This interview focuses on a chapter from A Nation at Risk +40, a report by the Hoover Education Success Initiative (HESI) that looks back at the birth, struggles, and future of the modern school reform movement. (Download the publication here.)
Thomas S. Dee is the Barnett Family Professor at Stanford’s Graduate School of Education, a research associate at the National Bureau of Economic Research, a senior fellow at the Stanford Institute for Economic Policy Research (SIEPR), and a senior fellow (by courtesy) at the Hoover Institution. His chapter deals with school staffing and measures to ensure teacher quality and career development. He spoke with Chris Herhalt about what has worked and what hasn’t.
Chris Herhalt: You cite the original A Nation at Risk report as saying that it found too many teachers being drawn from the lower echelons of achievement. Is this still the case today? And will it ever not be the case, given what teachers earn?
Thomas S. Dee: I think that still fundamentally characterizes the state of the teacher labor market. Without fully thinking it through or articulating it, we’ve essentially moved toward an equilibrium where we are paying teachers very little, and policy has put an emphasis on hiring more and more of those teachers to reduce class sizes instead of thinking about other possible equilibria, such as paying teachers more and asking them to meet rigorous standards of professional accountability commensurate with professional compensation. That’s why programs like the one in District of Columbia Public Schools (DCPS) are so critical; that program in particular established a compelling proof point for an alternative approach.
Herhalt: Another thing you mentioned were “drive-by” teacher evaluations. In my own experience, the extent of an evaluation of the average elementary school teacher was a principal or administrator poking their head in for a few minutes and then leaving because they had too much to do. That was all the accountability measurement most teachers faced.
Dee: That’s accurate. What you end up with is the vast majority—it’s nearly universal—of teachers rated as doing well. But that points to one way A Nation at Risk was so prescient when it focused on this issue. Over the past several decades, we’ve gotten much better high-frequency “big” data on students, and one of the major lessons from those data is that a good teacher really matters. Just one year of access to a high-quality teacher has demonstrable positive effects on a student’s lifetime trajectory—for example, future success in the labor market. We’ve also learned that there’s considerable variation in teacher effectiveness. There are some great, life-changing teachers, and there’s a left tail of teachers who aren’t really helping to realize students’ potential.
I want to stress that because it contrasts with the drive-by evaluations you just described. The way we rate and serve as teachers fails to capture that considerable variation in teacher effectiveness. The concern is that it creates disincentives for recruiting, retaining, and encouraging high-quality teachers. The way they’re evaluated, the way they’re paid on the single salary schedule, which you might be familiar with, are largely unmoored from the impact that they are or aren’t having on students’ lives.
Herhalt: Moving toward the carrots-and-sticks responses you discuss, you find that efforts to offer merit or incentive pay to a better-performing teacher were rarely matched with efforts to put teachers on notice or root them out if they were underperforming. You cite policy choices, teachers’ unions, and lots of other reasons for this. In your view, is it good enough to have a “carrot” but no stick if you can’t have a stick because of political pressure?
Dee: Well, it depends on the nature of the carrot and what it’s doing. For example, one compelling policy would be to offer substantial pay to teachers we know to be really high-performing, for them to go teach in schools with underserved students who need the very best teachers. Targeted incentive to teachers with demonstrated effectiveness can matter. For example, the IMPACT system in DC Public Schools included powerful financial incentives for highly effective teachers in hard-to-staff schools and there’s evidence that this was effective. Similarly, the Talent Transfer Initiative sought to identify teachers and pay them to teach in hard-to-staff schools, and they found that that was effective. It was also really hard to arrange. You had to give teachers a powerful incentive and offer it to many qualified teachers to find some willing to accept the challenge. When they did, it mattered for kids. So, carrots can work. But again, the details matter.
Herhalt: In the current post-COVID context, is it hard to improve teacher quality when there’s such a shortage of people willing to do the job?
Dee: The issue of so-called teacher shortages is a frustrating one because the indiscriminate “sky is falling” rhetoric we often hear is clearly at odds with basic, incontrovertible facts. This sometimes surprises people but it would actually be more accurate to say that, in the aggregate, public schools currently have a teacher surplus. That’s due in no small part to the pandemic exodus of students from public schools, which has not yet been matched by reductions in staffing. I’ve been collecting data and studying this student exodus in collaboration with the New York Times and other journalists. More recently, I’ve worked with the Associated Press to ask, “Have those students come back? Where did they go?” We’ve been tracking private school, home school enrollment, and demographic change. One important implication of this analysis is that this enrollment loss has been persistent. We are seeing the results of that now, especially as pandemic federal aid is set to expire. Many schools are experiencing serious financial pressure and it’s driving rhetoric around closing schools. But we’ve never had more teachers per student than we have now. So, when people talk about generic “teacher shortages,” I think they’re obscuring the reality of what’s going on.
Now, that’s not to say districts don’t face hiring challenges, but the challenges we have are the challenges we’ve always had: how do we get teachers in the hard-to-staff subjects such as STEM and special education? How do we get the needed teachers in hard-to-staff schools? This is where I would argue that a single salary schedule is an impediment, and that union-negotiated rules that don’t allow us to offer differentially higher pay to high-quality teachers in high-need schools are a problem.
Herhalt: So, where is this rhetoric coming from?
Dee: It’s hard for me to say. Again, there are staffing problems—hard-to-staff schools, hard-to-staff subjects. But if you look at the national data on pupil-teacher ratios, they’ve been declining over time, and that decline accelerated during the pandemic because schools lost a lot of students but haven’t yet laid off teachers.
We do not have a generic teacher shortage. We have a shortage of the right kinds of teachers in the right subjects and in the right schools.
For example, to elaborate on that in a California-specific way, California is in the process of expanding universal pre-K to all four-year-olds. We call it transitional kindergarten, and there are both opportunities and issues with that. But one of the challenges is staffing those classrooms. This is true even though at the same time, California public schools have lost something like 400,000 students but have not yet really been laying off teachers.
I think we need a stronger emphasis on recruiting and retaining really effective teachers. As a quick aside, I co-authored a Brookings Institution paper in 2017 with Dan Goldhaber that unpacks the issue of teacher shortages from the pre-pandemic period but is, I think, still relevant. It discusses the promise of initiatives such as “grow your own” programs, alternative pathways into the teaching profession, and creating real licensure reciprocity across states.
So, the general rhetoric around teacher shortages gives me a “this is why we can’t have nice things” vibe, because it’s just unmoored from basic facts.
Herhalt: You’ve singled out two programs, IMPACT in the District of Columbia and the Teacher Excellence Initiative (TEI) in Dallas, as benefiting student outcomes. Both of those projects involve measures that teachers’ unions vehemently oppose, such as merit pay and improvement or departure plans for underperforming teachers. Given the opposition, do these reforms have any hope of scaling?
Dee: My honest and unfortunately cynical answer is probably not. When I presented research on all the positive effects of the IMPACT teacher evaluation system, I would often say this is a proof point that will likely go wholly unreplicated in American public education. And that wasn’t quite true, but it was close. There are a few other indicators of promise. Recent research on an innovative reform in Dallas [in which Hoover senior fellow Eric A. Hanushek participated] is a particularly important example. However, the political consensus needed to adopt, nurture, and sustain such high-profile reforms at scale has always been difficult to achieve and is especially so in today’s political climate.
Given that, I think there is an important role right now for a “grass-roots” approach and research that shines a light on promising local innovations. However, to achieve that, our schools need more capacity to identify promising, evidence-based practices and to evaluate and refine their implementation. I’ve recently written an essay for the Education Success Initiative at Hoover on such continuous improvement efforts in public schools. There are serious challenges to realizing that vision of organizational change (such as staff capacity, political will, and data systems). However, I think it is important to confront those barriers if we hope to scale improvement across a complex federalist system of 13,000 school districts.
The teacher-assessment reforms and research we’ve seen outside of DC underscore the promise of continued replication and evaluation. This includes both the initiative in Dallas and in the few places that have taken up TAP as well. The idea behind continuous improvement is to have people on the ground taking up innovation and then nurturing it through cycles where they carefully plan an implementation, they assess whether it’s working, and they adapt. The importance of implementation fidelity is something I would stress about DCPS’s seminal IMPACT initiative. There was a popular impression—understandable given the press coverage—that DCPS Chancellor Michelle Rhee was creating chaos in the district. But when you look at IMPACT, it was an incredibly well-implemented and effective program.
Those implementation challenges are not trivial. You need good data systems, and you need raters who can do something more valid, nuanced, and actionable than the drive-by assessments that are the norm. Implementation matters. I think we have seen the high-water mark of prominent top-down education reforms like No Child Left Behind that didn’t really focus on implementation. To be clear, my collaborative research work shows reforms like NCLB were far more successful than people commonly acknowledge. However, I also think there’s something important in the idea of thinking big by thinking small. That is, we can identify and promote promising innovations in education through policy making and research with a practice-oriented, grass-roots focus.
Herhalt: Is there a role for some sort of central organizing body to offer better training, or more customizable training, given that you cite school districts spending $18 billion a year on this anyway? Are there opportunities to carry out better professional development?
Dee: Yes, I think the opportunities are considerable and they have developed just in the past few years. If you had asked me ten years ago, I would have said the state of teacher professional development was a national embarrassment. Lots of time and money was going into it, with little or no evidence of efficacy. In the paper, I mentioned a review that screened more than a thousand studies of teacher professional development and found fewer than ten that met the most basic criteria for high-quality research design.
Since then, the literature has grown and higher-quality evidence has come out of it. I was surprised to see evidence that professional development can have positive impacts on teacher performance. I don’t think we know yet what’s generalizable about that, though—the kinds of design features that make it effective—but there is broad enthusiasm around the idea that it can’t just be half a day in August, after which the teacher never connects with the program again. It makes much more sense for the professional development to have strongly job-embedded features, multiple touch points that guide teachers’ day-to-day practice in evidence-aligned ways.
As we move toward a better understanding of what kind of professional development works better for whom, there’s absolutely a role for some type of clearinghouse or other way of disseminating that information. But the challenges of consistently delivering high-quality professional development at scale remain considerable, and so there’s always a question of whether it’s better nurtured at a more grass-roots level, or possibly mediated by technology, to consistently deliver that high-quality support.
Herhalt: What are the “four persistent challenges of teaching”?
Dee: That’s portraying content, managing student behavior, enlisting student participation, and knowing what students understand. This dovetails with our understanding of high-quality teaching. It’s not just the lucid presentation of academic content, the first of those four, but also making sure the classroom is well managed and that students are engaged. The last one is hitting a pause button and giving students opportunities to say what they are or aren’t understanding.
