When it comes to policies that help the needy attend college, simplicity rules, and financial aid programs seem to be more effective when they link money to academic performance and/or support services for students.

Those are the key findings of a paper, "Into College, Out of Poverty? Policies to Increase the Postsecondary Attainment of the Poor," published this week by the National Bureau of Economic Research. The paper, by written David Deming of Harvard University and Susan Dynarski of the University of Michigan, was supported by the Robin Hood Foundation as part of its forthcoming book, Targeting Investments in Children: Fighting Poverty When Resources are Limited (University of Chicago Press).

The scholars' paper does not present any new data; it instead reviews 17 studies by other scholars on a range of federal, state and philanthropic programs and policies that aim in one way or another to reduce what needy students must spend on college to try to increase their access to college -- and, increasingly, to keep them enrolled. The list of programs is long: Pell Grants, subsidized loans, tax incentives and veterans' benefits at the federal level (as well as the Upward Bound college prep program), tuition subsidies and need- and merit-based grant programs at the state level, and foundation programs like the Gates Millennium Scholars.

Deming and Dynarski examined the programs to try to gauge which are most effective, dollar for dollar, at increasing college going and persistence. They draw attention to some comparatively little known programs -- underscoring the "very encouraging" results in student retention, for instance, produced through MDRC's "Opening Doors" project, which provided small grants and significant services at community colleges in Louisiana, New York and Ohio.

But the authors' most intriguing conclusions are probably the ones they draw about the country's most highly visible programs: federal Pell Grants and state initiatives like Georgia's HOPE Scholarship program. While Pell Grants are narrowly targeted at the scholars' chosen goal -- drawing low-income students into higher education who would not otherwise have gone at all -- they fall short of their promise because they require so much paperwork (especially on the part of applicants who are often intimidated by the Free Application for Federal Student Aid). "If targeted students are deterred by administrative hurdles, these programs will not work as well as intended," the authors write.

Many advocates for low-income students have criticized Georgia's merit-based HOPE program (which has no income cutoff and goes to students with high high-school grade point averages) because its funds have gone disproportionately to students from higher up on the income scale. But the program's simplicity -- potential recipients know that they can get free tuition at any public college in the state if they get a 3.0 in high school -- makes it highly effective in getting low-income students into college, even if it means that large numbers of students who are not in the target population of needy students also benefit.

"In sum, the best evidence for effective financial aid on educational attainment comes from simple, broad-based programs," the authors write. If enough students from low-income families enter college because of HOPE and programs like it, they argue -- a threshold that they say previous studies show the programs clearly pass -- it appears that "a simple, broad-based aid program can increase social welfare."

The authors go out of their way to say that they do not mean to imply that Pell is not effective -- only that it is far less effective than it could be, and that the changes Congress and the Obama administration are considering to simplify both the process for applying for federal aid and the criteria the government uses to award the grants would strengthen it significantly. "We try to focus on what can be done pretty simply to improve the program," Deming said in an interview.

Donald E. Heller, director of Pennsylvania State University's Center for the Study of Higher Education and another financial aid researcher, said he did not read the study as suggesting that the Georgia program is better or more effective than Pell. But he did say he thought the paper's authors "missed the boat" by failing to point out that HOPE's effectiveness as a tool to help the poor get a higher education could be greatly enhanced if it added a financial need component.

"They say that the reason that HOPE has such a large impact is because it is very transparent," Heller said. "But I'm surprised they didn't say that if you did something that was as simple to understand, but was based on need, it would have an even bigger impact" from a social welfare standpoint.

"Just imagine if you took the $460 million a year Georgia spends and said something equivalent, like 'If you're from income below $x, you will go to college for free,' " Heller said. "You would get some kind of multiplier if the state put it in need-based aid instead."