Why Teachers Stay: What Research Reveals About Retention

When Harvard Education Press authors Doug Larkin and Suzanne Poole Patzelt set out to study the relationship between teacher pay and retention, what they found surprised them. It turns out supportive culture is even more valued than salary. 

“Without fail, no matter what school we went to, what state we were in, that was always the number one response,” Poole Patzelt says. “We did nothing to put that at the top. That was far and beyond the number one reason why teachers stayed was because of who they were working with.”

She adds, “We are relational organisms. We rely on relationships and other people.”

Pay, Larkin explains, mattered but differently than we often assume. Teachers generally felt their compensation was adequate. What didn’t hold up was the idea that increasing pay would directly increase effort or retention. “It doesn’t fit that behavioral logic,” he says. “If we pay teachers 10% more, they’re going to work 10% harder. That’s not what was happening here at all.” Instead, what consistently surfaced were collegial cultures where teachers felt supported rather than scrutinized.

In their new book, The Reasons Teachers Stay, they draw on a six-year longitudinal study of schools, districts, and communities in the United States with high rates of teacher retention. In the districts they studied — spanning rural, suburban, and urban communities — a defining feature was a “real lack of teacher isolation.” Teachers shared resources. They kept doors open. Administrators fostered trust. Poole Patzelt notes that many of the top retention factors were intertwined: Supportive leadership strengthened teacher relationships, and those relationships reinforced a broader culture of care.

Each district operated within its own cultural and political context. Still, the strongest schools resembled what Larkin calls a “healthy ecosystem for teachers,” where induction went beyond onboarding and new teachers were not left in “sink-or-swim” environments.

To make sense of these dynamics, Larkin introduces the “teacher embeddedness” framework, a way of understanding retention not as a single decision, but as the accumulation of many small connections. He shares a metaphor from elementary schools where principals are duct-taped to a wall as part of a reading challenge. One strip of tape does nothing. But layer enough pieces together, and they hold someone in place. “Each little thing you can identify,” he explains, “is another piece of tape that holds that teacher in place.”

In this episode of the Harvard EdCast, Larkin and Poole Patzelt introduce the “teacher embeddedness” framework and gain better insight in why understanding what keeps teachers in the job might be the biggest shift a district can make.

Transcript

[MUSIC PLAYING]

JILL ANDERSON: I'm Jill Anderson. This is the Harvard EdCast.

[MUSIC PLAYING]

Doug Larkin and Suzanne Poole Patzelt show the biggest reason teachers stay isn't pay — it's supportive relationships. They initially set out to study the link between teacher pay and retention. Drawing on a six-year study across 13 districts and four states, they identified 10 factors that influence retention. Supportive relationships and system level support came out on top. They described strong schools as ecosystems built on trust, low isolation, shared resources, and leadership that supports teachers instead of policing them.

I wanted to know more about whether schools can replicate these practices in ways that lead to longer teacher careers. I asked, why do relationships matter so much for teachers?

SUZANNE POOLE PATZELT: Without fail, no matter what school we went to, what state we were in, that was always the number 1 response. We did nothing to put that at the top. That was far and beyond the number 1 reason why teachers stayed was because of who they were working with. I mean, we are relational organisms. We rely on relationships and other people. We're both science teachers, also.

JILL ANDERSON: Yep.

SUZANNE POOLE PATZELT: So thinking in terms of how we are as biological beings, that is something that we know has sustained humans forever. I don't know if that was too shocking, but it was shocking to see how much more than these other factors, I think, that it mattered.

JILL ANDERSON: Right. Pay gets the big headline. So it felt kind of surprising, because there is so much emphasis on, well, if we just pay teachers better, which I'm sure we all agree, we should pay teachers better. But going to work every day, I think no matter what profession you're in, it depends who you're working with.

DOUG LARKIN: I think there's a subtlety there in some of our research finding about pay. That was one of the things that our NSF program officer put at the top. When we were having our negotiation to fund this research, they said, we really want you to look at the relationship between teacher pay and retention.

I think, at the end, we have to say that the fact that the pay was deemed adequate. It wasn't like, we're getting paid $500 more than this district. Though there was some of that, there was enough of that that was wrong. There was a lot of times, teachers thought they were getting paid more than the district over, and it was not correct, but that they felt that they were adequately compensated.

And let's remember, at least in the United States, teachers' pay is one of the only salaries that's publicly negotiated across the whole United States. People vote on teacher salaries, like the public gets a say as what teachers are going to get paid. And so if anything is set at an acceptable level, it's because that's what society has deemed, this is what we should pay teachers.

Now in some places that's more than others, because of organizing and other things, but it doesn't fit that behavioral logic. That if we pay teachers 10% more, they're going to work 10% harder. That's not what was happening here at all. One of the findings from our research about collegial relationships being so highly valued is because this speaks to some of the other reasons people get into teaching. And we can look at teaching historically as one of the caring professions. Historically, it's been underpaid as women's work.

JILL ANDERSON: Right.

DOUG LARKIN: And so efforts to raise teacher pay have been linked tightly to social organization, working with colleagues to collectively do better for the profession and for a particular school or job. So to me, it was surprising to find it in every school district that we visited.

So in this research, what we did is we identified school districts that were doing a good job retaining teachers, and we went there. When we went to these places and we saw how strong that was, it was striking. It also resonated as true with us as teachers. But I kind of thought, well, that was just the places I was working. Everywhere is not like that.

And it is, right. Those are the reasons people want to stay. And it doesn't mean people will stay, but it makes it more likely that people will when they have those supportive, collegial relationships and they're working in a culture of support.

Part of those relationships is simply the sharing of resources and knowledge. And none of the districts that we saw were districts where people hid what they did from one another. They weren't these districts where people closed their doors and they teach in isolation.

In fact, a defining feature of the districts that we visited was a real lack of teacher isolation. We hardly saw any teacher isolation at all. And I think that speaks to the types of collegial relationships that existed in these high-retention schools.

SUZANNE POOLE PATZELT: It's hard for us even to sift some of these factors completely apart. And that was a big process, too, when we were analyzing our findings. But just like the resources were part of the relationships as well, those supportive administrators, which we saw was very close to the top, right, supported the relationships between the teachers. So it was like that support was not just like, oh, I'm supporting these individual teachers one-on-one, they were supporting having those sustained relationships between the teachers themselves. So that's interesting too. It's like some of these other factors were subfactors of the relationships between the teachers. So that is really interesting.

JILL ANDERSON: I want to just look a little bit deeper at that, because I know you spent a long time doing this research. Six years is a pretty good amount of time — 13 districts, four states, and it was different — rural, suburban, urban. You were really spreading yourselves out. What were the common threads that these districts had that maybe other districts could be lacking?

DOUG LARKIN: What was interesting is that it wasn't a formula. Each district and each school were situated in their own particular context. In writing about this research, we've been pretty careful not to lay out, here's how you do it, right.

JILL ANDERSON: Formula.

DOUG LARKIN: If we can just talk about new teachers for a moment, because this work actually started with a look at supporting new teachers and new teacher retention, and it sort of spread beyond that. But when we look at induction programs for new teachers, which, depending on the district, they can be something as simple as meeting once a month to talk about issues. There was one district that did a bus tour for their new teachers and took them around the district and ended up at a public housing place where they had a barbecue for them, as a way to forge links between the new teachers and the communities. We saw some districts had dedicated coaching. And it was a position that was a new teacher coach that moved between schools in the district to support new teachers, specifically to have someone who is not responsible for evaluating teachers giving them formative feedback.

There were other structures. We saw a lot of districts assign new teacher mentors. And while those were important, very few teachers cited those as the reason for staying. More common was to cite those informal mentors, the teacher next door, the teacher they had lunch duty with. One of the interviews we did was teachers who were sitting together in the hallway on their hall duty. Some of these types of relationships, these informal relationships, they existed because there was a capacity for them in the school, or teachers found them themselves. Some of those structures were there. But the other piece of it was that culture of support in the school, this idea like, OK, you are going to make mistakes, it's hard as a new teacher, it wasn't a sink or swim environment. And I've been in those environments and I've seen what those can look like. And we can uniformly say that each of these schools was like a healthy ecosystem for teachers. I work in teacher preparation. I would encourage them to go work in a school like this. So the induction wasn't just onboarding. It wasn't just, OK, here's your copier code.

JILL ANDERSON: Right.

DOUG LARKIN: Here's how you take attendance. Here's how you call out [INAUDIBLE]. There was that stuff, but it was much more than that.

SUZANNE POOLE PATZELT: We came into the research thinking, we really want to look at these induction and mentoring practices so that they could be replicated. So everything we're saying now, we actually were hoping to do something of not necessarily formula, but how can we get these practices more universally so that teachers are being supported? Because obviously where they're staying, these must be somebody's doing something good with induction and mentoring.

And although everything that Doug said is true, that they did have some different types of practices, it wasn't just onboarding. That wasn't what teachers were saying. They didn't come out and say, oh, I'm here because the induction was so good. Those first years of support I got were fabulous.

The replication part is hard, because it isn't cookie cutter. And every place we went had totally different cultural situations and contexts. But it's, how do we make room for the informal relationships to flourish? But it's almost like, when we tell new teachers — I'm also in teacher prep — but when we tell our new teachers to really try to support those relationships in their classrooms, and they have to give up a little bit of control, and that's a little bit scary, and we're like, you want to foster conversation and discussion, but that means you need to do really good planning, structure your lesson appropriately, but then you need to step back a little bit and let your students actually discuss and wonder and ponder about things and struggle.

And I'm seeing now the mirror of that with admin and teachers. You want good planning, you want structures in place, but you also want to, within those structures, allow this freedom for your teachers to be able to build those relationships and to explore and to wonder and to challenge each other. That's the type of administration that was there that allowed that for their teachers, that then those teachers were able to explore and do things in their classrooms, which is, when we go back to what brings us to the classroom in the first place, no teacher is going into the teaching profession because we're going to make $300,000. We know that. So that's not the reason why people go in. So thinking that salary is going to be the biggest reason why they stay makes sense. We go in because we want to see students learning.

JILL ANDERSON: If a school leader could change something tomorrow to strengthen that sense among new teachers, or even teachers who have been at it for many years, what do you think it should be?

DOUG LARKIN: We have to remember that in the US we've created a culture of schooling that's quite different from that of the rest of the world. When I was a high school teacher, I taught six periods a day, and then I had hall duty and lunch duty — and maybe I had one prep in there — to grade and do everything, right. So we have to start with the fact that teachers' days are very intense. We don't treat any other profession that way.

SUZANNE POOLE PATZELT: We hear it. We hear about teachers not being able to go to the bathroom at some point during the day.

DOUG LARKIN: Absolutely. Or you get used to eating lunch at 10 in the morning.

SUZANNE POOLE PATZELT: Yeah.

DOUG LARKIN: Like, whatever. Some of the schools where we went, the administrators did what they could to mitigate some of those things. There are other things, like supporting teachers' professional growth. So it could be paying for professional development, paying to go to conferences, graduate study, all those things. But administrators start getting nervous when you say that stuff, because they're like, I don't have money for any of that.

One thing that administrators can do right off the bat is demonstrate ways that you value the expertise that your teachers have. So, for example, any teacher who's listening to this or watching this has sat through a professional development day where they're like, I could be writing lesson plans, I could be grading papers, I could be putting up bulletin boards, I could be doing anything other than listening to this person speak at me.

But there is a movement throughout the country for teacher-led professional development, where there are PD days built into the school year. And so rather than hiring an external consultant to come in and talk to your teachers about the latest educational fad, give teachers that time to present something to their colleagues. Let the other teachers choose which ones they want to go to.

And maybe — this is radical —  but maybe let teachers go work on their bulletin boards or their grading with that professional development time. And let them choose how to best use and manage that time. If teachers are autonomous professionals with agency over their lives and careers, then let's respect that. And let's find ways to build that into their work world.

And another way is involving teachers in hiring decisions. As teachers, think about, who are we going to hire to teach this first grade class or to be the new biology teacher? Bringing teachers into that process demonstrates that we value their expertise.

SUZANNE POOLE PATZELT: I think another piece to add to that, too, is this, it's not about the I gotcha culture. And so examining what administrators can really do is examine how their school functions, the ways they're evaluating teachers at the moment, and see how can we shift this to, instead of an I gotcha culture, to how do we support you in what you're doing well.

Not that this is enough at all, but putting notes in teachers mailboxes or giving an actual, valuable resource to them during those Teacher Appreciation weeks, those little things become the norm. And the culture, I think, that makes teachers feel valued and supportive, and why it's not always about salary, either.

JILL ANDERSON: Can you explain the teacher embeddedness framework a little bit? And how that may be influential for teachers' decisions to stay

DOUG LARKIN: I've been fascinated with this phenomenon that happens in elementary schools with principals around. They set a goal for 10,000 hours of reading for their students. And if the students do it, in the 80s they used to let you put a pie in the principal's face or something goofy like that.

Now they duct tape the principals to the wall. I don't know if you've seen this. The principal stands on a stool and they give each kid a long piece of duct tape. And kids walk by, and eventually there's enough duct tape on the principal that they pull the stool away, and the principal is like a spider cocoon wrapped up on the wall.

I love the visual metaphor of that. That one little piece of tape, but then a bunch of these little pieces of tape end up holding, literally holding an educator in place. So we think of the teacher embeddedness framework in that way. Where it's not a formula where, if you do this, then somebody will stay.

But each piece of that embeddedness framework, we envision as a three-by-three grid. On one end, you have the domain of the school or the organization, you have the teaching profession, and then you have the community. And then going across, we have what we call fit. What's the fit for the teacher, with the school, with the profession, and with the community?

We have links, which is that human social tendency that we have to make connections with other people. So again, what kind of links does a teacher have within the school organization, within the profession, and within the community.

And then assets. What are the things that a teacher values, like tangible things and intangible things? A short commute is an asset, again, within the organization, within the profession, and within the community.

And each little thing you can identify in that three-by-three grid is like a piece of that duct tape that holds the educator in place. And we know that there are shocks. We know that things happen in schools. We know that there are critical incidents that maybe make teachers think about, am I going to stay or leave? I don't know if this job is for me here at this school, if teaching is for me? Do I want to stay where I'm at geographically or within this community?

So sometimes those shocks and those critical incidents are very large. Sometimes they're not even critical incidents. Sometimes your partner gets a job on the other side of the country, and so you all move.

JILL ANDERSON: Right.

DOUG LARKIN: So that's going to pull through the tape. But the thing that we're really looking at here is those teachers who leave, but they didn't have to. So we think about each web of that teacher embeddedness framework. Each strand, each thing in the box there is another piece of tape that holds that teacher in place.

And it doesn't mean that they're definitely going to stay. But to us, it's a much better way of thinking about teacher retention. And it fits the data better, frankly, than just simple job satisfaction. There are some satisfied teachers who leave, and there are teachers who are not satisfied at all but stay because they kind of have to, for lots of reasons. And so job satisfaction doesn't really go very far in explaining why teachers stay and why they don't.

But the job embeddedness and teacher embeddedness, which, in our work, we really want to highlight this job embeddedness work done by organizational psychologists, particularly at the University of Washington back in the late '90s, early 2000's. And we've taken a teacher version of that with this teacher embeddedness. But we feel it offers a nice explanatory and predictive framework. That's why we say, if an administrator can increase the strength of a teacher's links and do it not just within the school organization, but across the profession, maybe with professional organizations and within the community, links in all three of those domains are going to help strengthen that teacher's embeddedness within their job.

JILL ANDERSON: I wanted to bring up the state and policy level, because I think it's interesting that that doesn't have as much weight as what's, of course, happening within the culture of the school. But I did want to know if policy plays any level on getting teachers to stay, and what that might be?

DOUG LARKIN: The short answer is yes, it absolutely does impact, particularly in that profession column. So two of the states that we did our research in, New Jersey and Wisconsin, during the time period under consideration for this study, had governors who were quite hostile to teaching.

In Wisconsin, they went as far as abolishing collective bargaining for teachers. And an interesting side effect is it actually —  it appeared to increase teacher retention, because it increased the links that teachers had with one another. Remember in 2011, those pictures of this full statehouse protesting against Governor Scott Walker? Those teachers were really supporting one another. And the impacts of some of those changes weren't actually felt for a number of years.

Same thing in New Jersey —  there was a governor who was quite hostile and made public, hey, about having confrontations with teachers and calling teachers lazy and diminishing their sense of public service.

That's one place. It's not quite policy, although it impacts policy. The bigger policies are the ones that allow for, in New Jersey and Pennsylvania and Wisconsin, where we looked at individual districts negotiated for their teacher salaries. And they're subject to state approval, of course.

But we compare that to North Carolina, another state that we looked at, where there is a single salary scale across the whole country, with just some minor cost of living adjustments based on urban versus rural. But the salaries were much lower in North Carolina.

And so the policy shaped the mobility environment for teachers, whether teachers were moving from one district to another. But it did a lot to shape the way that teachers felt about their fit into the profession. And it had a lot to do with particularly the teacher evaluation component of how teachers interacted with administrators — and I think, at least in our work, those administrators who saw teacher evaluation as an opportunity to support teachers through that process rather than, as Suzanne mentioned, a gotcha.

So I think there's a standpoint in all policy, all policy takes a position. At least in our work, we saw that it definitely did have an effect, but the local context adapted to that. And teachers took it as the normal in their district.

And so I think teachers in New Jersey had a very different normal than teachers in North Carolina. And teachers who moved from one state to another probably felt that as almost a cultural shift within that state policy environment. And remember, in the US we have really 50 educational policies. We don't really have a national educational policy beyond that of some limited accountability for federal funding. That had an impact, but it was really the local environment, the local school and district organization that had the strongest impact.

And we have to mention residential patterns, demographic patterns, which districts were racially segregated, which ones were more multicultural. And there were some districts that were economically quite segregated, even within the district itself. We had others, like in New Jersey, we have very small districts, and some are much wealthier than others. But New Jersey has had a very strong commitment, enforced by some State Supreme Court cases around educational equity.

There have been a lot of people fighting for educational equity in New Jersey for many years, and that's also had an impact. One of the districts, Mulberry is the pseudonym that we use in our writing about it, but they've been the beneficiary of extra funding. And I think one of the things we've been able to do in our work is to document what that funding has purchased, and it's really purchased quite a lot in terms of retaining teachers.

SUZANNE POOLE PATZELT: The flip side of that, a little bit, is as a country we've really had ups and downs of equity being priority, and if that increases funding in places or takes away funding from other places. But what we did see throughout all of the places we went is, again, going back to that administrative support, they were almost sheltering as best as they could. There are teachers from some of these shifts and shocks that come due to who's in office and who's not and what priorities. Our research was during one window of time. And so we can only speak to that window.

But it seems as if, and we will always have these shocks —  because we are a 50 state nation with lots of different points of views-- that those administrators were doing their best. And one of those things that we saw was particularly supporting teachers of color, so providing places where teachers of color felt safe and supported and can flourish. And that was important. And even if they didn't have high populations of teachers that were of color, still thinking about how do we make it supportive for those teachers who are teaching students who are diverse, and how we can make equity a priority.

So I think, yeah, policy shifts will always be happening. Some are bigger than others and have a bigger impact than others. But what we saw was administrators doing their best to be a barrier between the chaos and their teachers. And I think that's something to be said, also.

JILL ANDERSON: You talked a little bit about things that you can do for that population within your school, or even if you're in a diverse community. Are the strategies different for teachers of color?

DOUG LARKIN: We have to start with the understanding that, whether we're talking about new teachers or experienced teachers, teachers of color have a qualitatively different experience in schools than white teachers. And I'm not placing a value judgment on that. I'm saying that so much of the research points to that starting fact.

So now knowing that, we extrapolate to the data that says that teachers of color, particularly new teachers of color, have different needs for support. That teacher that was so helpful to me, a white guy starting, might not be so helpful to the new teacher of color. And so if I'm an administrator, I have to recognize that.

We drew a lot from this particular study in Portland, Maine that was done by Doris Santoro, Julia Hazel, and Alberto Morales, where they were commissioned to look at some of the issues that were occurring in Portland public schools for teachers of color. And they identified two big factors. And we kind of saw an absence of these factors in the districts that were supporting teachers of color.

The first was what they called the smog of racism, that racism just got in the way of so many things, like an experienced teacher supporting a white teacher more than they might support, a teacher of color. The other thing that they found was what they called the impenetrable wall of whiteness, that this whiteness served a function that put up barriers for teachers of color.

We can very confidently say that in the districts we saw that were doing a good job at supporting new teachers of color, there was an absence of those things, at least reported to us. We don't want to say there was a complete absence, because we didn't interview every teacher in the district. But that smog of racism and that impenetrable wall of whiteness was missing from the districts. So I think at the very least, an administrator who is looking to support their teachers of color might look to see, OK, well, what barriers exist for teachers of color in our district, and particularly for new teachers of color, again, who may have different reasons for being there than some of the newly hired white teachers?

Obviously, talking and having conversations is one way to get that information, but another is just making sure that we're accurately assessing what the needs of new teachers are, and just not assuming that, OK, this teacher is going to need help with class management. Guess what? That new teacher might not need so much help with class management. They might need more help with navigating the curriculum. There can be all kinds of things.

And we called it a district-level race consciousness. There were two districts in particular that were almost like what we would call a Historically Black College University, HBCU-like environment, where there was an atmosphere of uplift, particularly for the Black students in those schools. And those students had a very different experience than students, maybe, in a predominantly white school would have. And as a consequence, the teachers had a very different experience in that school.

And so it's worth thinking for administrators like, where are the barriers for your teachers of color? And it's worthwhile to think about the particular needs of that group of teachers. From the data, it doesn't appear that those teachers leave at any higher rates than other teachers, but there are fewer teachers of color across all subject areas in the US. It is important that the demographic profile of our teaching workforce reflects the demographic profile of the student population. Whenever we lose a teacher of color, we're losing an opportunity to build our multiracial, multiethnic democracy a little bit further.

JILL ANDERSON: Doug Larkin is a professor in the Department of Teaching and Learning at Montclair State University. Suzanne Poole Patzelt is an assistant professor in the Department of Alternative Programs and University Partnerships at Touro University. They are the authors of “The Reasons Teachers Stay: Lessons from High-Retention Schools, Districts, and Communities.”

I'm Jill Anderson. This is the Harvard EdCast, produced by the Harvard Graduate School of Education. Thanks for listening.

[MUSIC PLAYING]