top of page

Be Notified of New Research Summaries -

It's Free!

Do Charter Schools Draw Students Away From Catholic Schools?

Writer: Greg Thorson
Greg Thorson
11 minutes ago
6 min read


Dougherty et al. (2026) examined whether new charter schools draw students away from nearby Catholic schools. They analyzed national data from 1998–99 through 2019–20 on about 8,400 Catholic elementary and middle (K–8) schools and 1,800 Catholic high schools. They found that the first charter opening within 5 miles reduced K–8 Catholic enrollment by about 10 students, or 3.5%, by the second year, and by about 50 students after eight years. It also raised the risk of closure within five years by more than 4 percentage points. Charter competition may account for about 17% of the national decline in Catholic enrollment. Catholic high schools showed no significant change.


Why This Article Was Selected for The Policy Scientist

School choice has expanded substantially over the past three decades, but research has focused mainly on how charter schools and private school programs affect participating students and traditional public schools. Much less is known about how new public options affect existing private schools, particularly Catholic schools, long the largest source of low-tuition private education in the United States. The question is timely as states continue to expand both charter sectors and private school choice programs, often with the goal of giving families more options. Published in Education Finance and Policy, a respected journal in education policy research, this study provides national causal evidence on the question, and one of its authors has also studied how magnet school openings affect nearby schools. Two decades of national data and a staggered difference-in-differences design, supported by a placebo test, provide a credible basis for causal inference.


Full Citation and Link to Article

Dougherty, S. M., Miller, A., & Yoon, Y. (2026). Charter school expansion, Catholic school enrollment, and the implications for school choice. Education Finance and Policy. Advance online publication.


Central Research Question

Does the opening of a nearby charter school reduce enrollment at Catholic schools, increase the likelihood that they close, and change the racial and ethnic composition of their students? The study also asks whether these effects differ between elementary and high schools and across states with different charter school and voucher policies.


Catholic schools enrolled more than 5 million students in the 1960s and 1970s and now enroll just under 1.7 million, with enrollment falling by more than 30% over the past 20 years. Over the same period, charter schools grew to more than 7,000 schools serving more than 6% of public school students. Catholic schools have historically been the largest low-cost private alternative to public schools: in 2023, average Catholic elementary tuition was $6,800, compared with $19,000 at nonsectarian private schools. Tuition-free charter schools may compete for many of the same families. The question bears on whether expanding public school choice adds to the options available to families or partly replaces existing ones.


Previous Literature

A substantial body of research finds that charter school expansion reduces enrollment in nearby traditional public schools. Slungaard-Mumma (2022) documents such declines in Massachusetts and North Carolina, as do Hicks and Lens (2022) in Los Angeles and Winters (2012) in New York City. Evidence on segregation is more mixed. Alcaino and Jennings (2020) find that charter growth increases racial and socioeconomic segregation at the county level, while Monarrez et al. (2022) find that charter schools modestly increase school segregation overall but reduce segregation between districts in metropolitan areas.


Research on charter schools and private schools is smaller and less consistent. Using Michigan data, Chakrabarti and Roy (2016) find an association between charter schools and lower private school enrollment, with no difference between Catholic and other private schools. In contrast, Waddington (2012), studying ten large Great Lakes cities, finds that nearby charter schools accelerated Catholic enrollment declines and closures, particularly in elementary grades. Lackman (2013) documents a 34% fall in Catholic enrollment in New York State in the decade after charter schools were introduced, and Buddin (2012) finds with national data that charter schools drew students from private schools, especially Catholic and urban schools. The present study extends this literature by providing national causal evidence on how charter openings affect Catholic school enrollment, closures, and racial composition.


Data

The study uses the National Longitudinal School Database, which compiles school-level data on public and private schools, including Catholic school data from the federal Private School Survey, for the 1998–99 through 2019–20 school years. The authors geocoded Catholic school addresses and measured their distance to 9,027 charter K–8 schools and 6,565 charter high schools. The final sample includes 8,369 Catholic K–8 schools and 1,784 Catholic high schools, analyzed separately because charter availability and Catholic tuition differ substantially by level. Of the K–8 schools, 2,818 experienced a first nearby charter opening during the period. Before that opening, they enrolled an average of 286 students, of whom 68% were white, 10% Black, and 15% Hispanic.


Outcomes include total enrollment, closure, the shares of white, Black, and Hispanic students, and an index of racial diversity. Because Catholic schools are not required to complete the Private School Survey, the authors use a conservative closure measure that treats a school as open unless its closure was verified. Controls include county characteristics from the 2000 Census and the Catholic share of each county's population.


Methods

The study employs a staggered difference-in-differences design, comparing changes at Catholic schools after the first charter school opens within 5 miles with changes at Catholic schools that have not yet experienced such an opening. Because openings occurred in different years, the analysis uses the Callaway and Sant'Anna (2021) estimator, which avoids biases that conventional two-way fixed effects models can produce when effects differ across groups treated at different times. Event-study models trace effects over time.


The authors test the results against distances from 1 to 20 miles, a panel limited to 1998–2002 to separate charter effects from the clergy sex-abuse revelations of 2002 and 2003, controls for nearby public school enrollment and composition, and alternative closure models. To test whether charter schools chose locations near already declining Catholic schools, they use proposed rather than actual charter sites in Massachusetts and North Carolina. Proposed sites show no enrollment effect, while actual openings do.


The design assumes that, without a charter opening, affected and not-yet-affected Catholic schools would have followed similar trends. Enrollment estimates before openings are small and statistically insignificant, which supports this assumption, but closure cannot be checked the same way because closed schools leave the data. The placebo test covers only two states, and the authors note that diocesan closure decisions may be related to charter locations in ways the data cannot test.


Findings/Size Effects

The principal finding is that a new charter school within 5 miles reduces enrollment at nearby Catholic K–8 schools. Enrollment fell by about 10 students, or 3.5%, by the second year after the opening. Averaged over the opening year and the following ten years, the decline was 6.8 students, about 2.4% of the pre-opening average, significant at the 0.1% level. The losses accumulated: a school exposed for eight years lost about 50 students. Across the 2,818 affected schools, this implies a loss of roughly 141,000 students, which the authors estimate is about 17% of the national decline of 824,951 Catholic school students between 1999 and 2019. Catholic high schools showed no statistically significant change.


Charter openings also increased the likelihood that K–8 Catholic schools closed. The risk of closure rose by about 1 percentage point on average in the years after an opening, significant at the 1% level. Over five years, the cumulative risk rose by more than 4 percentage points, implying about 113 additional closures among the affected schools. Alternative models that compare schools within the same county show a similar positive association.


Effects on racial composition were small. The white share of students fell by 0.6 percentage points from a baseline of 68%, significant at the 5% level. The Black and Hispanic shares rose slightly, by 0.2 and 0.4 percentage points, but these changes were not statistically significant, and the diversity index showed no net change. Because a small decline in the white share also appears before openings, the authors describe these results as suggestive.


Effects appeared somewhat larger in states without caps on the number of charter schools and in states with private school voucher programs. In voucher states, K–8 enrollment fell by an average of 12.2 students after an opening, compared with 5.8 students elsewhere. However, the differences between these groups are not statistically significant.


Conclusion

The study provides national evidence that new charter schools reduced enrollment at nearby Catholic K–8 schools and increased their likelihood of closing, with no comparable enrollment effect at Catholic high schools. The effect on an individual school is modest at first but accumulates over time, and for schools that operate on thin margins it appears sufficient to contribute to closures. Charter competition may account for roughly one-sixth of the national decline in Catholic enrollment, leaving most of it to other forces, including the long-term movement of Catholic families from cities to suburbs.


The authors interpret the results as consistent with families choosing tuition-free alternatives, noting that new charter schools have no performance record and that nearby charters were not, on average, higher performing on later test scores. They describe a "paradox of choice": if a new charter school leads a nearby Catholic school to close, the number of options available to families may not increase. They recommend that policymakers consider effects on Catholic K–8 schools when authorizing new charter schools.


The findings reflect an era of mostly small or means-tested voucher programs and may not apply to states with newer universal programs. Because the data describe schools rather than students, the study cannot show where departing students enrolled or how their learning changed. Research tracking student movement among Catholic, traditional public, and charter schools would clarify how choice policies reshape local school systems.

Comments

Rated 0 out of 5 stars.
No ratings yet

Add a rating
Screenshot of Greg Thorson
  • Facebook
  • Twitter
  • LinkedIn


The Policy Scientist

Offering Concise Summaries*
of the
Most Recent, Impactful 
Public Policy Research

*Summaries Powered by ChatGPT

bottom of page