U.S. Department of Health and Human Services (HHS)
Office of the Assistant Secretary for Planning and Evaluation (ASPE)
Office of Human Services Policy (HSP)
Contract: 233-02-0086; Task Order 23
Quinn Moore, Anu Rangarajan, and Peter Schochet
Mathematica Policy Research, Inc
This report is only available in PDF format"
We would like to thank those whose efforts made this report possible. Laura Chadwick, from the Office of the Assistant Secretary for Planning and Evaluation (ASPE), the current project officer for the study, and Susan Hauan, the former project officer for the study, have both provided invaluable guidance throughout the course of the study, as well as very helpful comments on both the substance and presentation of material in this report. We would also like to acknowledge the guidance of Julia Isaacs (now at the Brookings Institution) during the early phases of this project. At Mathematica Policy Research, Robert Wood carefully reviewed a draft of this report and provided very useful comments. Carol Razafindrakoto and Julie Sykes provided outstanding programming assistance in constructing the large and complex data files and in writing the computer programs to conduct the many analyses that were performed for this study. Finally, Jill Miller, Bill Garrett, and Jennifer Baskwell expertly produced the report, and Patricia Ciaccio and Leah Hackleman provided valuable editorial assistance. We gratefully acknowledge these contributions.
With the passage of the Personal Responsibility and Work Opportunity Reconciliation Act of 1996 (PRWORA), policymakers and researchers have recognized the importance of employment and earnings as a key way out of poverty and dependency for single mothers. During the middle and late 1990s, the Temporary Assistance for Needy Families (TANF) program, aided by a strong economy, helped move many recipients off the welfare rolls and into work. Several studies have looked at the labor force transitions of former welfare recipients, but few have focused on earnings and income progression, poverty dynamics, and the pathways out of poverty for single mothers (including those who have never received welfare). Single mothers represent a group vulnerable to extensive contact with poverty, and it is important to discern their prospects for long-term self-sufficiency.
This study seeks to broaden knowledge about the extent to which single mothers remain out of poverty and the factors most strongly associated with their continued economic progress. We used the 2001 Survey of Income and Program Participation (SIPP) data, collected by the U.S. Bureau of the Census, to examine the income and employment experiences of single mothers who exited poverty. We identified single mothers who exited poverty during 2001, and tracked their experiences over the subsequent two-year period.
In our study, we started by focusing on the following two sets of questions:
- Who are single mothers who exit poverty and what are their reasons for exiting? How common are poverty and poverty exits among single mothers? Do the characteristics of single mothers who exit poverty differ from those of single mothers who do not exit poverty? What proportion exits poverty because of a change in employment and earnings or a change in family composition?
- How long do single mothers who exit poverty stay out of poverty? What are the main reasons these single mothers reenter poverty? How long are their subsequent poverty spells? How common is poverty cycling?
Based on the duration of their non-poverty and poverty spells during the two-year follow-up period, we found that single mothers who exit poverty can be classified into three groups: 28 percent who exit poverty and never return, 56 percent who cycle in and out of poverty, and 16 percent who return to poverty and stay poor. Poverty dynamics differ markedly across these three distinct groups of single mothers. Thus, the remainder of our study focused on the following two sets of additional questions:
- Who are single mothers who exit poverty and stay out of poverty? How do their characteristics differ from those who cycle in and out of poverty and those who return to and remain in poverty? What are the characteristics of single mothers in each group? Do the reasons single mothers exit poverty differ across these groups?
- What are the income, employment, and welfare receipt experiences of single mothers across the three groups? How do their experiences differ? What proportion of months do they spend out of poverty? What are their employment and welfare experiences? What factors are associated with their economic success?
The remainder of this introductory chapter is in four sections. First, we present background information for the study and review the related literature. Second, we briefly summarize data, samples, and key definitions for the study, and third, discuss our analytic approach. Fourth, we provide a brief summary of study findings. We conclude with a roadmap to the rest of the report.
Background and Related Literature
Poverty has been a persistent and relatively common phenomenon in the United States, particularly among single mothers and female-headed families. In a recent assessment of poverty levels and trends using data from the Current Population Survey (CPS) for the period from 1967 through 2003, Hoynes et al. (2004) report that the point-in-time poverty level among the general population remained around 13 percent from 1980 through 2003, despite large increases in the gross domestic product (GDP) during the same time period. Poverty rates for single mothers were even higher. While the annual poverty rate for individuals in families with married parents was 7 percent in 2003, 40 percent of individuals in families with single-parent heads were poor (Hoynes et al. 2004). Similar patterns are found when poverty is measured on a monthly level. For instance, Naifeh (1998) presents descriptive results from the 1993 panel of SIPP that suggest that nearly half of individuals in female-headed households have experienced at least two months of poverty during a two-year analysis period, compared to 14 percent of those in married couple families.
Findings from the Poverty Dynamics Literature
A large literature on poverty dynamics offers valuable context to our study. The literature finds that most people who experience poverty have short spells, but the relatively few long spells represent the majority of the poor at any given time. The seminal work of Bane and Elwood (1986), based on evidence from the 1970 through 1982 panels of the Panel Study of Income Dynamics (PSID), shows that 45 percent of poverty spells end within a year, but more than half of the poor in a given survey year are in the midst of a poverty spell lasting 10 years or more. Similar patterns have been found in studies using monthly poverty measures using the SIPP data (Ruggles and Williams 1987; Eller 1996; Naifeh 1998) and in descriptive evidence presented in studies that use other methodological approaches (Rank and Hirschl 2001; Stevens 1994; Stevens 1999; McKernan and Ratcliffe 2002). Studies that look at poverty patterns by gender of household head find that female-headed households have much higher incidence of poverty, higher rates of recidivism, and longer poverty spells (Bane and Ellwood 1986; Stevens 1999; McKernan and Ratcliffe 2002).
Although the literature on the dynamics of poverty exits is well developed, few studies have investigated the dynamics of poverty recidivism. One exception is Stevens (1999), whose study of poverty persistence also includes duration analysis of non-poverty spells. Using data from the 1968 through 1989 waves of the PSID, the author finds that poverty recidivism is common — more than half of those who leave poverty return within five years. Stevens (1999) finds that individuals in female-headed households are significantly less likely to exit poverty and significantly more likely to reenter after having exited. These differences translate into much greater poverty exposure among those who live in female-headed households, leading Stevens to conclude that "even a short period spent in a female-headed household significantly increases poverty persistence."
The literature examining trigger events associated with poverty entry and exit has emphasized the importance of changes in income and employment status. Bane and Ellwood (1986) find that a change in the household head's earnings is by far the most common event associated with poverty entry and exit among all households. These findings are supported in multivariate analysis based on data from the PSID and SIPP presented in McKernan and Ratcliffe (2002), which shows that employment events, rather than changes in household composition or disability status, are the most important factors in poverty entry and exit.
Studies, however, have also found large differences in events associated with poverty dynamics for female-headed households/families compared to the general population. Bane and Ellwood (1986) report that, while nearly 60 percent of poverty spells in two-parent households begin with a decline in the head's income, only 14 percent of poverty spells in female-headed households begin this way. The most common event associated with poverty entry for female-headed households is a transition from married to single, rather than a change in income by a household member. Multivariate analysis presented in McKernan and Ratcliffe (2002) also supports the finding that transitions to and from female-headed households are important in explaining poverty transitions.
Findings from the Post-PRWORA Welfare Studies
The well-being of single mothers has been a topic of much interest among policymakers and researchers, particularly with the passage of PRWORA. Because increased employment was a central goal of the 1996 welfare reform, the labor market participation of single mothers has been a primary focus of the welfare literature. Several nationally representative studies have linked welfare reform and other policy changes to increased employment among single mothers (Meyer and Rosenbaum 2000; Blank and Schmidt 2001). Similarly, state studies of welfare leavers uniformly find high rates of employment. In a synthesis of ASPE-funded state studies, Acs and Loprest (2001) report that the median employment rate of single parents one year after leaving welfare was about 70 percent.
Despite high average levels of employment among former welfare recipients, evidence of a large degree of employment churning exists. For example, in a longer-term study of TANF recipients in Wisconsin, Wu et al. (2005) found that nearly 40 percent of their sample had patterns of unstable employment levels over a six-year follow-up period. Johnson and Corcoran (2003) also present evidence of a high degree of job instability and limited mobility to better quality jobs among welfare participants in Michigan. Wood et al. (2004) examine the poverty patterns of former welfare recipients in a study of current and former TANF recipients in New Jersey and find that, after leaving poverty, transitions back into poverty are common. Of the people in their sample who had left poverty by the second year after entering TANF, more than half returned to poverty during the ensuing three years.
Our review of the findings of the relevant literature highlights the facts that (1) incidence of poverty is much higher among our target population of single mothers, and (2) the factors underlying poverty transitions are somewhat different for single mothers than for the general population. The review of research from the recent welfare literature on the employment and poverty patterns among current and former welfare participants suggests a great deal of heterogeneity in the experience of mothers participating in welfare who leave poverty. However, little is known about the success, or lack thereof, of single mothers following poverty exits, particularly in the post-PRWORA period. The objective of this study is to help fill this void by using longitudinal data on a large, nationally representative sample of single mothers to present a detailed picture of their income, employment, and earnings during the period following poverty exit and to identify factors associated with successful exits from poverty.
Data, Samples, and Key Definitions
This study used as its primary data source the 2001 SIPP, a large, multipanel, longitudinal survey that collected demographic and socioeconomic information on a nationally representative sample of U.S. households. The data cover the period from late 2000 through 2003. SIPP provides detailed monthly measures on labor force participation (for those age 15 and older), income, participation in public programs, and household composition.(1) This section provides a brief overview of some of the key data and sample decisions made for this study; the Appendix provides a more detailed description of these decisions.
Selecting the Poverty Measure. We used the official U.S. Census Bureau poverty measure as the primary one for the study. This measure has several well-documented shortcomings, but it is the one most commonly used in research examining poverty-related issues, and it offers a straightforward and easily understood method for gauging poverty. The use of the official poverty measure also facilitates comparisons of our study results with those of previous research.
Under the official measure, a family is "poor" if its total family income is less than its money income threshold (based on the Office of Management and Budget's Statistical Policy Directive 14). Money income includes earnings; cash assistance (such as TANF benefits, unemployment compensation, or Supplemental Security Income [SSI]); child support; educational assistance; pension income; and interest and dividends. Income does not include non-cash benefits such as food stamps; Medicaid; public housing subsidies; and the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC).(2) There are 48 possible poverty thresholds that vary according to the size of the family and the ages of its members. These thresholds do not vary geographically, but they are updated annually for inflation using the Consumer Price Index for All Urban Consumers (CPI U).
We measured poverty at the family level largely because the family is the basis for the official definition of poverty. A SIPP family is a group of two or more persons related by birth, marriage, or adoption and living together; a SIPP family does not include cohabiters. We also considered using the subfamily — a nuclear family unit — rather than the family in defining poverty and found that this change made little difference in the size of our sample (see Appendix).
Analysis Sample and Unit of Analysis. The primary sample for the analysis consists of single mothers who, during the first 12 months of the panel period (roughly 2001), were in poverty in one month and exited poverty during the following month. If a sample member had more than one poverty exit during the one-year window, we used the first exit for the analysis. We defined single mothers as those women who, during the month before spell exit, were (1) older than 15 (so that employment information is available), (2) living with a related child younger than 18, and (3) unmarried. We included single mothers who were cohabiting with a partner, whose spouses were absent, or who were in school during the panel period.(3) This definition of a "single mother" included all single female family heads with related children under 18 in the family, rather than only single female family heads living with their own children under age 18. Thus, for instance, our sample included a small number of single grandmothers living with their grandchildren, and single aunts living with their nieces and nephews. Overall, we have a sample of 615 single mothers who exited poverty during the first 12 months of the panel period.
A key decision for the analysis was how to deal with changes in family composition. For example, a woman might move out of the family she is living in and move in with others, or other people might join her family. In our analysis, we treated the single mother as the unit of analysis and tracked her poverty status in whatever family she was in during any given month using the income of that family, even if its members changed from month to month. Thus, our analysis focused on obtaining estimates of, for example, the duration of non-poverty spells of single mothers who exited poverty rather than the duration of non-poverty spells of the families that contained these women.
Defining Poverty and Non-Poverty Spells. Another key decision for the study was how to define poverty spells. In data sets that try to capture monthly income, "transitional" poverty spells are sometimes observed — that is, in a given month, a family can have a sudden income dip due either to real changes in income or to measurement error. Such dips can lead to more people being "ever" poor than "consistently" poor. It is also possible that for those with incomes near the poverty threshold, small deviations in income could lead to frequent changes in poverty status. We conducted analyses to determine the extent to which such churning is present in our data and what steps could be taken to measure the "true" income volatility of single mothers.
We observed a high degree of poverty churning in the sample and were concerned about "noise" in the monthly poverty timelines. Thus, we explored several strategies to smooth poverty spells: (1) doing nothing, (2) closing "near-threshold" spells, (3) closing one-month spells, and (4) using a three-month moving average of family income. Based on these explorations, we elected to smooth volatility by closing all near-threshold spells, where a near-threshold spell is defined as a spell in which income is within 10 percent of the poverty threshold for the duration of the spell. Our main results, however, are not sensitive to the income-smoothing strategy.
Our study used data from SIPP to examine the income and employment experiences of single mothers who exited poverty during the 2001 to 2003 period. In particular, we identified a sample of 615 single mothers who exited poverty during 2001 and tracked their experiences over the subsequent two-year period. In this section, we briefly describe our approach to conducting the various analyses described above. A detailed discussion of our methodological approach is provided in the Appendix.
Descriptive Analyses. Our study is primarily based on descriptive analyses in which we tabulated means and distributions of variables of interest. All figures were calculated using longitudinal sample weights in SIPP to make the findings representative of the U.S. civilian population in April 2001. Descriptive methods were used in analyses to (1) calculate poverty rates and poverty exit rates among single mothers in 2001; (2) profile single mothers who exited poverty in 2001 and how they compared to other single mothers; (3) profile the three groups of single mothers who exited poverty based on their poverty and non-poverty spells during the two-year follow-up period; and (4) compare the income, poverty, and job experiences of the three groups over the follow-up period. Where relevant, we conducted t-tests or chi-square tests to gauge whether or not observed differences in variable distributions across different groups of single mothers are statistically significant.(4)
We also used descriptive methods to identify key events that might have triggered sample members' initial poverty exits and their subsequent poverty reentries. In our analysis, we identified trigger events over a two-month window and a four-month window prior to the poverty exit and examined the proportion of sample members who exited poverty via various pathways.
Finally, we used life-table methods to look at the duration of non-poverty spells as well as subsequent poverty spells. These methods account for right-censored spells — that is, spells that were ongoing at the end of the follow-up period.
Multivariate Analyses. We conducted multivariate analyses to examine the relative importance of individual and job characteristics associated with successful poverty exits. In these analyses, we also included as model covariates national- and state-level data on economic conditions, poverty levels, and welfare rules. We estimated logit models where we modeled, as a function of covariates, the probability that a sample member spent a significant fraction of the two-year follow-up period out of poverty. We also estimated multinomial logit models to examine factors associated with membership in each of the three groups of poverty exiters: (1) single mothers who exited poverty and stayed out of poverty, (2) those who returned to poverty and remained poor, and (3) those who cycled in and out of poverty.
Summary of Key Findings
Both poverty and poverty exits are common among single mothers. More than half of single mothers in 2001 experienced poverty for at least one month during the year, and, among those poor, more than half exited poverty by the end of the year. Single mothers who exited poverty in 2001 were more disadvantaged than single mothers who were never poor in 2001, but were somewhat less disadvantaged than poor single mothers who did not exit poverty in 2001.
The most common reason associated with exiting poverty for single mothers is an increase in their own employment or earnings. In 2001, about three-quarters of single mothers who exited poverty experienced an employment or earnings increase soon before their poverty exit. Changes in family composition and in the earnings of other adult relatives in the family are much less common events associated with poverty exits for this population.
Many single mothers who exit poverty soon return to poverty (in 2001, the median non-poverty spell duration was about eight months and about 60 percent of non-poverty spells ended within one year). However, subsequent poverty spells for these women tend to be even shorter, suggesting that many poor single mothers exit poverty again. Thus, we find considerable cycling in and out of poverty for this population.
Based on the duration of their non-poverty and poverty spells during a two-year follow-up period, we found that single mothers who exit poverty can be classified into one of three groups: (1) 28 percent who exit poverty and never return — poverty leavers; (2) 56 percent who cycle in and out of poverty — poverty cyclers; and (3) 16 percent who return to poverty and stay poor — poverty returners. Poverty dynamics differ markedly across these three distinct groups of single mothers at the time of their poverty exits and afterward.
At the time of their poverty exits, the poverty leavers tend to be less disadvantaged than the other two groups of single mothers in their basic demographic and human capital characteristics. For instance, they are more likely than the other two groups to be high school and college graduates and less likely to have health problems. Poverty leavers also have older and fewer children, and are more likely to live with other adult relatives. They are also less likely to receive public assistance. It is interesting, however, that in the months leading up to the poverty exits, there are no differences across the groups in their employment, earnings, and poverty levels.
The poverty leavers, however, are more likely than the other two groups to exit poverty because of an employment or earnings increase. Furthermore, the members of this more highly educated group tend to have substantially higher earnings gains when they exit poverty, primarily because they tend to work more than the other groups and to find higher-paying jobs. Thus, initial job quality at the time of poverty exit is associated with more successful outcomes for this population.
Differences across the three groups extend into the post-poverty-exit period. Poverty leavers tend to experience higher earnings growth over time than the other two groups, and move to better jobs with greater benefits. In contrast, the earnings of the poverty returners tend to diminish over time, and these women tend to become increasingly reliant on public assistance.
Overall, a mixed picture emerges about the economic success, in the medium term, of single mothers who exit poverty. Nearly 30 percent remain out of poverty, and only about 15 percent return to poverty and are unable to exit again. The majority of single mothers who exit poverty cycle in and out of poverty with short spells of non-poverty followed by even shorter spells of poverty. Over time, these cyclers experience some income growth, spend more time out of poverty than in poverty (nearly half stayed out of poverty for at least three-quarters of months over a two-year follow-up period), and become slightly less reliant on public assistance. However, their economic progress is slow.
Our key finding is that employment and earnings play a primary role in the economic outcomes for single mothers who exit poverty. Increases and decreases in single mothers' earnings are, by far, the most common reasons for their poverty exits and reentries. Education levels, health status, and the number of adult relatives in the family are important factors associated with economic success for this population, partly due to their effects on employment and earnings. Given that many single mothers are able to find jobs, but many are not able to sustain or advance in these jobs, policies providing employment support for single mothers who exit poverty could help improve their economic success.
In examining these findings, it should be kept in mind that the 2002-2003 period in which we track income and employment experiences of those who exited poverty was marked by weaker economic conditions than in earlier years. For instance, after a booming economy in the late 1990s and 2000, when the national unemployment rate was below four percent, the unemployment rate during the 2002-2003 period had crept close to six percent. The weaker economic conditions these single mothers faced could have affected their experiences, and made it more difficult for them to remain employed and stay out of poverty. The fact that the unemployment rate did not make a difference in our multivariate analyses, combined with the fact that the economic downturn was fairly mild, suggests that the poverty patterns for these single mothers might not differ much at other times. Furthermore, a thorough understanding of the issues raised in the study can provide insights into policy initiatives for single mothers, such as work supports for former welfare recipients, including the Earned Income Tax Credit, child care subsidies, and approaches to improve job retention and advancement.
Roadmap to the Rest of the Report
Chapter II of the report compares the characteristics of single mothers who exit poverty to those of other groups of single mothers. This chapter also examines key events associated with poverty exits. Chapter III reports results from the spell analysis, including the length of non-poverty spells, reasons that single mothers reenter poverty, and the duration of subsequent poverty spells.
Chapters IV and V focus on the three groups of single mothers who exited poverty: those who remained out of poverty for the rest of the follow-up period, those who cycled in and out of poverty, and those who returned to poverty for the remainder of the period. In Chapter IV, we compare the characteristics of single mothers in these three groups and their reasons for exiting poverty. Chapter V focuses in more detail on the income and employment experiences of these three groups, as well as their job characteristics and public assistance receipt over the two-year period. Chapter V concludes with the findings from our multivariate analyses that describe factors — measured at the time of poverty exit — that are associated with economic success for the study population.
1. In addition to SIPP data, in our multivariate analyses, we utilize basic national- and state-level data on economic conditions, poverty levels, and welfare rules.
2. Because food stamps are often considered a "near-cash" benefit, we experimented with including food stamp benefits in our definition of cash income; however, this led to only minor changes in our sample. See the Appendix for more discussion of issues related to income definition.
3. The decision to include single mothers with a cohabiting partner is based on our focus on family-level poverty rather than household-level poverty. Consistent with this approach, we do not include cohabiters in constructing poverty thresholds, nor do we include their income as available family resources. This strategy is equivalent to assuming that cohabiters consume only their own resources and do not contribute toward or consume the resources of the family.
4. All standard error estimates used in these tests account for clustering and stratification in the SIPP design.