Monday, January 6, 2003 Merit-Aid Programs Like Georgia's HOPE Scholarships Can Distort Students' Incentives, Scholars Say By DAVID GLENN Large-scale merit-aid programs like Georgia's HOPE scholarships can affect students' academic behavior in unexpected and perhaps undesirable ways, scholars said Sunday during a panel discussion at the annual meeting of the American Economic Association. In-state students at the University of Georgia -- anxious about maintaining the grade-point average required by the HOPE program -- are now significantly less likely than their pre-HOPE predecessors to take full academic loads during their first year, according to Christopher M. Cornwell and David B. Mustard, economics professors at the university. Roughly two-thirds of in-state students took full course loads during the years just before HOPE was introduced, but that rate fell sharply after the program began, they said. In a paper written with Kyung Hee Lee, a doctoral candidate in economics at Georgia, they also found that course-withdrawal rates among first-year students have increased by 4.7 percentage points since HOPE began. The HOPE program, introduced in 1993 and financed through lottery-ticket sales, pays for full tuition and fees at any public two-year or four-year institution in Georgia, or provides roughly equivalent sums of money to students who attend private institutions in Georgia. To receive a HOPE scholarship, a student must be a Georgia resident and maintain a grade-point average of at least 3.0 in both high school and college. Since 1996, there have been no income restrictions on the program; even students from wealthy families are eligible. There are also no time limits, so students are under no particular pressure to complete college in four years. Fourteen other states have adopted merit-aid programs modeled after Georgia's, but most of the newer programs include time limits. Mr. Cornwell and Mr. Mustard, a full professor and assistant professor, respectively, emphasized that their findings probably underestimate the changes in students' behavior. Their data set ends in 1997, just after the income restrictions were lifted and the HOPE program became nearly universal among in-state students at Georgia. In the final year covered by their study, 59 percent of first-year, out-of-state students, who are not eligible for HOPE, took full course loads; only 44 percent of first-year, in-state students did so. No such difference existed before the HOPE program began. A second paper presented at the panel, by Melissa Binder and Philip T. Ganderton of the University of New Mexico, found similar patterns in their own state's merit-aid program, which is known as New Mexico Success. Students at the University of New Mexico have registered for and completed fewer credit hours per semester in response to that program's incentives. These scholars also found that the program has created "worse student-institution matches." Attrition rates have risen at New Mexico, suggesting that the program has attracted students who might have been better off at two-year institutions. "The program has disproportionately attracted higher-income, lower-ability students to the University of New Mexico," they wrote. Lower course loads may or may not injure students' education, Mr. Mustard said in an interview, but they certainly pose challenges for the university. The state government allocates funds to the University of Georgia according to the aggregate number of credit hours generated there each year. The per-student rate of annual credit hours at Georgia has declined sharply during the last decade, according to Mr. Mustard, forcing administrators to scramble for economies. "I think this is a good example of government regulation with unintended consequences," Mr. Mustard said. He said that it is important for scholars to examine other hypothesized changes in student behavior. Are students maintaining their grades by choosing relatively easy courses and majors? Are they petitioning more frequently for changes in their grades? The patterns revealed in the paper are concentrated among students with grade-point averages between 2.7 and 3.3 -- that is, among students on the verge of gaining or losing a HOPE scholarship. Students with averages above 3.5 have not tended to change their course loads. Two other papers presented at the panel explored whether the HOPE program has induced Georgia colleges and universities to raise their tuition and fees more quickly than comparable institutions in other states have. The HOPE program provides a natural experiment for assessing the "Bennett hypothesis," the 1987 claim by William J. Bennett, the former U.S. secretary of education, that "increases in financial aid in recent years have enabled colleges and universities blithely to raise their tuitions." The first of these studies -- by Benjamin Scafidi, Ross Rubenstein, and Gary T. Henry of Georgia State University, and Amy Ellen Schwartz of New York University -- found no evidence that nonselective Georgia institutions, either public or private, had raised their tuition and fees at unusually high rates. They did find evidence, however, that selective private universities in Georgia had responded to HOPE by increasing tuition faster than their peer institutions. This is not necessarily an illegitimate choice, Ms. Schwartz said. Colleges "might try to siphon off some of this money in order to improve the quality of the education they offer," she said, rather than simply expanding their student bodies. In the second paper, Bridget Terry Long, an assistant professor of education at Harvard University, likewise found that selective private institutions in Georgia -- particularly those with high numbers of HOPE-scholarship recipients -- had raised their tuition at unusually fast rates in response to the program. Ms. Long also discovered that HOPE reduced public Georgia universities' rate of increased spending on instructional services by 5 percentage points. It may be, she said, "that now that public colleges have this great [price] advantage in attracting students, they have less incentive to invest in quality." Scholars have debated the effects of HOPE and its progeny for several years. A discussant at the panel, John H. Bishop, of Cornell University, strongly defended such programs, saying that HOPE has produced a 7- to 8-percentage-point gain in the rate of Georgia high-school graduates who go on to college. Mr. Bishop, who has studied Michigan's merit-aid program, also argued that the Georgia program "was well-targeted to get kids to work harder in high school," because a 3.0 high-school average is required to begin college with a HOPE scholarship. But other observers worry about what they see as the programs' regressive nature, since most of their beneficiaries are from well-off families. When he arrived at the University of Georgia, in 1997, Mr. Mustard said, he took several students to lunch and asked them how they happened to choose to go to college there. According to Mr. Mustard, the first student answered, "Well, I was bribed. It was between Georgia and Duke or Vanderbilt or something, and my parents said, if you choose Georgia, you'll have the HOPE scholarship, we'll buy you a brand-new SUV." "And then I went around the table that day," Mr. Mustard said, "and all five or six students had basically the same story." In 2002, he and Mr. Cornwell wrote that, relative to comparable states, Georgia's total number of car registrations had risen significantly after the HOPE program began. In a preliminary draft paper describing their research, they wrote that "a doubling of a county's HOPE scholars would, on average, lead to a 2 percent rise in the number of registered cars." The draft report, titled "Merit-Based College Scholarships and Car Sales," is available on a Web site maintained by the professors. "Is this a good use of public resources?" asked Ms. Binder at the panel. "It seems like a textbook example of a policy that makes economists go crazy. It doesn't make economic sense to give someone money to do something that they're going to do anyway." _________________________________________________________________ You may visit The Chronicle as follows: http://chronicle.com _________________________________________________________________ Copyright 2003 by The Chronicle of Higher Education