Table of Contents | Previous | Next |
Chapter 3
The Implementation of Jobs First: The Time Limit
Chapter 2 begins the discussion of the implementation of the Jobs First program, examining the program’s services, mandates, and message. This chapter completes the discussion, focusing on the program’s 21-month time limit on Temporary Family Assistance (TFA) benefits.1 Specifically, after summarizing the key findings, the chapter discusses the time-limit message that recipients heard and describes the time-limit review process that occurs during each individual’s twentieth month of cash assistance and at the end of each six-month extension of benefits. It presents information on how quickly Jobs First group members reached the time limit and on the outcomes of the time-limit review process, including the proportion of recipients who received six-month extensions. It then provides information on various characteristics of people who received multiple extensions to their benefits and on the status of individuals who left the welfare rolls because of the time limit. Finally, the chapter describes the services that are available after the time limit, including the Safety Net program that serves individuals who are denied extensions.2
As noted in Chapter 2, although MDRC conducted field research visits in Manchester and New Haven in late 2000, Jobs First’s implementation was studied most extensively during the first four years of program operations, from 1996 though 1999. Thus, most of the information on the time limit was covered in more detail in MDRC’s interim report, published in 2000.3
I. Summary
Just over half of the Jobs First group reached the time limit within the follow-up period for this study. Twenty-nine percent reached the time limit 21 months after their random assignment date; that is, they received TFA benefits continuously and were never exempt from the program. Another 24 percent reached the time limit between 22 and 48 months after random assignment. Thus, 47 percent of the Jobs First group did not reach the time limit in the four-year period. Most of these individuals left welfare, at least temporarily; others were exempt from Jobs First and its time limit for at least part of the period they received benefits.
Fewer than half of those who reached the time limit had their benefits discontinued at that point; most of those whose cases were closed were working and had income above the payment standard (the maximum grant for their family size). Just over half of those who reached the time limit were granted a six-month extension at that point because they had income below the payment standard and were deemed to have made a good-faith effort to find a job.
During the 36 months following the time limit, very few individuals whose grant was closed at the time limit returned to the rolls, and most of the individuals who received an extension at the time limit left welfare. A small minority, however, received additional six-month extensions and continued to receive benefits throughout the period. Not surprisingly, those who received multiple extensions were more disadvantaged than other sample members when they entered the study. In all, among recipients who reached the time limit, 63 percent were granted an extension at some point, but the vast majority (about 86 percent) were off the rolls 36 months later.
II. The Time-Limit Message
As discussed in Chapter 2, some components of Jobs First rely partly on communication to produce the desired effect. In the case of the program’s time limit on cash assistance, it would be possible to track recipients’ months of assistance and call them in for exit interviews as they approach the time limit without informing them beforehand about the limit. But the time limit is intended to spur recipients (and the system) to focus on self-sufficiency well before month 21, and its ability to do so depends on whether and how staff describe it to recipients. Overall, results show that the Jobs First group heard the message but did not necessarily change their behavior because of it.
A. Perceptions About the Time Limit
Table 3.1 presents sample members’ perceptions about the time limit on cash assistance receipt. It uses data from the Interim Client Survey, rather than the Three-Year Client Survey, because the former was administered 18 months after random assignment, before respondents could have reached the time limit. (Some of the same questions were asked at the three-year point; the results are shown in Appendix A.)
The table shows the percentage of survey respondents in each site who reported that they were (or, if they were not on welfare, that they had been) subject to a time limit on cash assistance. This question was asked of all respondents who reported that they had received cash assistance at some point since random assignment. The results show that staff in both sites successfully informed recipients about this key element of the Jobs First model: Approximately nine-tenths of the Jobs First group respondents reported that they were subject to a time limit (about three-fourths reported being subject to a 21-month time limit).4 Further analysis using administrative records showed that one-quarter of those who reported that they were not subject to a time limit were correct; that is, the records show that they were exempt for most of the follow-up period. Another one-third received benefits for only a short time (fewer than six months) and thus had little exposure to the program.
| Measure | Manchester | New Haven | Full Sample | |||
|---|---|---|---|---|---|---|
| Jobs First Group |
AFDC Group |
Jobs First Group |
AFDC Group |
Jobs First Group |
AFDC Group |
|
| Is/was there a time limit on how long you are/were allowed to receive cash assistance from AFDC/TFA? (%) | ||||||
| Yes | 91.5 | 32.9 | 89.1 | 19.6 | 89.3 | 23.0 |
| 21 months | 69.6 | 20.9 | 77.2 | 8.8 | 75.2 | 11.9 |
| Another length | 16.3 | 8.8 | 9.9 | 6.4 | 11.2 | 7.1 |
| Don't know length | 5.6 | 3.2 | 2.1 | 4.4 | 2.9 | 4.1 |
| No | 8.2 | 61.3 | 8.9 | 75.4 | 8.8 | 72.1 |
| Don't know | 0.3 | 5.7 | 2.0 | 4.9 | 1.8 | 4.9 |
| Sample size | 76 | 65 | 246 | 233 | 322 | 298 |
| SOURCE: MDRC calculations using Interim Client
Survey data. NOTES: These items were asked of the 620 survey respondents who reported receiving cash assistance since random assignment. Estimates were adjusted using ordinary least squares, controlling for pre-random assignment characteristics of sample members. Distributions may not add to 100 percent because of rounding. Results in this table were weighted to make them more representative of the full sample. These results may not represent true impacts of the program; the groups being compared may have differed in their background characteristics because these questions were asked only of respondents who reported that they had received cash assistance. Thus, significance tests were not conducted. |
Table 3.1 also shows that a little less than one-fourth of AFDC group members reported that they were or had been subject to a time limit. However, only about half of these people — 12 percent of all AFDC group respondents — reported being subject to a 21-month time limit. Further analysis (not shown in the table) found that almost half of the AFDC group members who reported that they were subject to a 21-month time limit were correct — that is, they were subject to Jobs First policies, usually because they had moved and received TFA in another part of Con-necticut where the AFDC group did not exist.5 It is not possible to say whether the others obtained erroneous information from staff, neighbors, the media, or some other source. It is worth noting that almost a quarter of the AFDC group members who incorrectly reported that they had a time limit had received welfare for less than six months during the 18-month follow-up period and thus had little exposure to the AFDC program message (of no time limit on benefit receipt).
This type of “contamination” narrows the treatment difference between the research groups and means that the impact analysis probably understates the impact of the time limit on recipients’ behavior in the period before they reached the limit. However, it is worth noting that the extent of contamination is low relative to other similar studies.6 Moreover, despite the contamination, there is a very large treatment difference between the two groups in this area.
B. Content of the Time-Limit Message
There are at least two key issues regarding the content of the time-limit message. First, how staff discussed the extension policy — particularly during the period before any recipients reached the time limit — may have shaped recipients’ views about whether the time limit was “for real.”7 In general, written materials produced by DSS explicitly stated that extensions would be possible in some situations. Discussions with staff in 1996 and 1997 suggested that they adopted one of two general approaches in describing the extension policy to recipients. Some workers said that recipients who cooperated with the program’s mandates would likely receive an extension if they could not find a job; they emphasized that recipients should comply with all program rules to ensure that they did not make themselves ineligible for an extension. Other workers were much less definite, saying that they did not know which recipients would receive an extension and that clients thus needed to make every effort to find a job in order to prepare for the possible loss of benefits. The staff survey results indicate that New Haven staff were much more likely to adopt the former approach, while Manchester staff adopted the latter. More generally, staff also reported that many recipients were initially skeptical that the time limit would be implemented (in fact, many staff said that they themselves were skeptical).
In small-scale telephone interviews conducted just a few months after individuals were randomly assigned, only a minority of Jobs First group members in each site said that their grant would definitely be discontinued if they reached the time limit; a larger proportion mentioned the possibility of receiving an extension. On the Interim Client Survey, administered in mid to late 1998 (people began reaching the time limit in late 1997), only about 32 percent of Jobs First group respondents said they believed that “nearly everyone” who reached the time limit would have her or his grant canceled, 48 percent believed that “some” recipients’ grants would be canceled, and 9 percent said they believed that “almost no one” would have their benefits canceled (not shown in a table).8 Thus, it appears that, from the beginning, most recipients understood that the time limit would not necessarily result in cancellation of their welfare grant.
Second, owing to the structure and generosity of the enhanced earned income disregard, DSS staff would be unlikely to urge recipients to leave welfare quickly in order to “save” or “bank” their available months of benefits. In some other states with time limits, this message can be particularly appealing to recipients who are working and receiving a small welfare grant, but there are few such recipients in Connecticut because of the disregard’s flat structure.9 In addition, during the study period, there was in theory no reason to bank months, because recipients could receive an extension either when they reached the time limit or at any point thereafter if they experienced an involuntary drop in income. On the 1997 staff survey, most workers — particularly in New Haven — said that they did not stress a banking message; staff were more likely to use the time limit to motivate recipients to cooperate with program rules or find a job.
C. Effect of the Time Limit on Jobs First Group Members’ Behavior
Figure 3.1 shows the responses to a number of questions from the Interim Client Survey that asked whether the time limit affected Jobs First group members’ behavior. (Appendix A presents responses to some similar questions from the Three-Year Client Survey.) Most respondents reported that the time limit had not motivated them to take the specified steps asked about. For example, only about one-third (35 percent) said that because of the time limit they decided not to apply for welfare when they could have. Somewhat surprisingly, respondents were most likely to report that the time limit caused them to enter an education or training program. It is not clear whether these respondents were referring to activities arranged by DSS or others that they entered on their own.
III. The Time-Limit Review Process
Jobs First includes provisions for six-month extensions of TFA benefits under certain circumstances. Extensions may be granted once a recipient reaches the time limit, at the end of an extension period, or if a recipient is initially denied an extension but later experiences an involuntary drop in income and seeks to return to TFA. During the study period, the program did not limit the number of extensions people could receive.10
[D] |
| SOURCE: MDRC calculations using Interim Client
Survey data. NOTES: These items were asked
of Jobs First group survey respondents who reported that they received
cash assistance since random assignment and who indicated that they
were subject to a limit. The total sample size is 289 (69 in Manchester
and 220 in New Haven). Estimates were adjusted using ordinary least
squares, controlling for pre-random assignment characteristics of
sample members. Results in this figure were weighted to make them more representative of the full sample. (a)Numbers represent the percentage of respondents who agreed a little or agreed a lot with specified statement. |
The key step in determining whether recipients will receive an initial extension is the exit interview, which is scheduled to occur during their twentieth countable month of benefit receipt. People who are granted a six-month extension are scheduled for another exit interview during the fifth month of the extension, if they are still receiving cash benefits at that point. These subsequent exit interviews generally follow the same process as the 20-month interviews. Until July 1998, exit interviews were conducted by DSS case maintenance workers; in some instances, employment services workers also participated. Since then, the interviews have been conducted by DSS Family Independence Representatives (FI-Reps). (DSS intake staff use a similar process when someone past the time limit reapplies for benefits.) After establishing that the recipient still meets TFA eligibility requirements, such as having a child in the home and not exceeding asset limits, staff assess each case by asking four basic questions (see Figure 3.2):
-
Is the recipient eligible for an exemption? If she is, her time-limit clock is suspended; she continues to receive benefits, but months of receipt do not count toward the 21-month time limit. If she is not, staff ask the second question.
-
Is the recipient’s family income equal to or greater than the payment standard (the maximum grant amount for her family size)? If it is, her benefits are discontinued (although she may reapply later if her income drops); if it is not, staff ask the third question.
-
Has she made a good-faith effort to find and retain employment? If she has, she is granted an extension. If she has not, staff ask the fourth question.
-
Are there circumstances beyond the recipient’s control that prevent her from working? If there are, she is granted an extension. If there are not, her benefits are discontinued.
In addition, during the exit interview, staff redetermine eligibility for Food Stamps and Medicaid and refer people to appropriate services. The following sections provide more information on the time-limit review process.
A. Determining Whether an Exemption Applies
After checking the accuracy of recipients’ time-limit clock counter, the first step in an exit interview is to determine whether they are eligible for an exemption from the Jobs First program. As noted in Chapter 1, those who are exempt are not required to participate in any employment-related activities, and their time-limit clock is suspended (they continue to receive TFA benefits, but months of receipt do not count toward the 21-month time limit). DSS staff reported that it is rare for someone to receive an exemption during an exit interview.
B. Measuring Income
If recipients are not exempt from Jobs First, the second step in the exit interview is to calculate their monthly income. If it is equal to or greater than the payment standard, the maximum grant amount for their family size, they are not eligible for an extension. (See Table 1.2 for the payment standards for three family sizes.)
[D] |
Income for this determination includes all household income, including earned income (minus a $90 work expense allowance) and unearned income, such as income from child support (but not income from TFA or Food Stamps). DSS staff generally measure earned income by examining recent pay stubs or other relevant records; many staff contact employers to verify earned income. As mentioned in Chapter 2, the earnings information recorded in the Eligibility Management System (EMS) is sometimes out of date. At the exit interview, many people who were assumed to be working report that they are no longer employed. (Until the exit interview, recipients have little incentive to report a job loss, because it generally does not affect their cash benefits.) When recipients report that they are no longer employed, the question is whether they quit the job without good cause or were fired for willful misconduct during the last six months of assistance; if either is discovered to be true, they are determined not to have made a good-faith effort to find employment (discussed below).
Recipients whose income is equal to or greater than the payment standard are usually eligible for up to two years of transitional Medicaid, since this benefit is provided to those who are employed when their case closes, and they may be eligible for Food Stamps. While people are in the Jobs First program, their cash assistance grant is counted as income in determining their Food Stamp benefit. However, the enhanced earned income disregard applies to the Food Stamp grant calculation — so all earnings are disregarded as long as recipients are earning below the federal poverty level. When recipients lose their TFA grant, the grant amount is no longer counted against their Food Stamp benefits, but they also lose the enhanced earned income disregard.11 These two changes work in opposite directions, but for many individuals they result in a lower Food Stamp benefit than before the time limit.
Figure 3.3 illustrates the drop in monthly income for families whose benefits are discontinued because their income is above the payment standard. The figure shows this drop for a parent working part time with earnings just above the payment standard and for a parent working full time with earnings just below the poverty level (in 1998, about midway through the study period). In the first scenario, a parent who works 25 hours per week at $6.25 per hour loses $542 of income after the time limit. In the second scenario, a parent who works 40 hours per week at $6.25 per hour loses $688. As the figure shows, parents who work full time experience a sharp decline in their Food Stamp benefit amount (from $247 to $101); in the Food Stamp benefit calculation, the loss of the earned income disregard — and the corresponding increase in the amount of earnings counted in the calculation — more than offsets the decrease in cash assistance.
For the examples in Figure 3.3, the income drop at the time limit is a secondary effect of the Jobs First earned income disregard policy. The figure illustrates how much income the same families would have under traditional AFDC rules (that is, without the disregard). Parents who work full time would be ineligible for cash assistance and thus would have the same income as the Jobs First participants after their grant was discontinued; parents who work part time would be eligible for a small cash grant. The horizontal dashed line at $790 shows that, under any of the scenarios, working parents even those working part time would have more income than parents who were receiving cash assistance and not working (although expenses might also be lower for nonworking parents).
[D] |
| SOURCE: MDRC calculations based on AFDC/TFA,
Food Stamp, and federal and State of Connecticut income tax rules
for June 1998. NOTES: Calculations do not
account for work-related expenses and assume the parent has no income
from sources not shown (e.g., child support, Supplemental Security
Income). The Food Stamp calculation assumes a monthly rental expense of $366. This calculation disregards 70 percent of net income, which includes the AFDC/TFA grant but excludes a $134 standard deduction and up to $250 of excess shelter costs. For clients in the Jobs First group, all earned income is disregarded before the time limit, and 20 percent of earned income is disregarded after the time limit. For clients in the AFDC group, 20 percent of earned income is disregarded. The Jobs First cash assistance calculation disregards all earned income before the time limit. The AFDC cash assistance calculation disregards $120 in earned income (in accordance with rules for the 5th through 12th months of employment), and applies fill-the-gap budgeting rules. The EIC amount reflects one-twelfth of the total annual credit, although most families receive the credit in an annual lump sum. Monthly net earnings are calculated by subtracting applicable payroll taxes from gross earnings. Federal and state income taxes do not apply at these income levels. Rounding may cause slight discrepancies in the calculation of sums. |
Recipients whose grant is closed because they are over income may be granted an extension later if their income drops below the payment standard and they have made a good-faith effort to find employment (before reaching the time limit and afterward). DSS training materials note that people may request an extension at any point from month 20 onward but do not emphasize that staff should remind them of this fact when they are denied an extension. Some DSS staff said that they routinely tell this to recipients, but others were much less direct.
C. Determining Whether Recipients Have Made a Good-Faith Effort
If recipients are not exempt from Jobs First and do not have income equal to or greater than the payment standard, then the third step in the exit interview is to determine whether they have made a good-faith effort to find and retain employment. Staff in Manchester and New Haven said that most of the recipients with income below the payment standard at their 20-month exit interview are determined to have made a good-faith effort and, thus, are granted an extension. Some recipients are denied subsequent extensions based on lack of good-faith effort (see, for example, Figure 3.6 on page 78).
1. The initial good-faith determination. The Jobs First program stipulates that recipients are assumed to have made a good-faith effort if none of the following conditions is true:
-
They failed the work test12 and had one employment services sanction during the first 20 months of TFA receipt; or
-
They received two or more employment services sanctions during the first 20 months of TFA receipt; or
-
They did any one of the following things without good cause during the last six months of TFA receipt: quit a job, refused a job, were fired from a job for willful misconduct, reduced hours of employment, or refused additional hours of employment.13
The first two good-faith conditions are relatively straightforward: Good faith is presumed unless there are specific indicators of program noncompliance (failing the work test and sanctions).14 Staff conducting exit interviews can easily check for such indicators using Connecticut’s automated case information system (EMS) and/or recipients’ case files. As discussed in Chapter 2, several factors made it difficult for staff to closely monitor recipients’ participation in employment activities during the pre-time-limit period, and few recipients have been sanctioned. In many cases, it may not be clear precisely what the recipients were doing while in the program, but if they had not failed the work test or had not been sanctioned, they would pass the first two good-faith conditions. In other words, these good-faith conditions are based on the absence of evidence of noncompliance rather than on evidence of compliance.
The third good-faith condition listed above, based on various employment issues, is less straightforward and often relies on staff members’ subjective assessments. If recipients, within the last six months of assistance, quit or refused a job, were fired from a job for willful misconduct, or reduced hours or refused additional hours of hours of work, the staff member must evaluate whether or not they had good cause for this behavior. DSS supervisors and staff reported that there is “a lot of gray” in making this good-cause determination and acknowledged that different workers might make different determinations based on the same evidence (this is also true for good-cause decisions made during the pre-time-limit period). At the same time, they were skeptical that more detailed rules would ensure uniformity across workers and cases, because there are too many variables to anticipate. In New Haven, supervisors met regularly, in part to review cases to try to develop consistent ways of determining good cause, and, in 2000, training on good-cause procedures was provided to staff.
Many cases require extensive investigation, including talking with former employers, but staff reported that it often remains unclear why recipients left a job, reduced hours, and so on. One important factor is whether the recipients are approved for unemployment insurance (UI) benefits, which generally are not available to those who voluntarily quit their job or were fired for willful misconduct (individuals are supposed to pursue other income sources, including UI, before turning to welfare). DSS rules state that good cause should be assumed if recipients are approved for UI but that further investigation is required if their UI application is denied (because the DSS definition of good cause is more expansive than the UI definition). The UI process may affect how employers describe a job separation: Employers have an incentive to characterize separations as voluntary quits or terminations for willful misconduct because their UI taxes may increase if former employees receive benefits.
Although DSS staff were familiar with their department’s policies regarding UI approval, it appears that some of them rely quite heavily on the UI determination; for example, some workers said that if a recipient’s UI application was denied because it was determined that she had quit voluntarily or been fired for willful misconduct, they would be very unlikely to grant good cause. Some DSS staff also expressed frustration that the state Department of Labor (DOL) staff sometimes approve UI benefits in situations in which, in their view, former employees clearly quit voluntarily or were fired for willful misconduct (benefits may be approved if the employer fails to show up for the UI determination hearing).
Some staff reported that they weigh recipients’ history and effort level: If recipients have been responsible and have tried to solve problems that have arisen, staff are more likely to conclude good cause than if they have been uncooperative. Others mentioned that when determining good cause, they weigh whether they would win a fair hearing. They noted that in order to win a hearing, DSS must have documentation of the noncompliance, which in some cases is difficult to obtain. For example, staff noted that some employers are willing to discuss job separations candidly by phone but are unwilling to put anything in writing, out of concern that their statements might be used as evidence in a lawsuit.
In New Haven, FI-Reps are required to keep a monthly log of all their conciliations and good causes granted. Although DSS managers reported that their goal is not to increase sanctions or grant closures, some staff reported that they feel pressure to less readily grant good cause. (As noted in Chapter 2, the New Haven office consistently has lower sanctioning rates than other DSS offices.) Manchester staff said that they felt no pressure in either direction.
Most staff members seem to err on the side of granting extensions in uncertain situations. For example, in one case that MDRC staff reviewed in 1998, the DSS staff member granted a good-faith extension even though the recipient had recently quit a job; the recipient said that she had quit because she was pregnant and could not lift heavy objects, and she produced a physician’s confirmation. In another case, a staff member granted an extension when a recipient had quit a job after becoming homeless. In New Haven, when a recipient’s income is below the payment standard, FI-Reps must discuss the case with their supervisor before assistance can be discontinued; the supervisor and the DSS office manager must sign off on each case closure.
2. Restoring good faith. Recipients who have been determined not to have made a good-faith effort to find and retain employment can restore good faith by successfully completing an Individual Performance Contract (IPC). As discussed in more detail in Chapter 2, this typically entails complying with an employability plan that prescribes work-related activities. By the time of the exit interview, some recipients have successfully completed an IPC and thus have restored good faith and will receive an extension; others have failed to complete an IPC and will not be granted an extension (unless there are circumstances beyond their control, as described in the following section). Recipients who are deemed not to have made a good-faith effort and who have not yet been offered an IPC are offered one at the exit interview. (IPCs are offered only during the first 21 months of TFA receipt, not during extensions.) As is noted in Chapter 2, most recipients who begin IPCs successfully complete them and restore good faith.
3. Good faith in the post-time-limit period. During extensions, recipients are officially subject to a one-strike policy: Those who fail to comply with any Jobs First requirements without good cause can lose their grant permanently. In practice, however, most staff give people a few chances before they record the official strike. During interviews, staff said that because permanent grant cancellation is “so harsh,” the noncompliance has to be very obvious, and that most recipients whose grants were closed during an extension were given several chances to comply. This remained true as of MDRC’s final field visit in late 2000.
Originally, Jobs First policy stated that if, at any point after reaching month 21, a recipient quit a job, refused a job, was fired for willful misconduct, reduced her hours of work, or refused additional hours, without good cause, she was not eligible for an extension. This applied whether the individual was receiving cash assistance benefits at the time or not. The state changed the policy in October 1999: When someone is off welfare after month 21 and is reapplying for benefits, her job behavior only in the previous 120 days in considered.
DSS staff reported that the difficulties of determining good cause for job separations after month 21 are similar to those at the initial extension determination. They also said that it is often difficult to sort out the circumstances regarding noncompliance with employment services requirements during extensions. Earlier in the follow-up period, this was in large part because no single DOL or Regional Workforce Development Board (RWDB) case manager was responsible for each case. Later, monitoring participation was complicated by the shift to a new subcontracted case management system, which took some time to fully implement. In general, the program seemed to be somewhat tougher during extensions than during the pre-time-limit period, but not as tough as the one-strike policy may sound.
Over time, the proportion of TFA cases that are in an extension has remained relatively stable. For example, in May 1999, 41 percent of the TFA cases that were subject to the time limit were in an extension; in May 2000, 39 percent; and in May 2001, 38 percent. The proportion of cases with multiple extensions, however, has increased over time. In May 1999, only 1 percent of time-limited cases were in their fourth extension or higher; in May 2000, 14 percent; and in May 2001, 16 percent. In interviews in late 2000, staff in New Haven mentioned a recent push to re-engage multiple-extension clients in employment services to help them find work.
D. Determining Whether There Are Circumstances Beyond Recipients’ Control
If recipients are not eligible for an exemption, do not have income equal to or greater than the payment standard, and have not been determined by DSS staff to have made a good-faith effort to find employment (and did not restore good faith through an IPC), then the fourth step in the exit interview is to establish whether there are circumstances beyond recipients’ control that currently prevent them from working; if so, they are eligible to receive an extension. DSS considers circumstances beyond control to be unusual or unexpected events, such as domestic violence, death of an immediate family member, or a fire or flood that results in loss of housing. Staff reported that they have had little experience with this policy; it is relevant only for recipients with income below the payment standard who are deemed not to have made a good-faith effort, and there have been few such cases.
Recipients who have income below the payment standard and are denied an extension are referred to Safety Net services. These services are provided by a contractor and are designed to prevent harm to the children of families whose cash grant is closed, by helping meet their basic needs. (Section VIII presents more information on Safety Net services and other services offered after the time limit.)
E. Dealing with Recipients Who Fail to Attend an Exit Interview
Recipients who do not attend an exit interview do not receive an extension. Exit interviews are typically scheduled around the middle of month 20; recipients who miss their scheduled interview usually have several weeks either to show up or to contact their DSS worker before TFA benefits and Food Stamps are discontinued. DSS usually sends a notice on about the twelfth day of month 21 informing recipients that they have failed to contact DSS and that their benefits will be discontinued. However, recipients who contact their worker before the end of month 21 can still qualify for an extension without any disruption in TFA or Food Stamp benefits. For those who do not, all benefits — TFA, Food Stamps, and Medicaid — will be discontinued (as they would be if a recipient missed any redetermination meeting), and transitional Medicaid will be granted only if EMS shows earnings for the recipient. Individuals can still request an extension at a later point, but they are treated as new applicants, and there may be an interruption in benefits.
Staff assume that most recipients who do not show up for their exit interview must have income above the payment standard and thus know that they would not qualify for an extension. This assumption is supported by a survey of individuals from six areas of Connecticut whose benefits were discontinued at the time limit (see Chapter 1): 77 percent of those surveyed who did not attend an exit interview reported that they were employed during their last month of benefits (although not necessarily earning above the payment standard).15 Such recipients may also assume that they are no longer eligible for Food Stamp benefits, although this may not be correct. During the early months that people started to reach the time limit, there were concerns that the exit interview notice did not make it sufficiently clear to recipients that they needed to attend the interview to continue their Food Stamp eligibility. Food Stamp benefits were discontinued for a number of clients because they did not attend an interview. DSS subsequently revised the notice and reinstated Food Stamp benefits for those whose benefits were discontinued.
There is concern that some recipients who do not attend their exit interview may have income below the payment standard and may not fully understand the purpose of the interview. They may assume that since they have reached the time limit, they must leave welfare; they may not know that it is possible to receive an extension (despite the fact that the letter scheduling them for the exit interview clearly notes this).
IV. How Quickly Jobs First Group Members Reached the Time Limit
Figure 3.4 illustrates how quickly Jobs First group members accumulated months of TFA benefits and how quickly they reached the time limit. The upper, solid line of the figure represents the percentage of the Jobs First group who accumulated at least 21 months of TFA benefits after random assignment, by the number of months that elapsed since random assignment. As the figure shows, 43 percent of the Jobs First group received 21 continuous months of TFA benefits after entering Jobs First. After 22 months of follow-up, 48 percent of the Jobs First group had received at least 21 months of TFA benefits, indicating that 5 percent of this group had left welfare for one month before returning. The percentage of the sample who received at least 21 months of TFA benefits increased slightly each month: By month 48, 58 percent of the Jobs First group had received at least 21 months of benefits. The fairly flat slope of the line implies that most people who left welfare did not quickly return and use up their remaining months.
Table 3.2 shows the number of months of TFA receipt for the Jobs First group by the end of the four-year follow-up period for this report. Of the 58 percent of the sample who had accumulated at least 21 months of benefits, some had received only 21 months (9 percent), while others had received benefits for almost the entire 48 months. As the table shows, 5 percent of the Jobs First group did not receive any TFA benefits, because their application was either withdrawn or denied.
[D] |
| Measure | Percentage of Sample |
|---|---|
| Months received TFA benefits | |
| 0 | 5.3 |
| 1-5 | 8.6 |
| 6-10 | 8.8 |
| 11-15 | 9.0 |
| 16-20 | 10.7 |
| 21 | 9.1 |
| 22-27 | 13.7 |
| 28-33 | 10.9 |
| 34-39 | 8.3 |
| 40-45 | 7.0 |
| 46+ | 8.6 |
| Months of TFA receipt counted toward the time limit | |
| 0 | 12.3 |
| 1- 5 | 8.2 |
| 6-10 | 7.8 |
| 11-15 | 8.8 |
| 16-20 | 10.7 |
| 21 | 12.2 |
| 22-27 | 13.9 |
| 28-33 | 9.9 |
| 34-39 | 7.6 |
| 40-45 | 5.1 |
| 46+ | 3.5 |
| Sample size | 2,396 |
| SOURCE: MDRC calculations using Connecticut
AFDC/TFA records. NOTE: Distributions may not add up to 100 percent because of rounding. |
The lower, dashed line of Figure 3.4 shows the proportion of the Jobs First group who reached the time limit: 29 percent reached the time limit by month 21 (that is, after continuous receipt), 34 percent by month 22, and 53 percent by month 48. The difference between the two lines on the figure represents the percentage of sample members who received 21 months of cash assistance but did not receive 21 countable months of cash assistance; they were exempt for some of their months on TFA. As the lines indicate, a substantial proportion of sample members who received at least 21 months of TFA benefits did not reach the time limit, although this proportion decreased over time (which suggests that exemptions expired over time).
It is important to note that reaching the time limit is not synonymous with losing cash benefits. In fact, an analysis presented in the next section shows that fewer than half of those who reached the time limit had their grant closed at that point.
V. After the Time Limit: Extensions and Case Closures
This section focuses on a subset of Jobs First group members who received at least 21 countable months of cash assistance and thus reached the time limit. Just over half of these sample members received an initial six-month extension of their benefits. All the extensions were granted because recipients had income below the payment standard and DSS determined that they had made a good-faith effort to find employment. Among recipients who did not receive an extension at the time limit, most had income equal to or greater than the payment standard; only one person was denied an extension for lack of a good-faith effort. Just 14 percent of the sample members were receiving TFA benefits 36 months after reaching the time limit.
The sample for the analysis discussed in this section is a subset of those who reached the time limit by March 1998.16 ) This sample was used for a similar analysis presented in the evaluation’s interim report. The March 1998 cutoff was chosen to allow at least 15 months of post-time-limit follow-up for each sample member in the interim report, and it allows 36 months of follow-up here. According to TFA administrative records, 353 Jobs First group members received at least 21 countable months of TFA benefits by March 1998. The three figures presented in this section explore a subsample of 100 randomly selected from these 353. Using EMS data and AFDC/TFA records, the analysis examines the outcomes of exit interviews for these 100 cases and traces their TFA activity during the 36 months following the time limit.
A. Outcomes of the Exit Interview
Figure 3.5 shows the initial outcomes for the subset of 100 Jobs First group members who reached the time limit. Just over half of those who reached the time limit (55 of 100) were initially granted a six-month extension of their TFA benefits.17 All these extensions were granted because the recipients had income below the payment standard and were deemed to have made a good-faith effort to find employment.
[D] |
Just under half of the recipients who reached the time limit (43 of 100) had their TFA benefits discontinued at that point. Most of these individuals (32 of 43) had income that was equal to or greater than the payment standard for their family size. In general, these people would have become ineligible for welfare earlier had it not been for the enhanced earned income disregard. Only one person had income below the payment standard and was deemed not to have made a good-faith effort to find employment. As noted earlier, those who do not attend an exit interview do not receive an extension; this occurred for 10 of the 43. (Most of these individuals were working at the time of their exit interview and may have assumed that they were not eligible for an extension.)18
Two of the 100 Jobs First group members were found to be eligible for an exemption from the program at their exit interview. In one case, the staff member realized that the recipient had not been granted the usual 12-month exemption when her baby was born, and she granted a retroactive exemption. In the other case, DSS granted a medical exemption.
In sum, of the 58 people who attended an exit interview and had income below the payment standard, only 1 had her benefits discontinued; 55 were granted extensions, and 2 received exemptions.19
B. TFA Status During the 36 Months After the Time Limit
Figures 3.6 and 3.7 track the TFA activity during the 36 months following the time limit for the 55 individuals who initially received an extension and the 43 who did not. The two recipients who were granted an exemption at their exit interview are not included in either figure.20
As illustrated in Figure 3.6, only 8 of the 55 recipients who initially received an extension stayed on the rolls in all of the 36 months after the time limit; the remaining 47 stopped receiving benefits at some point. As the middle box shows, 16 of 55 left TFA but later came back on assistance. About twice as many 31 of 55 left and did not come back on the rolls. Among all those who left, 5 had their grants closed because they were deemed not to have made a good-faith effort to find employment (2 from the middle box and 3 from the box on the right). (Two others were denied an extension for lack of good-faith effort when reapplying for benefits.) Most others had their case closed at a subsequent exit interview because their income rose above the payment standard or during an extension because their income rose above the poverty level. Thirty-six months after reaching the time limit, only 12 of 55 were still receiving cash assistance.
[D] |
[D] |
Figure 3.7 shows TFA activity for the 43 recipients who were not granted an extension at their 20-month exit interview. Recipients whose cases are closed because their income is over the payment standard (as well as those who fail to attend the exit interview) may be granted an extension later if their income drops below the payment standard and they have made a good-faith effort to find employment. However, of the 43 people whose cases were closed at the time limit, 37 never returned to the rolls in the subsequent 36 months. Thirteen of the 37 applied for TFA at some point but did not start to receive benefits; most were found to be financially ineligible or did not complete the necessary paperwork, and 2 were found not to have made a good-faith effort to find work. Of the 43 people whose cases were closed at the time limit, only 6 came back on TFA at some point, and, by the end of the follow-up period, none were still receiving benefits (1 of the 6 was denied a subsequent extension because she was found to have quit a job with no good cause).
In sum, of the 100 Jobs First group members who reached the time limit, 63 were granted an extension at some point, but only 14 were receiving TFA benefits 36 months later (12 of the 55 who initially received an extension and 2 who received an exemption at the time limit). In other words, the vast majority (86 of 100) of those who reached the time limit were off the welfare rolls 36 months later. A total of 11 of the 100 recipients were denied an extension because they were deemed not to have made a good-faith effort to find work.21
C. Post-Time-Limit Exits
Using results from the 100 cases examined above, as well as administrative records of TFA payments and case information from EMS, MDRC estimated what proportion of the Jobs First group exited TFA for various reasons after the time limit. As noted earlier, 53 percent of Jobs First group members reached the time limit within the four-year follow-up period for this report. About one-third of Jobs First group members ever left because of the time limit the vast majority because they had income over the payment standard. About 5 percent had their grant closed during an extension because they were deemed not to have made a good-faith effort to find employment. (Under Connecticut’s definition, recipients whose grant is closed because of noncompliance during an extension are not considered to have left TFA because of the time limit.) Another roughly 10 percent left the rolls after month 21 but for reasons unrelated to the time limit. About 5 percent of the Jobs First group reached the time limit but never left TFA.
VI. Characteristics of Cases with Multiple Extensions
As noted above, a substantial proportion of Jobs First group members who reach the time limit receive at least one 6-month extension. Although the analysis shows that most people who receive extensions leave TFA in the subsequent 36 months, some receive multiple extensions and remain on TFA. As noted earlier, statewide statistics for May 2001 showed that about 16 percent of the cash assistance recipients subject to the time limit (about 2,000 families) were in at least their fourth extension.
Table 3.3 presents selected characteristics of Jobs First group members at the time of random assignment, by the number of countable months of TFA received in the four years following random assignment. The right-hand column includes people who received at least 40 countable months of TFA; this group presumably received at least four 6-month extensions (just 9 percent of the Jobs First group).
The table shows that Jobs First group members who received at least four extensions were more disadvantaged than other sample members when they entered the study. They were less likely to have a high school diploma, had longer histories of prior cash assistance receipt, and were less likely to have been employed in the quarter of random assignment. They had more children, which is not surprising, since families who reach the time limit with income below the payment standard are usually granted extensions, and larger families have a higher payment standard. Multiple-extension group members were also substantially more likely to be black22 and to live in public or subsidized housing.
VII. Outcomes for Families After the Time Limit
Table 3.4 presents some measures from the Three-Year Client Survey for Jobs First group members who left the TFA rolls because of the time limit.23 For this analysis, only recipients who exited prior to their survey interview are included in the “closed because of time limit” category.
The table shows that individuals who left the welfare rolls because of the time limit were more likely to be employed in the month before their interview and earned more in that month, compared with the full survey sample. This is not surprising since, as noted earlier, most recipients whose case was closed because of the time limit were working and had income above the payment standard. They also were more likely to have health insurance coverage for themselves and their children, and they were slightly less likely to have reported facing severe hardships in the prior year.
| Characteristic | Number of Countable Months of TFA Received | |||
|---|---|---|---|---|
| 0-20a | 21 | 22-39 | 40-48 | |
| Demographic characteristics | ||||
| Race/ethnicity (%) | ||||
| White, non-Hispanic | 43.1 | 47.3 | 33.1 | 19.3 |
| Black, non-Hispanic | 33.2 | 31.3 | 46.3 | 53.8 |
| Hispanic | 22.6 | 19.9 | 20.3 | 25.9 |
| Other | 1.1 | 1.4 | 0.3 | 1.0 |
| Marital status (%) | ||||
| Never married | 56.9 | 64.8 | 71.3 | 76.4 |
| Married, living with spouse | 1.8 | 1.0 | 1.1 | 0.5 |
| Married, living apart | 18.3 | 16.7 | 10.1 | 11.6 |
| Separated or divorced | 21.7 | 16.4 | 17.1 | 11.1 |
| Other | 1.3 | 1.0 | 0.4 | 0.5 |
| Average number of children (%) | ||||
| None | 8.7 | 9.3 | 7.8 | 3.5 |
| 1 or 2 | 73.6 | 67.2 | 63.4 | 58.5 |
| 3 or more | 17.7 | 23.4 | 28.8 | 38.0 |
| Employment status | ||||
| Employed in quarter of random assignment (%) | 43.8 | 45.7 | 36.3 | 26.0 |
| Education status | ||||
| Highest degree/diploma earned (%) | ||||
| GEDb | 12.5 | 13.6 | 12.1 | 8.1 |
| High school diploma | 48.9 | 51.1 | 43.7 | 39.1 |
| Public assistance status | ||||
| Total prior AFDC receiptc (%) | ||||
| None | 24.8 | 14.3 | 12.7 | 5.7 |
| Less than 2 years | 27.6 | 22.6 | 18.9 | 16.1 |
| 2 years or more but less than 5 years | 19.5 | 19.4 | 25.0 | 20.7 |
| 5 years or more | 28.1 | 43.7 | 43.4 | 57.5 |
| Resides in public or subsidized housing (%) | 28.2 | 36.0 | 43.4 | 47.4 |
| Percent of total sampled | 41.8 | 12.2 | 31.5 | 8.5 |
| Sample size | 1,001 | 293 | 754 | 204 |
| SOURCE: MDRC calculations using Background
Information Form data and Connecticut's unemployment insurance (UI)
earnings records. NOTES: Invalid or missing values are not included in individual variable distributions. Rounding may cause slight discrepancies in the calculation of sums and differences. aDoes not Include 144 sample members who received more than 21 months of TFA but were exempted at some point.(back) bThe General Educational Development (GED) certificate is given to those who pass the GED test and is intended to signify knowledge of basic high school subjects.(back) cThis refers to the total number of months accumulated from one spell or more on an individual's own or spouse's AFDC case. It does not include AFDC receipt under a parent's name.(back) dThe percentages do not sum to 100 percent because of the exclusion of 144 sample members who were exempted.(back) |
| Outcome | Full Sample | Closed Because of Time Limit |
|---|---|---|
| Employed last month (%) | 62.6 | 77.5 |
| Earnings last month ($) | 672 | 801 |
| Average number living in household | 3.6 | 3.5 |
| Average number of children in household | 2.0 | 2.1 |
| Respondent lives with at least one other adult (%) | 42.8 | 31.6 |
| Respondent is currently married and living with spouse (%) | 9.4 | 6.0 |
| Respondent currently married and living with spouse (%) | 22.4 | 16.2 |
| Respondent owns a car, van, or truck (%) | 42.3 | 39.3 |
| Respondent has debt (%) | 65.2 | 62.4 |
| Respondent was ever homeless and living on street in past year (%) | 2.3 | 1.7 |
| Respondent didn't pay full amount of rent or mortgage in past year (%) | 35.0 | 38.5 |
| Respondent lives with family/friends and pays part of rent or mortgage (%) | 10.2 | 10.3 |
| Respondent has ever moved since random assignment (%) | 66.3 | 63.2 |
| Average amount of respondent's savings ($) | 163 | 144 |
| Respondent has no health insurance (%) | 11.9 | 3.4 |
| Children have no health insurance (%) | 3.6 | 1.7 |
| Two or more housing problemsa (%) | 18.0 | 20.2 |
| Four or more neighborhood problemsb (%) | 25.2 | 24.8 |
| Food insecure with hungerc (%) | 20.9 | 15.5 |
| Four or more material hardshipsd (%) | 16.0 | 12.9 |
| Three or more social services usede (%) | 26.4 | 26.5 |
| Three or more severe hardshipsf (%) | 12.4 | 8.3 |
| Sample size | 1,249 | 117 |
| SOURCE: MDRC calculations from the Three-Year
Client Survey data. NOTES: For the analysis presented in this table, Jobs First group members were defined as having left because of the time limit if their TFA case closed upon reaching the time limit or at the end of an extension. aHousing problems include the following: leaky roof or ceiling; broken plumbing; broken windows; electrical problems; roaches/insects; heating system problems; and broken appliances.(back) bNeighborhood problems include the following: unemployment; drug users or pushers; crime, assault, or burglaries; run-down buildings and yards; and noise, odors, or heavy traffic.(back) cThe six-item Food Security Scale recommended by the United States Department of Agriculture was used to measure food security. The items in the scale include questions about food consumed and the kind of things people resort to when money allocated for food is exhausted. The scale ranges from 1 to 6; two or more affirmatives indicate food insecurity, and five or more affirmatives indicate food insecurity with hunger.(back) dMaterial hardships include the following (all over the prior year): could not pay full amount of rent or mortgage; evicted for not paying rent/mortgage; could not pay full amount of utility bills; electricity or gas turned off; telephone disconnected; unmet medical needs; and unmet dental needs.(back) eSocial services include the following: rental assistance programs; utility assistance programs; prescription drug assistance programs; food banks; soup kitchens; and secondhand clothes.(back) f"Severe hardships" are based on the categories above and include: four or more neighborhood problems; two or more housing problems; four or more material hardships; three or more social services used; and food insecure with hunger.(back) |
VIII. Services After the Time Limit
This section describes the various services that are available after people reach the time limit. The type of service depends on the situation: People who are in an extension receive employment services; those whose grants have been discontinued and who have income equal to or greater than the payment standard may be eligible for transitional rental assistance; and those whose grants have been discontinued and who have income below the payment standard are eligible for Safety Net services.
A. Employment Services for People Who Received an Extension
As mentioned, in mid-2001, a substantial proportion (38 percent) of the recipients subject to the time limit statewide were in an extension. When recipients receive an extension of TFA benefits, whether at the time limit or at a later point, DSS requires them to participate in employment services that are intended to help them find a job or increase their earnings to exceed the payment standard. As noted earlier, TFA benefits can be discontinued for recipients who do not comply with any part of their employment program during an extension period, and such individuals are generally not eligible for further extensions. Of the 63 sample members examined earlier who received an extension at some point, 6 had their grant subsequently closed for lack of a good-faith effort.
Through mid-1998, DSS contracted with the Connecticut Council of Family Service Agencies (CCFSA, the organization that provides Safety Net and IPC services) to provide employment services to people in an extension. Because the contract was not large enough to cover all extension participants, both research sites supplemented the contract: The Manchester and New Haven DSS offices provided some services, and the New Haven DSS office established a second contract with a local employment services provider. As discussed in Chapter 2, in mid-1998, overall responsibility for providing employment-related services to cash assistance recipients shifted from DSS to the state Department of Labor (DOL). Then services for individuals in extensions were provided by DOL, the RWDBs, and their contractors. In late 1999, the RWDBs began subcontracting with outside agencies to provide case management services to cash recipients, including those in extensions.
Interviews with staff indicated that recipients who were in an extension received the same basic employment services that recipients in the pre-time-limit period received, including job search assistance and short-term training. Also, staff said that, as in the pre-time-limit period, participation among recipients in extensions was not closely monitored. Some staff said that they imparted to recipients in an extension a heightened urgency for achieving self-sufficiency and that they were less likely to place these individuals in longer-term activities. Recipients in an extension were sometimes referred to Project SOAR, which, as discussed in Chapter 2, worked with individuals to help them engage in program services. (As noted in Chapter 2, recipients beyond month 21 of assistance were not eligible for an Individual Performance Contract.)
B. The Transitionary Rental Assistance Program
As noted earlier, most of those whose benefits were discontinued at the time limit had income equal to or higher than the payment standard. In addition, many of those who initially received an extension later left the welfare rolls because their income rose above the payment standard. Individuals whose TFA grants are closed and who are over income may be eligible for the Transitionary Rental Assistance Program (T-RAP).24 This program is intended to soften the blow from the loss of assistance, which for some families may be significant, especially if their Food Stamps also decrease. T-RAP provides up to one year of monthly rental assistance, paid directly to the landlord. To be eligible, individuals cannot be living in public or subsidized (Section 8) housing and must be living in privately owned rental housing.
T-RAP was originally administered by Community Action Agencies under contract to DSS, and it is currently administered by a for-profit contractor.25 DSS is responsible for informing clients about the rental assistance program, but individuals must contact the T-RAP agency themselves. Funding for the program is limited, so rental assistance is not available to all eligible families.26
C. Safety Net Services
DSS contracts with the United Way, which then subcontracts with CCFSA to operate the Safety Net program, which provides services to people whose welfare grant is discontinued at or after the time limit because they were deemed not to have made a good-faith effort to find employment. (Cases that are closed because of the 60-month time limit are not eligible for Safety Net services.) The primary purpose of Safety Net is to prevent harm to the children of families whose cash benefits are “permanently” closed because of the state’s 21-month time limit. (People in Safety Net can return to welfare only if they qualify for an exemption, or if their case was erroneously closed, or if they have circumstances beyond their control, as discussed earlier.) Safety Net’s secondary purpose is to foster the economic self-sufficiency of these families.
The Safety Net program links families with existing community services to ensure that their basic needs are met. If appropriate community resources are not available, the Safety Net agency may provide vouchers and services. Through most of the study period, Safety Net services were not time-limited; families could receive services indefinitely, and if they left Safety Net, they could return at any time, as long as their income remained below the welfare payment standard. Beginning in 2000, the state limited most families to 18 months of Safety Net services.
Drawing on interviews with staff and a mini review of cases files,27 this section describes the implementation of the Safety Net program in Manchester and New Haven.
1. The referral process and initial contact. As noted earlier, most of the individuals whose grants were discontinued were over income; thus, relatively few individuals have been referred to Safety Net services. Analysis of referral records from the United Way/Infoline and CCFSA showed that about 5 percent of the Jobs First group (about 120 of 2,400) had been referred to Safety Net by mid-2001. Statewide, referrals to Safety Net increased over time until early 2000 and decreased thereafter.
The referral process for the Safety Net program is similar to that for IPC services: DSS sends the names of those whose family income is below the payment standard and whose cash assistance has been discontinued either at the time limit or at any point thereafter to Info-line.28 Infoline verifies the accuracy of the referrals, attempts to contact the individuals and collect some basic information, and then faxes the referrals to the CCFSA central office and the appropriate regional office. (Infoline also refers people to other social services, as necessary.)
If individuals are in crisis (for example, homeless or about to be evicted), CCFSA attempts to contact them immediately. In trying to contact people, CCFSA case managers must make phone calls, write one certified letter and one regular letter, and make three to six home visits. Staff reported that many individuals are difficult to reach, because their telephone has been shut off or they have moved.
During the first home visit, CCFSA case managers conduct an assessment covering such issues as participants’ employment and education history, their children’s school and health status, and the family’s financial circumstances (for example, whether the rent and utilities are paid up to date). During a later visit, they conduct an in-depth employment barrier assessment covering such issues as domestic violence, substance abuse, and mental health. According to CCFSA staff, many Safety Net participants have multiple barriers to steady employment.
2. Available services. Safety Net services encompass three main areas: basic needs assistance; case management; and referrals to employment services, counseling, and other services. To help families meet their basic needs, Safety Net provides rental assistance, utility payment assistance, food vouchers, clothing, and bus passes or tokens to help people get to employment activities. There are no monetary limits on any of this assistance, but rental assistance is limited to six months of payments. Between July 1999 and June 2000, over half (58 percent) of Safety Net participants statewide received at least one basic needs payment, and the average expenditure per family was about $1,200. Rental assistance accounted for almost two-thirds of the basic needs payments during that period.29
Intensive case management is a key component of Safety Net. CCFSA case managers visit families at home, and they are available during nontraditional hours, including evenings and weekends. These staff members work with about 20 cases on average (including some Safety Net cases and some IPC cases), which allows them to work closely with people. Staff who were interviewed said that typically they meet with Safety Net participants once every week or two but that they meet several times a week with participants who are experiencing a crisis.
In each of five regions of the state, CCFSA also employs one or two Child and Family Consultants (CFCs), who are available to consult on various clinical issues. CFCs typically have a master’s degree in social work. They are available to discuss cases with case managers and advise appropriate actions or to visit participants at home. When CFCs visit a home, their primary purpose is to evaluate recipients’ problems and refer them to the appropriate clinical services. They are called in on all cases that scored above a certain level on the in-depth assessment of employment barriers and at any time that an issue such as domestic violence, child neglect, or substance abuse becomes apparent. During field research in 1999, CCFSA staff estimated that CFCs were consulted for about one-quarter of Safety Net cases.
CCFSA case managers develop a service plan with each participant. Generally, unless someone has severe barriers, employment is the ultimate goal. Case managers often give participants one-on-one job search assistance, such as help with résumés, how to look for a job, and what to say and do at a job interview. Case managers help those who are employed to develop specific strategies to increase their earnings. For those who have a job and earn above the payment standard, CCFSA sometimes provides transitional case management services — but no financial assistance — for three to six months to help them retain the job.
Case managers also refer Safety Net participants to various services and agencies in the community for counseling, job search assistance, basic education, or job skills training. Case managers refer people both to services at CCFSA agencies — for example, Catholic Family Services in Meriden, outside New Haven, has a licensed mental health clinic and operates a weekly job club — and to services offered at other agencies in the community.
3. Participants’ economic status. CCFSA staff said that if they think someone qualifies for public assistance, they try to help her get benefits. For example, if they believe that an individual is incapacitated and should have received an exemption from the Jobs First program or that the reason she quit a job should have been considered good cause by DSS at the exit interview, they contact DSS and advocate for her. If they succeed, the Safety Net case is closed (as noted below, this has occurred many times). Sometimes case managers help Safety Net participants get Food Stamps, Medicaid, or Supplemental Security Income (SSI).
When asked how they thought that people in the Safety Net program make ends meet without the help of cash assistance, staff noted various means, including part-time work; help from family members, friends, and romantic partners; food pantries, shelters, and other agencies; SSI and child support; and perhaps selling drugs or prostitution. (MDRC has no way to determine the prevalence of any of these means.) Staff in one office have observed that the people who have family support tend to fare the best.
Among Jobs First group members who were identified as having been referred to Safety Net, about 20 percent returned to TFA within a year of their referral. (This is roughly consistent with the number of cases statewide that left Safety Net to return to TFA, as discussed below.) About 70 percent of the Jobs First group members who were referred to Safety Net worked during the next year, but their average earnings were low. (A similar percentage worked in the year prior to Safety Net referral, suggesting that people did not necessarily go to work because their benefits were cancelled.)30
4. Length of time in Safety Net. Staff said that the length of time that people receive Safety Net services varies widely, with some staying in the program for a week and others staying for well over a year. Of the 13 cases that were reviewed in 1999 as part of a mini review of case files for which service start and end dates were available, case duration ranged from about one month to one year and averaged five and a half months. This range partly reflects the diversity of people who are served by Safety Net: Many have multiple barriers to steady employment and receive extended services, while a minority quickly find a job or increase their hours and become ineligible because their income rises above the payment standard. As noted above, in 2000 the state limited Safety Net services to 18 months for most families.
According to CCFSA data, between July 1999 and June 2000, 1,324 cases were served by the Safety Net program across Connecticut; 374 of these cases were closed by the end of the period. Approximately one-third of the closed cases were employed and had income above the payment standard, and another third withdrew from the program (because they refused or no longer needed services or they moved). About 16 percent of the closed cases had their TFA benefits reinstated, and 8 percent began receiving SSI benefits. Just under 10 percent had their case closed because they did not comply with their Safety Net service plan.31
As mentioned, the primary purpose of Safety Net is to ensure the safety of the children in families. CCFSA case managers who think that a child is in danger of neglect or abuse will try to work with the family to remedy the situation. If the situation does not improve, they will contact the Connecticut Department of Children and Families (DCF). If at any point DCF removes a child from the home, the Safety Net case is closed. (Statewide, this occurred in 11 cases between July 1999 and June 2000.)
1In contrast to Chapter 2, this chapter focuses specifically on the implementation of the Jobs First time limit and thus presents little information on the AFDC program.(back)
2While the policies discussed in this chapter have been implemented statewide, MDRC’s observations and data are limited to the two research sites: Manchester and New Haven. The time-limit review process may operate differently in other areas of the state, and the numbers presented do not represent statewide results.(back)
4As the table shows, 11 percent of Jobs First group respondents reported that they were subject to a time limit, but gave a response other than 21 months when asked the length of the time limit. The most common incorrect responses were 24 months (the time limit in several other states) and 18 months (these respondents may have misunderstood the question and reported the number of months they had used, rather than the number they were allowed).(back)
5Over the four-year study period, about 8 percent of AFDC group members were subject to Jobs First policies for at least one month. The majority of these are people who moved and received welfare in a nonresearch district. The others were erroneously subjected to Jobs First policies.(back)
6In MDRC’s evaluation of Florida’s Family Transition Program (Bloom et al., 2000a), 29 percent of AFDC group members reported that they were subject to a time limit. The corresponding figure was 66 percent in the Abt Associates’ study of Delaware’s waiver program (Abt Associates, 1997).(back)
7Once recipients began to reach the time limit, the grapevine presumably started to shape the views of those remaining on the rolls: They could learn whether recipients who reached the limit were receiving extensions or not.(back)
8This question was asked of respondents who had received welfare since random assignment and believed that they had a time limit.(back)
9A banking message might make more sense for a recipient
who receives a substantial amount of child support, which would be budgeted
against her grant (after a $100 monthly disregard).(back)
10Beginning in October 2001, most recipients were limited to three
extensions.(back)
11Some staff mentioned that recipients who took advantage of Jobs First’s higher vehicle exclusion policy to obtain a reliable car may have found, after their TFA benefits were discontinued, that this vehicle disqualified them from receiving Food Stamps.(back)
12As noted in Chapter 2, the work test was a checklist used early in the follow-up period to verify that individuals had been searching for a job.(back)
13These things, along with noncompliance with program activities, can result in an employment services sanction at any point.(back)
14This is not to say that the original decision about whether to sanction recipients or to grant good cause was necessarily straightforward. For example, if a recipient missed some sessions of a program employment activity, it may be difficult to evaluate whether the reasons for not attending constitute good cause.(back)
15See Hunter-Manns, Bloom, Hendra, and Walter, 1998.(back)
16This method of defining the sample is not based on a uniform follow-up period; those randomly assigned earlier had more months to reach the time limit and be included in this sample, and those randomly assigned later had fewer months. For example, in order to be included in this sample, people randomly assigned in June 1996 had to reach the time limit within 21 months of random assignment, but people randomly assigned in January 1996 had 26 months to reach the time limit.(back)
17For this analysis, sample members who received cash assistance during month 22 (and were not exempt from Jobs First) were considered to have received an extension. Some of these recipients did not attend their exit interview when it was first scheduled, but they came into the office and were granted an extension in time to prevent losing a month of assistance. Conversely, those who did not receive cash assistance during month 22 were considered to have had their grant discontinued. Analysis (not shown) of TFA administrative records showed that the proportion of those who received an extension on reaching the time limit did not vary substantially over time; thus, the 100 cases from the first six months of random assignment can be considered representative of the Jobs First group as a whole.(back)
18This conclusion is based on a combination of UI-reported earnings data and data on reported earnings obtained from EMS.(back)
19 The two recipients who received exemptions are assumed to have had income below the payment standard.(back)
20The woman who received the retroactive exemption subsequently received extensions of her benefits based on good-faith effort and was receiving TFA 36 months later. The second case received a medical exemption for six months, then received extensions based on good-faith effort, and finally received another medical exemption near the end of the follow-up period.(back)
21One case was closed for lack of good-faith effort upon reaching the time limit, six were closed for noncompliance during an extension, and four were denied an extension for lack of good-faith effort when reapplying for benefits.(back)
22A regression analysis found that the association between race and receiving multiple extensions is considerably weaker after controlling for other factors.(back)
23For the analysis presented in Table 3.4, Jobs First group members were defined as having left because of the time limit if their TFA case was closed on reaching the time limit or at the end of an extension.(back)
24This program was previously called the Time-Limited Rental Assistance Program.(back)
25John D’Amelia and Associates began operating the rental assistance program in January 2001.(back)
26Data on receipt rates or payments were not available.(back)
27In 1999, MDRC reviewed the Safety Net case files of 20 people who were referred to Safety Net services from the Manchester and New Haven DSS offices.(back)
28People may also be referred to Safety Net services if they are initially denied an extension for being over income, subsequently reapply for benefits after their income drops below the payment standard, and are denied based on a lack of good-faith effort. People who do not show up for their exit interview may have income below the payment standard but generally are not referred to Safety Net services. Safety Net staff reported that sometimes an individual in the community will hear about Safety Net services and will call to inquire whether she is eligible. Staff confer with DSS, and if the individual has exhausted her TFA benefits and has income below the threshold, she can receive services.(back)
29Connecticut Council of Family Service Agencies and United Way of Connecticut-Infoline, 2000.(back)
30Although some sample members who were referred to Safety Net were included in the Three-Year Client Survey, the number who were interviewed after their Safety Net referral was too small to allow for a reliable analysis.(back)
31Connecticut Council of Family Service Agencies and United Way of Connecticut-Infoline, 2000.(back)
| Table of Contents | Previous | Next |