Location. Location. Location.
That's the theory in real estate, and new research suggests that the same theory may apply to graduation rates, too. Attend a more selective institution and you are more likely to graduate. That may not seem shocking, if you assume that better students attend more competitive institutions. But the new study focuses on the impact of the "10 percent" admissions system in Texas and was done in a way that challenges the theory of "minority mismatch," in which some critics of affirmative action say that graduation rates for minority students would be better if they attended institutions they could enroll at without any special admissions system in place.
The key finding is that minority students in Texas are significantly more likely to graduate if they enroll at a competitive institution through the 10 percent plan than if they enroll at a less competitive, and theoretically easier, institution. In fact the only minority students who don't appear to benefit from 10 percent are those who are below the top decile of their high school classes and who might have previously won admission to a highly competitive institution, but now frequently lose their spots and end up at other institutions. These students see a decline in graduation rates.
The percent plan idea originated as a law in Texas to respond to court rulings against affirmative action, but has been used elsewhere with different cutoffs. In Texas, those in the top 10 percent of their high school classes are assured admission to the public university of their choice -- regardless of standardized test scores.
The idea behind the percentage plans is that black and Latino students, on average, don't do as well on standardized tests as do white and Asian students. In addition, Texas is a state with many high schools that are overwhelmingly Latino or overwhelmingly black. Since every high school has a top 10 percent, eliminating the testing requirement meant that these largely minority high schools were going to end up producing good numbers of Latino and black students who would be admitted -- without consideration of race in ways that might offend courts or critics of affirmative action -- to such competitive institutions as the University of Texas at Austin and Texas A&M at College Station.
In many respects, the plan has been a major success in Texas,  helping the flagship institutions to admit more minority students than they would have been able to otherwise -- at least while the state was under a court order not to use affirmative action. But ever since the U.S. Supreme Court ruled in 2003 that public colleges could consider race in admissions, University of Texas officials have been pushing to get rid of 10 percent and to instead rely on other admissions strategies (including affirmative action). In the 2007 legislative session, the university was expected to win its fight, but at the last minute, the 10 percent system survived. 
This year, UT officials are again asking for the admissions system to be changed, with William Powers, the president at Austin, telling the Texas Associated Press Managing Editors  last week that 81 percent of freshmen are now admitted through 10 percent, leaving the institution with too little control over whom to enroll. “We’ve lost control of our entering class because we don’t have any discretion on the admissions,” Powers said. In California, where those in the top 4 percent are assured University of California admission, a faculty panel is recommending that up to 9 percent  be admitted that way (although in a key difference from Texas, the California 9 percent plan would guarantee a spot somewhere in the university system, not on a particular campus).
With these debates going on, the new research may challenge several assumptions. The study was conducted by Kalena E. Cortes, an assistant professor of education at Syracuse University, and was presented last week at the annual meeting of the American Economic Association. Cortes used data from Texas on admission of students from various high school ranks to the state's more competitive and less competitive colleges, and then tracked six-year graduation rates.
Her findings go directly to a fear that some have had about the 10 percent plan and that others have about affirmative action generally -- namely that it could end up hurting the minority students it is supposed to benefit. According to this "minority mismatch"  idea, minority students who earn admission to competitive institutions (either through a percent plan or more traditional affirmative action) are likely to do less well than they would have if they had enrolled at less competitive institutions. Advocates for this position say that minority students would be more likely to graduate and excel if they ended up at institutions without any mismatch risk. The mismatch argument is popular with some and criticized by others because of its political potency: It allows people to criticize affirmative action not for its its impact on white students, but on minority students.
But Cortes found evidence to rebut this assumption.
She found that minority students who attended selective colleges are 38 percentage points more likely to complete college within six years of enrollment than are the minority students who enroll at other colleges. While she found that some of the gap is based on student characteristics and high school characteristics, excluding those elements still left a gap of 21 percentage points.
While similar data were found for non-minority students, Cortes found that the benefits that relate to attending the more elite colleges appear to be clear factors in spurring more minority (and white) students on to graduation. That leaves "no evidence," she writes, for the mismatch theory. "After adjusting for observable characteristics, there is still a remaining gain from attending a selective college for both minority and non-minority students."