Editor’s note: This commentary is by Art Woolf, an economics professor at the University of Vermont. A slightly different version was first published in The Vermont Economy Newsletter.
According to the U.S. Department of Education, Vermont’s per pupil spending was about $17,400 per year in school year 2010-11. That’s 45 percent above the U.S. average. The National Education Association estimates that Vermont’s per pupil spending in 2011-12 was 71 percent above the U.S. average. Vermont spends a lot on education and more than all but a small handful of states.
How well do Vermont students perform, given this high level of spending? The recent Picus report to the Vermont Legislature, most legislators, the Vermont Department of Education, and just about everyone in the state believes that Vermont students’ performance is better than almost any other state in the nation. This article challenges that assessment by looking at the only measure of education performance that exists on a state by state basis, the National Assessment of Educational Progress (NAEP) test results. We use the 2013 NAEP math and reading test scores for fourth- and eighth-grade students and examine the percent of students who scored proficient or better, meaning they mastered what should be expected at their grade level.
We don’t think anyone would say that Texas schools are better than Vermont’s, so we’ll compare Vermont performance to Texas, as well as to the U.S. average.
The first thing to note is that Vermont’s student body is very different from the U.S. and from Texas in several dimensions, as the table below shows. Vermont has almost no minority students. More than one-third of all students enrolled in American schools are black or Hispanic, as are nearly two-thirds of Texas students. Vermont is a slightly above-average income state while Texas is slightly below average, but the differences are not huge, so on average, students in Vermont, Texas, and the U.S. come from households with similar average incomes. However, Vermont has a lot fewer children living in poverty than either Texas or the nation. Finally Vermont spends more, on a per student basis, than just about any other state in the nation. In contrast, Texas’ spending is well below the U.S. average.
If we look at students at the aggregate level, Vermont’s students perform better on all four NAEP tests than students in U.S., and much better than Texas. But, as noted earlier, Vermont’s student body is very different from Texas’ or the U.S. as a whole so that comparison is misleading. Minority student populations, especially blacks and Hispanics, have worse education outcomes than white students, and it is important to account for that.
A smaller share of white fourth-grade students in Vermont achieved proficiency on the NAEP math and reading tests than white students in Texas or the U.S. as a whole. On the fourth-grade test, there were so few black students in Vermont that no statistics are reported, and on all tests, that’s also true for Vermont Hispanic students so we can’t compare those students.
On the eighth-grade tests, we find similar results. In the aggregate, Vermont students’ performance looks exceptional compared to both Texas and the U.S. But that disappears when we control for race and ethnicity. On the eighth-grade math test, Texas’ white and black students outperform Vermont’s white and black students. Both states, however, do better than the national average for both black and white students.
On the eighth-grade reading test it also appears that Vermont students do much better than students in Texas or the U.S. as a whole. But again, when we look more deeply at the data, Vermont’s white students don’t do as well as white students in Texas and also slightly underperform the nation. Black students in Vermont do better than black students in Texas or the U.S.
Vermont’s student spending is well above the U.S. average and double that of Texas. Yet Vermont’s non-minority students, which is virtually all students in Vermont, do worse than non-minority Texas students on every one of these tests at both grade levels. They also underperform non-minority students nationwide on three of the four NAEP tests. And Texas’ better performance occurs in a state that spends far less than Vermont. We’re not surprised by that: Most economic studies find no relationship between spending and student performance.
Vermont’s education system’s performance is, at best, about average for the nation and a case could be made that it’s not as good as the national average. That’s far below what most Vermont policymakers, parents, and taxpayers believe to be the case. Vermont does not get a better performing education system in return for its very high education expenditures. We should all be concerned about why that is the case.