As President Obama advocates for a dramatic expansion of publicly funded preschool, there is growing debate about whether preschool pays off academically for kids.

Most controversial is the so-called “fade-out” effect that has surfaced in research, showing that students who attend preschool begin kindergarten more prepared than control groups of students who did not, but they lose their edge on reading or math tests within the first few years as other children catch up.

The large-scale National Head Start Impact studies, for example, showed no reliable difference in how students performed by the end of third grade on multiple measures of development, regardless of whether they attended the federally funded pre-kindergarten program.

“Everyone is concerned about this phenomenon,” said Deborah A. Phillips, a psychology professor at Georgetown University, at a panel discussion about early education research hosted Tuesday by the Cato Institute, a libertarian think tank.

Phillips was a co-author of a policy brief published in the fall that reviewed dozens of studies over 50 years. The brief found consistent short-term returns for preschool participants. But they also found consistent fade-outs of the short-term gains.

“There’s a very healthy scientific debate about what is going on,” Phillips said.

Among the questions she raised about these studies: Do preschool impacts not last because preschool programs don't work? Are elementary and middle schools not sustaining the gains somehow? Are the wrong things being measured? Or are there long-term advantages, not visible now, that may be come evident?

A handful of small-scale programs, including the landmark Perry Preschool and Abecedarian studies, found positive long-term differences in how preschool-educated students fared, including how much school they attended, how much money they made, how likely they were to smoke cigarettes or commit crimes or get pregnant as teenagers.

The mixed findings have inspired some to say that a broad expansion of preschool is unmerited.

Grover J. “Russ” Whitehurst, director of the Brown Center on Education Policy at the Brookings Institution, said it makes more sense to target investments more closely to the most disadvantaged students and those learning English as a second language who, studies show, tend to benefit the most.

“I’m in favor of government investment in services for families” in need, he said at the panel. “But how are we going to make an investment that really works rather than an investment that makes us feel good?”