Berends Writes "The Current State of School Choice in the United States" in Phi Delta Kappan

This article first appeared in Phi Delta Kappan.

Over the last 30 years, the school choice movement has been one of the most prominent large-scale reform efforts in American education. Arguably, only the national movement to promote rigorous learning standards and test-based accountability has received more attention and investment from state and federal policy makers (Berends, Primus, & Springer, 2020; Polikoff, 2021).

In 2001, for instance, when the United States Congress authorized the No Child Left Behind Act (NCLB), it specified that if a low-performing public school fails to improve over time, families must be given the choice to enroll their children in another nearby school. During the Great Recession of 2008, President Barack Obama called for an expansion of another form of school choice — the option to enroll children in charter schools — as part of the $4-billion Race to the Top stimulus package (Education Week, 2013). More recently, Betsy DeVos, the U.S. Secretary of Education throughout the Donald Trump administration, made it her top priority to expand school voucher and tuition tax credit programs, which are designed to help parents send their children to private schools. And in early 2020, as the COVID-19 pandemic left the economy in a shambles, Congress passed stimulus packages that provided states with billions of dollars in discretionary funding, which gave governors considerable leeway to invest in school choice programs (Hess, 2020).

However, while Americans have always believed that parents should have some power to decide how they want their children to be educated (Berends, 2021), they can easily become confused by the many and varied forms of school choice that are now under discussion (from vouchers and charter schools to magnet schools, tuition tax credits, inter- and intra-district public school choice, virtual schools, and homeschooling), and by the competing ways advocates have interpreted the existing research into these approaches.

In recent years, two specific school choice models  — charter schools and voucher programs (which are often associated with education tax credits ) — have received a great amount of attention and support from policy makers and philanthropists, and each has been promoted as a means of providing better learning opportunities to students in urban settings that lack high-quality options. But what does the research actually tell us about the effects of these models?

Charter schools

Charter schools are public schools (funded by the government, though they may also receive support from private donors), but they have a different governance structure than traditional public schools, in that they are run by parents, educators, community groups, or private organizations that have been given permission to do so (i.e., a charter). In theory, because charter schools have greater autonomy and flexibility than regular district schools, they are likely to be more innovative and effective. In return for that autonomy, though, they are held accountable to state and federal performance standards and student testing requirements, and if they fail to meet those standards, they can lose their charter and be forced to close. Also, it’s important to note that when a charter school has more applicants than there are seats available, it is required to hold a lottery to determine which students can enroll.

Numbers and enrollments

The first charter school opened in 1992 in Minnesota, shortly after that state passed the nation’s first charter school law. By 2019, there were more than 7,500 such schools across the country, serving more than 3.3 million students, with the highest enrollments in California, Texas, Florida, Arizona, Michigan, and Ohio (National Alliance for Public Charter Schools, 2021). Most of this growth has occurred in the last two decades. In the 2000-01 school year, 1,941 charter schools served 458,664 students; by 2018-19, the number of schools had almost quadrupled, growing to 7,533 schools serving 3,318,836 students. In large part, the expansion has been fueled by bipartisan support for charter schools by policy makers at the federal, state, and local levels.

Just over half of charter schools are located in urban areas, about one-fifth are in suburban locales, and the rest are in rural or small-town areas. Critics have long expressed concern that these schools may be leading to increased racial segregation (Orfield & Frankenberg, 2013), but the picture is complicated. Data from the National Alliance for Public Charter Schools and the U.S. Department of Education’s Common Core of Data show that the percentage of charter public students who are white or Black has declined somewhat over the last two decades, while the percentage of charter students who are Latinx has increased significantly. In 2000-01, for instance, Black students made up 32% of enrollments, but in 2018-19, only 25% were Black. Over the same period, white students declined from 41% to 31% of the total, and Latinx students increased from 19% to 33%. The percentage of charter students who classified themselves as Asian or some other racial/ethnic group increased over time as well (from 3% to 4% for Asian students and from 1.5% to 5% for students classifying themselves as some other racial/ethnic group).

By comparison, white students made up 47% of students enrolled in traditional public schools in 2018-19, while 27% of students were Latinx, 15% were Black, 5% were Asian, and 6% belonged to some other racial/ethnic group. In other words, students enrolled in charter schools were disproportionately Black (25% vs. 15% in traditional public schools) and Latinx (33% vs. 27%), and white students were underrepresented (31% vs. 47%). These compositional differences are due to many social factors within American society, not the least of which is that charter schools are predominately located in urban centers that have disproportionate numbers of students of color attending public schools, whether traditional or charter.

Over time, the percentage of charter school students who are economically disadvantaged has increased as well. In the 2000-01 school year, 29% of charter school students were eligible for free and reduced-price lunch; in 2018-19, the eligibility rate was 58% (compared to a rate of 52% for students attending traditional public schools; Irwin et al., 2021).

Types of charter schools

Not only have the overall numbers of charter schools and students grown, but we’ve also seen rapid growth in the numbers and size of education management and charter management organizations (EMOs and CMOs), which operate much like school districts without borders in the sense that they run a network of charter schools that are located in multiple areas. The main difference between them is that CMOs (such as Green Dot Schools, KIPP, YES Prep, and Aspire) are nonprofit organizations, while EMOs (such as Imagine Schools, Academica, and National Heritage Academies) are for profit.

While EMOs and CMOs have received great amounts of attention from the news media, more than half of all charter schools are not affiliated with either CMOs or EMOs. Still, the growth of these networks is noteworthy. From 2007-08 to 2018-19, for instance, CMOs grew from managing 11.5% of the nation’s charter schools to managing 29% of them, while EMOs increased their share from 10% to 18%. Meanwhile, independent charter schools declined from 78% of the total to 53%.

During the COVID-19 crisis, when schools and districts began to shift from in-person to online instruction, many observers speculated that this might lead to significant increases in charter school enrollments — given their autonomy and flexibility, perhaps they had a competitive advantage in implementing online learning strategies. Yet, at this point, virtual schools make up only about 2% of all charter schools. Moreover, recent research casts doubt on the idea that virtual charter schools are better able to provide effective online instruction (Fitzpatrick et al., 2020).

Voucher programs and education tax credits

In voucher programs (sometime called choice scholarships), parents are allocated public funds that they can use to pay some or all of the cost of sending their children to a school of their choice, whether public or private, religious or nonreligious. Because vouchers tend to be means-tested (i.e., lower-income parents are given priority), they could be viewed as a form of progressive social policy. For the most part, however, efforts to expand voucher programs have come to be identified with free-market economics and efforts to use public funds to pay for religious education, and they continue to be a lightning rod for partisan debates (see Berends, Cannata, & Goldring, 2011). And while many researchers and policy makers expected that careful evaluations of existing programs would settle disputes about vouchers’ effects on student performance and the overall health of public school systems, findings have turned out to be nuanced and mixed (for reviews, see Gegenheimer & Springer, 2020; Waddington, 2020).

Somewhat less familiar than voucher programs, education tax credits go to corporations, individual families, or both. At the corporate level, businesses receive state tax credits when they make donations to nonprofit organizations called scholarship granting organizations (SGOs), which use the donated money to fund private school scholarships for students. At the individual level, parents can receive a tax credit or deduction from state income taxes for approved educational expenses, which typically include private school tuition, books, supplies, computers, tutors, and transportation (though some programs have income restrictions for family eligibility or set the amount families can claim on their taxes). However, because education tax credit programs have emerged relatively recently, researchers have not yet gathered much data on their effects.

The number of school voucher programs has increased dramatically over the last two decades. In 2000, there were just five such programs in operation in school districts and states across the U.S.; by 2010, the number had increased to 12, and by 2021, it had climbed to 29 (EdChoice, 2021). State-sponsored individual tax credit programs increased from five in 2010 to nine in 2021, and education tax credit scholarships have increased from nine in 2010 to 26 today.

Participation in voucher programs

Reflecting an increase in policy efforts to expand voucher programs at the district and state levels, the number of students participating in voucher programs across the nation has increased significantly in the last decade, though the total number of students receiving vouchers remains a tiny fraction of the total number of students in the U.S. (about 0.5%). When the first voucher program began in Milwaukee, Wisconsin, in 1991, 341 students used a voucher. Since then, 16 states have adopted voucher programs, and in 2021, a total of 248,825 students used a voucher (EdChoice, 2021).

Over the past 15 years, several statewide voucher programs have emerged (see Berends, 2021). For instance, Ohio launched its Educational Choice Scholarship Program in 2006 with 3,169 students and now has expanded it to include 34,487 students in 2021. Indiana’s Choice Scholarship Program, one of the largest of its type in the U.S., allows students in low- and middle-income families to receive vouchers to attend private schools. The program began in 2011 with 3,919 students, and today that number has increased to 36,707. Louisiana’s statewide voucher program began in 2009 with 640 students and increased to 6,405 in 2021, though only about one-third of eligible students participate.

What the research says

Supporters of these approaches to school choice have often made the case that they allow low-income students at low-performing schools to move to a higher-performing charter or private school, which will give them better opportunities to learn. But as these programs have expanded over the last two decades, what has the evidence shown? Do students perform better in schools of choice, as measured by their test score gains relative to their counterparts who remained in traditional public schools?

Unfortunately, the research findings have been less than conclusive — “unfortunate” because this provides fodder for continued partisan debate about whether to scale up these approaches. Moreover, the research has been limited in scope, in that it has focused almost exclusively on students’ test-score gains and graduation rates, as though those were the only meaningful indicators of success. To get a clearer picture of the pros and cons of charter schools, voucher programs, and other school choice models, we’ll also need to collect data on a range of other outcomes, such as students’ behavior, engagement, motivation, health, educational and occupational expectations, and long-term success in reaching their goals. In addition, we’ll need to collect more information about the learning environments, teaching conditions, and social organization of the given schools in order to figure out how, exactly, they are similar to and different from one another (see Berends, 2015, 2020).

That said, it’s worth sharing some highlights from the existing research into the effects of charter schools and vouchers on student achievement:

Research on charter schools

Charter schools are the fastest-growing area of school choice, and the evidence base has grown quickly as well, amid efforts by the federal government and states to scale up the approach. Overall, studies have shown mixed effects on achievement — some positive, some negative, and some null (see Austin & Berends, 2018; Betts & Tang, 2019). That mix of findings isn’t all that surprising, though, given the many variations in types of charter schools, operators of charter schools, and kinds of oversight of charter schools in different states (Fitzpatrick et al., 2020).

At the same time, some studies, using randomized research designs, have shown significant, substantial, and positive effects on academic achievement for students who’ve been chosen, via lottery, to enroll in charter schools compared with similar students who did not win an admissions lottery. This has been the case particularly in urban areas where it has long proven to be difficult to implement meaningful educational reforms. For example, comparing students who won or didn’t win lotteries to attend charter schools in the Harlem Children’s Zone, Will Dobbie and Roland Fryer Jr. (2011) found that the benefits of attending charter elementary schools were large enough to close the racial achievement gap across subjects — i.e., students gained about 7 percentile points in both mathematics and English language arts. Large effects of attending charter schools have been found also in studies of students in Boston (Abdulkadiroglu et al., 2011; Angrist et al., 2011).

More recent research has started to provide some information about not just the overall impact of attending particular charter schools, but also about the specific conditions within those schools that may promote student achievement, particularly the kinds of curriculum and instruction that are most likely to affect student learning (see Berends & Dallavis, 2020). In short, researchers have increasingly come to ask not whether this form of school choice itself is beneficial for students but, more specifically, whether particular school models have hit on particularly effective approaches to instruction, assessment, organizational practices, teacher recruitment and development, and so on (Berends, 2020; Oberfield, 2017). In many ways, this hearkens back to the original vision of charter schools as laboratories of innovation (see Gleason, 2019; Ladd, 2019). Looking at the organizational and instructional conditions of charter schools appears to be a step in this direction, although “innovation” is often difficult to define (Austin & Berends, 2018).

Research on school vouchers

With the expansion in the number of voucher programs, the research addressing the effects of these programs has increased as well. And here, too, findings have been mixed (see Gegenheimer & Springer, 2020; Waddington, 2020). A number of voucher studies at the elementary and middle school levels have focused on programs in specific cities — Milwaukee, Charlotte, Cleveland, Dayton, New York City, and Washington, D.C. — and in general, they’ve found either modest positive effects on student test scores for certain subgroups of students and for certain years of program participation, or they’ve found no effects at all (see Berends, 2021).

However, recent studies of statewide voucher programs in Louisiana, Ohio, and Indiana have shown negative effects on student achievement for students in elementary and middle school grades. Following students who won and lost the lottery for a scholarship, Jonathan Mills & Patrick Wolf (2017) investigated the Louisiana program through its second year. In mathematics in Year 2, they found that students who won the voucher lottery scored 12 percentile points below those students who lost the voucher. “The magnitude of these negative estimates,” the researchers wrote, “is unprecedented in the literature of random assignment evaluations of school voucher programs” (p. 2). Negative effects of state voucher programs have been found in evaluations of the Ohio EdChoice Scholarship Program (Figlio & Karbownik, 2016) and Indiana’s Choice Scholarship Program (Waddington & Berends, 2018) as well.

Refocusing the research

Have the two most prominent approaches to school choice — charters and vouchers — led to the achievement gains their supporters have promised? Unfortunately, the research doesn’t provide a clear answer. Additional research will be critical to understanding the various effects of school choice reforms, particularly as the landscape of choice reforms continues to expand. My hope is that such research can make significant contributions to educational policy debates in this area.

But I’d like to make an important qualification to my call for more research: It’s time we move past the sort of “horse race” measurements that are designed to tell us which students perform best — those who attend charters or those who attend traditional public schools, those who do or do not use vouchers, or those who do or do not benefit from tax credits. As we’ve learned in recent years, research that aims to determine the effectiveness of school choice writ large is unlikely to result in conclusive findings, since schools of choice vary in their quality, design, and many other factors.

More important will be to learn about the conditions under which particular options and school models are effective (see Cohen, 2003). Our goal shouldn’t be to accumulate bits and pieces of evidence about school choice in general but to help inform policy debates by providing more comprehensive and systematic information about the specific resources and educational opportunities that appear to benefit our most underserved and vulnerable students, putting them on course for success in school and beyond.


Note: A more detailed version of this article appears in the Handbook of Urban Education, 2nd Edition, edited by H. Richard Milner IV & K. Lomotey (Routledge, 2021).



Abdulkadiroglu, A., Angrist, J.D., Cohodes, S.R., Dynarski, S.M., Fullerton, J.B., Kane, T.J., & Pathak, P.A. (2011). Accountability and flexibility in public schools: Evidence from Boston’s charters and pilots. Quarterly Journal of Economics, 126 (2), 699-748.

Angrist, J.D., Cohodes, S.R., Dynarski, S.M., Fullerton, J.B., Kane, T.J., Pathak, P.A., & Walters, C.R. (2011). Student achievement in charter schools. Center for Educational Policy Research, Harvard University.

Austin, M. & Berends, M. (2018). School choice and learning opportunities. In B. Schneider (Ed.), Handbook of the sociology of education in the 21st century. Springer.

Berends, M. (2015). Sociology and school choice: What we know after two decades of charter schools. Annual Review of Sociology, 41 (15), 159-180.

Berends, M. (2020). Social perspectives on school choice. In M. Berends, A. Primus., & M.G. Springer (Eds.), Handbook of research on school choice (2nd ed.). Routledge.

Berends, M. (2021). The continuously evolving landscape of school choice in the United States. In H. Richard Milner IV & K. Lomotey (Eds.), Handbook of urban education (2nd ed.). Routledge.

Berends, M., Cannata, M., & Goldring, E.B. (Eds.). (2011). School choice and school improvement. Harvard Education Press.

Berends, M. & Dallavis, J. (2020). School organization in the charter school sector: What we know about what charter schools do. Center for Research on Educational Opportunity.

Berends, M., Primus, A., & Springer, M.G. (Eds.) (2020). Handbook of research on school choice (2nd ed.). Routledge.

Betts, J.R. & Tang, E. (2019). The effect of charter schools on student achievement: A meta-analysis of the literature. In M. Berends, R.J. Waddington, & J. Schoenig (Eds.), School choice at the crossroads: Research perspectives. Routledge.

Cohen, D. (2003). Workshop on understanding and promoting knowledge accumulation in education: Tools and strategies for education research. Washington, DC: National Academy of Sciences.

Dobbie, W. & Fryer, R.G. (2011). Are high quality schools enough to close the achievement gap? Evidence from a social experiment in Harlem. American Economic Journal: Applied Economics, 3 (3), 158-187.

EdChoice. (2021). The ABCs of school choice: The comprehensive guide to every private school choice program in American, 2021 edition. Author.

Education Week. (2013). Obama’s second term: What’s ahead for education policy. Editorial Projects in Education.

Figlio, D.N. & Karbownik, K. (2016). Evaluation of Ohio’s EdChoice Scholarship Program. Fordham Institute.

Fitzpatrick, B., Berends, M., Ferrare, J.J., & Waddington, R.J. (2020). Virtual illusion: Comparing student achievement and teacher characteristics in online and brick-and-mortar charter schools in Indiana. Educational Researcher, 49 (3), 161-175.

Gegenheimer, K. & Springer, M.G. (2020). Voucher outcomes. In M. Berends, A. Primus, & M.G. Springer (Eds.), Handbook of research on school choice (2nd ed., pp. 226-237). Routledge.

Gleason, P. (2019). Let the search continue: Charter schools and the public interest. Point/Counterpoint. Journal of Policy Analysis and Management, 38 (4).

Hess, R. (2020, April 16). What coronavirus and the CARES Act mean for school budgets. Education Week.

Irwin, V., Zhang, J., Wang, X., Hein, S., Wang, K., Roberts, A., York, C., Barmer, A., Bullock Mann, F., Dilig, R., & Parker, S. (2021). Report on the Condition of Education 2021 (NCES 2021-144). U.S. Department of Education, National Center for Education Statistics.

Ladd, H.F. (2019). Experimentation falls short as a justification for more charter schools. Journal of Policy Analysis and Management, 38, 1074-1076.

Mills, J.N. & Wolf, P.J. (2017). Vouchers in the Bayou: The effects of the Louisiana Scholarship Program on student achievement after 2 years. Educational Evaluation and Policy Analysis, 39 (3), 464-484.

National Alliance for Public Charter Schools. (2021). Charter school data dashboard

Oberfield, Z. (2017). Are charters different? Public education, teachers, and the charter school debate. Harvard Education Press.

Orfield, G. & Frankenberg, E. (2013). Educational delusions: Why choice can deepen inequality and how to make schools fair. University of California Press.

Polikoff, M. (2021). Beyond standards: The fragmentation of education governance and the promise of curriculum reform. Harvard Education Press.

Waddington, R.J. (2020). The social context of vouchers. In M. Berends, A. Primus, & M.G. Springer (Eds.), Handbook of research on school choice, (2nd ed., pp. 202-213). Routledge.

Waddington, R.J. & Berends, M. (2018). Impact of the Indiana Scholarship Program: Achievement effects for students in upper elementary and middle school. Journal of Policy Analysis and Management, 37 (4), 783-808.