR: Paper presented at the National Association for Welfare Research and Statistics 40th Annual Workshop.

Kalil, A., Seefeldt, K. S. & Wang, H. (2002). Sanctions and material hardship under TANF. Social Service Review, 76 (4), 642-662.

Katz, K. S., El-Mohandes, A., Johnson, D. M., Jarrett, M. Rose, A., & Cober, M. (2001). Retention of low income mothers in a parenting intervention study. Journal of Community Health, 26 (3), 203-218.

Kim, R. Y. (2000). Factors associated with employment status of parents receiving Temporary Assistance for Needy Families. Social Work Research, 24(4), 211-222.

Kohn, M., Flood, H., Chase, J., & McMahon, P.M. (2000). Prevalence and health consequences of stalking--Louisiana, 1998-1999. Morbidity and Mortality Weekly Report, 49(29), 653-655.

Lee, B. J., Slack, K. S. & Lewis, D. A. (2004). Are welfare sanctions working as intended? Welfare receipt, work activity, and material hardship among TANF recipient families. Social Service Review, 78(3), 370-403.

Lindhorst, T., Mancoske, R. J. & Kemp, A. A. (2000). Is welfare reform working? A study of the effects of sanctions on families receiving Temporary Assistance to Needy Families. Journal of Sociology and Social Welfare, 24 (4), 185-201.

Loprest, P. J., & Zedlewski, S. R. (1999). Current and former welfare recipients: How do they differ? Washington DC: The Urban Institute's Discussion Paper on Assessing the New Federalism.

Louisiana Department of Social Services. (1998a). The facts about welfare and food stamps in Louisiana. Baton Rouge, LA.

Louisiana Department of Social Services. (1998b). Family Independence and Temporary Assistance Program (FITAP). [Online] Available at: http://www dss.state.la.us/departments/ofs/Family_dndependence_Temporary_ .html.

Louisiana Department of Social Services. (2003). TANF state plan. Baton Rouge, LA.

McElveen, J. Mancoske, R.J. & Lindhorst, T. (2000). The impact of welfare reform on Louisiana's families: Year two panel study report. Report to the Louisiana Department of Social Services. New Orleans, LA: Southern University at New Orleans.

Measurement Excellence and Training Resource Information Center. (2004). Critical review of Center for Epidemiologic Studies Depression Scale (CES-D). [Online.] Available at: http://www.measurementexperts.org// instrument/instrument_eviews.asp?detail=12

Moffitt, R. & Roff, J. (2000). The diversity of welfare leavers. Welfare, Children, and Families: A Three-City Study Working Paper, no. 00-01. Baltimore, MD: Johns Hopkins University.

Newschaffer, C. J. (1998). Validation of Behavioral Risk Factor Surveillance System (BRFSS) health related quality of life measures in a statewide sample. Washington, DC: U.S. Department of Health and Human Services, Public Health Service.

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

Pavetti, L. (2003). Review of sanction policies and research studies. [Online.] Available at: http://www.aspe.hhs.gov/hsp/TANF-Sanctions03.

Pedhazur, E. (1997). Multiple regression in behavioral research: Explanation and prediction. 3rd Ed. New York, NY: Harcourt Brace College Publishers.

Personal Responsibility and Work Opportunity Act (PRWORA). (1996). Public Law 104-193. Congressional Record, p. H7796.

Radloff, L.S. (1997). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement. 1(3), 85-401.

U.S. Census Bureau. (1998). Poverty thresholds: 1998. [On-line]. Available: http:// www.census.gov/hhes/poverty/threshold.html.

U.S. Department of Health and Human Services. (1999). Temporary Assistance for Needy Families (TANF) Program: Second annual report to Congress. Washington, D.C.: U. S. Department of Health and Human Services, Administration for Children and Families.

U. S. Department of Health and Human Services. (2004). Welfare caseload reductions. [On-line]. Available: http://www.acf.dhhs.gov/news/stats/caseload .htm.

Westra, K. L. (2000). Arizona cash assistance exit study. Phoenix, AZ: Arizona Department of Economic Security

Note

Funding for this research comes from the Louisiana Department of Social Services, Office of Family Supports. All opinions expressed and any errors are those of the researchers and not those of the Department.

TARYN LINDHORST

School of Social Work

University of Washington

RONALD J. MANCOSKE

School of Social Work

Southern University at New Orleans

Table 1
Odds Ratios (Standard Errors) from Multinomial
Logistical Regression Analysis Predicting
TANF Status ([section]).

                                Voluntary     Timed    Sanctioned
                                 leavers       off      vs. on
                                 vs. on      vs. on       TANF
                                   TANF       TANF
Demographics

Age                              .97          .98       .98
                                (.02)        (.02)     (.02)
African American                1.23         1.87      1.76
                                (.58)       (1.35)    (1.22)
Children less than 5 years       .55 ***      .52 *     .68
                                (.12)        (.15)     (.17)
Rural residence                 1.37         .53       1.05
                                (.51)        (.32)     (.51)
Human Capital Variables
Graduated high school           1.44         1.14      1.20
                                (.42)        (.46)     (.45)
Currently employed              3.24 ***     3.15 **    .76
                               (1.10)       (1.37)     (.38)
Ever married                    1.35         1.44      1.14
                                (.47)        (.67)     (.51)
Disabled recipient               .62          .33       .42
                                (.25)        (.23)     (.22)
Disabled child                   .69          .23 *     .74
                                (.25)        (.16)     3.00
Current domestic violence        .77          .63       .96
                                (.43)        (.53)     (.62)
Recent poverty                   .87          .97      1.11
                                (.08)        (.12)     (.13)
Childhood poverty               1.13         1.37      1.11
                                (.38)        (.62)     (.13)
-2 log likelihood            -360.57 **
Pseudo [R.sub.2]                 .08

([section]) Comparison group is those who remaining on TANF.

* p =.05, ** p =.01, *** p =.001.

[X.sup.2] (36, n = 309) = 65.85, p < 01.

Table 2
ANOVA of Differences between Welfare Recipients, Voluntary,
Timed Off and Sanctioned Leavers (SD in parentheses).

Variable                            Receiving      Voluntary
                                      TANF          Leavers

Available Resources

Monthly financial resources         $944.86         $849.71
                                    (459.71)        (486.91)
  % Working ([double dagger])         35.0            50.0
                                       (.48)           (.50)
  % Receiving Food Stamps             88.0            70.0
                                       (.33)           (.46)
Material Hardships
  % Insufficient food                 33.7            30.2
                                       (.47)           (.46)
  % Housing problems                  31.5            39.8
                                       (.44)           (.49)
  % Unable to obtain                   6.5            23.9
    medical care for parent            (.25)           (.43)
  % Unable to obtain                   2.8             5.8
    medical care for child             (.17)           (.24)
  % Needed, but didn't                 7.4            22.7
    receive Medicaid                   (.26)           (.42)

Health and Mental Health

  % Current major depression            .40             .31
                                       (.49)           (.47)
  # of days of poor                    9.54            7.68
    physical health                  (11.03)          (9.71)
  % Ever experienced                  32.4            22.7
    domestic violence                  (.47)         (42)
  % Ever experienced                  16.8             8.0
    sexual assault                     (.38)           (.27)

Variable                              Timed        Sanctioned
                                       Off

Available Resources

Monthly financial resources         $550.66         $820.36
                                    (420.24)        (480.17)
  % Working ([double dagger])         33.0            45.0
                                       (.48)           (.50)
  % Receiving Food Stamps             84.0            73.0
                                       (.37)           (.45)
Material Hardships
  % Insufficient food                 36.7            35.4
                                       (.49)           (.48)
  % Housing problems                  37.9            44.4
                                       (.49)           (.50)
  % Unable to obtain                  29.0            20.1
    medical care for parent            (.46)           (.41)
  % Unable to obtain                   6.7            10.9
    medical care for child             (.25)           (.31)
  % Needed, but didn't                22.6            27.1
    receive Medicaid                   (.43)           (.45)

Health and Mental Health

  % Current major depression           .39             .42
                                      (.49)           (.49)
  # of days of poor                   6.81            8.93
    physical health                  (8.55)         (10.76)
  % Ever experienced                 41.9            42.9
    domestic violence                 (.50)           (.50)
  % Ever experienced                  9.6             8.9
    sexual assault                    (.29)           (.29)

Variable                              ANOVA
                                    F (3,272)

Available Resources

Monthly financial resources           5.77 *** (a,b)

  % Working ([double dagger])         1.83

  % Receiving Food Stamps             3.61 ** (c)

Material Hardships
  % Insufficient food                  .20

  % Housing problems                   .92

  % Unable to obtain                  5.22 ** (a,c)
    medical care for parent
  % Unable to obtain                  1.32
    medical care for child
  % Needed, but didn't                4.45 ** (c,d)
    receive Medicaid

Health and Mental Health

  % Current major depression          .75

  # of days of poor                   .86
    physical health
  % Ever experienced                 2.54 ([dagger])
    domestic violence
  % Ever experienced                 1.46
    sexual assault

([double dagger])  Mean proportions are
reported as percentages for
ease of interpretation

([dagger]) p = .10, * p = .05,
** p = .01, *** p =.001.

Post-Hoc Bonferroni differences:

(a) Timed Off and Receiving TANF

(b) Timed Off and Voluntary Leavers

(c) Receiving TANF and Voluntary Leavers

(d) Receiving TANF and Sanctioned