Q&A: Economist Discusses Study Analyzing Voucher System
Although the issue of allowing parents to choose among public and private schools has provoked a great deal of heat, there have been few empirical analyses that have shed much light on the subject.
In one of the first studies of its kind, Charles F. Manski, a professor of economics at the University of Wisconsin at Madison, used an econometric model to examine the effects, particularly on low-income youths, of a voucher system. Analyzing the assumptions of both choice advocates and critics, Mr. Manski concluded that the nation "should not rush to implement voucher programs.''
While the educational effects of the policy appear neither uniformly positive nor negative, he said, the study also showed that vouchers "would not come close'' to equalizing educational opportunity.
Mr. Manski discussed his study with Associate Editor Robert Rothman.
Q. Why did you choose to conduct a quantitative study of this type?
A. I have been very frustrated looking at the debate on choice. It has been highly rhetorical. As I started reading the basis for the rhetoric, it seemed to me extraordinarily meager. I had been upset for quite awhile about the strong implications drawn from [Politics, Markets, and America's Schools, by John E. Chubb and Terry M. Moe], which has a little bit of analysis in the middle and an incredible polemic at the end. In America 2000, there is a definitive statement by the [Bush] Administration that this is the way to go. Here, unfortunately, as in so much of education, there is so much rhetoric, so little analysis. ...
The original argument for choice is the textbook economic argument. As any economist knows, textbook arguments are valid only under specific conditions. If the conditions don't hold, things get muddier. ...
You can't say economists can tell us private markets work better than public schools. You would flunk any economics exam if you answer like that. The only way to make progress is to stick your neck out and put numbers down.
Q. Could you explain your methodology?
A. The problem is, you want to forecast what would happen if the current system of school finance were to change. The usual way economists work in making a forecast is to model the behavior of the relevant actors involved, put them together, and determine what the outcomes might be.
[One set of] actors are youth and families. They have to decide what school to go to; that depends on cost and the quality of the school. ...
What gets tougher is dealing with the other actors, public and private schools. We don't know much at all about how they behave. There are implicit assumptions [choice] advocates make. To give the advocates the benefit of the doubt, I say el11lprivate schools act like standard competitive firms and set tuition levels to draw as many students as possible.
For public schools, it's really very controversial. That's almost at the heart of the debate about choice. Put somewhat crudely, choice advocates basically assume public schools maximize bureaucratic waste. ... One set of simulations makes that assumption. But we don't have any empirical basis for believing that's the way they behave. [So,] to balance [the picture], I did a set of simulations with the opposite assumption: that public schools behave efficiently.
The purpose of doing the simulations is not to come to sharp conclusions about what choice would do. It [depends on] what we know about the behavior of schools. Maybe it would improve things, maybe it would hurt things. We can't say right now. I don't want to close off the discussion of choice and say it's a bad idea. Rather, [I want] to cool things down.
Q. You found that, if public schools are wasteful, voucher systems encourage them to spend more efficiently. Doesn't that confirm one of the principles of voucher advocates?
A. Yes, it does that, [and it's] certainly not negligible. But there is a mix of results, some of which favor vouchers, and others that don't look so good. ...
One on the negative side is that a voucher system would draw more-motivated students out of public schools. ... That would exacerbate the problems of the public schools.
Q. Your state is experimenting with the first large-scale voucher system by proving funds for low-income youths in Milwaukee to attend private schools. On the basis of your study, do you think that experiment should continue?
A. It should continue. But I think only a limited amount could be learned from that experiment.
The Milwaukee experiment is very small--1,000 kids are eligible for vouchers, in a school system with about 100,000 enrolled students. Think about the critical argument: that the introduction of vouchers would induce public schools to do a better job. If you think about that argument in this experiment, in which the maximum amount the public schools could lose is 1 percent of their enrollment, then basically, the experiment is too small to provide any evidence of how public schools would react. ...
What should be done is a much larger-scale choice experiment. Really, something like taking an entire community--the whole city of Milwaukee--and opening it up. ...
It can't be a short-run experiment, either. ... You have to let it run an absolute minimum of 10 years to really understand what happens. ...
That leads to a larger question: What's the rush? There's always a crisis mentality in education. Things have got to be done right now, and demonstrate results in a very few years. That's totally unrealistic. That leads to bad long-term results. We lurch from one thing to another.