This interview focuses on a chapter from A Nation at Risk +40, a report by the Hoover Education Success Initiative (HESI) that looks back at the birth, struggles, and future of the modern school reform movement. (Download the publication here.)
Eric Hanushek, the Paul and Jean Hanna Senior Fellow in Education at the Hoover Institution, writes about how to make sure school spending is used to measurably improve student achievement. He spoke with Chris Herhalt about how schools and students can recover from the learning losses of the COVID-19 pandemic.
Chris Herhalt: Is it a coincidence that one of the states with the highest per-pupil funding [Connecticut] also scored highest in the nation in your study about learning loss due to COVID? What about the other Northeastern states that spend $15,000-plus per pupil?
Eric Hanushek: What you see when you look at achievement across the United States is that some of the states in the Northeast are the highest-achieving states in the nation as well as the highest-spending, but if you look across the overall pattern for all the states, you don’t see a high correlation between spending and performance. If you pay attention to the education and background of the families, the impact of spending is not significant.
Herhalt: In your chapter, you offer seven guiding principles for effective school funding. The fourth one really struck me: that school-funding formulas should minimize unproductive “gaming” by avoiding rewards for things that are easily manipulated by school personnel. Can we talk about some real-world examples?
Hanushek: Lots of people worry about how accountability will affect cheating on tests, and there is a little bit of cheating that people do for accountability. But for instance, if you reward graduation from high school—graduation is a bit subjective, and you can manipulate the graduation rate if you are rewarded to do that. We’ve seen that in states.
Herhalt: Is that sort of the game where suddenly the thresholds for passing change, or the assignment rules change?
Hanushek: Exactly. For most states, there are requirements of passing a certain number of courses, but whether people pass these courses or not is largely dependent upon local districts that can, in fact, change the grading of courses and allow for exceptions and so forth.
Herhalt: Your sixth principle is that schools should gather relevant programmatic and performance data. What is the best method for gathering this data? What should they be looking for, and is this goal best carried out through standardized testing and looking at those results?
Hanushek: Evaluation of schools and of personnel in schools has been very controversial, in part because you don’t want to reward a teacher for just having a good selection of kids in the class. You don’t want to reward the teacher if they get a classroom where the parents are really helping a lot on education.
The basic idea is that you want to evaluate and reward schools and personnel—principals and teachers—on the basis of whether kids are learning. This is the fundamental issue. And in that, you want to look at the value added by the school or the teacher. How much is the teacher adding to what the kids know from other ways, either prior schooling or from outside factors such as families and neighborhoods? In all cases, you want to try to look at what the school did as opposed to what other people did.
Now, you can do that with tests if you have sufficient testing. Today all schools, by federal mandate, test kids in grades three through eight and then once again in secondary school. But in order to judge schools or kids, you have to know over time what individual kids are doing and follow that. If you look at kids learning in grade three, but we haven’t seen what their performance was in grade two, it’s hard to judge what they knew before the beginning of the third grade. Similarly, in high school, you very seldom have information on the performance of kids at any point because you have a bunch of specialized courses.
So, the best way to evaluate schools is to use test score information if it’s available, although that’s only available for a minority. Also, what we have now are fairly sophisticated measuring systems for observations by supervisors and others from outside who can make judgments about which teachers are doing the best job and which aren’t, and which schools are doing the best job. Washington, DC, is a good example. More than ten years ago, they introduced a very elaborate evaluation system that used test scores if it was possible to calculate the value added on the basis of those scores. But only about 20 percent of the teachers had test scores that could be used, if they were in the right subject matter—math and reading, as opposed to specialized courses—and they had before and after measures for their class. The remaining 80 percent of teachers were evaluated by multiple observations by outside observers, trained observers, or supervisors, and it turns out that that’s a pretty good method.
Herhalt: You point out that NAEP scores started falling in the United States in 2012, and the conventional wisdom in the discourse is, “Oh, it was COVID, it was the lockdowns. That’s when things started to decline.” You’re saying this happened eight years earlier. Why was that?
Hanushek: I think one of the main reasons is that we changed the federal accountability system for schools. We had No Child Left Behind, which was started in 2002. It was not a very good law, frankly, but it did hold schools accountable for student performance. Huge pressure built up to change the law, and we passed a new accountability statute in 2015, which basically allowed all the states to decide how to measure performance and what to do about that. It turns out that several states backtracked on their accountability systems and what they were doing after this law came into effect. So, I think changes in the accountability system have been a big factor.
Herhalt: In your chapter you say, “it’s not how much money, it’s how the money is spent.” This theme addresses the variability across jurisdictions that are doing almost the same thing on the funding side but getting different results. Today, how big a role do you think technological capital—tablets, educational software, apps—might play in helping the 2020 cohort recover from learning losses?
Hanushek: Well, I think that we’ve had the hope that technology would save our schools for quite a while. We had the computers that were going to save the schools. Before that, we actually had TVs that would save the schools if we could broadcast in material. And we’ve been very slow to see any definitive results of adding technology. While I still believe technology offers great hope, it doesn’t work for everybody and the programs we have don’t work for everybody. Some people can use them very effectively; others can’t. I don’t think that’s where we’re going to see all the gains.
There’s one aspect of computers that I think is starting to be more effective: the providing of regular feedback to teachers about where kids are, allowing teachers to adjust to individual kids and tailor learning programs to them. But the answer still comes down to teachers. Personnel are the key. We learned that during COVID, when we had people sitting at home: it was not as effective as when students were in class with teachers. That’s just pretty consistent information. So, I look to improving the effectiveness of the teaching corps as the biggest hope for how we could improve our schools.
Herhalt: Do you think school districts need some sort of educational Marshall Plan to recover from COVID?
Hanushek: No. We’ve been trying the pure-money solution for a long, long time. We’ve tried to do all kinds of things, and in fact schools have been quite generously supported. The increase in spending for schools over the past fifty years has been very dramatic, and it’s been unrelated to performance. Now, that having been said, I believe money can make a difference if it’s spent well, and that there are places where spending has been useful—but we haven’t found a way to make sure that just providing money to schools leads to better outcomes. We have to be more sophisticated than that. As an economist, I believe that a large part of the answer is providing incentives to people in the schools to do what we want them to do: teach kids and get kids learning more.
Herhalt: In your view, is the best accountability measure something that tracks improvement?
Hanushek: Just knowing the level of performance doesn’t tell you who’s led to that difference or whether the individual teachers or schools have been very productive in that. On the other hand, if you look at the growth in performance of students, you can see that teachers can change the pattern of growth. They can provide more learning, and that’s what we should be paying attention to. If all students are just doing as well as they would have done based on outside-of-school factors, we shouldn’t reward that. Put another way, if my department chairman came in and said, “We’re going to pay you based upon the overall performance of your students,” the first thing I would do would not be to fine-tune my reading list. It would be to try to select the right students.
Herhalt: A lot is made about the teachers’ unions being in the way of reform, but many times a district wants to incorporate some new accountability measure and the union is receptive to it. In your chapter, you point out that even states with no collective-bargaining arrangements still run into inertia problems. Do you think too much is made about unions and how they resist change?
Hanushek: Well, I think that they’re a very important ingredient in stopping change. The unions have not been very helpful. Again, you can see their response to the pandemic, which I think was not very good. They kept schools closed. They would go crazy if Safeway were to close the grocery stores, because it’s important that those people be there to serve them, but then teachers wouldn’t go into the classrooms. And that was largely union-driven; there were strikes based on that. But unions aren’t the only factor, and I completely agree with the notion that there’s a lot of inertia in the system that you can see in states that don’t have collective bargaining, where the system itself is pretty stuck on where it is. Simply put, the people in the system like what they see, for the most part. They would rather be paid more, and so on, but they’re there because they like the job and they don’t want the job to change.
