From Inside Higher Ed:

One of the major justifications for state financial aid programs based on academic merit – grants that studies show go disproportionately to students from affluent families — is that they keep those students, who might otherwise go elsewhere for college, in the state during and after their undergraduate years.

But a paper published by the National Bureau of Economic Research by Maria D. Fitzpatrick and Damon Jones, public policy professors at Cornell University and the University of Chicago, respectively, calls that argument into question. Fitzpatrick and Jones looked at the migration patterns of students in states that adopted merit-aid programs and found that the scholarships do not have a sizable effect on migration patterns. The majority of the spending on these programs goes to students whose education and migration behavior is not altered.

“The magnitude of our results suggest that only a small fraction of the eligible population responds to merit aid by changing educational or migration decisions,” the authors of the report wrote. “We find that programs targeted to at least 30 percent of a cohort alter the behavior of at most 2 percent of a cohort at the margin.”

According to the paper, 28 states had some sort of merit aid program for resident students for the 2010-11 academic year, awarding a total of $3.9 billion. Since being introduced in the early 1980s, non-need-based aid has grown as a share of overall state aid, making up 29 percent of all state aid in 2010-2011, according to the College Board’s annual “Trends in Student Aid” report.

Several states — particularly those in the Southeast, such as Georgia, South Carolina, Arkansas, Mississippi and Tennessee — distributed the majority of their state aid based on criteria other than financial need, according to the College Board report.

Because they needed samples large enough to judge the effects of the policy, the authors only looked at programs where awards were given to at least 30 percent of a cohort. That left a list of 15 states where the authors could measure the effects of the programs.

The authors looked at whether the introduction of merit aid programs had any effect on college-going rates, completion rates, and the residency of students aged 24 to 32.

Previous research, most of which has focused on a small number of colleges, has found that states that offer such programs tend to see a slightly higher percentage of students attending in-state universities. Fitzpatrick and Jones found the opposite. Bachelor’s-degree completion actually dropped slightly in their data.

Fitzpatrick and Jones also wanted to see if the merit aid programs affected students’ decisions after college.

“If college provides location-specific human or social capital, students who are induced to remain in their home state for college may remain in the state after graduation,” the two wrote. “On the other hand, if the skills acquired during college are relatively portable geographically, states with merit aid policies may find themselves paying to train the future workforces of other states.”

Fitzpatrick and Jones found that the percentage of students whose behavior was affected by the merit aid program was an order of magnitude smaller than the percentage of students eligible for the program. So if 30 percent of students in a state were eligible for the program, the percentage of students who under different circumstances would have relocated but instead remained in the state was less than 3 percent.

The report’s authors note that a better understanding of how the marginal retention affects the overall workforce and economy is needed to know whether the policies truly paid off. “More work is needed to determine whether the increased retention of high skilled workers leads to increased economic growth for the state,” Fitzpatrick and Jones wrote. “Even if it does, further research will also be required to determine how distorting migration between states effects the economic growth of the country as a whole.”