Table of Contents | Previous | Next |
Appendix C
Background Information on the Leaver Studies Cited in Chapter 6
Chapter 6 includes findings from eight studies of people who left welfare because of a time limit. This appendix provides background information on each of the studies. Information is first presented on the leaver studies in Connecticut, Florida, Massachusetts, North Carolina, Ohio, South Carolina, Utah, and Virginia. The second section presents information on MDRC’s evaluations of Connecticut’s Jobs First program and Florida’s Family Transition Program (FTP), which are also discussed in Chapter 6.
Survey-Based Studies of TANF Leavers
Connecticut
| Sponsor/Research Firm
Connecticut State Department of Social Services MDRC (survey conducted under contract by Roper Starch Worldwide) |
Sample Families in 6 sites whose cases closed in September or October 1997 when they reached the 21-month time limit on cash assistance. (The Connecticut post-time-limit study included only families who left TANF because of the time limit.) Interviews were conducted 3 and 6 months after cases closed. The 3-month survey was fielded between January 1998 and April 1998. The 6-month survey was fielded between April 1998 and July 1998. Only individuals who were still off welfare at the time of the interviews were surveyed. |
Sample Size 3-month survey: 421 6-month survey: 448 Response Rate Mode of Administration Computer-Assisted Telephone Interviewing |
Florida’s Family Transition Program (FTP)
| Sponsor/Research Firm
Florida Department of Children and Families and private foundations MDRC |
Sample All FTP participants who reached the time limit during certain calendar periods (Novem-ber 1996 through May 1997 for those subject to a 24-month time limit, and June 1997 through February 1998 for those subject to a 36-month time limit). In-person interviews were conducted around the time benefits expired, and then 6, 12, and 18 months later. The 18-month interview was a lengthy open-ended discussion conducted by an ethnographer. |
89 people received final welfare checks during the two periods, and 70 completed the initial interview. 57 completed the 6-month interview, 49 completed the 12-month interview, and 54 completed the in-depth 18-month interview. Response rates (based on the initial 70 respondents) were 81%, 70%, and 77%. |
Massachusetts
| Sponsor/Research Firm
Massachusetts Department of Transitional Assistance Center for Survey Research (CSR) at the University of Massa-chusetts at Boston |
Sample Households that left welfare between December 15, 1998 (when the first families reached the state’s 24-month time limit), and April 30, 1999. Approximately 2/3 of the fielded sample consisted of households that had reached month 24 of time-limited benefits (time-limit leavers). The remaining 1/3 were households that left welfare for various reasons, such as earnings, sanctions, and changes in family status (non-time-limit leavers). Respondents had to have been off welfare for at least 2 months. Respondents were interviewed 6 to 16 months after they left welfare. Individuals were included in the study regardless of whether or not they were receiving TANF at the time of the interview. |
Sample Size Time-limit leavers: 460 Non-time-limit leavers: 210 Response Rate Mode of Administration: Computer-Assisted Telephone Interviewing, with some in-person interviews |
North Carolina
| Sponsor/Research Firm
North Carolina Department of Health and Human Services MAXIMUS |
Sample Families leaving welfare. Surveys were administered separately to those leaving because of the time limit and those leaving for other reasons. Time-limit leavers: The first cohort of families to reach the 24-month time limit in August 1998. Two rounds of surveys were conducted. The first round was administered between De-cember 1998 and March 1999 (4 to 7 months after families reached the time limit). The second round was administered between September 1999 and December 1999 (13 to 16 months after families reached the time limit), and was targeted only at the round 1 respondents. Non-time-limit leavers: Families in 8 counties who left Work First for any reason for at least 1 month between December 1998 and April 1999. Interviews were conducted between June1999 and February 2000 (approximately 6 months after respondents left welfare). |
Sample size Time-limit leavers: Round 1: 247 Round 2: 221 Non-time-limit leavers: 1,878 Response Rate Non-time-limit leavers: 70.1% Mode of Administration Computer-Assisted Telephone Interviewing |
Ohio
| Sponsor/Research Firm
Ohio Department of Job and Family Services and the Cuya-hoga County Board of County Commissioners Center on Urban Poverty and Social Change at Case Western Reserve University |
Sample Randomly selected individuals in Cuyahoga County who left welfare in October 2000. Leavers are defined as individuals who received assistance in the first month of a quarter and then were not part of the caseload in the second and third months of the quarter. The fielded sample included 300 time-limit leavers (individuals with a time-limit closure code) and 150 non-time-limit leavers. Interviews were conducted 6 and 13 months after individuals’ initial exit. The analysis also uses administrative records. |
Sample Size Time-limit leavers: 204 Non-time-limit leavers: 86 Response Rate Mode of Administration |
South Carolina
| Sponsor/Research Firm
South Carolina Department of Health and Human Services MAXIMUS |
Sample Stratified random sample of families who left welfare between October 1998 and March 1999. Time-limit leavers are those who left because of the time limit, according to the state’s data system. Non-time-limit leavers are those who left due to earnings, sanctions, or “other” reasons. The survey was fielded after sample members had been off welfare for approximately 10 to 14 months. |
Sample Size Time-limit leavers: 292 Non-time-limit leavers: 780 Response Rate Non-time-limit leavers: 73% Mode of Administration Computer-Assisted Telephone Interviewing |
Utah
| Sponsor/Research Firm
Utah Department of Workforce Services University of Utah School of Social Work |
Sample Families who had received welfare for 36 months or more and whose cases had been closed for at least 2 months as of February 2000. Time-limit leavers: Families whose cases were closed because they had reached Utah’s 36-month time limit in December 1999 (the first month in which individuals reached Utah’s time limit). Non-time-limit leavers: Families whose cases were closed due to increased income and other reasons – primarily sanctioning. Respondents were interviewed between Febru-ary 2000 and May 2000 (approximately 2 to 5 months after the time-limit leavers had their cases closed). |
Sample Size Time-limit leavers: 133 Non-time-limit leavers: 274 Response Rate Mode of Administration |
Virginia
| Sponsor/Research Firm
Virginia Department of Social Services Mathematica Policy Research, Inc. |
Sample Time-limit leavers: Families whose cases closed because of the time limit between February 1, 1998, and June 30, 1998 (Cohort 1), and families whose cases closed because of the time limit between February 1, 1999, and June 30, 1999 (Cohort 2). Because of the staggered implementation of the state welfare reform program, Cohort 1 was selected from very limited areas of the state, while Cohort 2 was drawn from districts comprising roughly half the state. The first round of surveys were administered 6 to 14 months after cases closed. A second round of surveys was administered to Cohort 1 families 18 to 24 months after cases closed. Administrative data on all cases are also available. Non-time-limit leavers: Data on time-limit leavers come from the Virginia Closed Case Survey – a study of cases that closed because of increased income or because of a sanction in late 1997. Interviews were conducted approximately 12 months after cases closed. |
Sample Size Time-limit leavers: 6-month survey: 751 18-month survey: 220 Non-time-limit leavers: 779 Response Rate Non-time limit leavers: 69% Mode of Administration Computer-assisted telephone interviews |
Evaluations
Connecticut’s Jobs First Program
Under a contract with the state’s Department of Social Services, MDRC conducted a large-scale evaluation of Jobs First, Connecticut’s welfare reform initiative. Welfare applicants and recipients in two welfare offices were randomly assigned to program and control groups from January 1996 through February 1997. Four years of follow-up data are available. As part of the evaluation, an analysis of welfare leavers was conducted using baseline demographic data and administrative records data on earnings, welfare, and Food Stamp receipt for 600 sample members randomly assigned to the program group from January through June 1996 who left welfare between study entry and March 1998.
Baseline and administrative records data are available for 477 program group members who, by March 1998, had left welfare for two or more consecutive months before reaching the state’s 21-month time limit (non-time-limit leavers) and 132 program group members who, by March 1998, had their benefits discontinued as a result of time limits (time-limit leavers). Administrative records data cover the quarter prior to exit through the third quarter after exit.
As noted earlier, there was also a separate study of time-limit leavers.
Florida’s FTP
There are several samples of welfare leavers available from a random assignment evaluation of Florida’s Family Transition Program (FTP) — a pilot program run in Escambia County from 1994 through 1999 — conducted by MDRC under a contract with the Florida Department of Children and Families. Importantly, FTP was a pilot program implemented two years prior to the implementation of PRWORA, and the findings do not reflect the effectiveness of Florida’s statewide welfare reform program. The study sample includes 1,405 single-parent cases randomly assigned to the program group from May 1994 through February 1995 who were subject to either a 24- or a 36-month time limit.
Time-limit leavers are FTP group members who received at least the time-limit amount (24 or 36 months) of TANF between date of random assignment and June 1999 (four to five years after study entry) and had their benefits fully terminated. Approximately one-quarter of the individuals subject to a time limit accumulated the time-limit amount of TANF. Of these, 237 (70 percent) reached the time limit, and 227 had their benefits fully terminated. Baseline and administrative records data are available for all 227 individuals who had their welfare benefits canceled because of time limits.
In addition, four-year client survey data are available for 136 time-limit leavers included in the four-year client survey sample. These 136 families had been off welfare for an average of 17 months at the time they were interviewed. However, since the survey was fielded based on random assignment date (not date of welfare exit), there is considerable variation in the length of time that these families had been off welfare.
Non-time-limit leavers are sample members who stopped receiving benefits before reaching the 24- or 36-month time limit within four years of study entry. Three-quarters (75.7 percent) of individuals in the FTP group left the program before reaching the time limit, primarily because of employment.1 Baseline demographic data are available for 954 individuals who received at least one month of welfare after study entry but did not reach the time limit within four years.
Four-year client survey data are available for 657 families who left welfare before reaching the time limit. Leavers were identified based on a survey question that asks respondents about welfare receipt in the month prior to the survey interview. Of these individuals, 84 percent had not received welfare in the year prior to the survey. The survey was administered from 48 to 61 months following respondents’ entry into the study. The response rate for the entire survey sample was 80 percent. Surveys were administered in person and by telephone.
As discussed above, there was also a separate small-scale survey of families who left welfare because of the time limit.
1Late in the follow-up period, the state implemented full-family sanctions. It is possible that some non-time-limit leavers had their cases closed because of noncompliance.(back)
| Table of Contents | Previous | Next |