Journal Issue: Opportunity in America Volume 16 Number 2 Fall 2006
Do the Costs of Education Differ by Family Background?
Education has various costs, the most obvious of which is the direct cost. For the 90 percent of U.S. K–12 students who attend public school, these direct costs may be minimal, but parents must still pay for such school supplies as notebooks, pencils, paper, and the like.22 Based on our estimates using data from the 2002 Consumer Expenditure Survey, families with children under age eighteen who are headed by a high school dropout spend roughly $34 a year on school books and supplies, whereas families whose head has a graduate degree spend roughly $85. These differences, however, are likely too small to generate significant differences in educational attainment.
Education also has psychological costs, information costs, opportunity costs, and borrowing constraints (the cost of obtaining funds). At the elementary and secondary levels, it is these costs that are likely to be important in explaining differences in schooling caused by family background.
Differences in Psychological Costs
Learning can be frustrating, and mastering new material and studying for tests can be time-consuming. Anything that increases these psychological costs for disadvantaged students relative to their more privileged peers (that is, makes them dislike school more) may help explain why they get less schooling.
As one example, systematic differences in the expectations of parents and teachers may raise the psychological costs for less advantaged students. A child from a poorer family may face different expectations from parents and teachers than a child from a more privileged family, even if the two children have the same “ability.” If these different expectations, in turn, affect the children's academic achievement, then expectations could be one reason why parental socioeconomic status affects schooling.
Data from the NELS indicate that more advantaged parents expect their children to complete more education than less advantaged parents do, although virtually all parents, regardless of socioeconomic background, expect their children to complete high school. If lower parental expectations cause children to have less confidence in their own ability, the children could face higher psychological costs. Although we are not aware of evidence that parental expectations causally affect children's academic achievement, some evidence exists that teacher expectations affect both student intelligence and achievement.
Robert Rosenthal and Lenore Jacobson's Pygmalion in the Classroom has been widely cited as providing just such evidence.23 The authors administered a baseline intelligence test to elementary students in a single school and then randomly assigned 20 percent of the students to be identified as likely to show a dramatic increase in intelligence over the next school year because they were “late bloomers.” The remaining students served as the control group. Rosenthal and Jacobson then told the teachers which students had been identified as late bloomers and later administered follow-up intelligence tests. They found that one and two years after being labeled, the late-blooming children had gained more IQ points than the control group. Rosenthal and Jacobson's study has spawned many more studies and has been much criticized, but a recent review of the research by Lee Jussim and Kent Harber concludes that teacher expectations do affect student intelligence, though the effects are likely small.24
A recent study by economist David Figlio also finds that teacher expectations affect academic achievement.25 Starting with the assumption that teachers' perceptions of a child's family background may be based on the child's name, Figlio assigns socioeconomic status rankings to student names. Because siblings' names are often assigned different rankings, Figlio can look for differences in treatment and outcomes among students with identical family background. He finds that teachers are more likely to recommend students with high-status names to gifted and talented programs than students with similar test scores but low-status names. In addition, using standardized test scores, he finds that children with low-status names score lower in mathematics and reading than their siblings with higher-status names.
Findings from both economics and psychology suggest that teacher expectations may indeed help explain why family background affects student achievement. If teachers have lower expectations for children from disadvantaged families, regardless of their ability, and if their perceptions about which children are disadvantaged are on average correct, then the lower expectations for disadvantaged children may raise the psychological costs of education relative to their more privileged peers and thus help explain why children of disadvantaged parents attain less education.
Differences in social or cultural identity may also generate differences in the psychological costs of schooling. In other words, those who drop out may feel more peer or family pressure not to continue in school. Again, however, one might ask why these cultural or social norms about education vary systematically with socioeconomic status. Cultural norms may vary because education helps determine socioeconomic status, so that disadvantaged children may feel pressure not to raise their own status through education above the average for the social and cultural group with which they most identify.
Another potential cost to completing more schooling is that of acquiring accurate information about the costs and benefits of more schooling. If students from more privileged families can get more or better information about the ramifications of their decision at a lower cost than those from poorer families (for example, a better understanding of the potential benefits to continuing in school, perhaps because of better family social networks), then they may get more schooling. Similarly, students who drop out may believe that the returns to schooling are much riskier than do students who continue, thus lowering their expectations of the value of a high school diploma. High school dropouts may also discount the future income benefits of more education at a much higher rate than those who graduate from high school, also leading them to have lower expectations of the value of more education.
Although such arguments could explain why some students decide to drop out of high school in spite of the seemingly large economic benefits of continuing, one needs to ask why perceptions of risk or discount rates vary systematically with family background. Further, low-income students appear to understand the potential economic benefits of college attendance about as well as more advantaged students. 26 Although research is far from conclusive, it suggests that a simple asymmetry in students' understanding of the costs and benefits of schooling is unlikely to fully explain differences in educational attainment.27
Opportunity Costs and Borrowing Constraints
Because students cannot work during the hours when they are attending school, they forgo income to attend school. In some families that income is a nontrivial share of family income. If instead the family could borrow money to allow the child to continue in school, then the increase in earnings from getting, say, a high school diploma would allow the family to repay the loan (and then some), assuming that interest rates are lower than the return to schooling. If credit markets are perfect—that is, if all families can borrow as much money as they need at the prevailing interest rate—then educational attainment should not vary by family background. If, however, poor families lack access to competitive credit markets and would have to borrow money at much higher interest rates, then the cost of continuing in school is higher for them than for wealthier families who do not need to borrow the money (or who can borrow it at competitive rates). In this case, students from wealthier families will complete more schooling than those from poorer families.
Whether borrowing constraints more generally explain differences in educational attainment, especially college attendance, by family background is an unresolved issue.28 There is, however, growing evidence from outside education that individuals, particularly teenagers, are credit constrained.29 Further, racial and gender discrimination in credit markets has long been documented.30 For example, researchers at the Federal Reserve Bank of Boston investigating racial discrimination in mortgage lending in the Boston area in 1990 found that the loan rejection rates of African American and Hispanic applicants were 8 percentage points higher than those of otherwise similar white applicants.31 Although race is certainly correlated with socioeconomic status, we know of no direct evidence of discrimination by socioeconomic status.
Overall, the evidence suggests that differences in the cost of education may help explain differences in educational attainment by family background. As we will show, many of these cost differences are potentially driven by variation in school quality by family background, which may also lead to differences in the value of schooling.