More money is not the answer for schools, suggested U.S. Education Secretary Betsy DeVos in Senate testimony Tuesday — but a wave of new studies show that additional money for schools can make a big difference for students.

In an exchange with Louisiana Senator John Kennedy, DeVos reiterated her view on the topic. Kennedy began by saying, “Do you find it at all strange that in America now, we seem to judge success in education by how much money we’re spending as opposed to whether our kids are learning?”

DeVos replied, “I do find that strange. In fact, in the last administration there was $7 billion invested specifically into schools that were failing … to improve them and there was absolutely zero outcome from that investment.”

“The notion that spending more money is going to bring about different results is ill-placed and ill-advised,” she said.

DeVos was relying on evidence from one specific program, but the broader research on money in education generally paints a more positive picture.

“There’s this notion out there that increased spending doesn’t help,” said Diane Whitmore Schanzenbach who has studied school spending and is the director the Hamilton Project at the Brookings Institution. “There’s good evidence that indeed increased spending does help — it increases student test scores and it improves later life outcomes.”

A few major national studies have reached that conclusion.

One analysis, published in the peer-reviewed Quarterly Journal of Economics, found that court-ordered increases in school spending caused students to attend college at higher rates and earn more money as adults. Another study, coauthored by Schanzenbach, showed that when states increased spending they saw substantial increases in scores on the federal NAEP exam.

Other national research has linked more spending to higher graduation rates and greater social mobility.

State-specific studies have pointed to similar results. Research in California, Massachusetts, Michigan, and Ohio have also found gains caused by additional education spending.

None of this research is bulletproof because money is never randomly assigned to schools, but the studies do make efforts to look specifically at the effects of additional spending itself.

Skeptics often point to the fact that education spending has increased while overall performance on federal high school exams has been flat, but according to Schanzenbach this approach to determining the impact of additional resources is problematic.

“If we want to think about what’s the effect of school funding, you can’t do correlations — you need a research design to isolate what happens when more money is pumped into the system,” she said.

In her testimony and in other instances, DeVos’s has highlighted the Obama-era school turnaround program, which produced some successes but did not seem to have much if any overall impact, according to a federal study of the program.

But Schanzenbach says this does not show that resources don’t matter. “That is an evaluation of a particular program, targeted at particular schools,” she said. “I don’t think we get to step back and say we learned something fundamental about the relationship between spending and student achievement.”