President Donald Trump and Secretary Betsy DeVos’s support for school choice, and especially private school vouchers, have brought to the forefront a long-standing debate, one nearly as old as the country. On one hand, opponents argue that schooling is a public good that should be governed by public means, especially the democratic accountability of school districts. Voucher supporters counter that we value individual freedom and that those on the front lines, families and school leaders, should make decisions that they believe are in the best interests of children.
Professor of Economics and Schleider Foundation Chair in Public Education - Tulane University
Nonresident Senior Fellow - The Brookings Institution
I believe this apparent conflict between public and private goals is exaggerated. Specifically, I argue that a “managed competition” approach would place more decisions in the hands of families and school leaders while providing certain roles for government. This approach would best support our values for individual freedom and public goals, such as democratic citizenship, social cohesion, and equitable access to quality schools.1
Schooling is a prime example of market failure
The roots of my argument may be somewhat surprising. Vouchers are often justified on the basis of free market economics, but this represents a simplistic view of the situation. In reality, the economics case against vouchers is stronger than the argument for it.
Economics tells us that markets work best when consumers have good information and many options to choose from, and when the decisions made by one family do not affect others. None of these conditions holds in schooling. Families expect many different things from schools, almost all of which are hard to measure and turn into practical information. We measure math, reading, and other skills with standardized tests because these are the skills we agree schools should provide—and we happen to be able to measure them reasonably well. But measuring creativity, social skills, citizenship, art, and music skills is much difficult.
Unfortunately, geography also stands in the way of free markets. We use most services only occasionally—haircuts, car repairs, and doctor visits. This is important because it allows us to worry less about distance. People often drive hours to get to the best doctor, but we have to get students to school every day, and families want schools to be accessible so they can avoid long commutes, attend sporting events and parent-teacher conferences, and pick up their children when they are sick. Except in high-density urban areas, families will have few options to choose from.
An additional condition for efficient free markets is that transactions between one consumer and producer do not affect other people. This also fails to hold with schools. Again, schooling is inherently public. One of the most widely accepted facts about education is that results depend not only on schools and families, but on peers and classmates. All parents worry about who their children interact with, and for good reason. The problem is that this gives incentives to schools to compete not on instructional quality, but on the types of students they can attract and retain, similar to country clubs. This is not a sound basis for competition. It is a zero-sum game.
Further, in almost no other market do we compel families to use a product or service. This means that, no matter how bad schools are in a given city or town, they will always draw a large number of consumers. Parents have to send their children somewhere. Since school buildings are only so large, most schools do not seek to expand, the better schools will not necessarily expand to meet high demand. This means the market will allow low-performing schools to continue operating.
The concern with markets is not just about efficiency, however. A key public goal of education is to level the playing field and provide opportunity to all children. But due to the fact that families pay so much attention to the kinds of students schools attract, private schools have a strong incentive to use admission requirements. In a free market, then, the most vulnerable students are likely to end up with fewest good options—not only because of geography and limited competition among schools, but because they are discouraged or excluded from entering the schools most successful in attracting strong students.
While thinking of schooling as a market can seem cold-hearted and detached from the caring environments we wish our classrooms to be, there is no denying that schooling is a market and acts like one. Indeed, it is precisely the unusual features of schooling described above that make it hard to think about it as a market. Of course, we should not just rely on theory to understand the problems with the schooling market. The same conclusion arises when we consider the evidence.
The evidence for a scaled-up voucher program is very weak
There is now a large and growing base of evidence about the effects of market-based school reforms, especially charter schools and vouchers, on student outcomes. I start with vouchers, which take a free market approach.
There are so many studies of vouchers that it is not necessary to rely on reviews across studies. In particular, one recent survey of evidence from the University of Arkansas showed that the average effect of U.S.-based voucher programs on student achievement is essentially zero.2
Scaled-up voucher programs like those previously advocated by Secretary DeVos show the worst effects. There have recently been four statewide voucher programs: Florida, Indiana, Louisiana, and Ohio. The Florida study3 is inconclusive, and the others show large negative effects. In some respects, the Louisiana results4 are more convincing because the results have been corroborated by two different sets of researchers and students were assigned to vouchers by lottery—the most rigorous way to evaluate vouchers. In terms of providing convincing results, the Indiana5 and Ohio6 programs, are somewhere in between, but these show negative results as well.
To see large negative effects is worrisome and highly unusual—many programs do not work in education, but few seem to make things worse. And all but one of these statewide program studies were released after the review showed no effect on achievement.
Voucher supporters argue that the results have been worse in the recent statewide programs because they have been “heavily regulated,” by which they mean the requirements that students be tested, that these results be made publicly available, and that schools must let in any student who is eligible for the voucher. The fact that this fairly minimal oversight is considered controversial or heavy-handed, however, only reinforces that private schools are designed to be exclusive and have little interest in external accountability.
Voucher supporters have also argued that the Louisiana and Ohio results look worse than they are because the tests have higher stakes for the traditional public schools in these studies, which serve as the comparison group. In this case, the scores of the public school students in the voucher studies may be inflated and make the results for private schools look worse than they are. This probably explains at least part of these unusual effects, but since earlier studies of vouchers were based on tests more like those used by private schools, this raises questions about why the earlier results that used difference tests still seem to show no effects on achievement.7
Only a handful of studies have examined the long-term effects of vouchers on measures like high school graduation and college entry. While a study of New York City8 shows no average effect on college-going, studies from Milwaukee9 and Washington, D.C.10, do show positive effects. There is not yet any evidence available on the long-term effects of the statewide program like those now being pursued across the country.
The evidence for managed competition is strong—and getting stronger
My argument is that a system of managed competition is much more likely than vouchers and tax credits to produce measureable results for all children. The evidence on charter schools is most relevant to the managed competition argument because charter schools have more government involvement. We also know much more about charter schools than vouchers, and the results are more promising.
While statewide voucher programs generate negative effects, statewide charter programs typically yield positive effects on achievement. The best and broadest-scale study covering lasting consequences in a number of states has also found positive long-term effects on high school graduation and college-going.11
The charter systems in many states started off with somewhat poor results, but there is a clear and consistent pattern of improvement over time.12 This is not especially surprising since charter schools were being created from scratch, while vouchers were being used mainly to send children to long-established private schools. As the charter sector developed, and as the government shut down low-performing charter schools, the results have gotten consistently better.
One of the advantages of managed competition is that it still leaves power in the hands of families, creating competition among schools. Research consistently shows that, whether we are talking about charter schools13 or vouchers14, competition does seem to improve student outcomes in traditional public schools.
We have also learned a great deal about how to make charter policies work better. Here, the case of my hometown of New Orleans becomes especially relevant. I have studied this for many years now through my role as the founding director of the Education Research Alliance for New Orleans at Tulane University.
In the wake of Hurricane Katrina, New Orleans experienced the most sweeping school reforms of any city in the country, if not the world. Our schools are now run almost entirely by charter schools and charter management organizations.
But New Orleans was, for several years after Katrina, perhaps the freest schooling market the country has ever seen. For reasons we might predict from the unusual nature of the schooling market, this period is widely seen as problematic:
- Rather than families choosing schools, schools were choosing students, sometimes called “cream-skimming.”15
- The enrollment process was time-consuming, chaotic, and created considerable uncertainty for families and schools. Schools did not know how many students would show up until just before classes started.16
- Student mobility and discipline incidents increased dramatically.17
- Many of the schools that initially opened soon failed.18
When the government stepped in to solve these problems, schools improved on almost every available metric. Achievement, high school graduation, and college entry continued to get better. Discipline incidents and student mobility, both of which had spiked during the free market period, began to drop when the state stepped in. Government oversight produced better outcomes with fewer unintended consequences than the market alone. This suggest at the very least that the government did not hinder improvement—and still left most of the key decisions to families and school leaders.
Managed competition defined
While I have focused so far on charter schools, this is not meant to be a charter proposal, and certainly I am not arguing that an all-charter system is best. Rather, the unique features of schooling combined with evidence on vouchers and charter schools point toward a general approach—managed competition—that will provide all families with real and better options. Managed competition includes five key elements:
- Accountability: It is important that some agency actively ensure that schools are of high quality, that families have good options to choose from, and that tax dollars are used wisely. With charter schools, this occurs through the charter authorization process, performance-based contracts, and, as a last resort, taking over low-performing schools. As the federal Every Student Succeeds Act recognizes, performance frameworks should be based on a wide variety of measures, not just student test scores.
- Accessibility: Given the unusual features of the schooling market, school choice is no guarantee at all that families will have real options. Broad access requires oversight of enrollment, discipline, and transportation policies to prevent intended, as well as unintended, exclusion of students.
- Transparency: Markets do not function without good information, and markets themselves often fail to provide that on their own. In addition to requiring data gathering and reporting, the government also needs to ensure that charter boards post their meetings publicly, provide audited financial statements, and ensure that publicly funded schools do not fall prey to conflicts of interest.
- Coordination: Some problems that may emerge (e.g., creating a supply of quality teachers) may be more easily solved through coordination by a government body that has some degree of authority.
- Enforcement: The rules and information necessary to make the market function are obviously of little use if they are not enforced. Moreover, the government needs to carry out its most basic function: protecting and enforcing students’ civil rights.
This managed competition approach provides many advantages. It shifts control down to those who are closest to students—their families and school leaders. It helps ensure that families have good options to choose from. It also creates healthy pressures on all schools to improve and meet students’ needs. Paul Hill and others have made similar observations.19
One possible response by those supporting free markets is that the complexity of schooling means that families will have better information than governments. This is partially correct. Families have information that others do not, but the government is arguably the only feasible source for some of the information that parents want, especially if parents want to compare schools as they are making choices. A system of managed competition, which combines both sources of information, is likely to yield the best results, and allow the government to carry out its fiduciary role to hold schools accountable for public funds.
More generally, managed competition provides the advantages of markets while avoiding the disadvantages.
School districts and state governments still have important roles to play
This proposal for managed competition is admittedly vague. There are many variations on the managed competition theme that could be effective. The above functions could be carried out by various combinations of state and local government agencies and nonprofits.
The role of school districts is perhaps most noteworthy given that these agencies still operate the vast majority of publicly funded school for more than a century. I believe districts still have important roles to play. There is a reason that local school districts have continued to operate for more than a century and that most families still like their public schools. They help build local neighborhoods and communities—the public aspect of schooling. They take advantage of economies of scale, which allows them to reduce administrative costs and push more funding into the classroom. There is something to be said as well for local democratic control in deciding the right mix of schools and providers to make available.
This does not mean school districts should be the main agency carrying out the five main functions of government. Having school districts serve as authorizer of independent schools is unlikely to help, given that such schools, such as charters, would then compete with existing traditional public schools. But making district schools available as an option for families helps address many concerns, including the possibility that the market alone will not serve some students well, and that the market leaves families with considerable uncertainty about what schools their children will be able to get into when they move to a new neighborhood. This means that school districts can be an important part of the managed competition mix.
It is difficult to imagine that these various roles for government under managed competition could be carried out at the state or federal levels; there are far too many schools to execute all these functions effectively. This means that some other local government agency would need to perform most of the key tasks. But again, there are many ways to design such systems, and we probably need different approaches in different locations. In many places, where traditional public schools are working well, a district-focused system may make sense. In others, where they are performing poorly, a more significant shift to managed competition will likely make more sense.
Market reform works better in urban locations
There is another important pattern in the evidence that may be more important than anything discussed thus far. While there are no achievement effects on average from vouchers, the more positive effects that have sometimes emerged with longer-term outcomes are all in urban locations: Milwaukee, New York City, and Washington, D.C. Indeed, this is exactly where we would expect a free market to work best, where traditional public schools appear least effective and where population density and transportation networks give more choice to families.
We see the same pattern with charter schools. A nationwide study of charter schools by Stanford researchers finds positive effects in urban areas and negative effects in rural areas, further reinforcing the conclusion that a broad-scale free market program will not succeed.20 Other studies of the effects of private schools (even aside from voucher programs) also suggest that their effects, where they arise, are concentrated in urban areas.21
An important implication of this pattern is that all of these types of market-based reforms, whether charter or voucher, seem to have more limited prospects in suburban and rural locations. This means that, for the vast majority of the country, we will need a different approach, and probably a strong reliance on traditional school districts.
Schooling is a highly unusual market. It is hard to identify other markets with such complex, hard-to-measure outcomes. It is hard to identify other markets where we have to travel to a specific location every day, limiting options and choice. It is hard to identify other markets where the value of the service so dependent on who else is in the building. This is why schooling is one of the clearest imaginable economic cases of market failure. If there is any market that would benefit from a role for government, it is the market for schooling.
The evidence to date should also make us cautious about broad-scale voucher and tax credit involving almost no government oversight. While voucher programs have produced some modest positive results in small-scale pilot programs with low-income and minority children in urban areas, they have failed when taken to scale in the way that some policymakers are now pursuing.
A system of managed competition, with varying designs in different types of locations, can provide the accountability, accessibility, transparency, coordination, and enforcement necessary to make this very unusual market work for all children.
- For a broader discussion of various goals of education and how they apply to schools and school choice, see: Henry M. Levin (2002). A Comprehensive Framework for Evaluating Educational Vouchers. Educational Evaluation and Policy Analysis 24(3): pp. 159-174.
- To be precise, the estimates are slightly positive but not statistically different from zero. See: M. Danish Shakeel, Kaitlin P. Anderson, & Patrick J. Wolf (2016). The Participant Effects of Private School Vouchers across the Globe: A Meta-Analytic and Systematic Review. EDRE Working Paper 2016-07. University of Arkansas. Downloaded January 10, 2017 from: http://www.uaedreform.org/downloads/2016/05/the-participant-effects-of-private-school-vouchers-across-the-globe-a-meta-analytic-and-systematic-review-2.pdf
- David N. Figlio (2011). Evaluation of the Florida Tax Credit Scholarship Program Participation, Compliance and Test Scores in 2009-10. Downloaded January 10, 2017 from: https://www.floridaschoolchoice.org/pdf/FTC_Research_2009-10_report.pdf.
- Jonathan Mills and Patrick Wolf (2016). How Has the Louisiana Voucher Program Affected Students. Tulane University: Education Research Alliance for New Orleans. Downloaded January 20, 2017 from: http://educationresearchalliancenola.org/files/publications/ERA-Policy-Brief-Public-Private-School-Choice-160218.pdf. This research has since been published in the journal Educational Evaluation and Policy Analysis.
- Mark Berends and R. Joseph Waddington (forthcoming). School Choice in Indianapolis: Effects of Charter, Magnet, Private, and Traditional Public Schools. Education Finance and Policy.
- David Figlio (2016). Evaluation of Ohio’s EdChoice Scholarship Program: Selection, Competition, and Performance Effects. Fordham Institute. Downloaded January 20, 2017 from: https://edexcellence.net/publications/evaluation-of-ohio%E2%80%99s-edchoice-scholarship-program-selection-competition-and-performance.
- One possible reason earlier voucher studies find no effects on achievement is that they are not held accountable for them, but this also reflects a lack of transparency even to the families who choose private schools. It is also worth noting that the Florida statewide study found no conclusive evidence of effects even though private schools in that case were allowed to the test of their choice.
- Matthew M. Chingos and Paul Peterson (2013). The Impact of School Vouchers on College Enrollment. Education Next 13(3), 59-64. Downloaded January 10, 201 from: http://educationnext.org/files/ednext_XIII_3_chingos.pdf. This study has been widely misinterpreted because the authors focused on the positive effects of just one sub-group (African Americans). Such interpretations are widely rejected by education researchers because, if you separate students into enough groups, someone is likely to experience a result by chance. Moreover, there were apparently large negative effects on other students that were not emphasized in the study.
- Cowen, Joshua M., David J. Fleming, John F. Witte, Patrick J. Wolf, and Brian Kisida (2013). School Vouchers and Student Attainment: Evidence from a State-Mandated Study of Milwaukee’s Parental Choice Program. Policy Studies Journal 41(1): 147-167.
- Patrick Wolf, Babette Gutmann, Michael Puma, Brian Kisida, Lou Rizzo, Nada Eissa, and Matthew Carr (2010). Evaluation of the DC Opportunity Scholarship Program: Final Report. Downloaded January 20, 2017 from: https://ies.ed.gov/ncee/pubs/20104018/pdf/20104018.pdf.
- Kevin Booker, Tim Sass, Brian Gill and Ron Zimmer (2011). The Effects of Charter High Schools on Educational Attainment. Journal of Labor Economics 29: 377-415.
- This can be seen in two ways. First, a study in Texas documents improvements over time in charter performance within the state: Patrick L. Baude, Marcus Casey, Eric A. Hanushek, and Steven G. Rivkin (2014). The Evolution of Charter School Quality. Downloaded January 20, 2017 from: http://harris.uchicago.edu/sites/default/files/Rivkin.paper_.pdf. The same conclusion can be seen by comparing the periodic CREDO studies over time. The CREDO studies include the majority of the nation’s charter schools.
- Julian Betts (2010). The Competitive Effects of Charter Schools on Traditional Public Schools. In Mark Berrends, Matthew Springer, Dale Ballou, and Herbert Walberg, Handbook of Research on School Choice (pp.195-208). New York Routledge.
- David Figlio (2016). Ibid.
- Huriya Jabbar (2015). How Do School Leaders Respond to Competition? Tulane University: Education Research Alliance for New Orleans. Downloaded January 10, 2017 from: http://educationresearchalliancenola.org/publications/how-do-school-leaders-respond-to-competition.
- Harris, D., Valant, J., & Gross, B. (2015). The New Orleans OneApp. Education Next. http://educationnext.org/new-orleans-oneapp/
- These results are based on internal calculations that have not been released.
- Whitney Bross and Douglas N. Harris (2016). The Effects of Performance-Based School Closure on Charter Takeover on Student Performance. Tulane University: Education Research Alliance for New Orleans. Downloaded January 20, 2017 from: http://educationresearchalliancenola.org/files/publications/Bross-Harris-Liu-The-Effects-of-Performance-Based-School-Closure-and-Charter-Takeover-on-Student-Performance.pdf.
- The managed competition approach I am proposing is similar in many ways to Hill’s portfolio model—indeed, the New Orleans system closely resembles his recommendations. What I am proposing is less specific, places greater emphasis on the potentially positive role for school districts, even in urban areas. See, for example: Paul Hill and Christine Campbell (2011). Growing Number of Districts Seek Bold Change With Portfolio Strategy. University of Washington: Center for Reinventing Public Education.
- This conclusion is based on the author’s compilation of results across state-specific studies by the Center for Research on Educational Outcomes at Stanford University.
- Derek Neal (1997). The Effects of Catholic Secondary Schooling on Educational Achievement. Journal of Labor Economics 15(1), 98-123.