[ Main page of Report | Contents of Report ]
The dramatic caseload decline since the mid-1990s has heightened interest in understanding the circumstances of families who have left welfare. Much of the research on this issue comes from the HHS-sponsored "leavers" studies, which tracked the status of leavers in 11 states and three counties (see Acs and Loprest, 2001, for a review). Other research has also compared leavers with other groups in the caseload, particularly people who did not leave welfare. The existing research on leavers is relevant since cyclers are a subset of groups who leave welfare and groups who return to welfare (recidivists).
[ Go to Contents ]
The weight of the evidence indicates, not surprisingly, that people who leave welfare are less disadvantaged and face fewer employment barriers than people who do not leave (Sandefur and Cook, 1998; Ver Pole, 2002; Miller, 2002). For example, Gleason et al. (1998) — using data from Camden, Newark, and the South Side of Chicago - found that mothers with less education and lower skill levels were slower to leave welfare than other mothers.
A key goal of the leavers' studies is to assess how people fare once they leave welfare. A summary of the studies by Acs and Loprest (2001) indicates that most leavers are working after they leave welfare. In both the first and fourth quarters after exit, for example, the median rate of employment across all state studies was 57 percent. However, the studies also show a fair amount of employment instability — the median proportion of people employed in all four post-exit quarters was 37 percent. Thus, job loss among welfare leavers may give rise to cycling back to welfare.
Another factor that may affect the number of families who eventually return to welfare is the receipt of non-welfare work supports. The existing studies show that only a slight majority of leavers receive non-welfare benefits in the quarters after exit (Acs and Loprest, 2001; Miller, 2002). Acs and Loprest (2001) find median rates of food stamp receipt in the four quarters after exit ranging from 40 percent to 50 percent. The corresponding rates of Medicaid receipt ranged from 45 percent to 57 percent. The authors also find that the proportion of leavers who receive these benefits at some point in the year after exit is much higher than the proportion who receive them in any given quarter, suggesting a fair amount of cycling into and out of these programs.
The evidence is mixed in terms of whether leavers are worse off economically, compared with before they left welfare. In examining material hardship, for example, some state leavers studies found that leavers were better off than before they exited welfare, while others found that leavers were worse off (Acs and Loprest, 2001).
Finally, the status of leavers may differ depending on why they left welfare. Bloom et al. (2002), using data from all the states and several welfare time limit evaluations, found that people who left welfare because they reached their time limit were struggling financially but were not experiencing more hardships than other leavers. On the other hand, people who left welfare because of sanctions appear to be worse off than other leavers (Loprest, 2002).
[ Go to Contents ]
One of the key outcomes examined for leavers is how many eventually return to welfare. People who return to welfare are not necessarily cyclers, since they may only return once. However, cyclers are a subset of this group of returners. The research on recidivists addresses three broad questions:
An important finding from the research on leavers is that most of them do not return to welfare (Rickman et al., 2001; Bruce et al., 2001; Acs and Loprest, 2001). Acs and Loprest (2001), for example, report that across the range of leavers' studies, a median of 27 percent of leavers returned to welfare at some point within the first year after exit (this number ranged from 17 percent to 38 percent across the sites). Loprest (2002), using data from the National Survey of America's Families (NSAF), found that a little over 20 percent of leavers had returned to welfare by the time of the follow-up survey, which ranged from one month to two years after their exit. Also, most of the individuals at risk of re-entry have been found to do so relatively soon after they exit welfare, such as within 6 months to a year (Julnes et al., 2000; Bruce et al., 2002; and Harris, 1996). The findings from this research suggest that the incidence of cycling is likely to be fairly low since cyclers, by definition, return to welfare numerous times.
The existing research shows that people who return to welfare are more disadvantaged and less employable than people who do not return. For example, people who return tend to have lower education levels, more children, and less work experience than people who do not return (e.g., Julnes et al., 2000; Rickman et al., 2001; Bruce et al., 2002; Harris, 1996). Other family circumstances, such as whether the family receives non-welfare supports, also predict who returns to welfare. For example, staying off of welfare is associated with the receipt of childcare subsidies, health insurance, food stamps, and help with family expenses (Loprest, 2002).
Thus, cycling on and off of welfare is likely to be related to personal and family circumstances, an individual's skill level, her work experience and employment patterns, and her receipt of supports.
[ Go to Contents ]
While limited, the current research on cyclers addresses the following questions: 1) What is the incidence of cycling among the caseload? 2) What are the characteristics and circumstances of cyclers? 3) Does cycling lead to longer or shorter stays on welfare? Researchers have also considered whether cyclers resemble other welfare populations (such as shorter- or longer-term recipients) in their background characteristics and employment and welfare behavior - or whether they should best be considered as a unique population.
Researchers have defined cycling quite differently. (See Table 1 for a summary of these definitions.) Nonetheless, the existing research indicates that cyclers make up a relatively small fraction of the caseload. Moffitt (2002), for example, defined cyclers as those with three or more welfare spells within a ten-year period. Using this definition and data from the National Longitudinal Survey of Youth (NLSY) from 1979 to 1996, he found that 20 percent of individuals who had ever been on welfare were cyclers. Ver Ploeg (2002) defined cyclers as recipients with three or more welfare spells within a nine-year period. Using data from a sample of adults in Wisconsin who received Aid to Families with Dependent Children (AFDC) in July 1995, she found that cyclers represented about 14 percent of the sample. Miller (2002) defined cyclers as individuals with only one short spell on welfare or those with two or more spells, who spent less than half of the observation period (three to five years) on welfare. Using this definition and data from several welfare waiver evaluations, Miller (2002) found that about 20 percent of a sample of new applicants and ongoing recipients were cyclers. Finally, Zedlewski and Alderson (2001) use data from the National Survey of America's Families (NSAF) and define cyclers as those who first received welfare more than two years prior to each survey (in 1997 and 1999) and who received welfare only intermittently in the two years prior to the survey. Using this definition, they find that cyclers were 20 percent of the caseload in 1997 and 23 percent in 1999, where the caseload includes all adults receiving benefits at the time of the survey.
The different ways in which researchers have defined cycling have affected their other findings on cycling, such as cyclers' background characteristics, typical length of welfare receipt, and relative disadvantage in the labor market. For instance, Moffitt (2002) finds that cyclers tend to have medium-length spells (of between 7 and 30 months), rather than short or long spells, suggesting that cyclers are a fairly welfare-dependent group. Similarly, Zedlewski and Alderson (2001) find that cyclers looked more like ongoing recipients than new entrants, in terms of employment rates and education levels.
Both Ver Ploeg (2002) and Moffitt (2002) concluded that over 80 percent of cyclers spent more than two years of the observation period on welfare, although the two researchers did not agree about whether cyclers experience very long spells of assistance. Moffitt (2002) found that 45 percent of them had a total time on welfare of more than five years, compared with only 18 percent found by Ver Ploeg (2002).(1) In contrast, Miller (2002) concluded that cyclers tended to be short-termers, receiving welfare benefits for less than two years during the observation period for her study. However, it should be noted that, for the larger purposes of her research, Miller (2002) intentionally sought to distinguish cyclers from long-term recipients.
This variation in findings across studies underscores the importance of comparing patterns of cycling in additional sites and additional welfare populations. This report helps meet this need. The report also provides more information about whether cycling is higher or lower since the implementation of PRWORA and also whether the incidence of cycling varies according to labor market conditions, welfare grant levels, and other site-related factors.
| Research Study | Cycler Definition | Sample | Data Sources | Cycler Estimate |
|---|---|---|---|---|
| Miller (2002) | Individuals who had only one spell on welfare that was six months or less or who had multiple spells but spent less than half of the observation period (three to five years) on welfare | A total of 36,449 new applicants and ongoing recipients from several welfare-to-work evaluation studies | Administrative Records | 20% |
| Moffitt (2002) | Individuals with three or more welfare spells within a ten-year period | A total of 514 women that received at least one month of AFDC from ages 20-29. | National Longitudinal Survey of Youth from 1979-1996 | 20% |
| Ver Ploeg (2002) | Individuals with three or more welfare spells within a nine-year period | A total of 48,216 single parent women that received welfare in July 1995 in the state of Wisconsin. | Administrative Records | 14% |
| Zedlewski and Alderson (2001) | Individuals who received welfare intermittently in two years prior to being surveyed and who first received welfare more than two years prior to being surveyed | 1,831 adults receiving TANF in 1997 and 850 families receiving TANF in 1999. | 1997 and 1999 National Survey of America's Families | 20% in 1997 and 23% in 1999 |
| Sources: See References for full citation. | ||||
[ Go to Contents ]
(1) As Moffitt, 2002, notes, there is a potentially important difference between his study and Ver Ploeg's. The base sample for his study is all people who ever received welfare at some point during a 10-year period, while Ver Ploeg uses all people who were on welfare at a point in time. The latter method will miss spells not in progress at the point in which the sample was drawn.
Main Page of Report | Contents of Report
Home Pages:
Human Services Policy
(HSP)
Assistant Secretary for Planning and Evaluation
(ASPE)
U.S. Department of Health and Human Services
(HHS)