In 2008, about 14 million children lived at or below the official poverty line in the United States (U.S. Census Bureau, 2008b). Numerous population-level studies have shown that poverty predicts negative developmental outcomes for children (Aber, Bennett, Conley, & Li, 1997; Brooks-Gunn & Duncan, 1997; Duncan, Yeung, Brooks-Gunn, & Smith, 1998; Guo & Harris, 2000; McLoyd, 1990). It is troubling that childhood poverty might not be a transitory experience. R.esearch findings have suggested that children who grow up in poverty are more likely to be poor in adulthood (Corcoran & Adams, 1997; Harper, Marcus, & Moore, 2003; McKay & Lawson, 2003; Rodgers, 1995). However, a few studies have documented that the intergenerational cycle of poverty is not certain; some children raised in poor families escape poverty in adulthood (Corcoran, 2001; Rodgers, 1995). For example, using a cohort of the Panel Study of Income Dynamics, Corcoran (2001) found that around 24% of children raised in poor families were poor in young adulthood, whereas the remaining 76% were not poor when they entered adulthood. Despite the finding of fairly high rates of discontinuity in economic status across generations, far less is understood about the mechanisms by which children from low-income families manage to escape poverty (Breen & Jonsson, 2005).
Studies have shown that education predicts continuity and discontinuity of low-income status across generations (Blau & Duncan, 1967; Bourdieu & Passeron, 1990; Downey, Paul, & Broh, 2004; Giroux, 1983; Grubb & Lazerson, 2004; Haller & Portes, 1973; Nieto, 2005; Sewell & Hauser, 1972). Blau and Duncan (1967) investigated the extent to which men's adult occupational status depended on their family background. They concluded that a child's educational attainment is an important determinant of that person's adult economic status, even after controlling for the father's occupational status. The positive association between educational attainment and adult economic status has been found consistently (Chen & Kaplan, 2003; Haveman & Smeeding, 2006; Kao & Thompson, 2003; Porter, 2002; Sewell, 1971; Wilson, 2001), which indicates that education promotes an escape from poverty across generations. High school graduates' mean earnings are $9,802 higher than those of high school dropouts (U.S. Census Bureau, 2008a), and high school graduates are more likely to be employed compared with high school dropouts (Caspi, Wright, Moffit, & Silva, 1998; Goldschmidt & Wang, 1999). Oxford, Lee, and Lohr (2010) reported that the odds of economic hardship in adulthood, defined as having an income-to-needs ratio below 1:85, were almost two times higher for those who did not complete their secondary education by age 19.
Yet for a large proportion of children from low-income families, public K-12 education fails to serve as a route out of poverty (Bowles & Gintis, 1976; Counts, 1932; Entwisle, Alexander, & Olson, 2003; Grubb & Lazerson, 2004; Lareau, 2000; Nieto, 2005; Oakes, 2005). The disparity in educational outcomes by socioeconomic status is well documented. The percentage of those who underachieve in educational outcomes remains much higher for children from low-income families compared with those who are not from low-income families (Heam, 1991; Morgan, 1996; Rouse & Barrow, 2006; Sewell & Hauser, 1972). Low-income status is negatively associated with high school completion (Haveman, Wolfe, & Spaulding, 1991; Laird, Lew, DeBell, & Chapman, 2006; Newcomb et al., 2002), a minimum requirement for many jobs (Rumberger, 1987). In 2006, 16.5% of children from families in the lowest income quartile dropped out of high school compared with 3.8% of those from families in the highest income quartile (U.S. Department of Education National Center for Education Statistics, 2007). Given the well-documented connection between educational attainment and adult economic status, it is likely that the disparity in educational attainment by socioeconomic status translates into continuity in economic adversity from one generation to the next.
However, educational experiences might not necessarily dictate continuity in poverty across generations. For example, Benner and Mistry (2007) demonstrated that some low-income children avoid academic failure despite economic adversity. Similarly, Downey et al. (2004) demonstrated that the gap in cognitive skill gains by socioeconomic status decreased when schools were in session compared with when they were not and argued that educational experiences have the potential to serve as an economic equalizer. Thus, there is some evidence that education might also serve to predict discontinuity in poverty across generations, at least for some children. This leads to the question of what factors predict such differences in educational experiences and adult income among low-income children.
Previous studies on educational outcomes have suggested that intrapersonal achievement-related characteristics are positively associated with educational outcomes (Akerlof & Kranton, 2002; Benner & Mistry, 2007; Eccles & Wigfield, 2002; Frome & Eccles, 1998). Specifically, it has been suggested that educational aspirations and expectations are important predictors of educational and occupational status in adulthood (Campbell, 1983; Haller & Portes, 1973; Kao & Thompson, 2003; MacLeod, 1995; Sewell, Haller, & Portes, 1969; Sewell & Hauser, 1972). Beal and Crockett (2010) reported that adolescent educational expectations were positively associated with educational attainment in young adulthood. Sewell and Hauser (1972) demonstrated that educational aspirations of high school seniors predicted their educational attainment at the age of 25, and educational attainment, in turn, positively predicted earnings at age 28. Feliciano and Rumbaut (2005) reported that educational expectations in high school increased expected occupational status at the age of 30.
Several theorists have suggested that educational aspirations and expectations change over the course of adolescence (Eccles, Barber, Stone, & Hunt, 2003; Gottfredson, 1981). Beal and Crockett (2010) reported that more than one-third of the adolescents in their study (N = 317) shifted to higher or lower educational expectation categories. Similarly, Cooper (2009) documented that educational aspirations fluctuated between 10th and 12th grade: In that sample of African American male students who aspired to get a bachelor's degree at 10th grade, 25.1% increased their educational aspirations (a graduate degree), whereas 24.2% decreased their educational aspirations (less than a bachelor's degree). Similarly, Fredricks and Eccles (2002) showed that children's belief in math competence declined from childhood to adolescence. However, to our knowledge, no studies have linked changes in educational aspirations and expectations over the course of adolescence with long-term academic outcomes and adult economic status, especially among children from low-income families. In extant studies, educational expectation has been modeled as a single point in time or averaged over study points.
Findings based on such analyses provide limited guidance toward the formation of intervention and policy (Feinstein & Peck, 2008). It is important to understand the longitudinal development of educational aspirations and expectations as they relate to achievement so that the timing of preventive interventions can be pinpointed. Thus, rather than using a single time point or cross-year average, we used a strategy (growth mixture modeling [GMM]) that identified groups of children following different trajectories of educational aspirations and expectations.
In the present study, we sought to understand how children from low-income families escape poverty in adulthood by analyzing a contemporary sample from a longitudinal panel study that followed participants from age 10 to 30. We hypothesized that trajectories of children's educational aspirations and expectations would predict educational attainment and economic status in adulthood.
Sample The data set we used is from the Seattle Social Developmental Project (SSDP). The SSDP is a theory-driven panel study that has collected extensive longitudinal information on a sample of 808 participants who attended schools serving disadvantaged neighborhoods in Seattle, Washington. The SSDP followed these participants during the years 1985 to 2005, from the time they were age 10 to age 30, and conducted periodic surveys regarding their experiences in family, school, and community to examine participants' social, psychological, and economic well-being. The longitudinal nature of these data makes it possible to investigate the predictive power of children's educational aspirations and expectations over an extended period and to examine their long-tern1 connection to adult economic status. The demographic data indicate that the sample is gender balanced and ethnically diverse. Of the 808 participants, 49% were female and 51% were male; 47% were European American, 26% were African American, 22% were Asian American, and 5% were Native American or other ethnic group identity. Over 52% of the participants were from economically disadvantaged families as indicated by eligibility for the National School Lunch/
School Breakfast Program across the fifth and seventh grades. SSDP sample retention rates were consistently high: Over 93% of the original participants were interviewed at each of the last five interview waves. Data collection procedures were approved by the Human Subjects Review Committee of the University of Washington, Seattle. Further details about the SSDP study and its underlying theory appear in several published articles (Catalano & Hawkins, 1996; Catalano, Kosterman, Hawkins, Newcomb, & Abbott, 1996; Hawkins, Catalano, Kostennan, Abbott, & Hill, 1999).
Childhood low-income status was derived from school archival data and defined by...