One of the most controversial questions in education has been whether preschool — and specifically Head Start — helps kids succeed as they move through elementary school.

Critics have long noted, and research has supported, that the benefits of Head Start fade in a few years. It’s an important question for an $8 billion federal program that provides support for nearly a million low-income children and their families.

This year brought several new studies, however, that found that — when done right — Head Start and other programs can give low-income students lasting benefits. It’s not only through elementary school: At least one study we wrote about found the benefits of preschool paying off for individuals, and society, into adult life.

All this research, however, was no blanket endorsement. Some of this year’s findings reinforced earlier studies showing the uneven quality of Head Start programs around the country.

And so the lessons from 2016 seem to reinforce the emphasis — by President Obama and others — on quality.

One of the most closely watched attempts in the country to provide universal, high-quality preschool has been in Oklahoma. In 1998, the state became one of only two states to offer universal preschool.

Today, the vast majority of Oklahoma’s programs are in public schools. The rest are run by child care centers or Head Start.

Deborah Phillips, a professor of psychology at Georgetown University, has spent more than a decade studying and tracking children in these programs. Her most recent findings were published in the journal Developmental Psychology.

Her study focused on Tulsa’s biggest Head Start program, which is run by CAP Tulsa, a nonprofit group that serves 3- to 4-year-olds. It looked at how the students in the program were faring years later.

And it found clear benefits for children who had gone through the program. In August, I talked with Phillips and asked her to summarize her findings.

Your research team wanted to know whether Head Start produced significant, long-term academic success for the 1,774 children in your sample. What did you find?

Children who attended Head Start had higher test scores on state math tests [by eighth grade]. They were less likely to be retained and less likely to display chronic absenteeism. These are highly consequential outcomes that we know are predictive of high school graduation, college enrollment, even earnings.

Low-income children who come out of [Tulsa’s CAP Head Start program] defy teachers’ and principals’ expectations of them based on their background.

We’ve been able to demonstrate that the Head Start model, with its strong family support component and comprehensive services for children, can give children a strong pathway through school and hopefully out of poverty in their adult lives.

Still, some of your findings are pretty worrisome. African-American boys in your study, for example, did not benefit very much at all from Head Start. Why?

We don’t have the answer to that, but it’s extremely troubling. We are digging very deeply into our data we collected in third grade and now, of course, we have eighth-grade data. So we’re going to look over time and see what’s going on.

In middle school, you tested kids in things like algebraic reasoning and geometry. Then you examined their vocabulary and reading comprehension. In math they did pretty well, but in reading the results were not good. Do you know why?

We can only guess why we’re not seeing the same impact on reading test scores. Maybe because children’s early home environment has a bigger impact on reading skills than on math skills. Although it is a little perplexing because we found very large, immediate impacts on children’s reading assessments. So that’s a puzzle.

On the bright side, you found that Latinos, including those who did not speak English at home, benefited significantly from Head Start by the time they got to middle school. And you say that’s important because Latinos are the fastest-growing group in Tulsa’s public schools.

But don’t English-language learners, or ELLs, face more difficulties in the elementary grades, especially in reading?

They shouldn’t as long as children are getting linguistic input, whether it’s in Spanish or English at home. English-language learners … in homes with very poorly educated parents benefited more than any other subgroup.

Why is middle school an important marker in terms of these kids’ progress?

The middle school years are a fascinating developmental stage. It’s when children are really beginning to solidify attitudes about school and its role in their future. In most studies that have followed children into early adulthood, it’s in the middle school years that you see major fade-out. So we wanted to look at middle school as an extreme test of the impact that Head Start has.

Fade-out of course is the single biggest criticism of preschool programs, including Head Start. Doesn’t the research show that the benefits don’t last beyond first grade?

There is some evidence that part of what happens when you see fade-out is that children come out of pre-K eager and ready to learn with strong initial skills. They arrive in elementary school and those skills are not advanced. Kids get the same lessons they got in preschool. So of course you’re going to see fade-out under those circumstances.

But as your own study acknowledges, the findings from Tulsa are not truly representative of most Head Start programs nationally.

There is tremendous variation in what children experience in Head Start. In Tulsa we have a very high quality Head Start program.

When we compared the quality of teacher-child interaction in the classroom [in Tulsa] to national data collected from 11 different states at the same time, we found that Tulsa’s Head Start teachers had [better] instructional support and spent more time on academic content compared to the other 11 states.

[Tulsa] pays its Head Start teachers the public school wage, so there’s no incentive for a teacher to jump ship to earn a higher wage. So Head Start is able to attract and retain talent.

Still, I would not describe the Head Start program in Tulsa as “exceptional,” because the national Head Start program has been moving over this decade in the direction that Tulsa has arrived, namely, requiring higher teacher qualifications.

Every teacher in Head Start nationally is now supposed to have a B.A. degree and be early-childhood certified. Performance standards have been substantially revised to ensure that these programs foster academic learning and social emotional development.

Do you worry about the skeptics who will question your findings and look for flaws in your methodology?

Many of us who’ve been studying preschool education are feeling that it’s time to turn the question away from, “Does pre-K work?” to “What is going on [when kids] transition from pre-K to the early elementary grades?”

I think our nation has arrived at a watershed moment when it comes to early education. We now have not only scientists, developmental psychologists, economists, neuroscientists but also police officers, business people, many politicians and parents believing in the power of preschool. And that kind of coalition will drive this agenda home.

So why is it so hard to replicate the success you’ve documented in Tulsa?

That’s the million-dollar question. You could say it boils down to money but it actually boils down to commitment. Money flows from commitment. We spend money on the things we care most about and I’m hopeful that these very encouraging results will reassure people that high-quality preschool is the first step in fulfilling the promise of America’s education system.

A version of this story was published on NPR Ed in August 2016.