Focusing on young children is one of the few things Donald Trump and Hillary Clinton seem to agree on—after all, who wouldn’t want to help little kids? But the political palatability of the issue could also be one reason early-education programs that may not actually help young children succeed are allowed to flourish.

New York City recently adopted a universal pre-k model, following the implementation of similar systems in places like Georgia and Oklahoma. The Georgia program, rolled out in 1995, was the first in the nation to provide all 4-year-olds with access to pre-k and is funded by the state’s lottery. Deeply conservative Oklahoma is a surprising site for such large-scale social program, but researchers say they have shown positive, long-term effects for the state’s young participants, particularly when it comes to mitigating achievement gaps in later years.

But in a new report published by Dale Farran and Mark Lipsey in Behavioral Science and Policy Journal, the researchers criticize the rapid adoption of widespread, publicly funded preschool programs. They assert that states are failing to adequately research early-education practices prior to implementing them, favoring the adoption of the feel-good issue without stopping to consider what data show is actually working.

Additionally, according to a different new policy briefing published by Princeton University and the Brookings Institution, federal and state governments spend about $34 billion annually on early-childhood education. This money, however, may be gravely misallocated. “Children are not well served by a perpetuation of magical thinking about the likelihood of profound effects resulting from poorly defined, state-run pre-K programs,” Farran and Lipsey, who both work at Vanderbilt University’s Peabody College of Education and Human Development, state in their report.

These findings piggyback on the landmark study the pair released last year, which suggests participation in pre-k does not lead to long-lasting positive gains for low-income children. The 2015 findings show that, although there are significant initial benefits for disadvantaged students who participate in Tennessee’s Voluntary Pre-k program, by the time these children reach first grade, the effects begin to fade. What’s worse is that by the end of third grade, the data show children who participated in the program had lower attitudes toward school and performed more poorly on statewide assessments. Farran and Lipsey’s new research supports their initial findings.This calls into question whether the push for more widespread early-education programming is worthwhile, especially against the current backdrop of rapid implementation and weak research standards.

In their new report, Farran and Lipsey also identify the lack of a common definition for what statewide pre-k is or should achieve as a barrier to success for the programs. Clinton has explicitly promised to “make preschool universal for every 4-year-old in America,” but Farran said these programs need to be put within the broader context of childhood development. “I am very concerned about universal pre-k, when it’s stated like that—as just one little section of the zero to 5 developmental period,” she said in an interview following a panel discussion at Brookings this week. “When you put public funding in and make a new grade in school, you are pulling kids out of what has been a very fragile public system … When you pull the 4-year-olds out, it means the cost of care for zero to 3-[year-olds] goes up.”

That’s not to say, however, that Farran is wholeheartedly against the establishment of a universal program. Instead, she emphasized the need to reevaluate how public dollars are applied to early-childhood education. Currently, Farran said, there is a disconnect between where there is space to put a classroom and where there is actually demand for that classroom. As a result, some public programs remain under-enrolled while others lack necessary infrastructure.

The results from the Tennessee study, however, have been met with significant pushback. Farran said critics have claimed the study is anomalous and questioned the researchers’ methods and parameters. William Gormley, a professor of public policy at Georgetown University, said during the panel that he believes in the legitimacy of Farran and Lipsey’s study, but his own research on pre-k in Tulsa, Oklahoma, paints a different picture. “Do high-quality pre-k programs, including large-scale programs, improve school readiness? Researchers have produced lots of interesting, rigorous studies that seek to answer this question, and they’ve reached a scholarly consensus, and that kind of consensus is rare.”

Gormley, Deborah Phillips, and Sara Anderson released findings in August of this year that show there are positive, long-term benefits for children participating in Tulsa’s Community Action Project Head Start Program. According to the study published by Developmental Psychology, participation in the CAP Head Start “produced significant positive effects on achievement test scores in math and on both grade retention and chronic absenteeism for middle-school students as a whole.” Studies like Gormley’s can serve as rationale for the expansion of universal pre-k, especially in the contemporary era of hyper-polarized politics, when any bipartisan issue is a welcome reprieve.

These results seem to directly counter Farran and Lipsey’s, despite the fact the voluntary pre-k system Farran studied in Tennessee is not universal like the Tulsa system Gormley examined. Gormley hypothesized that what is going on in Tennessee can be attributed to something happening in kindergarten through third-grade classrooms: “It is possible that the k-3 teachers have not upgraded their pedagogy in response to this big influx in students who have gone through a high-quality pre-k program, and when that happens, based on research … we have reason to worry because students may be getting redundant instruction.”

If this is the case, and the findings from Tennessee are the unique results of an elementary-education system that isn’t keeping pace with the state’s high-quality pre-k program, the question is no longer whether universal pre-k provides long-term benefits. Instead, the burden falls on the education system as a whole to improve overall quality and solve what both sides agree is at the heart of this issue: closing the achievement gap. In order to succeed in this, Farran said pre-k as a prescriptive solution needs to be evaluated. “[Policymakers] assume that the achievement gap, if you remediate it at kindergarten entry, will then take care of itself across time, and we don’t have any evidence that that’s true. … We need social science evidence to determine if our solution is working. We must not become wedded to the solution and forget the problem we were working on.”