Eric Hanushek: The quality of the teacher is the most important element of schools

Your field of research, the economics of education, can be quite unusual for many people. How did you become interested in it?

A long time ago when I was going to write a Phd thesis, there was a major study in the United States that said that it didn't appear that schools had much influence on children and on their achievement. And I couldn't believe that. And so I ended up writing a thesis on schools and the impact on achievement. It was almost unbelievable that there was basically nobody in the economics profession that was looking at schools and what they did. That surprised even me because we spend a lot of money, every country in the world spends a lot of money on its schools and it is an important sector.

Erik Hanushek at the Hungarian Academy of Sciences Photo: mta.hu / Szigeti Tamás

What do you think was the most surprising discovery of your career in the field of educational economics?

I think there are two things that I would note. One surprise that actually spoke about in Budapest is how important the skills of the people are for economic growth. It looks like differences in the rate of economic growth of countries are very closely related to the skills of the people and then by implication, what goes on in schools. And secondly, what I found in the United States, which turns out to hold around the world, is that there is no close relationship between what is spent on schools and the performance of schools. And so there is a big gap between understanding how best to use funds in schools to get achievement and what schools are doing.

But then what are the most effective policies which can be used for ensuring the quality of education?

Well, I have one simple answer: the quality of the teacher is the most important element of schools. It doesn't really surprise anybody when you say that. It sounds almost trivial to say, well, it is important how effective the teacher is. But it turns out that there are huge differences among teachers in what they do for kids. I think that it became really obvious during the COVID period when kids were out of school for a long time without teachers, they just didn't do as well. Even if they had computer technology and so forth.

What makes a good teacher? I think you recognized from your studies that several characteristics which were considered important are not actually crucial.

You are correct. We often measure teachers by how much experience or how much education they have. And it turns out that neither of those is very closely related to effectiveness. We have never been very good at describing why somebody is good, but, on the other hand, we have a pretty good understanding of who is good. So you can go into a school and figure out pretty quickly who the most effective teachers are in the school and who the least effective teachers are. But it is a little bit difficult to describe why, because teaching is just a very complicated activity. As a teacher, you have to interact with the kids and know what they know, and what they need to know. You have to know things yourself and have the right background. The teacher’s personality and other things enter into it, and that is why we have never described very well the difference between a good and bad teacher. But we know one when we see an effective teacher.

Are the results of the children the best metrics of the quality of teacher?

I think even that is a little complicated because we know that the teacher is not the only person adding to the education of the child. The child, the student, him or herself is very important. But we also know there is just no doubt that parents have a big influence on the learning of their children. And so to look at the effectiveness of teachers, you really have to try to separate the influences that are outside the school from what the teacher is doing. And we have ways of doing that statistically, we can do that for research. But the principal or a headmaster in a school knows who the good teachers are and the other teachers know who the good teachers are. They can observe it, they might not be able to pinpoint exactly why one teacher is better than the other, but there is no doubt about who he is.

In Hungary, there is a very serious shortage of teachers, mostly in science domains. The majority of them are over 50 and there are hardly any young teachers. What educational policy solutions would you propose to address this problem?

I think this is common around the world, that there are shortages, particularly of math and science teachers, but some other specialized teachers as well. In the United States, language teachers and teachers for special education are in shortage. And there is one common answer that an economist gives, that is you should pay more to the people that you need. But on the other hand, you don't want to just raise all salaries because there are some people that aren't in shortage. So the answer is clear to economists, but it is not that clear to schools. School policy doesn't go along with that very often, but the answer to an economist is that we should pay the math and science teachers more because we want to get high quality math and science education.

Your new research on basic skills shows that a significant proportion of children are not acquiring basic skills. What can be the effect of this fact on their opportunities in later life?

The common standardized testing we have does a pretty good job of measuring a set of skills that is important for a modern economy. We know that in basically every country in the world, people who know more earn more. That is because they have more opportunities – employers want people with more skills and they pay more for them so people's incomes are higher if they have a better background in schooling. But the reason for that is not entirely clear. I think that the largest part of this is that people with more skills are more adaptable. They are able to adjust to different circumstances, to make decisions under uncertainty and to change. When an economy is growing, the character of the labor market and the jobs is changing. And people who have more skills are better able to adapt to the changes that come about in a growing economy.

Hungary has been performing very badly at the PISA test year after year. What could be the best policy to improve this performance?

Yes, actually I was surprised to see that Hungary had done it badly over time, even getting worse over time. And I think that the answer still comes back to having highly effective teachers in the classroom. This often translates into that we have to increase the requirements for getting a teaching certificate and we have to deal with the quality of the teacher training institutions and so forth. But there is very little evidence that that actually works. Again, for the reasons we talked about, it is hard to define what makes a good teacher. So in my opinion, you just have to make decisions based upon how well people do in the classrooms, and keep the good teachers and maybe not the poorer teachers.

Besides the teachers, people think that the size of the class has a huge effect on the quality of the education. So in your opinion, what is the ideal size of classes, for example, in elementary school?

- Well, if you listen to the policy discussion people, when they talk about the optimal class size, it is always fewer than we have today. So people always think, well, we'll do better if we just make it smaller and smaller. But the evidence is actually very weak that class size has a very big impact. It is clearly the case that the impact of class size is much, much less than the impact of an effective teacher or the differences in teachers. I was just at a conference where we were talking about how to deal with the pandemic and so forth. And I had a suggestion that nobody agreed with, and that is that you provide incentives for your most effective teachers to teach larger classes, to teach more intensively. And you give smaller classes to poor teachers. Now, that goes against what everybody argues, which is that we should reward the good teachers by giving them small classes. But that is exactly what I think you should not do.

Every class has one or two children who have very great skills, they are talented. In Hungary the main educational dilemma has been for a century whether it is more important to nurture the talent or just teach all the children with the same effort. What is your suggestion?

This is a basic question. First, I think we are getting to a point where we can deal with the heterogeneity within a classroom better. We can deal with these differences better by bringing in some technology and having sort of more individual pace instruction with the help of technology. It turns out that both the education of the top people, who we might call rocket scientists, and the average students are equally important. The rocket scientists are important because they help to define new productive ways of doing things. But having a high base level of skills is also important. And it is hard to say exactly which is more important. But we do know that rocket scientists are more productive if they have good people to work with. Some countries concentrate only on the very top people. My example is usually India. They have a very steep pyramid to get to just the top people. And I think that that is a mistake for two reasons. One is that it doesn't provide the overall workforce that would be most productive in the country. And secondly, large portions of those people now work next to me in Silicon Valley. They have left India and they work here very productively in the United States but they are not helping India develop.

You mentioned the technology. Do you think that the emergence and very rapid development of AI, the fact that AI will presumably be able to take over many white-collar jobs, may change the basic skills required in the near future and that schools should therefore teach differently?

We are just in this era of trying to figure out whether artificial intelligence replaces all of us, whether it replaces you and me. My view right now is that artificial intelligence is just another development of industry. We have gone to different production technologies. We went from agriculture to manufacturing to information technologies. The difference appears to be the speed with which things are changing. But in the past, we also had new inventions. We have robots coming in and so forth. It has increased the value of more skilled people, because in fact they do better with this new technology. It is the same with your work and my work. I mean, we have changed what we do over time, and I think that that is what artificial intelligence is going to do for us. It is going to change some very routine things that are going to go away, but it is not going to substitute for you interviewing me.

I hope so.

I think so. Well, for example, the computer can read x-rays better, it turns out that machines do that at least as well if not better than doctors. But that doesn't mean that we are going to get rid of doctors, they are only going to do different things. That is my general view also on the teaching function. I don't see in the short run that artificial intelligence will eliminate teachers, but it could change what they do for just the reasons we are talking about. You could see that artificial intelligence could help guide the instruction of the rocket scientist in the classroom while also dealing with the more basic skills of other students.