Welfare use as a life course event: toward a new understanding of the U.S. safety net.

Rank, Mark R. ; Hirschl, Thomas A.

Few behaviors are as stigmatized in American society as that of welfare use. Survey research has repeatedly documented the public's considerable animosity regarding means-tested welfare programs and their participants (Gilens, 1999; Horan & Austin, 1974; Kluegel & Smith, 1986; MacLeod, Montero, & Speer, 1999; Tropman, 1998). At the heart of this opposition is the belief that welfare recipients are largely undeserving of such assistance. For example, regarding cash assistance programs for the working age poor people, Gilens wrote (1999):

While no one factor can fully account for the public's opposition to welfare, the most important single component is this widespread belief that most welfare recipients would rather sit home and collect benefits than work hard themselves. In large measure Americans hate welfare because they view it as a program that rewards the undeserving poor. (pp. 2-3)

Accentuating this belief is the pervasive image that those who use welfare are predominately ethnic minority populations, ofter plagued by alcohol or drug problems, have large numbers of children, and remain on the dole for years at a time (Gans, 1995; Quadagno, 1994). For the most part, Americans perceive the use of welfare as something that happens to someone else and atypical of the American experience.

In this article we empirically analyze this assumption through two straightforward research questions--How likely is it that Americans will at some point during their adulthood turn to welfare assistance? In addition, what is the extent of time that Americans use the social safety net? The application of a life table methodology allows us to answer each of these questions.

Such questions appear particularly relevant for social work. The profession has had a long and noteworthy association with government programs for the needy--from daily casework, to administering public assistance programs, to advocating the importance of a safety net in contemporary society. Yet this involvement has often been devalued and marginalized, partially as a result of the perception that welfare and its recipients are simply not a part of the mainstream American experience. An empirical understanding of the fact that most Americans will at some point in their adult lives use a means-tested program is a vital component in breaking this mind-set. As we discuss in the last section, such an understanding can lead to a greater appreciation regarding the wider relevancy of a social safety net, as well as the role of social workers in assisting individuals and families through periods of economic vulnerability.

Current Knowledge of Welfare Participation and Dynamics

To qualify for various means-tested welfare programs, households must fall below certain income levels (for example, to receive food stamps, families must generally be at or below 130 percent of the official poverty line). In addition, public assistance programs do not allow households to hold assets beyond a certain monetary level. In general, these are set quite low. The result is that to receive benefits from a social safety net program, individuals and families must be either below or not far removed from the poverty line, with a minimum accumulation of assets. Income and asset guidelines vary from program to program and, in some cases, from state to state within the jurisdiction of a program.

Welfare in the United States consists of either in-kind assistance or cash. In-kind programs provide specific resources such as food, housing, and medical assistance. Major in-kind programs include food stamps, Medicaid, and housing assistance. Cash programs provide the recipient with a monthly check. These include Temporary Assistance to Needy Families (TANF), which replaced Aid to Families with Dependent Children (AFDC) in 1996; Supplemental Security Income (SSI); and general assistance.

Participation by the U.S. population varies from program to program. For example, using data from the Survey of Income and Program Participation (SIPP), the U.S. Bureau of the Census (1999) estimated that in an average month in 1994, 11.3 percent of the U.S. population received Medicaid, 9.7 percent received food stamps, 5.5 percent received AFDC or general assistance, 4.7 percent received housing assistance, and 2.0 percent received SSI. The percentage of the U.S. population that received any means-tested welfare program in an average month in 1994 was 15.2 percent.

As the observed time is lengthened, the percentage of the U.S. population turning to welfare assistance increases. For example, in an early and highly influential study of poverty and welfare receipt, Duncan (1984) focused on the likelihood of individuals receiving public assistance (defined as AFDC, general assistance, SSI, or food stamps) in a single year versus a 10-year period. Duncan found that although only 8.1 percent of the population received any welfare income in 1978, one-quarter of the population participated in one or more welfare programs at some point during the period from 1969 to 1978. Reasons for turning to welfare were related to income or job loss, family changes and dissolution, and health concerns.

Other longitudinal studies have further analyzed the length of time individuals remain on public assistance. The vast majority of these studies have focused on female-headed families receiving AFDC (Bane & Ellwood, 1994; Boisjoly, Harris, & Duncan, 1998; Harris, 1996; Moffitt, 1992; O'Neill, Bassi, & Wolf, 1987; Plotnick, 1983; Rein & Rainwater, 1978; Sandefur & Cook, 1998). Using several large longitudinal data sets, such as the Panel Study of Income Dynamics (PSID), the National Longitudinal Survey of Youth (NLSY), and state welfare caseload data, these analyses have shown that most periods of welfare use are of fairly short duration (for example, two or three years) with a substantial degree of recidivism. For example, Bane and Ellwood estimated that of all women beginning a period of welfare receipt, 69 percent would exit from public assistance after four years. However, 26 percent of those leaving the welfare system would return within three years.

This body of research has also shown that people experiencing longer spells of welfare use often have characteristics that place them at a distinct disadvantage vis-a-vis the labor market (for example, individuals with work disabilities, low education, greater numbers of children, or residence in inner-city areas). The results from these studies largely mirror the findings that have been gathered regarding the length and duration of poverty spells (Bane & Ellwood, 1986; Blank, 1997; Devine & Wright, 1993; Duncan et al., 1995; Stevens, 1999; Walker, 1994).

A Life Course Perspective

An alternative approach to conceptualizing and measuring welfare use is to place welfare participation in the context of the adult life course. Welfare receipt then is analyzed as a life course event that may or may not occur as individuals age through adulthood.

The life course approach raises a number of questions not previously addressed. To what extent is the use of welfare a "normal" life course event? How does the life-cycle risk of short periods of welfare use differ from the risk of longer periods? What is the total number of years during adulthood that individuals will use the welfare system?

The concept of the life course has had a long and distinguished history in the social and applied sciences (Settersten & Mayer, 1997). It has provided a framework for thinking about how individual lives unfold (Elder, 1995). The term itself refers to "social processes extending over the individual life span or over significant portions of it, especially [with regard to] the family cycle, educational and training histories, and employment and occupational careers" (Mayer & Tuma, 1990, p. 3). In addition, as Settersten and Mayer argued, "While these dimensions describe the primary activities across life, a more complete picture of the life course must also include more marginal periods and events--such as brief periods of training, second or part-time jobs, periods of unemployment or sickness" (p. 252). Welfare use would appear to be an appropriate addition to this list.

The analysis in this article builds on our earlier work that has examined the probabilities of poverty across the life cycle (Rank & Hirschl, 1999a, 1999b, 1999c, 2001a, 2001b). This research has shown that the life-course risk of poverty is quite high. For example, between the ages of 20 and 85, two-thirds of Americans will experience at least one year of impoverishment (Rank & Hirschl, 1999c). For those encountering poverty, it is generally for only one or two consecutive years. However, once people experience poverty, they are quite likely to encounter poverty again (Rank & Hirschl, 2001c). This body of work has established that brief periods of economic vulnerability are very much a part of the adult American life course. On the basis of these findings, we anticipate a substantial life course risk of welfare use.

Our focus is on the likelihood of Americans using public assistance, broadly defined as the major means-tested programs in the United States (food stamps, Medicaid, SSI, AFDC, and other cash welfare). Thus, to what extent will Americans at some point find themselves economically strapped and having to use government assistance? Or, to what extent does the welfare system touch the lives of U.S. adults?

Method

Data

The Panel Study of Income Dynamics is a nationally representative, longitudinal sample of households interviewed annually since 1968 (see Hill, 1992, for a detailed description of PSID). It constitutes the longest running panel data set in the United States and was designed to record income dynamics over time. In addition, it tracks yearly household participation in an array of welfare programs.

The PSID initially interviewed approximately 4,800 U.s. households in 1968, collecting detailed information on roughly 18,000 individuals. The PSID tracks these individuals annually, including children and adults who leave their original households to form new households (for example, children leaving home, separations, divorce). Thus, the PSID is designed so that in any given year the sample is representative of the entire nonimmigrant U.S. population. Throughout the analysis we used the sampling weights to ensure that the PSID sample will accurately reflect the U.S. population.

The PSID interviews a primary adult in each household to obtain information about each member of the family. From 1968 to 1972, these interviews were conducted face to face. Since 1973 the majority of interviews have been carried Out by telephone. Consequently, recent PSID waves consist of approximately 92 percent telephone interviews and 8 percent face-to-face interviews (as a result of respondents not having a telephone or having circumstances that prevent a telephone interview being feasible). The original response rate in 1968 was 76 percent. Since 1969 the annual response rates have ranged between 96.9 percent and 98.5 percent (PSID , 2001).

Although the PSID is without question the best available data set for the life table analysis presented in this article, it is not without drawbacks. In particular, the cumulative nonresponse rate and the lack of representativeness of the immigrant population is problematic. (For a more detailed discussion of the issues of data quality within the PSID, see Duncan & Hill, 1989; Hill, 1992; and Kim & Stafford, 2000.)

We used both the household and individual information from the initial wave of 1968 through 1997. Thirty years of longitudinal information are embedded in the analysis. Our analytical strategy was to use the household income and demographic information on individuals to construct several life tables that estimate the use of welfare across the adult life span. For conceptual clarity we confined our analysis to the prime working years of adulthood (ages 20 to 65).

Life Table Approach

The life table is a technique that demographers and medical researchers often use. Although primarily found in studies of mortality, it can be applied to other areas of research as well (Namboodiri & Suchindran, 1987). The life table shows the extent to which a population encounters a specific event across an interval of time. Our time intervals cover each year that an individual ages. During those years, we can calculate the proportion of the population experiencing an event (in this case, the use of welfare) for those who have yet to experience the event. The cumulative proportion of the population experiencing an event across adulthood can be calculated from the proportions at each age. These cumulative proportions allow us to gauge the percentage of the overall population that will participate in a welfare program during adulthood (for a more detailed description of this approach, see Rank & Hirschl, 2001c).

The process of arriving at the specific proportions is as follows. For each wave (or year) of the study we have information regarding the age of an individual as well as whether the household received any welfare benefits during the year. If the household did not, this is noted, and the individual is allowed to continue to the next year. If the individual did receive welfare, this is also noted, but the individual (as well as any other household members) is then removed from any further analysis. In other words, once the event of welfare receipt has occurred, the individual is no longer at risk of using welfare for the first observed time and is excluded from the calculations of proportions at later age intervals. Each age interval, therefore, contains a large number of individuals who have not experienced welfare receipt and a much smaller number of individuals who have. From these numbers the overall proportion of the population experiencing a first observed period of welfare use at each specific age is cal culated. Finally, from these age-specific proportions we can generate the cumulative proportions that span the adult life cycle.

To estimate the occurrence of various consecutive and total years of welfare use, we used a similar approach. For example, in our life tables that estimate welfare use in two or more consecutive years, if an individual first used a welfare program at age 25 and then again at age 26, he or she would be counted as experiencing two consecutive years of welfare receipt at age 26.

Individuals may contribute anywhere from one to 30 person-years in the life table. For example, a woman in the PSID study who turned 20 years old in 1980 and then in 1985 experienced a year in which she received benefits from a welfare program would have contributed six person-years in our analysis. In this case, she would be included in the estimates for ages 20, 21, 22, 23, 24, and 25. Period effects are smoothed out within and across the age intervals.

To extend our analysis beyond the 30 years of data points, we allowed individuals to enter our life tables at the ages at which they entered the study, rather than simply age 20. For example, an individual who was age 30 in 1968 (the start of the study) would be included in our 30-year-old age-specific proportions and then followed accordingly (although obviously he or she would be excluded from the 20 to 29 age-specific proportions). This procedure enabled us to extend the life table estimates out to age 65. In addition, it allowed us to use the full array of data found in the PSID, which ensured ample sample size for all age categories from which we derive our estimated probabilities.

A consequence of this approach, however, is that it introduces left censoring into the analysis. Left censoring occurs for individuals who enter the study in midstream and for whom we do not have information as to whether the event (in this case, welfare use) has occurred before the age of entry. To adjust for any bias introduced as a result of this procedure, we estimated a series of correction factors that were applied to each life table using the procedure described in Rank and Hirschl (1999b, 1999c). This allowed us to detect and correct for the fact that left censoring is present in our analysis.

Measurement

Our measure of welfare use was constructed from a series of questions asked by the PSID interviewers as to whether the household has received any cash or in-kind public assistance at some point during the previous calendar year (1967 through 1996). Cash programs include AFDC, SSI, and other cash welfare such as general assistance. In-kind programs consist of food stamps and Medicaid. From 1967 to 1996, Medicaid, food stamps, and AFDC were available in the United States (although the PSID only began to ask questions about Medicaid receipt in 1977). SSI was included in the PSID from 1975 onward. (Information about means-tested program participation in the PSID has been shown to be quite accurate, reflecting positively on the ability of respondents to recall information about welfare receipt; see Duncan & Hill, 1989.)

For each of the cash programs (AFDC, SSI, and other cash welfare), respondents were asked, "Did you receive any income in [prior year] from [the specific program]?" Respondents who were married were also asked a set of questions if their spouse had received any income from these programs. With regard to food stamps, respondents were asked, "Did you (or anyone else in your family) use government food stamps at any time in [prior year]?" Finally, Medicaid use was assessed through the question, "Is anyone in your family living there covered by (Medicaid or Medical Assistance/Medi-Cal)?" Interviewers were instructed to be careful not to have respondents include Medicare in their answer regarding Medicaid receipt. These procedures ensured that respondents would not confuse one program with the other or combine them in their response.

If a household received any public assistance, the individuals in the household were counted as receiving welfare. Earlier research has demonstrated that assistance from welfare programs is typically shared and benefits all members of a household (Edin & Lein, 1997; Rank, 1994; Seccombe, 1999). Cash programs and food stamps are obvious examples, but even in the case of one family member receiving Medicaid, the other members of the family would typically benefit in that they do not have to absorb the costs of medical expenses for that individual.

Finally, we note that a household's use of welfare during any given year might consist of varying months that they actually received welfare. For some, that time may constitute the entire year, whereas for others, it may represent only one or two months. In both cases, however, the individuals in those households were counted as receiving welfare during the calendar year.

Results

Table 1 displays the age-specific proportions, cumulative proportions, and cumulative percentages of the U.S. population receiving welfare between ages 20 and 65. For reasons of space, these are reported at five-year intervals (the age-specific proportions beyond age 20 are the five-year averages). We divided the analysis into receipt of a cash program, an in-kind program, and any welfare program (cash and/or in kind).

Looking first at the cumulative percentages, we can see that for 20 year olds, 8.3 percent of the population were beneficiaries of a cash welfare program, 12.6 percent received an in-kind assistance, and 14.2 percent received a cash and/or in-kind program benefit (which is roughly equivalent to the earlier reported average cross-sectional participation rates by the U.S. Bureau of the Census, 1999). By the time Americans have reached age 35, 22.4 percent have received cash benefits, 38.1 percent have received in-kind benefits, and 38.7 percent have received benefits from either type of program. For people at age 50, the percentages rise to 30.1 percent, 50 percent, and 50.9 percent, respectively, and by age 65, 37.6 percent of Americans will have used a cash welfare program at some point during their adult years, 64.2 percent will have used an in-kind program, and 65 percent will have used a cash and/or in-kind program.

Americans are much more likely to have received benefits from an in-kind program during the life course than from a cash program (Table 1). In fact, virtually every American who receives welfare receives benefits from at least one in-kind program. Put a slightly different way, of the two-thirds of Americans who received public assistance, we calculated in a separate analysis that 63 percent received Medicaid, 52 percent received food stamps, 13 percent received AFDC, 10 percent received SSI, and 14 percent received some other cash welfare benefit. Individuals can receive benefits from several programs at once (such as AFDC, food stamps, and Medicaid) or only one program (such as Medicaid). Clearly, Medicaid and food stamps have the farthest reach in terms of use across the adult life course. Rather than being targeted to one category of the population (for example, SSI to people with disabilities or AFDC to female-headed families), these programs are more widely available as a result of fewer eligibility res trictions. This, in turn, results in greater use across the adult years. In addition, AFDC or SSI recipients automatically receive Medicaid.

Looking at the age-specific proportions for receiving benefits from any welfare programs, we can see that they are higher at the early ages (partially because this is the starting point for our analysis) and the later ages. These proportions reflect a U shaped pattern for the age specific risk of using welfare for the first time. That risk is higher at the early ages, declines to a low point during the 40s, and then starts to increase during the 50s and 60s. This U-shaped pattern is consistent with patterns of cross-sectional rates of welfare use by age (U.S. Bureau of the Census, 1999) and cross-sectional poverty rates by age (U.S. Bureau of the Census, 2001). These dynamics largely reflect the fact that the prime earning years are generally in the 40s and early 50s, reducing the need for welfare.

In particular, people age 20 and those in their 60s have much higher age-specific rates of using welfare for two different reasons. Age 20 is the start of the analysis; therefore, these individuals reflect the overall cross-sectional rate of 20 year olds, which is quite high. However, by age 21, individuals experiencing welfare receipt are removed from the analysis, leaving the population as a whole less at risk. In addition, younger adults are much more economically vulnerable than their middle-aged counterparts. Those in their 60s experience higher age-specific rates because they are much more likely to be receiving Medicaid for the first time (particularly at age 65).

The majority of Americans will use a public assistance program at least once during their adulthood (Table 1). Two-thirds of adults will turn to a means-tested program for some type of assistance between the ages of 20 and 65. These findings are consistent with our earlier analyses addressing the likelihood of poverty across the life course that has revealed that the majority of Americans will experience poverty or near poverty at some point during their adulthood (Rank & Hirschl, 1999c, 2001a, 2001b). The fact that two-thirds of adult Americans will use a social safety net program is emblematic of these life-course patterns of poverty and economic vulnerability.

Table 2 shows shows the risk that individuals face in terms of the amount of time they will experience welfare receipt. The top panel of Table 2 shows the likelihood of encountering one, two, three, four, and five or more consecutive years of welfare receipt. The bottom panel illustrates the total number of years that welfare was received (regardless of whether the years are consecutive).

This is calculated for one, two, three, four, and five or more total years. Again, for reasons of space, only the cumulative percentages are reported at five-year intervals. (Readers interested in obtaining the full set of life table calculations for Tables 1 or 2 may contact the first author.)

Between the ages of 20 and 35, 38.7 percent of Americans received welfare in at least one year, 23.3 percent in two or more consecutive years, 16.8 percent in three or more years, 11.3 percent in four or more years, and 8.5 percent in five or more years. By age 50 the percentages are 50.9 percent, 31.3 percent, 23.2 percent, 15.8 percent, and 12 percent, respectively. Finally, by age 65, the corresponding percentages are 65 percent, 41.4 percent, 31.8 percent, 21.1 percent, and 15.9 percent, respectively. Consequently, the top panel of Table 2 reveals that although the likelihood of experiencing one, or two consecutive years of welfare use across the adult life course is considerable, there is a sharp drop in the probability of experiencing a higher number of consecutive years of welfare use.

The bottom panel of Table 2 tells us a somewhat different story. The total number of years in which welfare will be received across adulthood is considerably greater. By age 65, 65 percent of the population will have received welfare in at least one year, 58.7 percent in at least two different years, 54.2 percent in three different years, 48 percent in four different years, and 40.3 percent in five or more different years. Thus, whereas the upper panel tells us that the number of consecutive years that welfare is used is generally quite short, the bottom panel informs us that such periods tend to reoccur across the span of adulthood. Accordingly, 90 percent of those who use welfare once will do so at least once more during their adulthood.

The patterns found in Table 2 are consistent with the earlier cited research on the dynamics of welfare periods. That is, most Americans who use welfare do so over fairly short intervals, whereas a minority will receive public assistance for extended periods of time. Yet once individuals use welfare, they are quite likely to do so again at several points in their lives. These patterns are also evident in the context of the adult life cycle (Table 2).

Summary and Implications

Our findings appear to cast considerable doubt on the notion that the use of welfare falls outside the mainstream American experience. In fact, two thirds of all Americans between ages 20 and 65 will at some point turn to a public assistance program. Such assistance is often in the form of in kind programs such as Medicaid or food stamps. It should also be noted that had we carried the analysis out to age 80 or 85, these percentages would be higher. From age 65 onward, there is an increased likelihood in the use of Medicaid or SSI or both.

Our analysis also indicates that the use of welfare across the adult years tends to occur over fairly short intervals of time. Although 65 percent of the population will encounter at least one year of welfare use, only 16 percent of the population will use public assistance for five consecutive years. Yet, it is also true that once individuals use welfare, they are likely to do so again; of those using a welfare program between the ages of 20 and 65, 90 percent will do so more than once. As a result, the total number of years that welfare is used across the life course is sizeable.

These findings have several important implications. First, contrary to much of the popular rhetoric, the use of the U.S. social safety net is widespread and mainstream. Although public assistance users are routinely vilified and portrayed as members of marginalized groups, in fact, most Americans will encounter the welfare system at some point during their adult years. Although it is true that the more disadvantaged members of our society will interact to a greater extent with poverty and the welfare system, the relevance of a social safety net is widespread.

It is interesting to contrast these patterns with the fact that the U.S. welfare state, and particularly its social safety net, has often been described in minimalist terms (Esping-Andersen, 1990). Compared with other Western industrialized countries, the United States devotes far fewer resources to programs that assist economically vulnerable people (Organization for Economic Cooperation and Development, 1999). For example, Noble (1997) wrote, "The U.S. welfare state is striking precisely because it is so limited in scope and ambition" (p. 3).

Although the U.S. welfare state may be minimalist in terms of the scope and level of benefits it offers to impoverished people, it is far from minimalist in the extent to which it is eventually relied on by the general population. This fact has gone largely unnoticed until now. The importance and relevance of the welfare state should be perceived as a mainstream issue, affecting the lives of millions of Americans.

Second, the fact that most life course episodes of welfare use are relatively short implies that the welfare state operates largely as an insurance policy against sporadic periods of economic uncertainty that occur across the life cycle. Although some individuals use public assistance for long consecutive periods of time, the majority of Americans turn to welfare programs for short-term assistance. This would suggest that the social safety net is operating to a large extent as most Americans indicate they would like it to operate (Gilens, 1999).

Yet, it is also true that the total number of years across the life course in which Americans will receive assistance from a welfare program is quite high. Our data revealed that approximately 40 percent of the population will use a welfare program in five or more separate years between the ages of 20 and 65. It should be noted that we are measuring only whether welfare is used at any point during the year, rather than looking at use across all months of the year. Nevertheless, the fact that two of five Americans will use a social safety net program in at least five different years during their working age adulthood is startling.

These findings appear relevant in the context of the welfare reform legislation of 1996 and the 2002 reauthorization. The welfare reform legislation established a five-year lifetime limit on the use of the TANF program (which replaced the AFDC program). What might our results imply in terms of this stipulation? Because AFDC receipt was a relatively small component of welfare use across the life course, TANF is likely to be a relatively small component of future welfare use. In addition, given the patterns revealed in this research, relatively few recipients are likely to use TANF for 60 consecutive months. On the other hand, across a period of 20 to 30 years, the patterns found in our data imply that the number of individuals who will eventually run into a five-year lifetime TANF limit would be far from trivial.

A third implication is that the patterns found in this article clearly underscore the existence of economic vulnerability in the United States. This is consistent with and builds on our earlier findings that two-thirds of Americans will fall below the poverty line between the ages of 20 and 85 (Rank & Hirschl, 1999c) and that three-quarters of Americans will experience poverty or near poverty (for example, have household income below 125 percent of the poverty line) (Rank & Hirschl, 1998). These findings suggest that our free market economic structure is associated with periods of economic uncertainty in the lives of its participants (see Rank, 2001). A consequence of this economic vulnerability is that the majority of Americans will turn to a safety net program to help them through these periods of economic turmoil.

Finally, what do these findings suggest for social work practitioners in direct and policy practice. Direct practitioners are familiar with the feelings of guilt, shame, and sense of failure that often accompany their clients' use of public assistance.

Understanding that the use of welfare is a life-course event experienced by a majority of the U.S. population can empower individuals and social workers to view this behavior in a radically different light. Economic vulnerability and poverty are a part of the American experience. The use of welfare can be comprehended as a commonplace behavior engaged in by millions of individuals and families to cope with periods of economic adversity. As Leisering and Leibfried (1999) wrote with respect to their life course analysis of poverty and social assistance in Germany:

Poverty is no longer (if ever it was) a fixed condition or a personal or group characteristic, but rather it is an experience or a stage in the life course. It is not necessarily associated with a marginal position in society but reaches well into the middle class. Poverty is specifically located in time and individual biographies, and, by implication, has come to transcend traditional social boundaries of class. (p. 239)

This knowledge constitutes a fundamental shift in understanding the use of social welfare, from viewing it as a consequence of individual failure or the lack of a strong work ethic to recognizing it as a byproduct of the economic structure in the context of the life course. Such awareness can inform and guide social workers and their clients about the context in which public assistance takes place.

Being aware of the life course patterns of poverty and the use of social welfare programs also adds an important but often neglected element to the concept of human development. Social workers frequently use this perspective as a guide for interpreting events and behaviors at various stages in their clients' lives. Economic vulnerability and the need for social assistance would appear quite prevalent across the stages of the life span, particularly during the early and late years of adulthood. Anticipating and planning for these periods of economic vulnerability might be built into a long-term, proactive economic strategy that social workers develop in conjunction with their clients. One such strategy might include building assets to partially offset future periods of economic insecurity (see Shapiro & Wolff, 2001).

The findings of this study also are important to social workers in policy practice. The profession historically has emphasized the importance of an ongoing government commitment to mediate the conditions of poverty and economic vulnerability. In advocating for a humane response to the vagaries of our economic structure, social workers often have championed the role of a safety net. Our findings suggest that such advocacy is well-placed. Although the merits of particular programs and approaches can be debated, the need for some type of government response is without question.

This study also demonstrated that a social safety net is applicable not only to disenfranchised individuals in society, but to the majority of Americans. Although policy analyses have shown that the bulk of government resources benefits middle- and upper-income groups rather than people who are poor (Abramovitz, 2001; Sherraden, 1991), the concept of a safety net has been framed as an issue of relevance to marginalized groups in society. Our results indicate that this outlook is incomplete and misleading. Those who will be directly affected by decisions at the national and state levels to contract (or expand) the social safety net are in fact the majority of the U.S. population. Social workers advocating and lobbying for safety net issues should use this information, particularly with respect to the changes that have occurred and will likely occur in social welfare programs.

The profession has the opportunity to shift the debate about the social safety net in a fundamental way--from an issue of "them" to an issue of "us." Although social work has admirably defended and supported downtrodden individuals in society, its voice has often been marginalized in the process. Understanding that downtrodden groups may one day include any number of us brings a very different reality and relevance to both the area of social welfare and the profession of social work. This allows us to recognize our frailty and shared connections as we navigate across the life course. Few professions are better suited to helping individuals and families traverse such a journey than that of social work.

Table 1

Life Table Analysis of Experiencing Welfare Receipt for the U.S. Adult
Population

 Recipient of a Cash Recipient of an In-kind
 Welfare Benefit Welfare Benefit

 Age Specific Cumulative Cumulative Age Specific Cumulative
Age Proportion Proportion Percentage Proportion Proportion

20 .0834 .0834 8.34 .1259 .1259

25 .0156 .1526 15.26 .0319 .2569
30 .0101 .1947 19.47 .0198 .3274
35 .0074 .2241 22.41 .0165 .3810

40 .0071 .2512 25.12 .0164 .4301
45 .0070 .2771 27.71 .0121 .4639
50 .0068 .3014 30.14 .0140 .5003

55 .0066 .3240 32.40 .0145 .5355
60 .0077 .3498 34.98 .0170 .5738
65 .0081 .3755 37.55 .0339 .6415

 Recipient of Recipient of Any
 an In-kind
 Welfare Welfare Benefit
 Benefit

 Cumulative Age Specific Cumulative Cumulative
Age Percentage Proportion Proportion Percentage

20 12.59 .1424 .1424 14.24

25 25.69 .0306 .2660 26.60
30 32.74 .0188 .3326 33.26
35 38.10 .0168 .3868 38.68

40 43.01 .0169 .4369 43.69
45 46.39 .0127 .4718 47.18
50 50.03 .0144 .5088 50.88

55 53.55 .0155 .5457 54.57
60 57.38 .0176 .5843 58.43
65 64.15 .0334 .6495 64.95
Table 2

Cumulative Percentages of the U.S. Adult Population Receiving Welfare
Benefits by Number of Years

 Years of Welfare Receipt

 1 Year 2 Years 3 Years 4 Years
 or More or More or More or More
Age % % % %

Consecutive Years Experienced

20 14.24 - - -
25 26.60 14.92 9.90 6.22
30 33.26 20.01 14.02 9.29
35 38.68 23.25 16.81 11.25
40 43.69 26.19 19.11 12.97
45 47.18 28.55 21.10 14.50
50 50.88 31.27 23.18 15.76
55 54.57 33.94 25.55 17.44
60 58.43 37.33 28.49 18.84
65 64.95 41.40 31.79 21.07

Total Years Experienced

20 14.24 - - -
25 26.60 18.92 13.90 9.84
30 33.26 26.52 22.28 17.91
35 38.68 31.93 27.42 23.33
40 43.69 36.48 32.37 27.82
45 47.18 40.55 36.54 31.37
50 50.88 44.23 40.50 35.18
55 54.57 48.21 44.11 38.83
60 58.43 52.71 48.35 42.56
65 64.95 58.66 54.20 48.00

 Years of
 Welfare
 Receipt

 5 Years
 or More
Age %

Consecutive Years Experienced

20 -
25 4.52
30 7.05
35 8.47
40 9.77
45 10.94
50 11.95
55 13.19
60 14.41
65 15.88

Total Years Experienced

20 -
25 6.72
30 14.22
35 18.14
40 21.94
45 25.34
50 29.16
55 32.22
60 35.48
65 40.34

Original manuscript received May 22, 2000

Final revision received September 7, 2001

Accepted October 20, 2001

References

Abramovitz, M. (2001). Everyone is still on welfare: The role of redistribution in social policy. Social Work, 46, 297-308.

Bane, M. J., & Ellwood, D. T. (1986). Slipping into and out of poverty: The dynamics of spells. Journal of Human Resources, 21, 1-23.

Bane, M. J., & Ellwood, D. T. (1994). Welfare realities: From rhetoric to reform. Cambridge, MA: Harvard University Press.

Blank, R. M. (1997). It takes a nation: A new agenda for fighting poverty. Princeton, NJ: Princeton University Press.

Boisjoly, J., Harris, K. M., & Duncan, G. J. (1998). Trends, events, and duration of initial welfare spells. Social Service Review, 72, 466-492.

Devine, J. A., & Wright, J. D. (1993). The greatest of evils: Urban poverty and the American underclass. New York: Aldine De Gruyter.

Duncan, G. J. (1984). Years of poverty, years of plenty: The changing economic fortunes of American workers and families. Ann Arbor, MI: Institute for Social Research.

Duncan, G. J., Gustafsson, B., Hauser, R., Schmaus, G., Jenkins, S., Messinger, H., Muffels, R., Nolan, B., Ray, J., & Voges, W. (1995). Poverty and social assistance dynamics in the United States, Canada, and Europe. In K. McFate, R. Lawson, & W. J. Wilson (Eds.), Poverty, inequality and the future of social policy: Western states in the new world order (pp. 67- 108). New York: Russell Sage Foundation.

Duncan, G. J., & Hill, D. H. (1989). Assessing the quality of household panel data: The case of the Panel Study of Income Dynamics. Journal of Business and Economic Statistics, 7,441-452.

Edin, K., & Lein, L. (1997). Making ends meet: How single mothers survive welfare and low-wage work. New York: Russell Sage Foundation.

Elder, G. H, (1995). The life course paradigm: social change and individual development. In P. Moen, G. H. Elder, & K. Luescher (Eds.), Examining lives in context: Perspectives on the ecology of human development (pp. 101-140). Washington, DC: American Psychological Association.

Esping-Andersen, G. (1990). The three worlds of welfare capitalism. Princeton, NJ: Princeton University Press.

Gans, H. J. (1995). The war against the poor: The underclass and antipoverty policy. New York: Basic Books.

Gilens, M. (1999). Why Amen cans hate welfare: Race, media, and the politics of antipoverty policy. Chicago: University of Chicago Press.

Harris, K. M. (1996). Life after welfare: Women, work, and repeat dependency. American Sociological Review, 61, 407-426.

Hill, M. S. (1992). The Panel Study of Income Dynamics: A user's guide. Newbury Park, CA: Sage Publications.

Horan, P., & Austin, P. (1974). The social bases of welfare stigma. Social Problems, 21, 648-657.

Kim, Y. S., & Stafford, F. P. (2000). The quality of PSID income data in the 1990's and beyond (Panel Study of Income Dynamics Working Paper). Ann Arbor: University of Michigan.

Kluegel, J. R., & Smith, E. R. (1986). Beliefs about inequality: Americans' views of what is and what ought to be. New York: Aldine de Gruyter.

Leisering, L., & Leibfried, 5. (1999). Time and poverty in western welfare states: United Germany in perspective. Cambridge, England: Cambridge University Press.

MacLeod, L., Montero, D., & Speer, A. (1999). America's changing attitudes toward welfare and welfare recipients, 1938-1995. Journal of Sociology and Social Welfare, 26, 175-186.

Mayer, S. E., & Tuma, N. B. (1990). Life course research and event history analysis: An overview. In K. U. Mayer & N. B. Tuma (Eds.), Event history analysis in life course research (pp. 3-20). Madison: University of Wisconsin Press.

Moffitt, R. (1992). Incentive effects of the U.S. welfare system: A review. Journal of Economic Literature, 30, 1-61.

Namboodiri, K., & Suchindran, C. M. (1987). Life table techniques and their applications. Orlando, FL: Academic Press.

Noble, C. (1997). Welfare as we knew it: A political history of the American welfare state. New York: Oxford University Press.

O'Neill, J. A., Bassi, L. J., & Wolf, D. A. (1987). The duration of welfare spells. Review of Economics and Statistics, 69, 241-249.

Organization for Economic Cooperation and Development. (1999). Social expenditure database 1980- 1996. Paris: Author.

Panel Study of Income Dynamics. (2001). On-line user's guide. Ann Arbor: University of Michigan.

Plotnick, R. D. (1983). Turnover in the AFDC population: An event history analysis. Journal of Human Resources, 18, 65-81.

Quadagno, J. (1994). The color of welfare: How racism undermined the war on poverty. New York: Oxford University Press.

Rank, M. R. (1994). Living on the edge: The realities of welfare in America. New York: Columbia University Press.

Rank, M. R. (2001). The effect of poverty on America's families: Assessing our research knowledge. Journal of Family Issues, 22, 882-903.

Rank, M. R., & Hirschl, T. A. (1998, August). The likelihood of poverty across the American adult life span. Paper presented at the annual meeting of the American Sociological Association, San Francisco.

Rank, M. R., & Hirschl, T. A. (1999a). The economic risk of childhood in America: Estimating the probability of poverty across the formative years. Journal of Marriage and the Family, 61, 1058-1067.

Rank, M. R., & Hirschl, T. A. (1999b). Estimating the proportion of Americans ever experiencing poverty during their elderly years. Journal of Gerontology: Social Sciences, 54B, $184-$193.

Rank, M. R., & Hirschl, T. A. (1999c). The likelihood of poverty across the American adult life span. Social Work, 44, 201-216.

Rank, M. R., & Hirschl, T. A. (200la). The occurrence of poverty across the life cycle: Evidence from the PSID. Journal of Policy Analysis and Management, 20, 737-755.

Rank, M. R., & Hirschl, T. A. (200lb). Rags or riches? Estimating the probabilities of poverty and affluence across the adult American life span. Social Science Quarterly, 82, 651-669.

Rank, M. R., & Hirschl, T. A. (2001c). The measurement of long term risks across the life course. Social Science Quarterly, 82, 680-686.

Rein, M., & Rainwater, L. (1978). Patterns of welfare use. Social Service Review, 52, 511-534.

Sandefur, G. D., & Cook, S. T. (1998). Permanent exits from public assistance: The impact of duration, family, and work. Social Forces, 77, 763-786.

Seccombe, K. (1999). "So you think I drive a Cadillac?" Welfare recipients' perspectives on the system and its reform. Boston: Allyn & Bacon.

Settersten, R. A., & Mayer, K. U. (1997). The measurement of age, age structuring, and the life course. Annual Review of Sociology, 23, 233-261.

Shapiro, T. M., & Wolff, E. N. (2001). Assets for the poor: The benefits of spreading asset ownership. New York: Russell Sage Foundation.

Sherraden, M. (1991). Assets and the poor: A new American welfare policy. Armonk, NY: M. E. Sharpe.

Stevens, A. H. (1999). Climbing out of poverty, falling back in: Measuring the persistence of poverty over multiple spells. Journal of Human Resources, 34, 557-588.

Tropman, J. E. (1998). Does America hate the poor? The other American dilemma: Lessons for the 21st century from the 1960s and the 1970s. Westport, CT: Praeger.

U.S. Bureau of the Census. (1999). Dynamics of economic well-being: Program participation, who gets assistance? In Current Population Reports (Series P70-69). Washington, DC: U.S. Government Printing Office.

U.S. Bureau of the Census. (2001). Poverty in the United States: 2000. In Current Population Reports (Series P60-214). Washington, DC: U.S. Government Printing Office.

Walker, R. (1994). Poverty dynamics: Issues and examples. Aldershot, England: Avebury.

Mark R. Rank, PhD, is professor, George Warren Brown School of Social Work, Washington University, 1 Brookings Drive, Campus Box 1196, St. Louis, MO 63130; e-mail: [email protected]. Thomas A. Hirschl, PhD, is professor, Department of Rural Sociology, Cornell University, Ithaca, NY.