New Research: Charter Schools and Segregation in Pennsylvania

New research published last week looks at the effect of charter school choice on racial segregation in Pennsylvania. Although the findings are not particularly encouraging, the research adds valuable new information to our understanding of charter school segregation.

The article (free, full text) was published in education policy analysis archives, an open-access online journal that has a strong academic reputation, especially in addressing policy issues of public importance. The research team was led by Erica Frankenberg, who has done a lot of research in school choice and school segregation. Here’s maybe the most important finding:

  • “Black and Latino students experience particularly large increases in racial isolation when transferring to charter schools.”

Not good. And, when you dig into it, it doesn’t get much better. This is all based on analysis of over 8,000 individual students from the state’s 10 largest cities who transferred from a traditional public school to a brick-and-mortar (i.e., not cyber) charter school between the 2010-11 and 2011-12 school years. In reviewing student transfers, this article does something different: it looks at whether there were diverse options available in the first place. Previous research has found that charters increase segregation, but hasn’t really addressed whether this is because parents were choosing same-race schools or because these were the only kinds of schools available within a reasonable distance. This study finds that it is actually a matter of choice.

So, first, here’s a breakdown of the transfers:

  • Black students – On average, transferred from a traditional public school with 68% Black students to a charter with 80% Black students.
  • Latino/a students – On average, transferred from a traditional public school with 41% Latino/a students to a charter with 53% Latino/a students.
    • In both cases, the increase in segregation was 12% points.
  • White students – Results were mixed: K-7 students moved to schools that were slightly less segregated, while students in grades 8-12 moved to schools that were slightly more segregated.

Using a 10-mile radius around each individual student, the article then organizes charter options according to their diversity (% White) in quintiles of 20% (e.g., 0-20% White students is one interval, 20-40% is the next, etc). (And, it feels like this is a good point to briefly salute the complexity of doing this for over 8,000 individual students.) Overall, “students of all races had charter schools of varying racial composition in their choice set.” The breakdown is interesting:

  • Compared to White students, higher percentages of Black and Latino/a students had choices in all five intervals of diversity. For example, about 95% of K-7 Black students and 91% of K-7 Latino/a students had charter school choices in the 60-80% White interval, compared to 81% of White students.
  • This held true for all intervals, across all 10 cities, with one interesting exception: At the high school level, fewer Black and Latino/a student had choices available in the most diverse quintile (40-60% White). Shockingly, just 0.32% of Black students and 1% of Latino/a students had high school choices in this interval, compared against 13% of White students.

The authors then try to understand why parents have chosen same-race schools when more diverse and/or opposite-race choice options exist. They look a three factors in particular: the diversity intervals mentioned above, distance, and overall school enrollment, using these to develop a model that would predict transfer decisions. The findings here illustrate what we’re up against when we’re talking about school choice and school segregation:

  • “Black students were about 99% less likely to attend a charter school with between 60-100% White students when compared to a charter school with 0-20% White students.”
  • “A Latino child was between 95 and 99% less likely to enroll in any of the choices where more than a fifth of the students were White.”
  • “The odds of a White student enrolling in a charter school with 80-100% other Whites is six times as high as that of the odds of them enrolling in a school with 0-20% Whites.”

Although distance is a factor here, it’s not the only factor – families didn’t simply choose the charter closest to them. In some cases, they opted for further distances, but still stuck to same-race schools.

There’s a lot to say about why this is the case. Although it’s not their primary purpose, the authors suggest a few plausible explanations:

  • Charter schools may not actively recruit for diversity. (See here for one exception, from New Orleans.)
  • Parents may not have access to information about all the charter school options that are available to them.

Although not noted in detail in the article, I think there’s a caveat here: for high school students, segregation may have been due to the limited choices available. Only very small percentages of Black and Latino/a families had high school options in the most diverse quintile (40-60% White), as noted above.

There’s also a difficult to reality to face: parents may have had choices, had the info they need and still deliberately choose a same-race school. Although this may be disheartening, it shouldn’t be surprising. There are many, many possible explanations – everything from the enduring racial fears of White families to risks that Black and Latino/a families face in sending their children to majority White schools. As a researcher, I almost always feel like more research is needed on whatever topic I’m writing about, though I try to avoid using that as a conclusion. But, this is definitely a situation where that’s the case – especially (as noted by the authors) qualitative research that seeks to understand the “why” behind parent choices. Interviews with parents and charter school leaders, reviews of charter school recruitment practices, policies about racial diversity, etc. I would love to do this.

Nonetheless, this study adds to a growing body of research that demonstrates strong connections between charter expansion and school segregation. With this now becoming more firmly established, it really shouldn’t be possible for charter advocates (however benevolent their intentions) to make their case without at least addressing concerns about segregation. In the short term, there are policies that can help stem these trends – as noted in the article, “weighted lotteries,” for example, can help “prioritize certain groups in charter school admissions that would enhance diversity.”

These kinds of policies are important, but it feels like they just tinker around the edges. The larger problems are the ones that are silhouetted by the numbers here – the risks (perceived and real) that parents navigate in their decisions about where to send their children to school. It highlights the important work of a network that I became connected to this week: Integrated Schools, which understands that policy change may not accomplish all that much before hearts, minds and public discourse begin to change. They work through grassroots networking to talk with parents, one conversation at a time, to understand barriers to integration. It’s hard work, but I guess I don’t see any other way.

5 thoughts on “New Research: Charter Schools and Segregation in Pennsylvania

  1. Pingback: News Roundup- Late July: Local Stories, Housing Policy and School Choice | School Desegregation Notebook

  2. Pingback: News Roundup- Late July: Local Stories, Housing Policy and School Choice –

  3. Pingback: SD Notebook is now a part of NCSD! | School Desegregation Notebook

  4. Pingback: Research Roundup, Feb 2018: More data on charters and school re-segregation | School Desegregation Notebook

  5. Pingback: Research Roundup, Feb 2018: More data on charters and school re-segregation –

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s