Exit exams don't cause increase in dropout rate
May 2, 2004
by Jay P. Greene & Marcus A. Winters
When Arkansas legislators passed an education accountability law earlier this year, they balked at requiring students to pass a test before receiving their high school diploma.
Opponents of that provision feared that this so-called "exit exam" would push the state's already low graduation rate even lower. Eventually the Legislature will reconsider this difficult question, and it will surely be faced with the same outcry as before. Next time, Arkansas should look at the facts. Recent evidence suggests that exit exams don't actually increase dropout rates.
A new study by the Manhattan Institute finds that adopting a high school exit exam has no effect on graduation rates. The study measured graduation rates in each state over the last decade and evaluated how they were affected by implementation of an exit exam. The results of the analysis showed no relationship between exit exam requirements and graduation rates.
This study is only the most recent of several analyses showing that exit exams do not lead to higher dropout rates. Most prominently, researchers at Stanford University recently found that implementing exit exams has no effect on student retention in high school. The evidence indicates that the popular idea that exit exams push students out of high school is mostly a myth.
Certainly it seems counterintuitive that exit exams would not affect graduation rates. One could point to the numerous media reports every year profiling students who would graduate if only they could pass their states' exit exams. But Arkansas' policymakers should focus on the total body of available evidence, not intuitions or anecdotal reports.
First, many of the students who don't pass exit exams would have failed to graduate anyway. In Florida, home of one of the nation's most difficult exit exams, state officials estimated that about 40 percent of the seniors in the class of 2003 who could not pass the state's exit exam had also not completed the necessary coursework to receive a diploma.
Furthermore, the number of students in Arkansas who truly would not be able to pass an exit exam is probably quite small. An analysis by the Fordham Foundation found that exit exams actually require surprisingly low levels of proficiency. Students are routinely given second, third and even seventh chances to pass the exams before they are finally denied a diploma. Between each administration of the test, students who have failed are provided with extra help specifically designed to get them past the test requirement. Given so many tries, eventually most students would be able to pass the tests, even if only by chance.
Still, there are at least some students who would not pass the tests, so why would we expect there to be no decrease in Arkansas' graduation rate? An exit exam would force schools to focus their time and resources on low-achieving students they previously ignored. This improved use of resources would cause some students to earn their diplomas who otherwise would have dropped out. Since research consistently finds no relationship between exit exams and graduation rates, the number of students positively affected by these exams appears to be roughly equal to the number of students who fail to graduate because they cannot pass.
Arkansas might support an exit exam even if it did lower graduation rates, in order to protect the value of its high school diplomas. Requiring students to demonstrate proficiency in order to graduate should protect the quality of their diplomas in the labor market. Fortunately, the evidence suggests that Arkansas doesn't have to choose between awarding more diplomas and awarding higher quality diplomas. Research shows that implementing an exit exam would allow the state to give higher quality diplomas to the same number of students as before. Thus, implementing an exit exam would allow Arkansas to have its cake and eat it, too.
Jay P. Greene is a senior fellow and Marcus A. Winters is a research associate at the Manhattan Institute's Education Research Office, which can be found at http://www.miedresearchoffice.org.
©2004 Arkansas Democrat-Gazette