This report analyzes the experiences of welfare “cyclers,” a group that has received relatively little attention in previous research on welfare dynamics. For this study, “cycling” is defined as receipt of welfare benefits during three or more discrete spells during a four-year observation period. The goals of this report are to understand the incidence of cycling and the types of families who cycle on and off the rolls, and, if possible, to shed light onto why they repeatedly return to assistance. The report also considers whether welfare cyclers appear to be more advantaged or more disadvantaged than other welfare recipients in the labor market. One view of cycling is that cyclers move on and off welfare, repeatedly, during transitional periods as they attempt to leave welfare. Eventually, cyclers may attain stable employment and leave assistance more permanently. An alternative view of cycling is that cyclers may work for pay only briefly and return to welfare for longer spells with little progress toward self-sufficiency.

To explore these issues, we compare welfare, employment, and other outcomes for cyclers to those of two other groups within the welfare caseload: short-term recipients and long-term recipients. For this study, a short-term recipient is defined as someone who had one or two spells and a total of up to 24 months of welfare receipt during the four-year (48-month) observation period. Long-term recipients are defined as sample members with one or two spells and a total of 25 to 48 months of welfare receipt during the observation period.

The report tracks the patterns of welfare receipt, employment, and other outcomes of 161,007 single-parent welfare recipients, aged 18 to 59, from five MDRC studies of welfare reform initiatives during the mid- to late 1990s. Three of these studies are experimental (random assignment) evaluations of welfare reform initiatives—Connecticut Jobs First, Florida Family Transition Program (FTP), and Vermont Work Restructuring Project (WRP). The other two are nonexperimental studies of the effects of welfare reform in large urban areas: Cleveland (Cuyahoga County) and Philadelphia Urban Change.

The period of sample intake for this study took place during 1993 through 1997 and varied in duration from one year to five years across the five sites. For the three experimental evaluations, sample intake occurred when sample members were randomly assigned to a program or control group. (Only program group members are included in the sample for most analyses). For the two Urban Change sites, the years of sample intake were chosen to maximize data availability and to cover a similar time period as the evaluation sites. Sample intake in Cleveland and Philadelphia took place when sample members were first recorded as receiving a welfare payment during these years.

As a result of the sample intake procedures for these studies, about 40 percent of the sample entered the study as new recipients, individuals who were just beginning their first observed spell of welfare receipt. The remaining 60 percent were ongoing recipients, individuals who entered the study in the middle of an observed spell of welfare receipt.

For each sample member in the five sites, the observation period began with the month of sample intake and ended four years (48 months) later. For the first members to enter the sample, most or all of the observation period took place during the years before their state welfare agency implemented their Temporary Assistance for Needy Families (TANF) regulations in response to the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA). The rest of the sample experienced TANF’s services, time limits on eligibility to receive welfare benefits, and other requirements throughout all or most of the observation period.

The findings from this report are as follows:

• Cyclers constituted a relatively small portion of the welfare caseload.

Pooled together and weighted equally by site, 8.5 percent of the sample became welfare cyclers during the four-year observation period. About 47 percent of sample members became short-term recipients and 45 percent became long-term recipients. Rates of cycling ranged from 3.8 percent of sample members (in Philadelphia) to 13.7 percent (in Florida FTP). In four sites, the incidence of cycling increased by about 2 percentage points for a large subsample with five years of observed data on welfare receipt (see Figure 1).

• Cyclers’ background characteristics differ from those of short-term and long-term recipients.

Cyclers’ background characteristics suggest compelling reasons for their subsequent pattern of welfare receipt. Among the pooled sample, the average age of cyclers at sample intake was 28 years—slightly more than two years younger than the average age of short- and long-term recipients. Cyclers also tended to start having children about a year to a year and a half earlier in their lives and were the most likely of the three groups to be parents of a child under the age of 6. Prior research suggests that young, single parents with young children often have a hard time holding steady employment. Such individuals may have difficulties sustaining employment given their family responsibilities, such as childcare.

Cyclers appear to be less disadvantaged than long-term recipients but more disadvantaged than short-term recipients in terms of their previous history of employment and welfare receipt. For instance, cyclers and short-term recipients recorded similar levels of finding employment before sample intake, but cyclers were slightly more likely to lose their jobs. Similarly, 58 percent of those who became cyclers during the subsequent observation period were ongoing recipients at their time of sample intake, compared with 49 percent for short-term recipients and 73 percent for long-term recipients.

• During the observation period, cyclers fared better in terms of employment and welfare receipt than long-term recipients, but not as well as short-term recipients.

On average, cyclers received 27 months of cash assistance within the four-year observation period, compared with 12 months for short-term recipients and 40 months for long-term recipients. At the end of the observation period, about 40 percent of cyclers were receiving welfare, only 6 percentage points below the level for long-term recipients. In comparison, almost no short-term recipient remained on assistance in month 48.

Nearly all sample members in each group found employment during the observation period. Over 90 percent of cyclers worked for pay during at least one quarter, the highest incidence of employment among the three groups. However, cyclers had somewhat greater difficulty maintaining employment than short-term recipients. For instance, about 35 percent of short-term recipients worked during at least three-fourths of the quarters in the observation period (an indicator of stable employment), compared with 28 percent of cyclers and only 12 percent of long-term recipients.

Partly as a result of their less stable employment during the observation period, cyclers’ average total earnings of $16,885 were $8,000, or 33 percent, below the mean for short-term recipients. On the other hand, cyclers earned nearly $7,000 or 60 percent more than long-term recipients (see Table 1).

Differences in job quality also contributed to this difference in total earnings. Among 4,285 survey respondents selected from the main research sample, cyclers who were working at the time of their interview received an average of $7.53 per hour in wages, about $0.70 less than their counterparts among short-term recipients. Cyclers employed at interview were also much less likely than short-term recipients to be enrolled in their employers’ medical plan. On both of these measures of job quality, the average for cyclers only slightly exceeded the level for long-term recipients.

Finally, cyclers demonstrated a greater propensity to combine work and welfare than both short-term and long-term welfare recipients, but were less likely than short-term recipients to rely on earnings alone. For example, for cyclers, quarters with both earnings and welfare receipt accounted for just over a third of all quarters in the observation period, compared with only 14 percent and 26 percent of quarters for short-term and long-term recipients respectively. Short-term recipients, however, had more quarters of earnings alone, compared with cyclers, an indication of greater self-sufficiency. Among the three groups, long-term recipients showed the lowest overall levels of employment and most often relied on welfare payments alone.

• On average, cyclers had less access to support (financial and otherwise) from other adults and were more likely to have additional children during the observation period compared with short-term recipients.

Among survey respondents, about one-fifth of cyclers reported that they were living with a spouse or partner at the time of their interview, compared with one-third of short-term recipients. Cyclers were similarly less likely than short-term recipients to be living with another wage earner at the time of their interview. Finally, a higher percentage of cyclers reported that they gave birth to or adopted another child after sample intake. Most likely, the birth of a child caused disruptions to the employment situation of cyclers and increased their need for financial support, compared with short-term recipients. Results for cyclers on these measures of household membership and support more closely resembled those for long-term recipients.

• During an additional follow-up year, cyclers did not “catch up” to short-term recipients in employment stability and self-sufficiency, although they continued to fare better than long-term recipients.

About 91 percent of sample members in four sites (not Connecticut Jobs First) had a fifth year of earnings and welfare data following sample intake. These data were used to compare welfare and employment outcomes for sample members who became cyclers, short-term recipients, or long-term recipients during the observation period.

At the end of year 5, about 40 percent of each group worked for pay and the majority of each group—even long-term recipients—no longer received welfare benefits. Each group demonstrated greater self-sufficiency during the post-observation period, as evidenced by their higher rates of employment without welfare and lower rates of combining work and welfare. Less positively, many cyclers continued to experience unstable employment in year 5. For instance, among cyclers who worked for pay at any time during year 5, only about 43 percent were employed during all four quarters, an important measure of employment stability. Furthermore, the rate for cyclers was more than 9 percentage points below the level for short-term recipients and was no higher than that for long-term recipients.

Again, cyclers exhibited stronger attachment to the welfare system than short-term recipients, but less attachment than long-term recipients, as represented by the percentage that received TANF benefits in the last month of year 5. In this month, cyclers were 17.5 percentage points more likely than short-term recipients to be receiving cash assistance, but 8.6 percentage points less likely than long-term recipients. In addition, nearly 30 percent of cyclers worked for pay at the end of year 5 and did not receive welfare benefits. This average exceeded the level for long-term recipients by 4 percentage points but was 7 percentage points below the level for short-term recipients. Cyclers were also about four times more likely than short-term recipients to combine work and welfare (11.8 percent to 2.9 percent) at the end of year 5—and did so only slightly less often than long-term recipients. In other words, fewer cyclers are able to sustain themselves with employment alone.

• The incidence of cycling increased during the years following passage of PRWORA.

In Cleveland and Philadelphia, the sites with the longest sample intake periods, rates of cycling were about 3 percentage points higher among welfare recipients who entered the research sample following passage of PRWORA, when compared with welfare recipients who entered the research sample previous to its passage. Furthermore, cyclers constituted a larger portion of the welfare caseload by end of the 1990s than earlier in the decade. For instance, in Cleveland, cyclers represented 18 percent of all recipients who received a welfare payment in December 1999, up from 10 percent in January 1993. A similar increase occurred in Philadelphia, although the incidence of cycling remained below 10 percent of the caseload at the end of the decade.

Additional, though indirect, evidence of the possible effects of PRWORA on welfare cycling was obtained by comparing the incidence of cycling among sample members in the three evaluation sites to those of their corresponding control groups. (As noted above, control group members were otherwise excluded from this analysis.) Program group members experienced welfare programs similar to TANF, whereas control group members experienced pre-TANF programs. However, in none of these three evaluations did the rate of cycling for the program exceed the rate for the control group.

It should be kept in mind that in the pre-/post-PRWORA comparisons in Cleveland and Philadelphia, members of the post-PRWORA groups entered the sample during a much stronger labor market than their counterparts in the pre-PRWORA group. In contrast, in the three evaluations, both program and control group members entered the sample at the same time and experienced the same (relatively strong) labor market. These findings suggest that PRWORA was one of several factors related to increases in cycling rates seen in Cleveland and Philadelphia. Most likely, the economic expansion of the late 1990s also played a role.

In summary, cyclers were shown to be a group in the middle—less disadvantaged in the labor market than long-term recipients, but less able than short-term recipients to attain stable employment and work without welfare. Furthermore, cyclers were the most likely to be parents of toddlers and preschoolers. However, compared with short-term recipients, cyclers had less access to financial and other support from a spouse or partner. For policymakers and administrators of state and local TANF programs, these findings suggest that cyclers should derive particular benefit from enhanced supports for work and post-employment services intended to promote employment retention and advancement.