Table of Contents | Previous | Next |
VI. The Relationship Between Sanctions and Work Participation Rates in Three Sites
Sanctions are intended to support program goals, but how they advance those goals depends on the details of sanction policy and implementation. One goal that sanctions are intended to support is achievement of a high work participation rate. While certain minimum rates are required by law, most states were not achieving the statutory standards (50 percent for all families and 90 percent for two-parent families, without the caseload reduction credit) prior to the DRA. A high rate helps agencies meet federal program requirements. In addition, a high participation rate may be important to agencies as an indicator of success in getting people employed. Sanctions can influence the participation rate both by inducing changes in behavior and by removing nonparticipating families from the participation rate calculation. Other policies or program requirements related to sanctions, such as applicant work requirements, also may influence the extent to which sanctions affect the work participation rate.
Either threatened or actually imposed sanctions may encourage TANF recipients otherwise not likely to participate in work activities to do so. Sanctions can thus increase the number of household heads participating in work activities, resulting in a higher work participation rate. If stronger penalties induce more recipients to participate, larger numbers of recipients might be expected to participate in work activities, and states with relatively more stringent sanctions would achieve higher participation rates. If this is the case, a shift from a partial to a full-family sanction would result in an increase in participation in work activities and a higher work participation rate, all else being equal. Behavioral change can also occur if information about work requirements and penalties for noncompliance lead some people never to apply for assistance or to leave the caseload. In those cases, sanction policies would increase the work participation rate by eliminating families that likely would have been noncompliant from the denominator of the participation rate calculation.
Sanctions can also affect the work participation rate through the interaction between sanction policy and how the participation rate is computed. As prescribed in the DRA, the rate is calculated by dividing a numerator consisting of “participants”—families engaged in federally acceptable work activities for the requisite hours per week—by a denominator that is a count of “total families.” Participants include families receiving TANF or state-funded assistance that counts toward maintenance of effort (MOE) requirements. States may exclude from the denominator families on the caseload that are sanctioned for failure to participate as long as they have not been sanctioned for more than three of the preceding 12 months. The Interim Final regulations of the DRA increased pressure on TANF agencies by requiring them to include in the denominator families under partial sanctions that remove parent needs from the grant.
The extent to which imposed sanctions affect the work participation rate thus depends heavily on state sanction policy. Most states with immediate full-family sanctions let little time elapse in removing noncompliant families from the caseload, lowering the work participation rate denominator. In these states, the only families included in the denominator are those participating in federally acceptable activities for the required hours per week (and thus in the numerator) and those participating in a manner that does not meet federal requirements but is acceptable to the state. Such policy clearly increases the work participation rate. In states with gradual full-family sanctions, the direct impact depends on the time elapsed to progress from a partial to full-family sanction, the time families remain in sanction status, and the likelihood of families being sanctioned more than once (when full family sanctions are usually imposed more quickly). If it takes three months to move from a partial to full-family sanction, the direct impact is the same as for a full-family sanction in that sanctioned families can be removed from the participation rate denominator for three months. If the transition from partial to full-family sanction takes longer than three months, families sanctioned for longer than three months would be counted in the denominator, but not in the numerator. In some states, the progression from partial to full-family sanction is faster for repeat sanctions such that the overall direct impact of sanctions on the work participation rate depends on the mix of first and subsequent sanctions. In states with partial-sanction policies, the direct effect of sanctions depends almost entirely on how long families remain in sanction status. If families remain in sanction status for short periods, the direct effect could resemble that of full-family sanctions. However, if families generally stay in sanction status longer than three months, sanctioned families could accumulate on the caseload and substantially reduce a state’s work participation rate.
The upfront work requirements for TANF applicants may also influence the effect of sanctions on the work participation rate. Many states set forth job search and other work requirements (e.g., work orientation) that TANF applicants must satisfy before enrolling in the program. Those who fail to comply with the requirements are denied TANF benefits; the effect is essentially the same as that for a full-family sanction for TANF recipients who fail to comply with program requirements. Applicant work requirements are intended to serve two purposes: (1) to ensure that those who do not truly need assistance or need only minimal help in obtaining employment do not enter TANF and (2) to ensure that those approved for TANF enter the program while already participating in activities. 1 However, applicant requirements can also divert from the caseload those who are not inclined to or cannot comply with work requirements. If applicant work requirements have this diversionary effect, many families that would otherwise be in the denominator of the participation rate, but not in the numerator, never enter the TANF rolls. In other words, the TANF caseload is “pared down” to families that enter the program already in the numerator of the participation rate or those that at least have proven their ability and desire to comply with work requirements. In this case, applicant work requirements would raise work participation rates. In fact, the work participation rate may be higher than in a state with an immediate full-family sanction policy because families on TANF must first demonstrate noncompliance before imposition of a full-family sanction. In addition, in states that impose significant work requirements as a condition of eligibility, sanction rates are almost certain to be lower than in states without such a requirement.
Three study sites offer opportunities to examine how the above theoretical relationships between sanction policy and observed work participation rates play out in reality. Texas, Georgia, and Los Angeles County recently implemented major changes in sanction policy at least in part to increase their participation rates. Texas shifted from a partial to a full-family sanction. Georgia eliminated a permanent ban from TANF for second or subsequent sanctions and revamped its applicant job search process. Los Angeles County implemented a home- visiting initiative that provides outreach primarily to TANF recipients at risk of sanction. We examine participation rates and related outcomes (for instance, employment rates) before and after the changes went into effect (1) to explore whether there is any correlation between policy and programmatic changes and changes in participation rates and (2) to consider the ways that policy and programmatic changes may influence the rates. 2 Without data on what would have happened to participation rates and related outcomes in the absence of these changes, we cannot draw rigorous conclusions about the effect of the changes on these outcomes. This is because we cannot distinguish the effects of changes in sanction policies from the effects of other factors, such as other changes in TANF policy or practice; economic influences on the behavior of low-income families; or changes in policy or practice in other programs serving low-income families.
A. The Texas Experience—Shifting from a Partial to a Full-Family Sanction
In September 2003, Texas shifted from a partial to full-family sanction with strict cure requirements. Before 2003, failure to comply with work requirements resulted in a benefit reduction equal to the entire adult portion of the grant. Repeat acts of noncompliance had the same effect on the TANF grant but were subject to progressively longer minimum sanction periods. The new policy requires termination of the TANF grant for one month for all clients who fail to meet their work activity hours. During that month, clients remain on the caseload in sanction status. After a second consecutive month of noncompliance, Texas drops clients from the TANF rolls completely. To cure a sanction, clients must perform one month of work activities; and, to return to TANF after being sanctioned off the caseload, clients must complete 30 days of work activities within 40 days of their TANF eligibility interview.
To explore the outcomes of the sanction policy change, MPR obtained and analyzed state TANF and Unemployment Insurance data for one case cohort that was on TANF one year before the policy changes and one case cohort on TANF in the month before the changes. The first cohort consisted of the 131,566 cases on TANF in August 2002; the second cohort consisted of the 138,916 cases on TANF in August 2003. The two cohorts are remarkably similar with respect to demographic characteristics and total time on TANF (see Table VI.1). We tracked the TANF, work participation, sanction, and employment status of each cohort for one year to determine the extent to which outcomes differ for clients subject to different sanction policies.
In the wake of its new policy, Texas has experienced a substantial increase in its work participation rate. Data from the Administration for Children and Families indicate that the average monthly work participation rate increased from 28.1 in FY 2003 to 34.2 in FY 2004 (a 21.7 percent increase). The administrative data obtained and analyzed by MPR indicate a similar pattern. Monthly participation rates among single parent TANF recipients in the partial sanction cohort hovered between 21 and 24 percent while monthly participation rates among single parent TANF recipients in the full family sanction cohort ranged between 34 and 40 percent (see Figure VI.1).
The increase in Texas’s work participation rate following the implementation of full-family sanctions demonstrates the complexity of tying changes in key outcomes of interest to policy changes. The increase in Texas’ work participation rate likely resulted from several different factors, the most important of which was a decline in the caseload. The number and proportion of cases participating in work activities was also higher immediately after the implementation of full-family sanctions than before, but declined over time to levels similar to those among cases subject to partial sanction policies. In addition, after the policy change was implemented, a number of families left the caseload for employment (and more did so than before the policy change), however, they do not affect the work participation rate because they are no longer part of the TANF caseload. The following sections discuss these three factors—changes in the caseload, changes in participation, and changes in employment—in more detail.
| Characteristic | Partial Sanction Cohort | Full Family Sanction Cohort | |
|---|---|---|---|
| Ethnicity/Race of Case Head | Non–Hispanic, white | 19.8 | 19.4 |
| Non–Hispanic, African American | 30.5 | 31.3 | |
| Hispanic, any race | 48.9 | 48.6 | |
| Other | 0.8 | 0.8 | |
| Education of Case Head | Less than high school diploma/GED | 59.6 | 58.7 |
| High school diploma/GED | 40.4 | 41.3 | |
| More than high school diploma/GED | 0.0 | 0.0 | |
| Number of Children in TANF Case | 0 | 0.7 | 0.7 |
| 1 | 41.6 | 41.9 | |
| 2 | 30.0 | 29.9 | |
| 3 | 17.0 | 16.9 | |
| 4 or more | 10.7 | 10.6 | |
| Age of Youngest Child in Case | Younger than 1 | 11.8 | 11.4 |
| 1 to 2 | 15.4 | 15.3 | |
| 3 to 5 | 12.2 | 12.4 | |
| 6 or older | 18.3 | 18.4 | |
| Type of Case | Single-parent | 58.1 | 54.6 |
| Two-parent | 4.7 | 4.6 | |
| Child-only | 37.1 | 40.8 | |
| Cumulative TANF Months since November 1996 | Fewer than 6 months | 25.1 | 25.8 |
| 6 to 11 months | 17.6 | 18.8 | |
| 12 to 24 months | 19.2 | 19.4 | |
| 25 months or more | 36.9 | 34.9 | |
| Sample Size | 131,556 | 138,916 | |
| Source: MPR analysis of Texas TANF administrative data and Texas Unemployment Insurance data. Note: Child age is calculated as of August 1 of the cohort year. Of the two cohorts, 55,522 cases in the partial sanction cohort and 59,076 cases in the full family sanction cohort were missing data for the youngest child’s birth date. |
[D] |
1. Changes in the Caseload
The increase in Texas’s work participation rate following the implementation of full-family sanctions primarily reflects a decrease in the caseload, the denominator of the rate. The number of cases on TANF in August 2002 (131,556) was similar to the number of cases on TANF in August 2003 (138,916). The number of single-parent cases on TANF in August 2002 and August 2003 was even more similar (76,489 and 75,845, respectively). A substantial number of cases in both cohorts went off TANF over the course of the year, but the trend was much more pronounced for the 2003 cohort (see Figures VI.2 and VI.3). Forty-eight percent of those on TANF in August 2002 were off the caseload one year later, while 62 percent of those on TANF in August 2003 had left the caseload one year later. Among single-parent cases, 56 percent of those on TANF in August 2002 were off the rolls one year later, while 76 percent of those on TANF in 2003 were off one year later.
[D] |
[D] |
While the advent of full-family sanctions may not account for Texas’s entire caseload decline, it likely accounts for a substantial portion of it. Though the economic climate was rather similar between 2002 and 2004, other policy changes and other external factors may have contributed to the decline. Given that the 2002 cohort was subject to the original partial-sanction policy for the entire follow-up period, none of the cases in the cohort was closed as a result of a sanction; rather, cases were closed because of financial or other ineligibility or because of voluntary exit from TANF. With the 2003 cohort subject to the immediate full-family sanction policy, cases in the cohort could be closed as a result of a sanction as well as for all of the same reasons affecting the 2002 cohort. The rate of sanctioning in Texas did not change when the state shifted to an immediate-full-family sanction policy (see Figure VI.4), although the percent of cases on the rolls and in sanction status declined dramatically. The reason is that, under the old policy, sanctioned cases could remain on the rolls indefinitely and thereby accumulated on the caseload. Under the new policy, sanctioned cases remain on the rolls for only one month and therefore do not accumulate on the caseload. Cases sanctioned under the new policy are closed in the following month if they do not come back into compliance. Thus, after the implementation of the full family sanction, the percent of non-exempt clients in sanction status each month is the same as the percent receiving a sanction each month. In contrast, prior to the implementation of the full family sanction, the percent of non-exempt clients in sanction status each month included those newly sanctioned as well as those that had been in sanction status for some time.
[D] |
We observed no demographic differences in the types of cases sanctioned before and after the policy changes took effect. In both cohorts, clients sanctioned some time during the one-year follow-up period were slightly more likely to have less than a high school diploma or GED, to have more than one child, and to have younger children (under age 6) than those never sanctioned during that period (see Table VI.2). The magnitude of the differences is similar across cohorts. Similarly, clients in the partial sanction cohort remaining on the caseload at the end of the follow-up period appear similar to clients in the full family sanction cohort remaining on the caseload at the end of the follow-up period. However, in the full family sanction cohort, substantially more of the families remaining on the caseload by the end of the follow-up period were child-only cases, reflecting the relatively higher rate of TANF exit among single-parent cases in the full family cohort versus the partial sanction cohort.
2. Changes in Participation
The Texas policy change may have also affected the work participation rate by increasing the number of recipients who participate in program activities (the numerator). The number of single-parent cases participating in work activities for at least 30 hours per week in August is about 28 percent higher in the 2003 cohort than in the 2002 cohort—23,170 versus 18,151 cases (see Figure VI.5). 3 It is possible that news about the upcoming policy changes inspired those imminently facing the changes (the August 2003 cohort) to increase their work activity relative to peers who did not face the changes (the 2002 cohort). In both cohorts, however, the raw number of single-parent cases participating for at least 30 hours per week declined over the course of the following year. The decline was much more rapid among the 2003 cohort. By the end of the follow-up period, 13 percent fewer TANF recipients in the cohort subject to full-family sanctions (the 2003 cohort) were participating in activities for at least 30 hours per week than recipients in the cohort subject to partial sanctions (the 2002 cohort). The trend in participation may be indicative of the fact that, over time, fewer clients in the full family sanction cohort versus the partial sanction cohort are required to participate in work activities (see Figure VI.6). If fewer recipients are required to participate in activities, then it is almost a certainty that fewer will.
| Characteristic | Cohort 1 | Cohort 2 | |||||
|---|---|---|---|---|---|---|---|
| Ever Sanctioned during Year | Never Sanctioned during Year | On TANF in July 2003 | Ever Sanctioned during Year | Never Sanctioned during Year | On TANF in July 2004 | ||
| Ethnicity/Race of Case Head | Non–Hispanic, white | 19.8 | 19.8 | 16.4 | 19.9 | 19.2 | 15.4 |
| Non–Hispanic, African American | 33.7 | 28.8 | 32.0 | 34.7 | 30.3 | 30.3 | |
| Hispanic, any race | 45.7 | 50.7 | 50.9 | 44.7 | 49.7 | 53.6 | |
| Other | 0.8 | 0.8 | 0.7 | 0.6 | 0.8 | 0.7 | |
| Education of Case Head | Less than high school diploma/GED | 61.9 | 58.4 | 65.1 | 62.1 | 57.8 | 64.3 |
| High school diploma/GED | 38.1 | 41.6 | 34.9 | 37.9 | 42.2 | 35.7 | |
| More than high school diploma/GED | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | |
| Number of Children in TANF Case | 0 | 0.3 | 0.8 | 0.6 | 0.2 | 0.8 | 0.7 |
| 1 | 36.4 | 44.4 | 40.7 | 39.2 | 42.6 | 42.3 | |
| 2 | 31.1 | 29.4 | 29.9 | 30.7 | 29.7 | 29.6 | |
| 3 | 19.2 | 15.8 | 17.0 | 18.4 | 16.5 | 16.5 | |
| 4 or more | 12.9 | 9.5 | 11.8 | 11.5 | 10.4 | 10.9 | |
| Age of Youngest Child in Case | Younger than 1 | 15.1 | 10.1 | 11.0 | 14.2 | 10.7 | 9.6 |
| 1 to 2 | 18.9 | 13.5 | 15.1 | 18.1 | 14.6 | 14.3 | |
| 3 to 5 | 13.1 | 11.7 | 12.5 | 13.1 | 12.2 | 12.5 | |
| 6 or older | 16.2 | 19.5 | 20.1 | 15.3 | 19.3 | 20.7 | |
| Type of Case | Single-parent | 92.2 | 39.6 | 49.5 | 91.6 | 44.9 | 34.5 |
| Two-parent | 6.3 | 3.8 | 3.0 | 6.3 | 4.2 | 2.3 | |
| Child-only | 1.5 | 56.6 | 47.5 | 2.1 | 50.8 | 63.2 | |
| Cumulative TANF Months since November 1996 | Fewer than 6 months | 28.3 | 23.3 | 16.8 | 31.2 | 24.4 | 14.7 |
| 6 to 11 months | 20.7 | 15.9 | 13.9 | 21.7 | 18.0 | 13.7 | |
| 12 to 24 months | 21.6 | 17.9 | 19.1 | 21.5 | 18.9 | 19.2 | |
| 25 months or more | 28.2 | 41.7 | 48.7 | 24.5 | 37.6 | 51.1 | |
| Sample Size | 46,445 | 85,110 | 68,226 | 28,732 | 110,184 | 52,194 | |
| Source: MPR analysis of Texas TANF administrative data and Texas Unemployment Insurance data. Note: Child age is calculated as of August 1 of the cohort year. Of the two cohorts, 55,522 cases in cohort 1 and 59,076 cases in cohort 2 were missing data for the youngest child’s birth date. |
[D] |
[D] |
3. Changes in Employment
Another question of interest is whether under the new sanction policy TANF recipients are more likely to work when they leave the caseload. For the cohort subject to the new full family sanction policies, the rate at which recipients left welfare with or for employment—49 percent, initially—is about ten percent lower than the rate for those subject to the old, less stringent partial sanction policies—30 percent, initially (see Table VI.3). While more people in the August 2003 cohort left TANF with or for employment than in the August 2002 cohort (16,567 versus 9,653 people, respectively), even more left TANF without work (25,713 versus 10,047 people, respectively). While 1.7 times as many people in the August 2003 cohort left TANF with or for employment than in the August 2002 cohort, 2.5 times as many left TANF without employment. 4
| Cases | Employed | Not Employed | |||
|---|---|---|---|---|---|
| Number | Percent | Number | Percent | ||
| Among Partial Sanction Cohort Cases that Went off TANF in Third-Quarter 2002 (N = 19,700) |
Employment status in third-quarter 2002 | 9,653 | 49.0 | 10,047 | 51.0 |
| Employment status in fourth-quarter 2002 | 9,239 | 46.9 | 10,461 | 53.1 | |
| Among Full Family Sanction Cases that Went off TANF in Third-Quarter 2003 (N = 42,154) |
Employment status in third-quarter 2003 | 16,567 | 39.3 | 25,587 | 60.7 |
| Employment status in fourth-quarter 2003 | 16,777 | 39.8 | 25,377 | 60.2 | |
| Source: MPR analysis of Texas TANF administrative data and Texas Unemployment Insurance data. Note: Employment data are based on Unemployment Insurance wage data records and are unavailable for first and second quarters of 2004. |
B. The Georgia Experience–Sanction Policy and Applicant Work Requirement Changes
Beginning in 2004, Georgia implemented revisions to its sanction policy and practices. Before that year, an initial failure to meet work requirements resulted in a 25 percent reduction of the TANF grant. A subsequent violation resulted in termination of the grant and a lifetime prohibition of further TANF assistance. However, TANF clients were entitled to a conciliation process for each act of noncompliance and could thus avoid imposition of a sanction. In 2004, Georgia revised its policy so that subsequent violations no longer make a family ineligible for TANF for life, but conciliation may occur only once in a TANF recipient’s lifetime. Currently, the first instance of noncompliance results in an official conciliation, or warning, with no effect on the TANF benefit. The second instance of noncompliance results in a 25 percent reduction in the TANF grant for three months. In the third month of the grant reduction period, a case manager sends the client a notice instructing the client to visit the TANF office for a meeting with the case manager in order to restore full benefits. If the client attends the meeting within 30 days of the end of the sanction period, full benefits resume. If the client does not attend, the case is closed and the client must reapply for TANF in order to receive benefits again. The third instance of noncompliance results in full case closure for a minimum three months, after which the client may reapply for benefits. The fourth instance of noncompliance results again in a 25 percent reduction in the TANF grant for three months, followed by case closure if the client does not attend a meeting with the case manager within 30 days of the end of the sanction period. The fifth instance of noncompliance results in full case closure for a minimum of 12 months, after which the client may reapply for benefits. The consequences for future instances of noncompliance are the same as the consequences for the fourth and fifth instances, in turn.
Georgia’s policy change was part of a larger initiative entitled “The Right Work the Right Way,” instituted by the Department of Human Resources. The initiative emphasized “welfare is not good enough for any family.” In addition to encouraging counties to engage more TANF recipients in work and work-related activities, the initiative encouraged counties to move more TANF applicants away from the welfare rolls and toward stable employment. While Georgia always had an upfront applicant job search requirement in place, many counties—in response to the initiative--modified the way in which they implement the job search process. For instance, in October 2005, DeKalb County institutionalized what was previously an independent job search requirement for TANF applicants. Formerly, DeKalb required TANF applicants to make a certain number of employer contacts per week and the county tracked job search activities through self-reported forms. As of October 2005, the county refers all TANF applicants to a for-profit contracted service provider and requires participation in a four-week job search program for 40 hours per week. The first week involves workshops in resume development, interviewing, problem-solving skills, use of technology, and similar topics. The second week focuses on a structured job search and in-depth assessments. The final weeks consist of independent job search with oversight and thorough documentation. Applicants who miss more than two days of the job search program are denied assistance.
Georgia’s TANF caseload has declined consistently and substantially since 2003 at the same time that its participation rate has been rising sharply (see Figure VI.7). The most drastic changes in the caseload have occurred since 2004. In June 2004, the caseload totaled 52,515 (including exempt and non-exempt adult cases as well as child-only cases) but had declined by 26 percent to 38,669 as of June 2005. Between June 2005 and June 2006, the caseload declined by another 24 percent and continued to decline the following year. By June 2006, the caseload was almost half of its 2003 total. Increases in the state’s participation rate have also been occurring since 2002 but have been most dramatic since 2004. Before the institution of the initiative, work participation rates according to federal data were extremely low—8.2 percent in FY 2002 and 10.9 percent in FY 2003. By FY 2004, participation rates had increased to 24.8 percent. According to state data, by April 2005, the participation rate had reached 63.7 percent and, by April 2006, 69.0 percent. 5
[D] |
To understand whether and how the 2004 initiative has contributed to the changes in Georgia’s caseload and participation rate, it is important to examine trends in sanction rates, TANF application and application approval rates, sanctions, and case closures. Thus, for each month from January 2001 through December 2006, MPR obtained data on the total number of applications received statewide, the percent of applications approved and denied, and the reasons for denial. We also obtained information on the number of sanctions imposed, the number of TANF cases closed, and reasons for case closure.
Anecdotal evidence suggests that, under the old policies, program staff perceived the consequences (i.e., a permanent ban from TANF) of sanctions as too punitive and therefore did not sanction families. Program staff either conciliated or disregarded most acts of noncompliance. Data confirm that the use of sanctions was relatively low in FY 2002 and FY 2003 but that sanction rates increased substantially thereafter. The number of TANF recipients with a first-level sanction nearly doubled between FY 2003 and FY 2004 and then more than doubled again by FY 2005, reaching its highest level since FY 2000 (see Table VI.4). Between FY 2000 and FY 2004, second-level sanctions throughout Georgia were relatively infrequent, never occurring more than 24 times per year. In most years, no county had more than two or three cases subject to a second sanction. While the number of second-level sanctions remained small relative to the caseload, however, FY 2005 saw a sharp increase in the frequency of second-level sanctions, from 24 in FY 2004 to 89 in FY 2005. These increases occurred as the TANF caseload was declining, resulting in a substantially higher sanction rate during the state’s implementation of its new initiative.
| Fiscal Year | TANF Recipients in First-Level Sanction (25 percent grant reduction) | TANF Recipients in Second-Level Sanction (grant termination) |
|---|---|---|
| 2000 | 476 | 6 |
| 2001 | 410 | 6 |
| 2002 | 168 | 19 |
| 2003 | 210 | 13 |
| 2004 | 380 | 24 |
| 2005 | 835 | 89 |
| Source: MPR analysis of Georgia administrative data. |
While sanction rates did in fact increase in the wake of policy changes, the number of case closures due to sanctions is still extremely low and does not account for the large decline in the TANF caseload. In addition, even if sanction policy changes have led program staff to implement sanctions more frequently (making consequences for nonparticipants more certain), the relatively small number of case closures suggests that the changes may not be significant enough to result in measurable changes in behavior. We have observed no net increase in the number of families meeting the federal work requirements since implementation of the state initiative and policy changes. While 8,750 families were participating in work activities at federally acceptable levels in October 2004, only 5,974 and 3,001 families, respectively, were doing so in October 2005 and January 2006. Thus, the increase in the participation rate is not due to an increase in the numerator of the rate, but rather a decrease in the denominator.
However, that the strong messages about work and program expectations communicated by program staff to applicants as a result of the 2004 initiative have been followed by a decrease in the number of families applying for TANF (see Table VI.5). The number of TANF applications steadily increased between FY 2000 and FY 2003 and then steadily decreased between FY 2003 and FY 2006. It is likely that changes in the implementation of the upfront job search requirement coupled with new messages are motivating more low-income families either (1) to work instead of accept welfare and all of its requirements, or (2) to refuse to agree to TANF work requirements even if unemployed. Moreover, recent years have seen a substantial decrease in the percent of TANF application approvals. While about half of the applications were approved between FY 2000 and FY 2003, the approval rate plummeted to less than one-quarter in 2006 after reaching a high of just under three-quarters in 2004.6 Application approval rates prior to the program changes were similar to approval rates in other states—53 percent on average in 2000 and 51 percent in 2001 (Abt Associates, Inc. 2003).7
| Fiscal Year | Number of TANF Applications | Percent of TANF Applications Approved |
|---|---|---|
| 2000 | 115,155 | 50 |
| 2001 | 130,648 | 51 |
| 2002 | 137,791 | 51 |
| 2003 | 146,279 | 51 |
| 2004 | 145,314 | 73 |
| 2005 | 132,309 | 48 |
| 2006 | 104,547 | 22 |
| Source: MPR analysis of Georgia administrative data. |
Once families have submitted their applications for TANF, the penalty for failing to comply with upfront job search and other work requirements has the same effect as a full-family sanction for noncompliant recipients. Increasingly, Georgia is denying TANF applications because applicants either cannot or do not want to comply with work requirements during the eligibility process. In every year between FY 2000 and FY 2006 with the exception of FY 2004, the three most common reasons for application denial were (1) applicants’ failure to cooperate in the eligibility process; (2) applicants’ voluntary withdrawal of their applications; and (3) applicants’ financial ineligibility. However, the relative frequency of these reasons has shifted over time (see Figure VI.8). While the rate of non-cooperation with the eligibility process remained relatively constant (at or slightly more than one-third), the rate of voluntary application withdrawals has increased substantially over time even as the rate of financial or other ineligibility has decreased.8 By FY 2006, almost 38 percent of TANF applicants did not become TANF recipients because of their own decision to stop pursuing assistance. With an additional 37 percent failing to cooperate with the eligibility process, fully three-quarters of applicants could not or did not want to comply with program expectations. While it is likely that much of this shift can be attributed to changes in client behavior, it is also possible that changes in coding and reporting practices among case managers also contributed.
Terminations for reasons other than sanction for failure to comply with work requirements have also contributed to changes in the caseload and participation rate. Case closures in recent years have occurred more frequently than in the past. While recipients may not leave TANF directly as a result of a sanction, they may decide that financial assistance is no longer worth the burden of program requirements; thus, they voluntarily leave the caseload. Indeed, case closure at families’ request or for failure to comply with the re-certification process represent an increasing share of all case closures (see Figure VI.9). At the same time, case closure due to changes in families’ financial circumstances that make families ineligible for assistance represent a decreasing share of all case closures. In 2003, the year before Georgia’s policy and philosophy changed, 42 percent of cases closed for financial reasons, and 10 percent closed when families requested closure or did not comply with re-certification. By 2006, these numbers were 22 and 37 percent, respectively. Once again, while it is likely that much of this shift can be attributed to changes in client behavior, it is also possible that changes in coding and reporting practices among case managers also contributed.
[D] |
[D] |
C. The Los Angeles County Experience–Efforts to Reduce Sanctions
In an attempt to improve its efforts to serve families required to participate in work activities, Los Angeles County has been particularly pro-active in examining who is sanctioned and why. In recent years, community advocates and policymakers have expressed interest in reducing the number of recipients in sanction status and enhancing participants’ capacity to comply with work requirements. They have been concerned that sanctioned TANF recipients are unable to comply with program requirements because they do not receive needed supportive services for substance abuse, domestic violence, and mental health problems. In response, in 2004, the Los Angeles Commission for Public Social Services requested systematic information on welfare sanctions and the county’s sanctioned population.
In March 2005, the county released a study the Department of Public Social Services conducted of a group of TANF recipients who had been sanctioned between April 2002 and February 2004. The study led to formation of the Partner’s Workgroup, which brings together the Los Angeles Commission for Public Social Services, the Department of Public and Social Service (DPSS) managers, and other program stakeholders. The Workgroup was charged with creating a Sanction Action Plan that would use the study’s findings and recommendations as the basis for a series of policy enhancements designed to reduce the county’s sanction rate. DPSS submitted the Sanction Action Plan to the Board of Supervisors in August 2005. The plan provided a blueprint for making changes throughout the CalWORKs and GAIN service systems to reduce the number of sanctions.
One of the most significant aspects of the plan was the creation of the home-visiting program, whose primary goal is to conduct outreach to and conciliation with nonparticipating clients before imposition of a sanction. Home visit workers mail a letter to each noncompliant client in danger of sanction advising the client that a home visit will occur if the client does not attend a required meeting to conciliate the impending sanction. If the client attends the meeting, the home visit is not necessary. If a home visit is in fact necessary, it usually takes place the day after the scheduled conciliation meeting. During the visit, home visit, workers attempt to identify reasons for client noncompliance and encourage re-engagement in activities. The home visit project was implemented on October 31, 2005, in five of Los Angeles County’s seven regions. Initially, the project targeted recipients who had previously been identified with a supportive services need, specifically mental health and/or substance abuse. Effective December 2005, the project expanded to serve all TANF recipients in the five regions; effective March 2006, the project expanded to serve all TANF recipients in the remaining two regions.
In the first year of project implementation, 41,233 TANF recipients were noncompliant and subject to home visits. Among the noncompliant recipients, 77 percent successfully resolved their noncompliance and/or sanction (see Figure VI.10). Among cases successfully resolving their noncompliance issues, only 10 percent required a home visit. Program staff could not contact 12 percent (either because the client was not home or because of an incorrect address) and thus the case was still pending, and another 11 percent were sanctioned; some of those sanctioned received home visits and some did not.
[D] |
The number of recipients sanctioned in Los Angeles County has been declining since implementation of the home-visiting project. In September 2005, 21,282 families were in sanction status. By November 2006, that number had declined by 26 percent to 15,662 families. The sanction rate (defined by the number of families in sanction status divided by the number of families enrolled in GAIN, or subject to work requirements) has also been declining, suggesting that the decline is not simply a result of a declining TANF caseload. In November 2005, the sanction rate was almost 43 percent (see Figure VI.11). Within one year, it had declined to 36 percent. It appears that the decline in sanctions can be attributed to both a reduction in new (initial and subsequent) sanctions, and an increase in the number of sanctions cured or resolved (for example, because the household head left the caseload). The average number of new sanctions imposed in fourth-quarter 2006 was nearly 16 percent lower than the average number of sanctions imposed a year earlier. However, the number of GAIN enrollees was declining during this period such that the rate of new sanctions declined, but not enough to account for the full decline in the number of individuals in sanction status. While the rate of imposition of new sanctions (defined by the number of new sanctions imposed divided by the number of non-sanctioned families enrolled in GAIN) hovered around 7.3 – 7.5 percent in fourth-quarter 2005, it fell to 6.3 percent in early 2006 and fluctuated between 5 and 6 percent for the remainder of the year.
[D] |
Although Los Angeles County has reduced the number of recipients in sanction status, the decline has not translated into a substantial increase in the county’s work participation rate. Many clients had their cases resolved in a manner that presumably would have a positive effect on the county’s work participation rate. Twenty-two percent agreed to participate in work activities, and another 7 percent were employed (see Table VI.6). At the same time, many other cases were resolved in a manner that would have an adverse effect on the county’s work participation rate. About one-third (33.5 percent) were found to be exempt from work requirements or to have good cause for not participating. Many of these cases likely would remain in the denominator of the participation rate. In fact, we observed little change in the county’s work participation rate since implementation of the home-visiting project. In FY 2005, the county’s work participation rate was 27.9 percent and, in FY 2006, it was 26.5 percent. The county now believes that noncompliant families may need more assistance and encouragement to begin participating in program activities after they have agreed to comply. As a result, the county is in the process of implementing a new case management system whereby case managers would spend more time with families to help them identify and resolve potential participation barriers. Case managers would contact clients as often as daily but at least weekly in addition to daily monitoring of program participation. Acknowledging that affecting the work participation rate can take time, the county is willing and eager to engage in a trial and error process to find an approach that benefits both individual recipients and the county.
| Cases | Number of Cases | Percent of Cases | |
|---|---|---|---|
| Successfully Resolved | Agreed to participate | 9,100 | 22.1 |
| Excused for good cause | 9,394 | 10.7 | |
| Exempt | 4,400 | 22.8 | |
| Employed | 2,978 | 7.2 | |
| Homeless | 2,051 | 5.0 | |
| Cash grant terminated | 1,694 | 4.1 | |
| Sanction imposed in error | 1,678 | 4.1 | |
| Other | 450 | 1.1 | |
| Sanctioned | 4,612 | 11.2 | |
| Unresolved (no contact) | 4,876 | 11.8 | |
| Total | 41,233 | 100.0 | |
| Source: MPR analysis of data provided by the Los Angeles County Department of Social Services |
1 The latter responds to pressures to increase participation rates while the former responds to budgetary pressures. With respect to the former, keeping those who need only minimal help finding employment off of TANF (and thus would likely be able to participate in activities and to count in the numerator of the participation rate) could have negative implications for participation rates. (back)
2 Appendix D contains more detailed tables of results from the analysis of participation and other outcome data in Texas and Georgia. (back)
3 Recall that the total number of single-parent cases on TANF in August was similar among each cohort. (back)
4 The analyses in this section are based on a comparison of data from the year before and the first year after policy changes took effect. Data for subsequent years may portray different, perhaps more positive, outcomes. (back)
5 Federal data on state participation rates are not available for years after 2004. (back)
6 Georgia began implementing its “the Right Work the Right Way” initiative in 2004; thus 2004 can be viewed as a transition year and an outlier with respect to percentage of applications approved. (back)
7 Data available from the Georgia Department of Human Resources at http://dfcs.dhr.georgia.gov. (back)
8 Once again, because Georgia began implementing its “the Right Work the Right Way” initiative in 2004, this year can be viewed as a transition year and an outlier with respect to application denials. (back)
| Table of Contents | Previous | Next |