It is painfully well known that in America, black students lag behind white students in essentially all measures of academic achievement. Especially distressing is that many black students who might otherwise succeed are discouraged from doing so by peers who accuse them of “acting white” in their pursuit of education. David Austen-Smith (Kellogg School’s Department of Managerial Economics & Decision Sciences) and his colleague Roland Fryer Jr. (Harvard University) show that rather than reflecting some inherent dysfunction in individuals or communities, this phenomenon could result from perfectly rational individuals responding in the best way possible to their social and economic realities.
Just how crippling is this achievement gap? The average black 17-year-old reads at the level of a typical white 13-year-old, according to the 2004 National Assessment of Educational Progress (Perie, Moran and Lutkus 2005). On every section of the Scholastic Aptitude Test (SAT) college entrance exam, black students graduating from high school in 2006 received average scores roughly 18 percent—about 100 points, nearly a full standard deviation—below white students’ scores (College Board 2006). Moreover, these achievement differences seem to be amplified in the most highly segregated areas.
Seeking explanations, many scholars have traced the roots of this gap back to the stigma of “Uncle Tom” among plantation slaves, describing “acting white” as a unique and powerful pressure that imposes unbearable social costs on blacks. Those who pursue behaviors associated with upward mobility, such as education, certain types of speech, and deference to authority, run the risk of alienating peers and losing valuable community support.
“I was a member of the MacArthur Foundation’s Social Interactions and Inequality Network, so I was thinking about this kind of question,” said Austen-Smith, the Earl Dean Howard Distinguished Professor of Political Economy. “Roland [Fryer] had an essay on social capital and another on ‘acting white.’ We combined our resources to ask, ‘What’s observable about ‘acting white?’ How does it affect education investment? How does that translate into broad patterns of wages?’”
The paper by Austen-Smith and Fryer in The Quarterly Journal of Economics provides evidence that in an interconnected system of individuals, employers, and peer groups, with each pursuing their own best interests, some individuals can receive social benefits large enough to outweigh benefits they might otherwise receive via education and wages. In a purely rational sense, these individuals prefer peer acceptance to the benefits of education.
“Not to discredit all the work that other fields have contributed to understanding this phenomenon, but our work doesn’t depend on empirically contingent issues like culture, or on invoking some psychological phenomenon,” said Austen-Smith. “It’s grounded on rational choice theory, a set of assumptions that have had great success in explaining human interaction. We treat this like more mundane questions, premised on individuals acting on their own best interests.”
Austen-Smith’s work is built upon the concept of economic signaling: Unequal access to information can impair the exchange of goods and services. This problem can be addressed if one party sends a signal to convey useful information to a second party. The second party can then adjust its behavior, such as by offering goods at a different price than if the signal had not been received.
Starting in the early 1970s, Michael Spence showed how education functions as an economic signal, work for which he would receive the Nobel Prize in 2001. Employers, he argued, will pay more for good workers, but they lack the information needed to tell in advance which workers are good and which ones are bad. Bad workers do not mind this confusion because they might earn the same pay as good workers without actually having to be good at their jobs. But good workers know that they deserve better pay than bad workers, so they send a signal to employers by investing in education. Even if investment in education does not actually lead to improved productivity, Spence explained, talented, motivated workers will still invest time and energy on education simply to signal their value to employers.
Since they are inherently less talented and less motivated, bad workers have to spend more time and effort than good workers to get educated. In this sense, education is “cheaper” for good workers, Spence reasoned. As a result, bad workers typically accept lower pay rather than pursue education, which is a relatively expensive investment for them. Recognizing that a worker’s signal, her educational choice, predicts her quality, employers offer better wages to workers who invest more in education. Over time, workers and employers pursue their most profitable education and wage strategies until neither party can gain any further advantage from changing strategy unless the other party also changes. This stalemated, mutually beneficial condition is one that Spence called a “signaling equilibrium.”
The key innovation introduced by Austen-Smith and Fryer is their incorporation of a second audience into this signaling scenario. This two-audience signaling model acknowledges that a person’s education can influence not only employers’ wage offers but also peer groups’ offers of valuable social support. Moreover, the response to a single educational signal can differ dramatically between employers—who have reason to value education—and social groups, who may be wary of more educated individuals.
Their model hinges upon calculating the payoffs that individuals enjoy as a result of their inherent abilities as workers, their investments in education, and their attractiveness to social peers. Employers increase the wages they offer as individuals increase their education without regard for individuals’ social types. Alternately, peer groups accept individuals based on social features without regard for their inherent abilities as workers.
“One of the features of the model is that it separates factors that appeal to the wage market from those that appeal to the peer group,” said Austen-Smith. “It’s quite possible to have highly successful people who are also highly popular.”
Such popularity and success is possible, but is it probable? A key feature of this model is that while employers can incrementally increase or decrease wage offers according to slight variations in individuals’ educational levels, social groups make a much more “black and white,” “in or out” distinction, either fully accepting or fully rejecting an individual, with no middle ground. Further, while the group’s decision to accept or reject is ideally based on an individual’s social compatibility with the group, the group’s estimate of social compatibility can be strongly influenced by the extent of the individual’s education. Even if an individual is welcomed into the social group, every increase in time and effort spent on education necessarily diminishes the time and effort remaining to spend with peers, thus limiting the benefit gained from group membership.
So a socially-accepted individual seeking a slightly higher wage could invest in slightly more education, but this minor increase could completely undermine his social acceptability and benefit. Depending on the individual and the level of education, the resulting social cost due to exclusion from the group might be surpassed by increases in wages. Alternatively, the individual’s increased pay may pale in comparison to the cost of losing social support. The unfolding of such processes toward equilibrium conditions leads to familiar patterns of educational achievement suggestive of the “acting white” phenomenon.
Interesting equilibrium conditions emerge from this model when individuals reveal their educational choices to employers and peers. Individuals who are naturally unappealing to peer groups invest in education without regard to social acceptance. Those who have natural talent and motivation invest more in education. Regardless of social attractiveness to peer groups, all individuals who exceed certain levels of inherent skill and ability as workers disregard peer pressures and seek higher wages through investment in education.
“Talented people will realize the high cost of turning their back on rewarding employment and will choose to leave the group,” said Austen-Smith. In describing those with low inherent abilities who are least attractive to employers, he added, “People who aren’t very economically talented are not going to invest in education anyway. So they don’t have much worry about peer pressure.”
But, Austen-Smith said, “People in the middle are most interesting. They could have success in the world, but will have a high price of leaving the social group. That very same act of getting educated is a signal to the peer group ‘I’m not like you.’” Indeed, it is in this middle range of economic ability where individuals pool on a level of education. This means that individuals whose modest natural abilities might lead them to pursue higher pay by investing more in their education, instead choose to under-invest in education to appear educationally similar, and thus, they hope, socially similar as well to peers who have lower inherent ability but who are welcomed into the peer group.
Tweaking the model in search of potential policy solutions, Austen-Smith and Fryer showed that if economic incentives such as wages are improved, this pooling region shifts, pulling some people away from the group and toward education. But individuals embedded more deeply in the group are then mired even further.
“Though they’re not explicitly linked in the model,” said Austen-Smith, “Economic and social factors influence one another, so policies can’t expect to succeed without addressing both.” He and Fryer cite several instances in which coordinated combinations of education and social intervention have been shown to be more successful than similar programs that address only one of those factors in isolation.
Austen-Smith acknowledged that this model may strike some as oversimplified. “Some might contest the assumptions that firms don’t care whether or not you’re a member of a peer group.” He also recognized that some social signals might not truly reflect an individual’s educational intent, imagining “stealth performers, who slack off at school and then go home at night and work very hard indeed.”
While acknowledging that the patterns reflected by their model have been demonstrated elsewhere using other methods, Austen-Smith and Fryer note that the strength of their approach is in its potential to strip this issue of racially and ethnically charged scapegoating and finger pointing. They demonstrate that any group of individuals, regardless of race, ethnicity, culture, or history could arrive at similar patterns of education simply through rational, self-interested participation in basic economic signaling.
College Board (2006). 2006 College-Bound Seniors: Total Group Profile Report. (accessed April 15, 2007).
Perie, M., R. Moran, and A.D. Lutkus (2005). “NAEP 2004 Trends in Academic Progress: Three Decades of Student Performance in Reading and Mathematics (NCES 2005–464).” U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics. Washington, DC: Government Printing Office (Accessed 15 April 2007).
Spence, Michael (1973). “Job Market Signaling.” Quarterly Journal of Economics, 87(3): 355-374.
Stair Steps to Heavenly Savings
How to design sales force incentive programs to unlock operational cost savings