School choice and charter schools improve outcomes for all students

The claim

School choice advocates argue that introducing market competition into public education will function as an accountability mechanism. When parents can exit failing schools and take their per-pupil funding to charter schools or private schools via vouchers, traditional public schools face competitive pressure to improve. In this framing, choice benefits everyone: the students who switch gain access to better schools, and the students who remain in public schools benefit as those schools reform in response to the threat of further exit. The mechanism is borrowed from consumer markets: competition raises quality for all participants, not just those who switch providers.

The mechanism

The competitive-pressure theory rests on two claims that can be tested separately: first, that charter schools and voucher-receiving private schools produce better outcomes than the public schools students would otherwise attend; second, that the presence of choice options causes improvement in the public schools students stay in.

The first claim — charter/voucher superiority — is an empirical question with a large evidentiary record. The second — competitive spillover — requires that public schools respond strategically to enrollment loss, which in turn requires that those schools have the organizational capacity, financial flexibility, and management autonomy to change in response to market signals. When enrollment falls, public schools lose funding but not necessarily fixed costs (buildings, administrators, tenured staff), which can concentrate disadvantage rather than stimulate improvement.

The full competitive logic also implicitly assumes that the entering schools will generally be better than the schools they displace or compete with. If charter quality is highly variable — as the evidence shows — then the competitive pressure mechanism may produce improvement at the top and deterioration in aggregate, depending on the distribution of new entrants.

The evidence

Charter school quality: the CREDO record

The Center for Research on Education Outcomes (CREDO) at Stanford University has produced the most comprehensive quasi-experimental assessments of charter school quality using a virtual control record methodology that matches charter lottery participants to demographically similar public school students. The 2023 National Charter School Study III, covering 41 states and roughly 1.6 million charter students, found that 26% of charter schools produce significantly better academic growth than matched traditional public schools in math, 29% produce significantly worse outcomes, and 45% are statistically indistinguishable. In reading, the distribution is similar. The average effect across all charters is modestly positive, but this average obscures enormous variance and masks the substantial share of charters that are producing worse outcomes than the public schools students would have attended.

Earlier CREDO studies (2009, 2013) found aggregate charter effects that were neutral to negative, with improvement over time concentrated in specific urban markets — notably Boston, Denver, New Orleans, and Newark — where strong authorizers closed failing charters and maintained quality control. The positive aggregate shift between 2013 and 2023 is largely explained by these high-accountability urban markets, not by the broader national expansion.

Voucher programs: the randomized evidence

Voucher programs are ideally suited to randomized evaluation because oversubscribed programs assign vouchers by lottery, producing clean treatment and control groups. The results from the two largest recent randomized evaluations are discouraging for the competitive claim.

The Louisiana Scholarship Program (LSP), enacted in 2012, provided vouchers to low-income students to attend private schools. Abdulkadiroğlu, Pathak, Walters, and Cohodes (2017) analyzed the program using lottery-based identification and found that voucher recipients scored 8 percentile points lower in math after one year than comparable students who remained in public schools. Reading effects were also negative, though smaller. The mechanism was straightforward: the private schools accepting vouchers were predominantly low-quality religious schools that had excess capacity because better-informed parents had already avoided them.

The Indiana Choice Scholarship Program evaluation by Waddington and Berends (2018) found negative math effects in the first year (approximately -0.15 standard deviations) with partial recovery by year three, but no significant positive effect relative to the control group by the end of the study period. Both programs served predominantly low-income students in states with weak regulatory oversight of voucher-receiving schools.

A partial exception is the District of Columbia Opportunity Scholarship Program, where some analyses find modest positive graduation rate effects. But the DC program serves a smaller population in a high-cost urban context with more selective participating private schools, and its generalizability is limited.

Michigan and Detroit: a natural experiment in weak authorization

Michigan’s 1994 charter law and its 2011 expansion represent one of the most studied cases of large-scale charter expansion under weak authorizer oversight. Michigan allows for-profit charter management, permits multiple competing authorizers (including universities with financial incentives to approve charters), and historically lacked meaningful closure mechanisms for low-performing schools. By the early 2020s, charter schools enrolled roughly 55% of Detroit’s K–12 students — the highest market penetration of any major US city.

The NAEP Trial Urban District Assessment placed Detroit last or near-last among large urban districts in 4th-grade reading in 2022 and in most prior assessment years. A 2016 New York Times investigation and subsequent academic analyses found that Detroit’s choice landscape had produced a fragmented system with neither sector — charter nor traditional public — holding clear superiority, while both competed for declining enrollment in a depopulating city with concentrated poverty. The competitive pressure hypothesis predicted improvement; the evidence shows stagnation in a high-choice context without meaningful quality control.

The competitive spillover: what happens to students who stay

The most direct test of the competitive claim is whether non-choosing students in traditional public schools improve when faced with charter competition. Hoxby (2003) found positive effects in Michigan using a geographic regression discontinuity design. Subsequent research has been far more skeptical. Bifulco and Ladd (2006) using North Carolina data and Sass (2006) using Florida data found weak or null competition effects. Bettinger (2005) found no consistent evidence of competitive spillover in Michigan. Epple, Romano, and Urquiola (2017) review of the broader competition literature concludes that evidence for positive spillovers to public schools is “fragile and inconsistent.” The mechanism requires public schools to have the organizational capacity to respond to market pressure, which is often absent, particularly in high-poverty contexts where departure of motivated families and concentrated disadvantage may actually worsen conditions for remaining students.

Cream-skimming and selection effects

Charter and voucher programs do not draw randomly from the public school population. Families with higher social capital, more information access, and greater motivation to navigate application processes are systematically over-represented among choosers. This means the comparison group left in traditional public schools is, on average, harder to educate — more likely to have disabilities, English language learning needs, or extreme poverty — while the schools serving them lose per-pupil funding. Several studies document that charter schools, particularly those in unregulated contexts, disproportionately counsel out or fail to re-enroll students with high service needs. Lacireno-Paquet et al. (2002) and Zimmer et al. (2009) document these patterns in multiple states.

Contexts where choice works

The evidence is not uniformly negative. Urban charter sectors in Massachusetts (particularly Boston), New York City, and New Orleans post-Katrina have produced positive lottery-based effects for participants, particularly low-income students of color. What distinguishes these contexts: strong authorizers with meaningful closure authority, non-profit CMO dominance, transparent performance reporting, and integration with district planning rather than adversarial displacement. The evidence suggests that well-regulated charter sectors in dense urban markets can produce real gains for a portion of students — but this is a far narrower and more contingent claim than the general competitive-improvement hypothesis.

Cross-national context

Finland, consistently among the top PISA performers, operates no private voucher system and has a negligible charter-equivalent sector. Its between-school variance in outcomes is among the lowest in the OECD — meaning school quality is relatively equalized, which is the stated goal of choice advocates achieved through an entirely different mechanism: equal funding, equalized teacher quality, and strong professional standards. South Korea and Canada, both high-performing, likewise do not attribute their outcomes to market competition in schooling. The Netherlands operates a long-standing voucher system but with tight national curriculum standards and funding equity requirements that bear little resemblance to the deregulated US models. No high-performing national system credits market competition among schools as a primary quality driver.

Who benefits

The political economy of school choice involves specific organized interests. The Walton Family Foundation has invested over $1 billion in charter sector development and choice advocacy. The DeVos family network — including the American Federation for Children — has funded state-level voucher legislation for decades. Arnold Ventures funds both pro-choice policy research and advocacy organizations. For-profit charter management companies (e.g., National Heritage Academies, Charter Schools USA, Imagine Schools) earn operating margins from public per-pupil funding. Real estate investors in cities like Chicago and New York have documented financial interests in school-driven neighborhood change.

These interests are not monolithic — some funders support only high-quality, regulated charters — but the political infrastructure for weak-oversight choice expansion is substantially funded by entities with financial or ideological stakes in reducing the scope of traditional public education.

The counter

The strongest steelman for school choice focuses on the students trapped in genuinely failing schools who have no exit option. In cities where low-income students are assigned to chronically underperforming schools with no alternatives, the moral case for providing exit options is powerful and not adequately answered by pointing to aggregate evidence. The lottery-based evidence from high-quality urban charter sectors shows that individual students do gain real academic benefits — effects equivalent to an additional year of schooling have been documented in the Boston and NYC studies. The aggregate competitive-spillover evidence being weak does not negate the individual-level benefit to those students.

The individual-choice claim is also more defensible than the system-level competitive claim: families with information and options can identify better schools for their children, and extending that capacity to low-income families who previously had none is a genuine equity argument. The problem is that this individual-level logic has been used to justify system-level expansion without adequate quality controls, producing a much weaker — and sometimes negative — aggregate result. The contested evidence reflects a real tension: well-designed choice programs in accountable contexts can produce genuine gains; poorly designed ones produce fragmentation without improvement.

References

Abdulkadiroğlu, A., Pathak, P. A., Walters, C. R., & Cohodes, S. R. (2017). Inputs and impacts in charter schools: KIPP Lynn. American Economic Journal: Applied Economics, 9(3), 1–36. https://doi.org/10.1257/app.20160036

Abdulkadiroğlu, A., Pathak, P. A., & Walters, C. R. (2018). Free to choose: Can school choice reduce student achievement? American Economic Journal: Applied Economics, 10(1), 175–206. https://doi.org/10.1257/app.20160634

CREDO. (2023). National charter school study III. Center for Research on Education Outcomes, Stanford University. https://credo.stanford.edu/

Epple, D., Romano, R., & Urquiola, M. (2017). School vouchers: A survey of the economics literature. Journal of Economic Literature, 55(2), 441–492. https://doi.org/10.1257/jel.20150679

Hoxby, C. M. (2003). School choice and school productivity: Could school choice be a tide that lifts all boats? In C. M. Hoxby (Ed.), The economics of school choice (pp. 287–341). University of Chicago Press.

OECD. (2023). PISA 2022 results: The state of learning and equity in education (Vol. I). OECD Publishing. https://doi.org/10.1787/53f23881-en

Waddington, R. J., & Berends, M. (2018). Impact of the Indiana Choice Scholarship Program: Achievement effects for students in upper elementary and middle school. Journal of Policy Analysis and Management, 37(4), 783–808. https://doi.org/10.1002/pam.22086

Zimmer, R., Gill, B., Booker, K., Lavertu, S., Sass, T. R., & Witte, J. (2009). Charter schools in eight states: Effects on achievement, attainment, integration, and competition. RAND Corporation. https://www.rand.org/pubs/monographs/MG869.html