A new study suggests that Tennessee might want to up spending and look toward its neighbor to the east if it wants to get prekindergarten right.

This week, Duke University released research showing that North Carolina’s investment in public pre-K programs led to better outcomes for its students. Its researchers found that the positive effects — including higher test scores, less grade retention, and fewer special education placements — grew or held steady over the years.

At first glance, the findings seem to contradict those released last year by Vanderbilt University researchers about Tennessee’s public pre-K programs. That study’s authors concluded that by third grade, the students who attended pre-K actually fared worse academically, calling into question how much return states and communities can expect from their investments in early childhood education.

“Money does matter,” said Helen “Sunny” Ladd, a co-author of the Duke study. “It’s important that (it’s) used well. But if something is important, like preschool, it seems to make sense to spend money on it and keep improving it.”

In fact, the Duke results echo what Vanderbilt researchers Dale Farran and Mark Lipsey have said all along: that pre-K’s promise can only be realized if programs are high-quality and work in tandem with other parts of the education system. They theorized that a lack of support before children turn 4, as well as the low-performing elementary schools that those children attend after pre-K, might be to blame for the Tennessee’s program’s disappointing results.

The Duke study looked at more than 1 million North Carolina public school students born between 1988 and 2000, and how their counties funded two early education programs: Smart Start, which focused on ensuring that infants through 4-year-olds entered school healthy and ready to learn, and More at Four, a pre-K program.

 By the end of fifth grade, children living in counties with average levels of early education funding saw a gain of more than six months of reading instruction and more than three months of math instruction, regardless of poverty level. The children also had significantly higher mean math and reading scores in grades three, four and five, and their odds of needing special education during elementary school were lower.

Ladd said several factors might put alumni of North Carolina’s pre-K programs at an advantage: Smart Start served as a foundation for pre-K, and North Carolina historically has had higher-performing elementary schools than Tennessee, which might help sustain pre-K gains. Tennessee also spends about $500 less than North Carolina per pre-K student, up to $10,000 less per classroom.

“The effectiveness of preschool undoubtedly depends both on what happens before and what happens after,” Ladd said, adding that it was impossible to separate out the effects of Smart Start on North Carolina students.

“There’s evidence from lots of studies across the country that preschool is important,” she said. “What the Tennessee study suggests is that maybe Tennessee needs to keep working on their program.”