‘A Diverse Teaching Force Is a Quality Teaching Force’: Researcher Seth Gershenson Talks About His New Book and the Changing Demographics of American Schools

See previous 74 Interviews: Author Jal Mehta on the value of teaching, Harvard scholar David Perkins on “playing the whole game,” and Professor Nell Duke on project-based learning and standards. The full archive is here. 

Take a moment to remember your favorite teacher — the one whose attention and encouragement made you work harder and achieve more than you thought you possibly could.

If you happen to be white, the odds are overwhelmingly high that your old mentor was too. By most recent estimates, the American teacher workforce is still close to 80 percent Caucasian. That proportion has declined over the last four decades, but not as quickly as the non-white share of the American school enrollment has grown; for the first time in U.S. history, no racial or ethnic group makes up a majority of students.

That representational disparity in classrooms means that many students are never taught by an educator who shares their cultural background. That’s a critical failing, as research has consistently shown the academic benefits of teacher diversity. Black students assigned to even one same-race teacher during their K-12 years are more likely to complete high school and take college entrance exams, according to one study.

One of the authors of that paper, American University professor Seth Gershenson, has just released a book on the topic through Harvard Education Press, Teacher Diversity and Student Success: Why Racial Representation Matters in the Classroom. Along with co-authors Constance A. Lindsay, a professor of educational leadership at the University of North Carolina, and Michael Hansen, a senior fellow at the Brookings Institution, Gershenson explores the evidence for student-teacher racial matching, examines the social and policy realities that prevent more non-whites from entering the profession, and offers several proposals for how district leaders can both overcome those obstacles and make the most out of the workforce they have.

In this discussion with 74 reporter Kevin Mahnken, Gershenson lays out his arguments for why, as he puts it, “a diverse teaching force is a quality teaching force.”

The interview was edited for length and clarity.

The 74: Let’s start with an overview of student-teacher racial match research. How clear it is that students of different racial and ethnic backgrounds benefit from having a same-race teacher, how widely accepted is that finding, and what is the magnitude of the effects?

Seth Gershenson: I think the evidence that students benefit from having a same-race teacher, particularly Black and Hispanic students, is very compelling and truly ubiquitous across all levels of schooling and across all contexts. There are studies at the pre-K level, studies in elementary school, studies in middle and high school, studies in college, and even studies in law school: Students, but especially students of color, benefit from having a same-race or same-ethnicity teacher.

They benefit in every possible dimension you could imagine measuring, whether that’s test scores and grades, behavior, reducing chronic absenteeism, or longer-run outcomes we really care about, like high school graduation and college enrollment. In some instances, [it even affects] choice of major in college. These effects are observed both in the short run in the long run, at all levels of schooling, and in different parts of the country. One study shows this is true using administrative data from Texas. Another study of mine uses data from both North Carolina and Tennessee. These effects have been documented at multiple colleges and state university systems. This is not unique to the U.S. South, or to particular states with large Hispanic populations, like California or Texas.

And these are sizable effects! In terms of test scores, the effects are maybe 5 or 10 percent of a standard deviation, which is similar in size to a class-size reduction or a general increase in teacher effectiveness. In terms of long-run outcomes, the effects are on the order of a 5-10 percent increase in college enrollment.

You mention in some of the book’s early chapters that the evidence around student-teacher racial match is one part of a larger picture of representation in society more generally, which also illustrates the impact of same-race encounters in medicine, law enforcement, and the courts. Can you put your own findings in that larger context?

The bottom line is that these effects happen for a couple of reasons. One of them is that implicit biases — and even explicit biases — against certain groups manifest in real-life interactions; when you’re dealing with somebody from the same backgrounds as you, those biases are less likely to arise. The flip side is that for underrepresented groups, there’s also what some people might call role-model effects, or just better communication that leads to better outcomes. It’s hard to sort out which of those two things is happening, and it’s probably a little of both.

In any situation with real stakes, anything that can improve the outcome is worth investigating. In all the examples you mentioned, a fundamental part of those interactions is that there’s very clearly someone who’s in a position of authority over the other: teacher/student, police officer/citizen, doctor/patient. There’s a hierarchy there.

The other thing is that trust and communication are central to good outcomes in all of those scenarios. This is also true in the workplace generally. Good, productive relationships at work require a little bit of trust and mutual respect and open dialogue. Those fundamental characteristics of relationships show up everywhere, but one of the first places they show up in most of our lives is in the classroom in school.

What processes are leading to better outcomes in these same-race interactions? On the one hand, they do sound like the product of what social scientists call homophily, the affinity of like for like. But on top of the general absence of bias, it seems that a big part of these effects can be chalked up to higher expectations among members of the same group.

One of the main mechanisms that drives these effects is expectations, which can create self-fulfilling prophecies. And for a long time, studying the effect of expectations was quite difficult for exactly that reason.

It could be that teachers’ expectations for a certain demographic group are lower than for whites. Maybe that is reflecting accurate information that comes from pre-existing differences in test scores, or pre-existing neighborhood segregation, or income differences at home. That’s not to say that those gaps should exist, but if there’s a certain group that’s been historically marginalized and discriminated against, then lower expectations for that group could be accurate in the sense that they’re reflecting the current state of the world.

But if those low expectations are inaccurate, they could cause students to not live up to their true potential and instead converge on those artificially low expectations. That’s what we mean by self-fulfilling prophecy: A student who could have had a good outcome encounters a teacher with low expectations, and whether it’s the student or the teacher who changes their behavior to meet those expectations, the student ends up realizing that lower outcome. That fundamental issue is what stymied research for so long because, yes, we see racial gaps in educational outcomes. And yes, we see racial gaps in teachers’ expectations. But are those accurate expectations that are just reflecting a history of discrimination, or are they biased expectations that look accurate after the fact due to the self-fulfilling prophecy?

In work I’ve done with Nicholas Papageorge at Johns Hopkins, we find pretty compelling evidence that teachers in general are optimistic about kids. But white teachers are less optimistic about Black students than Black teachers are, and that lack of optimism really hurts Black students’ outcomes in terms of college-going and college graduation.

At a few points in the book, you argue in favor of a redefinition of teacher quality that incorporates the value of representation and the potential it brings for better learning outcomes. Can you get into the specifics of that?

We make that argument in the book, in very plain terms. The phrase we repeat often is that a quality teaching force is a diverse teaching force.

Almost everyone agrees that teachers are one of the most important parts of a good school, and identifying effective teachers is tricky. But for a long time, policymakers, researchers, principals, and parents have collectively agreed that experience is a good thing. What we are simply saying is that, sure, that’s true, the evidence backs that up. But with all this evidence that same-race teachers matter, it suggests that access to a representative set of teachers is just as important. The effects are the same size as the effect of ten years of experience. We’re not saying to get rid of other measures of quality, we’re saying to widen the definition of quality to include that diversity piece.

One encouraging thing is that, for over 30 years, the number of non-white teachers across the country has markedly increased. But you also note that the student population has diversified just as quickly.

It’s important to think about teacher diversity in terms of representation instead of an absolute number. Absolutely, the number of non-white teachers in the U.S. has gone up; however, it has not gone up at the same rate as the share of non-white students. So in relative terms, we’ve been treading water and sometimes falling behind. Right now, the national public school enrollment is a smidge under 50 percent white. However, the teaching force remains 80-85 percent white; it’s a little different in some districts, but those are what the raw averages look like. So there’s obviously a long road ahead to achieving a truly representative teaching force.

Isn’t there a self-perpetuating problem in the fact that, since non-white students are less likely to encounter same-race role models at school, they’re less likely to consider a career in education? And given that research suggests that the children of teachers are more likely to become teachers themselves, there’s also a role-modeling issue at home, right?

That’s right. I looked at families with teacher parents in a paper I recently published with a graduate student at American University named Alberto Jacincto. And what we show is that, all else being equal, the children of teachers are significantly more likely to enter teaching than the children of non-teachers in similar households. This is true for white households, it’s true for Hispanic households, and it’s true for Black daughters. It’s not true for Black sons, which highlights the particular lack of Black male educators. Like you said, this is just one example of how the status quo perpetuates racial imbalance.

I’m currently working on a project to figure out the effects of same-race teachers on students eventually entering teaching, and I suspect we’ll see something similar to what we see with parents. But those are two examples of how there’s this chicken-and-egg problem that’s really hard to resolve with minor tweaks here and there. I hope we’re fairly bold in the book about laying out some initiatives that go beyond simple tweaks and push on a variety of fronts.

While a lasting diversification of the profession is a long-term project, are there also ways we can work with existing white teachers to improve their interactions with non-white students — whether that involves “cultural competency” training, professional development around academic expectations, or something else?

That absolutely has to be part of the focus, both now and in the future. And it’s broader than just white teachers because these issues of representation really span race and ethnicity. It’s all teachers, not just white teachers, who would benefit from better-designed teacher training programs. That’s true in terms of content, but also with respect to better-designed mentorship and student-teaching assignments early in their careers.

One of the things we know about implicit bias from social psychology is that part of the problem is just the lack of interactions with members of other groups. That’s something that can be facilitated in a number of creative ways, ranging from virtual reality student-teaching opportunities to after-school and summer school opportunities. So there’s a lot of ways that teachers of all backgrounds can get practice working with diverse classrooms in somewhat lower-stakes settings than in the real-deal classroom as a head teacher.

It’s very common to hear from teachers who feel they weren’t totally ready to take on some aspects of the job. I don’t think I’ve heard one say, “I never trained to be a white teacher in front of an all-minority classroom,” but it sounds like that relative lack of experience could explain some of the challenges that novice teachers face.

Absolutely. And it’s not just the students, it’s also interacting with parents. A lot of student-teaching happens in relatively affluent suburban school districts, even though that’s not where the majority of teachers are going to initially teach. In some sense, they’re practicing the wrong thing and gaining experience in something like the ideal setting. It’s a little like the analogy of a pilot doing a test flight on a clear, sunny day with no wind, and then going out on a cloudy day during hurricane season. That practice on the sunny day might help a tiny bit, but it’s just a different thing.

We could very easily provide that training in schools that are more similar to the settings where teachers are ultimately going to teach. And this has the added benefit of not just preparing the teachers better, but also providing an extra resource to those schools that could really use the help of a student-teacher.

What about expanding the pipeline? You and your co-authors look at a number of options for achieving a more racially representative teaching workforce, from alternative credentialing pathways to “grow-your-own” programs that try to identify future educators. But you also write about the need to add a rung to the professional ladder by encouraging some existing school employees to transition to head teacher roles.

The simple fact is that teachers are expensive — even if they’re underpaid, arguably — and there are cheaper substitutes like care professionals and teacher’s aides that a lot of districts rely on. Because they are lower-paying and require fewer educational credentials, the ranks of those aide and paraprofessional jobs are much more diverse.

Whether it’s because of teacher’s unions or long-standing district policies, the gateway into teaching is fairly difficult to cross for an outsider who hasn’t prepared in a school of education. That gatekeeping has often kept paraprofessionals and teacher’s aides from even considering becoming full-time teachers. We argued in the book that that’s a real oversight, and there’s a lot of talent in that pool that already knows the students and schools, already has relationships with the principals and teachers, and so on. Right now, we’re just not fully tapping on that resource.

So even setting aside the diversity concerns, it’s worth considering which of those paraprofessionals might be interested in becoming teachers. But when you add in the diversity factor, it’s really a no-brainer. And making it easy for teaching assistants to hop onto that career ladder is fairly easy to do. We could count some of that time spent as a teacher’s aide as their student-teaching, which is a way to allow people to earn a living while they’re completing the necessary coursework and credentials. And if we really want to push this, we can even let that experience accrue toward the salary chart, which is usually based on experience in school.

I want to ask about another type of classroom representation that isn’t discussed as often: gender matching. The teaching workforce is almost as disproportionately female as it is disproportionately white, and girls have definitely tended to outperform boys both on short- and long-term academic outcomes for decades now. Is the comparative lack of male teachers something that researchers and policymakers need to be taking more seriously as an impediment to male students?

There are a few points to make about achievement gaps based on gender. The weird thing about the gender gaps is that they’re not just smaller than the racial gasps, but the racial gaps are always in the same direction for every outcome. The gender gaps are much more nuanced.

In terms of high school graduation and college enrollment and college graduation, there is a growing advantage in favor of female students. However, within college, there are differences in choice of major that favor male students, who are more likely to pursue higher-paying majors like those in STEM fields. With respect to test scores, some tested subjects in K-12 have a male advantage, and some subjects have a female advantage. And at the end of the day, there’s still a huge gender wage gap. So even though for some educational measures, the high school graduation rate and college enrollment rates have reversed to favor women, there are still many educational measures as well as labor market measures that favor men.

Even though the gender makeup of the teaching force might be changing a tiny bit, it’s been very, very stable over the past 100 years. So you can’t tie the change in female high school graduation to a change in the percentage of female teachers. Something else is driving that. That’s one reason we don’t focus on the gender gap. I’m not saying it’s not important. But I think that’s a separate book with a whole set of separate issues.

Long story short, I think that racial representation and closing racial gaps is a first-order issue that ranks higher than gender representation in the teaching force. Other people, sometimes in good faith and sometimes not, have asked about other types of representation in the teaching force, like disability status or LGBTQ status or religious status. And I do think that diversity and representation is important for all those questions of who’s marginalized and who would benefit from role models and mentors who look like them. But the bottom line is that there’s a very long, ugly history of racism and segregation in the U.S. that we think lifts this to a higher priority.

Help fund stories like this. Donate now!