TECHNOS Interview: Caroline M. Hoxby - Interview

When parents have a choice among schools, are their children likely to get better teaching? Do smaller classes raise the achievement of young students? Harvard University professor Caroline M. Hoxby, a member of a new generation of economists, provides answers that surprise and infuriate many educators. Only a few years out of graduate school, she and her colleagues are applying the tools of traditional economic analysis to education. What she has found out about the effects of school choice, class size, funding equalization among a state's districts, and other issues challenges educators and policymakers to rethink long-held assumptions. Underlying Hoxby's studies are two fundamental questions: If we want better, higher-achieving schools, what reforms are most efficient? And, what incentives produce improvement at the least cost?

The Interview

You are called "the economist of education." What particularly interesting challenges or concerns does education present to an economist?

They're twofold. First, education is the major road for economic advancement for individuals. Second, we have a truly education-based economy. We've always had a very skill-based economy in the United States, even in the 19th Century, but it's become even more education-dependent over time. That's the way the economy is moving -- to doing things that are technologically advanced. Our economy depends so much on whether the education sector functions well or functions poorly, and that's a good reason to study it.

The problem for many years -- until my generation of economists -- was that economists felt that they lacked the tools to study the education sector. There are tools that we've used for almost 100 years to understand how firms work with one another in the private sector, but they weren't developed initially for use with nonprofit organizations or semi-governmental organizations like schools. Schools are somewhat like government because they're supported by taxes. On the other hand, there isn't just one government school; there are many government schools, and they interact as firms do in a marketplace. So you have to use some knowledge from the study of government -- what we call public economics -- and some knowledge from the study of firms in order to understand schools.

You use tools that have been applied to two separate sectors and show that they apply to a sector that has characteristics of both.

That's right. The real breakthrough was having people learn about the education sector. It wouldn't be obvious how to adapt the tools if you didn't know how schools actually work, especially the way schools are financed, which turns out to be very different from most other types of government finance. But once you've done the homework and adapted the tools, moving ahead becomes quite straightforward.

What new perspective can this approach bring to education?

First, we look hard at incentives and ask, "What incentives does this school face? Is its behavior in line with the incentives?" We need to think about what makes its budget change. Does its budget go up when it attracts students? Is it important to its budget that property values remain high in its area? What happens when it has to compete with a private school? These determine the incentives a school faces and its behavior. Looking at incentives is the first important thing that we've contributed.

The other contribution is trying to understand competition in K-12 schooling -- or the lack thereof. It's changed a lot over the 20th Century. There used to be more competition than there is now, and being able to detail what's happened to competition has really been important for our understanding of what's happened in the marketplace.

How did competition among schools work traditionally?

Our choices among public school districts mimic the choices we normally make as consumers, except that instead of shopping at one supermarket rather than another, you have to choose among public school districts by choosing a residence. This gives districts indirect price incentives because school budgets depend on property values. If a school district doesn't attract very many parents because it isn't providing very good value for money, then property values in the district fall. As a result, the school budget for the district falls. So you can imagine a school superintendent or principal thinking, "It's not in my best interest not to attract parents to live in my school district, because I'm the person who's ultimately going to be facing a tighter budget. I won't get a salary increase myself or get to hire the teachers I want to hire, and I'll have to cut the program that's my special baby." In the United States there's a very big range in the number of school districts serving a metropolitan area, all the way from just one school district for the entire area to some metropolitan centers that have more than a hundred separate districts. In areas with maximum competition, achievement is about 5 to 7 percent higher, and school spending is about 8 percent lower. There are efficiency improvements, so you get more achievement for less money.

Isn't there an issue of equity? People who can afford higher property values also benefit from higher-achieving schools. If you are limited to cheaper housing options, you are probably going to be in a school district that is not as high-quality as those in some of the wealthier suburbs.

There is an equity issue, and that's why a lot of the school reforms are targeted toward improving the school-choice opportunities of low-income people and poor people. Rich people in the United States have a lot of school choice, and upper-income people (not just the very rich, but people above the median income) who live in one of these metropolitan areas with a lot of districts have a lot of choice, and they have much better schools. It's not just the money. A lot of inner-city schools spend as much money as rich suburban schools. The difference has to do with the fact that parents in inner cities lack control over their schools. Their schools don't have to pay any attention to them because they have no other choices. A lot of school-choice reforms are oriented toward making school choices available to poorer people. Rich people are not going to be very much affected one way or another by any school-choice reform.

The ideal is to have equal educational opportunity for children. Proposed and existing voucher programs are means-tested. The plans move toward directing more vouchers, more choice, toward poorer children, and I think that's very important. The ideal voucher plan would have less money directed toward people who already have a lot of choice and are using private schools already.

You have said elsewhere that a program would only work if every child in a district had a voucher.

Yes, that's right. You don't want to get in a situation where 1,000 kids get to leave, and the rest get left behind. But there's no reason why everyone has to have the same size voucher. Right now, in school finance, neither the states nor the federal government give the same amount of money to school districts for every type of student. They give school districts more money for poor kids. So it's appropriate that vouchers be bigger for poor kids.

You have also said that voucher programs could be structured to encourage certain socially desirable goals, such as racial integration. How would this work?

The idea is that some things that we are already trying to do can be achieved even better and more efficiently with vouchers. Let's say you have a city that's trying to achieve a desegregation order, like Kansas City. Suppose the ratio of blacks to whites in the school district is 70 percent blacks to 30 percent whites. You structure the voucher program so that if a school that accepts vouchers has a ratio of blacks to whites that is also 70 percent to 30 percent, it gets a big voucher. If its enrollment moves away from that ratio in either direction, the vouchers of all its students decrease. In this case, every parent in that school, whatever race, has an incentive to help the school achieve the desegregation plan. The closer you get to the desirable ratio, the better off everybody is in terms of money.

Isn't there a danger of schools becoming selective and adopting a quota system to keep their enrollments balanced? Couldn't a child be "selected out" of all available choices?

No, I don't think so. It's very reasonable with vouchers to require admissions by lottery, or on a first-come, first-served basis, or something like that. You want the school to attract a pool of applicants that is representative of what the district is like. Then you would achieve the district's goal, even if you ran a lottery. And that should be the goal: for the school to be attractive to the whole population of people.

Wouldn't it be enough just to have public schools competing with each other? Why have private schools in the mix?

There are two reasons for thinking about the importance of private schools, the first being political and the second about resource use. The political reason is that once you send your child to a private school with a voucher, you're not really subject to the actions of elected officials. I think this is important for insuring that an education policy that's working well for parents remains there, because the political interest groups, who influence policymakers, are not really interested in kids. If there were a policy that was good for kids but bad for them, they would use all their political clout to make sure the policy disappears. Many people think that the reason for having vouchers and not just public charter schools is that the charter schools continue to fight, say, with teachers unions, often for their very existence, seven or eight years after the law authorizing them is passed. It is less likely that this would happen with private schools using vouchers, because they're not as subject to political control.

The second reason is that private schools have made quite different use of resources than public schools have. They've paid teachers differently, and they often pay less for certain types of supplies, buildings, and things like that. They appear to be good bargainers about inputs. It is difficult to get public entities to achieve that kind of efficiency. Maybe they would if they were subject to more competition, but right now we see that private schools are much more efficient.

Doesn't the independence of private schools also mean that the people who are involved with them are less invested in the public school system that has to support much of the U.S. economy?

What you want is to have people invested in the education of children in America. Loving the public education system for itself is not very sensible. We have parts of the public education system in the United States that routinely graduate kids from high school who literally cannot read or do arithmetic. I find it very hard to feel devoted to supporting this institution for itself. I would rather support the kids.

But you are supporting it. As long as you pay taxes, you are supporting the institution. So, if you care where your tax money goes, you do have an indirect interest in ...

Of course. Everybody who pays taxes to support public schools has an indirect interest in making sure that these schools are actually using their tax dollars well. But the key thing is to get people to support the education of American children as opposed to the system. If a bunch of private-voucher schools do a great job of educating all the kids in the metropolitan area, I don't think you should care very much whether the public system is supported or not, especially if there are no kids using it any more. The key thing is whether kids are being educated.

Would a voucher system work if the educational system were structured so that every child simply received the equivalent of the state's or the district's per-pupil funding -- $5,000, $6,000, $7,000, or whatever -- to use for education wherever he or she wished?

This wouldn't be a bad place to start. The first thing you could do is give each child the per-pupil spending currently associated with him or her. Special education and bilingual education children would automatically get a higher amount than other kids, because they really do have more per-pupil spending associated with them. Poor children should take more money than rich children, because right now they have money that they attract to the school district. If vouchers started out that high, there would be a large supply of charter schools and private schools to receive them. A program like this would be very easy to implement because all you have to do is to figure out how much the per-pupil spending is for each child and hand each one a voucher equal to that amount. It wouldn't be a tough, complicated calculation to do. It's an intriguing thought.

If you were able to initiate a voucher reform program that you felt would be most effective, what would some of its features be?

I would set up what I call an ideal voucher system. It would be a universal voucher; and it would be highly means-tested, so that probably the voucher would be zero for children of rich parents and generous for poor students -- at least as big as the per-pupil spending they're receiving now, including the amounts that come with them because they are poor. It would include the existing supplements for bilingual education and special education, so that special-education children would not go to a school with their vouchers and find that the school rejects them because the vouchers aren't enough to implement individual education programs. The system would also include anti-sorting controls -- what some people call anti-cream-skimming controls -- so the voucher would be a lot smaller if the school did admissions testing or interviews to exclude students who might be more expensive to educate. The voucher would also be the same size as per-pupil spending right now, because we want to induce people to enter the market for schooling. There are lots of education entrepreneurs out there who are very serious about improving schooling, but if you offer vouchers that are only $1,800 to $2,000, they cannot really supply schooling.

It's important to say that you want everyone to exercise choice. Suppose you find, for instance, that poor parents in an inner-city district don't appear to be doing anything with the vouchers, that they're the only ones left in a public school because everyone else has exercised choice. Once the public school gets down to some core like that, there ought to be something built into the program that says, "Everybody needs to exercise choice now. We don't care whether you want to exercise choice or not, because everyone else in your school has shown that your school is really not succeeding, so the fact that you're still there means you need to get motivated."

What have you found out about the effects of lower class sizes?

In looking at all Connecticut elementary schools over about a 25-year period of time, I found that class-size reductions did not have a significant effect on student achievement. I found this evidence by looking for variations in class size that occur almost arbitrarily when a school's enrollment changes from one year to the next. You can look at two kids who are going to the same school, one year apart, and they end up in classes of different size because of arbitrary variation.

Did you come up with some conclusions about why there was so little effect?

I think there's very little effect because there's no incentive for the teachers to do anything with the smaller classes. They're not being evaluated, and no one says if you don't produce achievement in three years, your class-size reduction will disappear. Nothing happens. It's an opportunity to improve achievement, but an opportunity that is not usually used.

So the inference is that if you want to get improvement, you have to offer incentives, with or without class-size reduction. Is that correct?

I think that's true in every field of endeavor. If you give people extra resources, you ought to make sure that there is some incentive to use the extra resources to improve their output, or they may do nothing different. We ought to say, "If you, as a school district, claim that class-size reduction is what you need to improve achievement, then we're going to pay for a class-size reduction for you for the next five years. If, after five years, you have not improved achievement by X amount, that class-size reduction money disappears. I think that it should be something as simple as that.

In a recent case in New York, a judge ruled that New York City schools were underfunded, and he ordered the state to equalize funding. But your studies of school funding equalization have found that often the results of these efforts are pretty negative.

People, including the judge in New York, don't understand school finance equalization very well. The goal is simple: Children should have equal resources spent on their education, regardless of their parents' ability to pay, and there should be equity among schools and educational opportunity. The problem is that typical school finance equalization judges in court cases show zero expertise about how taxes work. They implement plans that are counter-productive because they don't understand how property taxes work, but property taxes are the way schools are mainly financed in the United States. They don't function like income taxes, and people make ridiculous plans because they don't understand the mechanism that they're trying to work with. It's like saying that we want to build a bridge from Boston to Cambridge, but we're not going to bother calling in an engineer. Then we're unhappy because our bridge collapses in the middle. The problem is not that people can't agree whether or not to build a bridge, or where to build a bridge; the big problem is that they don't call in engineers! In the case of school finance equalization, an economist is like an engineer.

What do you suggest?

It's quite easy. Property taxes cannot be used, period, to do equalization. They do not work. Property taxes are attached to specific pieces of land, by definition. Traditionally, the taxes support local schools, which are therefore responsive to the local taxpayers. If you use property taxes for equalization, you take money from people who own bigger houses and who live in districts with better schools, and you give it to people in smaller houses where the schools may not be so good. That's the nature of equalization. The difficulty is that now each piece of land associated with a big, nice house has a tax bill attached to it that no longer supports local schools, but goes into the pool for equalization. You've said, "This piece of property, forever after, will be associated with a tax liability." The property market then says, "If this house comes with a big tax liability for all future time, then the house will be worth less because people won't pay as much for a house with a big tax liability that does nothing to guarantee the quality of local schools." The people who make up school finance equalization systems never realize that the property market will undo what they intended to happen through redistribution, but that always happens, and it often happens quite quickly.

The largest source of money for the property tax pool suddenly becomes worth less, so your total pool shrinks.

Exactly. People who were previously property-poor suddenly become property-richer because their properties are attached not to a future stream of taxes but rather to a future stream of subsidies; their local schools are now better funded than they were before. No economist thinks that you should ever use property taxes for redistribution. It's just silly. You should use income taxes or sales taxes for redistribution, something that depends on how much income a person has, not how much tax revenue is associated with a house. If I'm a rich person, and you tax me, there's no mechanism by which my income automatically begins to shrink, but if you tax a rich house, the house's value starts to shrink automatically.

What is your opinion of the current education reform plans of the Bush Administration ?

There are two major components to his plan, as I see it: standardized testing and what are called Title I vouchers. My sense is that both of these things are smart ideas, given the fact that the federal government has not provided more than 7 percent of K-12 education spending in the last 20 years, and its discretionary spending is only 1 or 2 percent of the total U.S. K-12 budget. Standardized testing is a good idea because most states are already moving toward it. The federal government is now encouraging all the states to test in grades 2 through 8 -- in other words, do it in a consistent set of grades. They're saying, "We want to be able to compare schools across states; and we want everyone to be able to understand what is happening to achievement in the United States." Right now we have some states that do testing once a year, and some do it once every two years. Consistency would be useful for research. This will help us learn about what's really happening, and it's a very small price tag. Standardized testing costs about $2.50 per student, compared to class-size reduction, which costs about $800 per student.

The second proposal, the Title I vouchers, is what I would call an exposure issue. Title I doesn't give very much money on a per-pupil basis. If you are a very poor student in a very poor school district in the United States, Title I might give as much as $800 or $900, but not a real voucher size. There are very few Catholic schools that would allow you to enroll for $900, and by definition you're very poor, so it's not clear where your parents would get the rest of the tuition money. As a voucher, it's almost useless. But it would make it obvious that there were some schools that were underserving their children repeatedly, year after year. That's the idea of the Title I voucher: to expose schools that are failing their children repeatedly, and to say that their parents should be given some option.

You don't think it's going to lead to any massive shift out of public schools ?

No, there's not enough money. But it could embarrass some very bad schools, and I think that's an important thing to do. Some schools deserve to have someone say, "It's terrible that you're taking this much money and that kids are not learning how to read."

Before we conclude, tell us about your forthcoming book.

The book that I edited is The Economics of School Choice, due out from University of Chicago Press this summer [2001]. I wrote two of the 16 essays in it, and the others are by leading economists in the field. It will be the book on school-choice reform for education policymakers.

What are your next projects?

I'm working on an ideal voucher program, and I'm also beginning a study on peer effects, the effect on students of having diversity in the student body -- that is, ability grouping versus mixed-ability classes, ethnic, racial, and economic diversity, and so forth.

Sybil Eakin, a resident of Bloomington, IN, is Contributing Editor of TECHNOS. See her profile of John Dewey in the Winter 2000 issue.

COPYRIGHT 2001 Agency for Instructional Technology
COPYRIGHT 2001 Gale Group