North Carolina’s public schools are currently producing better results than you might think, according to a recent analysis of independent testing data. But there’s bad news, as well: North Carolina’s rate of improvement has lagged behind the national average since the turn of the 21st century.

Matthew Chingos is a senior fellow at the Urban Institute, which published his study last fall. Chingos analyzed state performance on reading and math exams administered by the National Assessment of Educational Progress (NAEP). A random sample of students in each state takes these NAEP tests every two years.

Whenever student test scores come out for schools, districts, or states, officials hasten to explain that there are many factors known to shape the results. They are quite right to do so. The characteristics of the families within which students grow up — household income, parental education, marital status, etc. — clearly affect student performance. Race and ethnicity exhibit statistical correlations with performance, as well, perhaps reflecting not only those family-background variables but also factors such as neighborhood effects, cultural norms, or discrimination.

In the end, socioeconomic explanations can take you only so far. Students who fail to learn what they need to learn to succeed in life are at great risk. Blaming their fate on poverty or a lack of role models doesn’t change that fate. They still lose — and so do the rest of us, as there are broad social costs when young people fail to live up to their potential.

Still, unless policymakers take student demographics into account when evaluating the success or failure of education programs, they will draw faulty conclusions. Some conclusions will be false negatives: policymakers will fail to detect the net benefits of a promising school reform. Some conclusions will be false positives: policymakers will ascribe a test-score gain to a school reform when it was really caused by something else.

In his paper, Chingos adjusted the NAEP scores for student background. It made a big difference in the state-by-state comparisons. Here are the top 10 states in student performance, in descending order: Massachusetts, New Jersey, Texas, Florida, Indiana, Maryland, North Carolina, Pennsylvania, New Hampshire, and Ohio.

While the first two states would have ranked at the top in any event, the adjustment dramatically elevates states such as Texas (3rd), Florida (4th), and North Carolina (7th) with large populations of non-white, low-income, or foreign-born children. Correspondingly, states such as Connecticut, Nebraska, and North Dakota that rank high in raw scores fall below the national average when one adjusts for the fact that they educate many fewer disadvantaged students.

I told you the Chingos report produced some bad news for our state, as well. Here it is. The vast majority of the performance gains North Carolina has posted on state NAEP tests actually happened during the 1990s. Since the turn of the century, most states have posted larger gains than North Carolina has, whether in raw or adjusted scores.

Sorry, partisan hacks, but there are no political points to be scored here. North Carolina’s NAEP performance began to level out at the tail end of the Jim Hunt administration. The trend has continued under Democratic and (so far) Republican governors and legislatures.

Despite North Carolina’s stagnation in NAEP improvement, our students continue to test about three months ahead of their national peers. It’s a start but not nearly good enough. Students in New Jersey and Texas test about six months ahead of the national average. Kids in Massachusetts test about nine months ahead. And there are plenty of students in those top-performing states who still aren’t learning what they need to know.

Spending more money on well-constructed school reforms is fine with me, but don’t fall into the trap of believing the primary issue is funding. Massachusetts and New Jersey spend more than North Carolina does, yes, but Texas and most high-performing countries in the world spend less.

Our ultimate goal should be to deliver the greatest educational value for the dollar in the world. The social and economic benefits would be enormous.

John Locke Foundation chairman John Hood is the author of Catalyst: Jim Martin and the Rise of North Carolina Republicans.