DR. GREENE: I'm glad to be here at JTS.
The question is not "should we have school choice?" We have school choice. We have always had school choice and likely always will have school choice. We just don't call it school choice. The school choice that exists now is choice for people who can afford to either purchase a private-school education or move to an area with a desired public school. This kind of choice has pretty much always been available and will likely always be available. In fact, if you think about residential school choice, which is the most common type of school choice, people choosing where to live, to get access to desired public schools, then essentially, public schools are private schools in the sense that they have a tuition. It is just built into the price of the house and is paid in advance. And so the reality is that there isn't such a stark choice here between having school choice and not having school choice. The real question is more subtle: It's who will have access to choice and under what conditions. And this is an important question because while we do have choice, the barriers to choice are quite high and sometimes too high for people, particularly if they are low-income families. I guess the relevant policy question that we might discuss is should we expand school choice to let more people have a range of options in their schooling, which some people already enjoy. I would answer yes: We should expand school choice. I have four reasons why I think we should do that.
First, I think it's equitable; second, I think it helps the students who get to choose; third, I think it helps the students who don't choose and who remain in traditional assigned public schools; lastly, I think it's a good idea because I think it actually helps promote integration and teaching of the civic values that we expect from schools.
Let me now go through each reason in much more detail. First, equity. If we agree that there already is school choice, that some people have access to it and other people don't, and that it's largely on the basis of income that some people have access to it and some people don't, then I think that this is one of the greatest causes of social inequity that is out there right now. We have lots of other policy areas where we recognize this tension, this difficulty of people being denied access to something desirable in society simply because they lack income. And we have lots of government programs that are specifically designed to make access to that kind of good more available to people even if they are low income. So, for example, we have housing vouchers. We recognize that not everybody has access to desirable housing and so we have a program called Section 8—housing vouchers—where the government helps subsidize the rent for people, so that they have an expanded set of options in their housing. We have food stamps—our vouchers for food—that expand the range of options for people who can't easily buy food on their own. And so we can understand school vouchers as essentially another part of this same social-equity agenda of trying to expand the options available to people who are low income.
The second reason why I think we should expand school choice is because I think that it helps students who get to choose. Now I think we don't want to get too much into the details of evidence, but let me summarize what the evidence says about the effects of school choice on students who get to choose.
Let's focus on eight random-assignment studies, which are the highest-quality studies that have been produced on the effects of school choice on those who get to choose—those who get the voucher. It's important to focus on these high-quality studies because random-assignment studies are like medical experiments: people are randomly assigned to receive a voucher—the pill in a medical experiment—or randomly assigned not to receive the voucher and return to public schools—they get the placebo in a medical experiment. And if chance is the only thing that distinguishes between those who get the pill and those who get the placebo, we reduce the likelihood that differences in students' backgrounds will cause differences in result. We can isolate, more clearly, that the school itself made the difference. And that's why it's so important to focus on these random assignments. There have been eight random-assignment experiments, eight random-assignment studies, in five different locations, five different experiments. Every one of those experiments finds positive results. Seven of the eight are statistically significant positive results. So, in general, when you step back and look at this, you have to say that there are eight high-quality studies that are all finding that students who get vouchers have higher test scores than students who are denied that option. And so you would have to conclude that there are some benefits here.
Now, to be clear, there is a lot of disagreement among the authors of these studies, and it is the nature of academics that we fight with each other. That's what we do for a living. But if we focus too much on those fights, we lose track of the general agreement that there are some benefits and certainly no harm. No one is finding statistically significant negative effects from these eight random-assignment experiments. The disagreements are over whether the benefits exist in reading and math; whether they occur after a long period of time or a short period of time; whether they are only available to students who are African American or students of different backgrounds. These are the kinds of disagreements that are out there. But if we forget those details and step back and look at the pattern of results, we see across-the-board consistent, positive results to the question, do students who get vouchers experience high test scores if they have expanded options.
Then there is the question of public schools: What happens to public schools if we expand educational choice. This is a particularly important question, because we know that no matter what, the vast majority of students are likely to remain in traditional public schools for the foreseeable future, and so it's conceivable that vouchers would be beneficial to the students who receive them but detriment also the students who remain in traditional public schools, which might make the policy undesirable. It might be helping some, but hurting others, and we might decide it's not worth it.
But, even on this question, what is interesting is that we have some pretty good evidence, not in the same confidence, but some pretty good evidence that public schools improve when they are faced with increased competition from expanded choice. I have done some research on this in Florida, San Antonio, and Milwaukee. Helen . . . has done a number of studies along these lines in Milwaukee, in Arizona, in Michigan. She is . . . at Harvard.
But you don't have to take my word for it. Actually, my colleague here, Henry, has done an excellent review of the literature on the effects of competition on school performance, which he did with Clive Belfield. In it, they looked at as much literature as they could find that met certain criteria for acceptable research and reviewed that research on this question. And they concluded that "a sizable majority of these studies will report beneficial effects of competition across all outcomes, with many reporting statistically significant correlations." In particular, they reviewed the question of, does school choice improve test scores at public schools, that is, expanding competition and improving test scores at public schools. They reviewed 206 analyses from twenty-five studies. From those, they found that thirty-eight percent of the analyses produced statistically significant positive results for the effects of competition. If there was no relationship, we would only expect to see about five percent statistically significant positive relationships.
They also looked at the question of whether graduation rates are improved when schools face more competition from choice. They found that of the fifty-two analyses they reviewed from six studies, forty-two percent produced statistically significant positive results. They also looked at the efficiency of public schools in response to competition: Do they produce better outcomes with less money? Are they able to do more with less? These questions might be quite relevant here in New York City, now that you are considering increasing expenditures considerably . . . add taxes considerably as part of the . . . lawsuit.
They found that of the sixty-four analyses they reviewed from thirteen studies that sixty-six percent showed statistically significant positive results on the question of school efficiency, public-school efficiency in response to competition. So rather than drain talent and resources and prevent public schools from improving, it appears that choice and competition helped motivate public schools to use the resources they have more effectively to improve the outcomes for the public-school students that they do have.
Lastly, we have the question of what does choice do to integration and civic engagement, because, as you heard in the introduction, schools weren't just made to produce economically successful people; a system of schools was created, in part, to convey to future generations values that were thought to be essential for the proper functioning of our democracy.
Interestingly, here, private schools have been demonstrating better results, both at producing integration among students and at conveying to students the civic values of tolerance and participation, which we actually expect from public schools and that we have created public schools to do. Private schools are able to pull off better integration, in part because they are unconstrained by politically drawn boundaries. School-district boundaries and school-attendance-owned boundaries often replicate and reinforce racial segregation in housing, and so, in part, our public schools look as racially segregated as they do because they reflect and reinforce racial segregation and housing. Private schools are not constrained by these political boundaries. They are constrained by the real limitations of travel and distance, but they are not constrained by political boundaries, and so they are able to draw students from across these political boundaries and produce a better mix of students from different backgrounds from across those boundaries. They are not wonderful in the integration that they produce, but they are less segregated on average than public schools.
In the study I did, I looked at twelfth-grade classrooms in public and private schools to see how many classrooms had either more than ninety percent white students or more than ninety percent non-white students. These would be segregated classrooms. I found that fifty-five percent of public-school twelfth graders were in classrooms that were racially segregated, compared to forty-one percent of private-school twelfth graders. Again, both are fairly segregated, but you have a lower level of segregation in the private schools than in the public schools.
In another study I did, I looked at lunchrooms and the racial mixing of students in the lunchrooms, and it seems lunchrooms were interesting to look at because lunchrooms reflect students mixing with each other in a social environment. After all, one can force integration in a formal setting, but it might produce the exact opposite of what is desired. It might produce hatred. But a lunchroom is a social environment, and perhaps that is a more consistent or a better measure or read on our social goals from integration. I found that seventy-nine percent of the private-school students that I observed in the two cities that I looked at were sitting in racially mixed groups during lunch, compared to forty-three percent of public-school students. You had a higher likelihood of having a child sit in a racially mixed group during lunch if they went to a private school than if they went to a public school.
Now, on the question of tolerance. There have been a number of studies by myself, by Patrick Wolff of Georgetown University, by David Campbell of the University of Notre Dame, which looked at the question of political tolerance and public and private schools. Social scientists measure tolerance by asking you to name your least-liked group and then ask whether you would let members of your least=liked group engage in certain political activities like marching in your town, running for elected office, giving a speech in your town, and things like that. And the more willing people are to let members of their least-liked group do these things, the more tolerant people are said to be.
Well, as it turns out, people who went to private school or people who go to private school are more likely to report tolerant attitudes; that is, they are more willing to let members of their least-liked groups engage in these political activities than people who attended public schools in a number of these settings.
In the Latino national political survey, this is a very large survey of adult Latinos, we were able to look at where students, where people went every year when they were younger for school. Overwhelmingly the number-one, least-liked group among Latinos in this survey was gays. They disliked gays. I should tell you that in most surveys the number-one disliked group is the KKK. There is basically a convention that we have agreed on which is that we don't like the KKK. But for the Latino survey the number one disliked group was gays.
Overwhelmingly the students, if they attended private school, went to Catholic school. The result of the study was that the more the Latinos went to Catholic school, the more willing they were to let gays run for office, hold demonstrations, give speeches in their town. Now why is this? It seems like it shouldn't be true. Well, maybe Catholic schools might be teaching things along the lines of hate the sin, love the sinner. They might be willing to distinguish between condemnation of behavior and respect for political rights of the people who engage in behaviors that they dislike.
Public schools, on the other hand, might not be teaching the things that we fantasize that they do teach. We did set them up to teach these values, but it doesn't mean that they do so effectively. Perhaps just as private schools appear to be more effective at teaching math than reading, perhaps they are also more effective at teaching other things, including political values.
And so I think that there are a number of good reasons here to believe that we ought to be expanding the range of options available. I think it's equitable, I think it's effective, and I look forward to hearing your comments and the comments of Professor Levin.