The role of postsecondary education in welfare recipients' paths to self-sufficiency.

The effectiveness of higher education as a route into self-sustaining employment has been demonstrated by decades of educational and economic research. Despite this, recent policy changes at the federal level have led to a widespread denial of access to postsecondary education for welfare recipients, arguably one of the most economically disadvantaged groups in society. Participants in the recently created Temporary Assistance for Needy Families (TANF) program--implemented as part of the 1996 Personal Responsibility and Work Opportunity Act--are strongly discouraged by requirements set out in federal law from pursuing postsecondary education as a route out of poverty. Instead, shorter-term options such as immediate job search, vocational training that lasts a year or less, and subsidized or unsubsidized employment are among the activities available to assist low-income parents transition to self-sufficiency.

In limiting postsecondary educational options to welfare recipients. TANF legislates some of the biases already inherent in societal institutions that steer disadvantaged and minority young women away from higher education. Research has shown that young people with lower socioeconomic status have worse educational outcomes, including lower rates of college attendance and graduation (Bainbridge & Lesley, 2002; Haycock, 2001; Sewell & Shah, 1967). Among those with low socioeconomic status, minority students in particular tend to have lower aspirations for college (Kao & Tienda, 1998). This may disproportionately affect young women, whose socioeconomic status has been shown to be even more important in their college-going decisions than intelligence (Sewell & Shah, 1967). Higher socioeconomic status students' comparatively better educational outcomes are due, in part, to both their more highly educated parents (Haveman, Wolfe, & Spaulding, 1991; Sewell & Shah, 1968) and the higher per-pupil expenditures in their schools (Elliott, 1998). Practices within schools themselves, such as tracking, dissuade some high school students, particularly those with low socioeconomic status and minority backgrounds, from attending college (Rosenbaum, 1976, 1980). Even efforts to de-track schools face hurdles in overcoming long-held attitudes about racial differences, social stratification, and intelligence (Oakes & Wells, 1998). Because welfare recipients disproportionately stem from low socioeconomic status families and minority backgrounds, they would face considerable challenges in pursuing postsecondary degrees, even in the absence of TANF regulations.

TANF's anti-higher education stance has the potential to affect both welfare recipients and their children. Research has shown that mother's educational attainment is strongly linked to children's educational achievement (Haveman et al., 1991; Sewell & Shah, 1968). The effects of this policy may reverberate for generations to come.

Welfare recipients are not all equally prepared for higher education and, as has been discussed for high school students (Rosenbaum & Person, 2003), should not be universally encouraged to attend. Barriers to employment such as mental health and substance abuse problems, low literacy, and learning disabilities are pervasive among the welfare population (Turner, Danziger & Seefeldt, 2004) and may place postsecondary education out of the immediate reach of a sizeable proportion of welfare recipients. Still, access to higher education can provide a second chance for some women on welfare, particularly the 44% of recipients who had their first child as a teenager (Wertheimer & Moore, 1998). Early parenthood has been shown to have negative effects on educational attainment for women (Ahituv & Tienda, 2004; Marini, 1984) and TANF provides another layer of discouragement for welfare recipients.

Research is suggestive of improved outcomes for welfare recipients who attend college while on aid, but few studies have addressed this issue specifically. No study has used longitudinal data to examine the extent to which simultaneous postsecondary education enrollment and welfare receipt results in improved outcomes over a longer time period. The simultaneity of these activities is important in the policy context as poor women are not restricted by federal law from attending college while they do not receive welfare. Studies have also not attempted to control for the inherent bias involved in examining the effects of college attendance for welfare recipients. One would expect that women who attend college while on aid possess some potentially unobservable characteristics, such as a high level of motivation or a familial expectation of higher education, that would lead them to achieve better outcomes than their non-college attending peers even in the absence of postsecondary enrollment. Overcoming this bias is important in quantifying the extent to which college attendance improves the outcomes of welfare recipients.

This paper adds to the literature on the effects of postsecondary education for welfare recipients by using 20 years of panel data from the National Longitudinal Survey of Youth (NLSY). I concentrate on the extent to which both college attendance and graduation are associated with improved outcomes--employment, welfare recidivism, and poverty--in one-year and five-year follow-up periods. To address the bias issue raised previously, I employ an instrumental variables approach, discussed fully in the theory section. Findings indicate that college attendance is most associated with improved employment outcomes, but graduation has an even larger effect on family poverty and return to aid five years post-welfare. However, the graduation rate among student welfare recipients is substantially lower than that for other students nationwide.

Background

In most states, even those that allow postsecondary education, TANF is essentially a "work first" program, aiming to move recipients into the workforce as quickly as is appropriate. The philosophy behind this approach is at odds with the long-standing human capital literature, which indicates that there are tremendous returns to schooling particularly at the postsecondary level (Grubb, 1993; Kane & Rouse, 1995; Leslie & Brinkman, 1988). For instance, Kane and Rouse (1995) show that having an Associates Degree leads to a 30% increase in annual earnings for women. The returns are even greater for a four-year degree, leading to a 51% increase in earnings for women. However, credits accumulated in postsecondary institutions, community colleges in particular, without a degree are not necessarily associated with improved labor market outcomes (Grubb, 1993). For welfare recipients specifically, having a higher level of education is associated with shorter welfare spells (Barrett, 2000; Blank, 1989), increased post-program employment and earnings (Michalopoulos & Schwartz, 2000), and better educational outcomes for children (Magnuson & McGroder, 2002). Although studies point to returns to education for welfare recipients, none specifically examines the effects of simultaneous welfare receipt and college enrollment. Rather, analyses of returns to education for welfare recipients generally rely on education levels measured at the start of one's time on welfare.

Although postsecondary education is not a key feature of the current federal welfare program, previous welfare programs have focused more specifically on human capital acquisition as a means to self-sufficiency. The Job Opportunities and Basic Skills (JOBS) program, initiated as part of the Family Support Act of 1988, encouraged welfare recipients to attend college while on aid. Through JOBS, all but three states had provisions in place for welfare recipients to attend college while they received cash assistance. In the mid-1990s, states began to experiment with alternative ways to serve welfare recipients and it became evident that shorter-term "work first" services--typically including assisted job search activities--also led to increased employment (Friedlander & Burt-less, 1995). However, the positive impacts associated with work first programs tend to decline over the long term, while the impacts associated with a more human capital approach tend to improve (Riccio, Friedlander, & Freedman, 1993). Opponents of postsecondary education for welfare recipients do not necessarily dispute the benefits of human capital approaches. Rather, they argue that allowing welfare recipients to attend college while on aid undermines the short-term focus of the TANF program (Friedman, 2001) and unfairly allows some low-income students to subsidize their college education with funds from cash assistance.

Most evaluations of welfare recipients' experiences in college in the TANF and pre-TANF eras report positive outcomes associated with college graduation, but these studies are largely descriptive. For instance, wages for graduates are generally higher than for non-graduates, and graduating from college is associated with higher rates of exit from welfare (Gittell, Gross, & Holdaway, 1993; Karier, 2000; Thompson, 1993). Psychosocial benefits from college attendance have also been reported. Butler and Deprez (2002) find that student TANF recipients in Maine report increased feelings of self-concept and independence, better job opportunities, and an increased ability to meet goals and set new ones as a result of their college attendance. Being enrolled in a college-based targeted support program for welfare recipients may increase students' probability of graduation (Gittell et al., 1993). These programs generally provide remedial assistance, counseling or case management, and camaraderie. Enrollment in a supportive program may also improve post-program outcomes, including earnings (Hollenbeck & Kimmel, 2002). Notably, welfare recipients who graduate from college cite financial aid as the primary form of assistance necessary to graduate (Gittell et al., 1993; Thompson, 1993).

Although findings from these studies provide suggestive evidence that attending college leads to improved outcomes for welfare recipients, one cannot draw causal conclusions from them. They have not dealt with the inherent biases associated with the selection into college attendance, and most do not control for other intervening factors. Careful attention to modeling and estimation is necessary in order to more precisely estimate the effects of college attendance and graduation on welfare recipients' outcomes.

Data

Data for this study come from the special geocode version of the 1979 National Longitudinal Survey of Youth (NLSY). The NLSY is a longitudinal data set that follows a nationally representative sample of nearly 13,000 young men and women from 1979 to 1998. Although the NLSY has been utilized extensively to study issues of welfare participation, it has been used less often by education researchers attempting to track postsecondary experiences. Still, these and other …