The Chronicle of Higher Education, July 3, 2006, Monday

Chronicle of Higher Education
Monday, July 3, 2006

High-School Exit Exams Are Associated With Higher Dropout Rates, Researchers Find
By DAVID GLENN
http://chronicle.com/daily/2006/07/2006070301n.htm

Since 1979, a growing number of states have required high-school students to pass exit examinations before they can receive diplomas. For nearly as long, scholars and policy makers have debated whether such exams do more harm than good.

Proponents of exit exams say that they improve learning and future employment by giving both students and school districts better incentives to succeed. Skeptics say that the exams needlessly prevent decent students -- who have otherwise completed all of their course work -- from receiving diplomas. They also warn that the exams could prompt some students to drop out of high school as early as the 10th or 11th grade, if they foresee that they will fail the tests.
The latest battleground over the issue is California, where on July 25 an appeals court will consider a lawsuit that claims the introduction of the state's new exit exam should be delayed because certain low-income districts allegedly do not teach much of the material on the exam.
Now two teams of scholars have written papers whose findings might provide support to those on the more-harm-than-good side. In a recent working paper, Thomas S. Dee, an associate professor of economics at Swarthmore College, and Brian A. Jacob, an assistant professor of public policy at Harvard University, report that students in states with relatively easy exit exams are roughly 4 percent more likely to drop out of high school than similar students in states with no exams. In states with relatively difficult exit exams, students are 5.5 percent more likely to drop out than their counterparts in states with no exams.

The effects are stronger among African-American men, Mr. Dee and Mr. Jacob found. In states with easy exit exams, black male students are 5.2 percent more likely to drop out of high school than their counterparts in states with no exit exams. In states with more-rigorous exit exams, they are 7.3 percent more likely to drop out than are their counterparts in states with no exit exams. (On the other hand, Mr. Dee and Mr. Jacob found strongly positive results for native-born Hispanic women, who are significantly more likely to complete high school and to enter college if they live in states with exit exams.)

"Our experience with this has been that it exacerbates achievement gaps," Mr. Dee said in an interview last week. "The more stringent exams seem to have more-serious effects in terms of reducing educational attainment."

Mr. Jacob added, however, that the jury is still out on whether exit exams have, over all, a positive effect on students' learning or on their ability to find jobs. (In their study, Mr. Dee and Mr. Jacob found tentative evidence that African-Americans' post-high-school wages are higher in states with exit exams, and that white workers' wages are lower. They emphasize, however, that those patterns might apply only to workers whose wages are very close to the statewide average, and more study remains to be done.) "It's possible that these policies are having beneficial effects that we just haven't been able to detect," Mr. Jacob said.

Mr. Dee and Mr. Jacob used data from the "long form" of the 2000 Census, which allowed them to work with an extensive amount of data. They looked at the experiences of nearly three million people who turned 18 between 1980 and 1998. The census data also allowed them to look at relatively recent high-school graduates; certain other recent studies of exit exams have been criticized because they rely on the National Education Longitudinal Study, which looked at students who were scheduled to graduate from high school in 1992 -- a long time ago in terms of evaluating policy.

The second new paper, which appears in the summer issue of the journal Educational Evaluation and Policy Analysis, reports that rates of high-school completion are lower in states with exit exams than in states without such exams. In what may be a consequence, states with exit exams have higher rates of General Educational Development test-taking.

"If exit exams are having an upside -- if, on average, kids are learning more or earning higher wages -- then they might be worthwhile despite causing dropouts," said the paper's lead author, John Robert Warren, an associate professor of sociology at the University of Minnesota-Twin Cities. (Mr. Warren wrote the paper with Krista N. Jenkins, a graduate student in sociology at the University of Pennsylvania, and Rachael B. Kulick, a graduate student in sociology at Minnesota.)

By contrast, Mr. Warren continued, "if the upshot of this policy is only to deny diplomas to some kids, and there's no real upside for anyone else, then we should realize that this is purely a punitive policy." Mr. Warren is now studying exit exams' effects on learning and future employment, and at this point he is skeptical that many strong positive effects exist.

Not everyone agrees. "How bad is it if there's a 1½-percent reduction in the rate of people who get a regular high-school diploma?" asked John H. Bishop, an associate professor of human resource studies at Cornell University, in an interview on Friday. Mr. Bishop said that most students who fail to pass exit exams have such poor skills that they are not likely to do well in the labor market, with or without a diploma. He argued that the benefits of such exams outweigh the costs borne by students who do not win diplomas.

"What counts is, Do these policies result in more people learning more?" Mr. Bishop said. "In the long run, it's knowing stuff, not having a high-school diploma, that will help you in the labor market." In a 2005 study, Mr. Bishop found that, over a period of roughly a decade, states that began to use exit exams raised their eighth-grade mathematics scores significantly, whereas states that cruised along with no exit exams did not see such gains. That finding suggests, Mr. Bishop said, that introducing an exit exam can have far-reaching consequences on a school system's effectiveness.