A new study of North Carolina public schools finds that black students in charter schools are more likely to have black teachers than their regular public school counterparts, and that the positive effect of “teacher/student racial match” on the test scores of black students is more pronounced in charter than in regular public schools.

Like most good analyses of charter and regular public schools, this report, written by economist Seth Gershenson and published by the Fordham Institute, is an opportunity to learn from the comparison between the charter and regular public school sectors. For instance, the fact that the “match effect,” which is fairly well-established in the literature (e.g., Dee 2005), is stronger in charter schools is fascinating, though a well-informed discussion of the reasons why this may be the case is well outside of my rather modest wheelhouse (there are some possibilities mentioned in the paper’s conclusion). 

I’d actually like to focus briefly on the first finding – that teacher/student racial match is more common for black charter school students. This is the descriptive and arguably less interesting part of the analysis, but it struck me because, like the paper's main finding about the magnitude of the "match effect," it too raises policy-relevant questions, in this case about why teacher diversity might vary between sectors.

Specifically, Gershenson reports that 35 percent of black charter school students had at least one black teacher between 2007 and 2012, compared with 22 percent of regular public school students. 

This may be due in part to within-school assignments (i.e., charter schools may be more likely to assign black students to black teachers). It may also arise if black charter school teachers teach in schools with larger shares of black students than black regular public school teachers (though North Carolina charters overall, unlike those in most states, serve larger proportions of white and non-FRL-eligible students than do regular public schools in the state). 

But the primary reason why racial match for black students is more common in charter schools is most likely the simple fact that charter schools employ more black teachers (as a proportion) than regular public schools, and so black charter school students are more likely to have black teachers.

The differences here are not enormous in absolute terms – roughly 14 percent of charter school teachers statewide are black, compared with about 10 percent of regular public school teachers – but they are certainly large enough to affect the frequency of racial match, at least for black students (white students were no more likely to have white teachers in charter schools than their peers in regular public schools).

Before discussing why this might be the case, it bears mentioning that, despite a couple of news article headlines implying that this is a nationwide phenomenon, it is possible that North Carolina is among the exceptions here. The evidence is admittedly partial, but, for instance, in our 2015 teacher diversity study, which presented detailed data for nine large U.S. cities, regular public schools employed a larger share of black teachers, and a smaller share of white teachers, than did charter schools in five of six cities with available data (the exception was NYC’s small [at the time] charter sector, where black teacher representation was three percentage points higher). There were also well-documented and large decreases in black teacher representation that coincided with charter proliferation in D.C. and New Orleans. 

If you know of any other breakdowns of charter versus regular public school teacher diversity, please leave a comment. Note that national comparisons are not really meaningful in this context. Across the entire U.S., charter school teachers are more diverse than regular public school teachers. Of course they are – charters are concentrated in cities. Meaningful inter-sectoral comparisons really should be made on a location-by-location (or even neighborhood-by-neighborhood) basis, in much the same way that studies of charter versus regular public school test score effects compare schools in the same districts or neighborhoods. State-level comparisons are only interesting in states, such as North Carolina, where the charter and regular public school sectors seem to be relatively comparable in terms of the characteristics of the areas in which they operate, and even then findings should be viewed with caution.

Again, our diversity report is hardly conclusive generalizable evidence on charter versus regular public school teacher diversity (and most of our datasets ended around 2010), but we can actually ask the big question generically: if we found that charter schools in a given location employed a more diverse teacher workforce than nearby public schools, why might this be the case (and what can this teach us about improving teacher diversity in all schools)?

We’ll start with the least satisfying possible answer – it might not have much at all to do with school policies or practices or governance structures. For example, it may simply be the case that charter schools are disproportionately located in labor markets with larger minority populations. This is a particular salient possibility when making aggregate (e.g., state-level) comparisons, as in North Carolina, or, perhaps, in large cities (especially those with smaller charter sectors, where a few schools might sway the overall averages). If, for instance, charters in a given city (or state) were concentrated in certain areas, and those areas were disproportionately black (or white), we might expect to find at least a somewhat larger share of black (or white) teachers working in charters due to little more than location. People tend to live closer to work.

But it is also more than plausible that concrete policies and practices matter, and could help explain any observed differences between sectors. This in turn could provide insight into improving teacher diversity in both sectors, since we’d have diversity variation in the same job, industry, and labor market. Put differently, if the same labor pool is applying for the same type of jobs in the same type of institution in the same place, why would one subset of employers exhibit more diversity than another?

Gershenson raises one possible explanation in the North Carolina context: the fact that charter schools in the state are allowed to hire non-certified teachers. The idea here is that passing rates on certification tests are lower for black teaching candidates, which may be excluding many potentially qualified black teachers from jobs in regular public schools. 

This is a compelling possible explanation, but if this were the major factor, we might expect to find a tendency toward overrepresentation of black teachers in charter schools in other states with laws permitting the hiring of non-certified teachers, all else being equal. The (admittedly scarce) evidence available doesn't support this possibility, though even if we had a large set of race and ethnicity distributions, by sector, all else is not equal – simple bivariate relationships could be misleading (the laws might boost diversity but not enough to overcome all the other factors that have an impact). In any case, it is possible that relaxing certification requirements might improve teacher diversity, but this outcome would of course have to be weighed against the possible costs in terms of teacher qualifications.

Similarly, I’ve heard charter opponents chalk up anecdotes about the lack of diversity among charter school teachers to their reliance on Teach for America and other alternative certification programs (which have been criticized for hiring disproportionately white cohorts of new teachers). This too would be a strong possible explanation if it were true, but I’m not aware of any systematic evidence supporting it (and TFA, to its credit, has reported improvements in the diversity of its teachers).

There may also be more informal institutional mechanisms at work here. For example, state-level oversight or authorizing entities may emphasize and monitor teacher diversity as an expectation, which could generate statewide improvement within charter sectors. But again, this is just a possibility, and such “pressure” is no less likely to be exerted similarly on regular public schools.

Beyond that, to the degree concrete policies and practices are having an impact on teacher diversity in any charter school sector, they are as likely to school- as they are sector- or state-level policies and practices. Part of the idea of charter schools is that they are more independent and autonomous. Districts can implement policies to improve teacher diversity, and these policies would presumably affect all (or at least most) schools in the district. This is less likely to occur within a charter sector, where each school (or chain) would maintain its own hiring practices. From this perspective, it might be as useful to compare teacher diversity within charter sectors as it is to compare between sectors.

(One more implication here: It’s possible that any charter/regular public school discrepancy in teacher diversity is as much about what regular public schools aren’t doing as it is about anything charter schools are doing.)

In any case, the first and most basic step is to compile a more comprehensive set of teacher race and ethnicity distributions, by sector, for more locations, particularly those with large charter sectors. Ideally, the U.S. Department of Education would collect the data for all states as part of its Civil Rights Data Collection. In the meantime, we recently published a report showing that most states do in fact collect teacher race and ethnicity data, including charter schools, so it’s entirely possible that the distributions are already out there for many locations. At the very least, we could get a sense of the situation beyond the highly incomplete evidence discussed above. And, perhaps eventually, we might find actual reasons why some charter sectors’ teachers are more or less diverse than others’ and more or less diverse than nearby regular public schools’.

This is really the question that matters – it’s the why rather than the whether. It’s kind of odd to use comparisons of charter and regular public school teacher race and ethnicity distributions to argue that one type of school is somehow better than the other. If there are concrete policies and practices that are responsible for why charter or regular public schools do better, whether in terms of diversity, test scores, or most anything else, then we’ve reached actionable policy conclusions, and that would be good news for everyone.