Without respect and safety, school improvement efforts are doomed to fail.
A few years ago, one of the coauthors, Jal, and a colleague were visiting a school in Chicago. It was our second day at the school, where we had been observing classes and talking to teachers and students. Things seemed to be going well. But as we walked down the hallway, a teacher sidled up to us. “You seem like nice people,” he said, “but I wanted you to know that word on the street is that you’re spies for Chicago Public Schools.”
Well, then.
What happened to us in Chicago was less about us (we hope) than about a sector rife with distrust. The schools in that district had a lot of experience with CPS and, as in many cities, much of it wasn’t positive. The late educator and longtime school superintendent Arlene Ackerman once said during a visit to Harvard that the first job of any new leader is healing; people have been burned so many times by broken promises that it’s difficult to convince them that this time will be different.
Between 2017 and 2020, the coauthors, along with Kim Frumin, Amelia Peterson, Rebecca Horwitz-Willis, and James Jack, set out to study a variety of methods—such as the Data Wise Improvement Process, instructional rounds, lesson study, continuous improvement, and design thinking—to bring some order and rationality to the world of school improvement. By developing a clear process and norms, creating ways to collectively look at data, and empowering local educators to act on what they learned, the architects of these initiatives hoped to create a more systematic approach to school reform.
As we studied how these methods were applied in different settings, we discovered that although they sometimes soared as intended, they more often fell short. What differentiated the places that succeeded from the ones that struggled? In our new book, A Chance to Solve Their Own Problems (Mehta et al., 2026), we argue that one crucial differentiator was what we came to call the relational container.
A relational container is a bounded social space where the norms, structures, and ways of interacting are carefully constructed to build trust, emphasize common humanity, and protect psychological safety. Relational containers can range in size from a small team to a network of educators from multiple schools or districts. They are particularly needed to protect educators from the prevailing norms in school districts that, as Jal saw during his school visit, are often rife with distrust, punitive accountability, and failed promises that have accumulated over many years.
So how does a leader build such an environment? Four elements are particularly important.
Much of what happens within school systems is profoundly disrespectful of teachers’ time, expertise, and priorities. Teachers are the most time-starved actors in the system, yet we fill their professional learning days with PowerPoints they didn’t ask for and that often have little relevance to their work. The more successful relational containers we observed respected teachers’ time, intelligence, and knowledge.
That respect started with how teachers were brought into the process. In two of the cases we studied, joining the reform effort was voluntary. For example, in one school, kindergarten and 1st grade teachers joined The Ohio State University’s Literacy Collaborative, which focused on how to teach early reading, because they saw teaching reading as a crucial part of their job and as something they wanted to improve. In another case, high school science teachers joined the Next Generation Science Partnership (NGSP) initiative and collaborated with local researchers because they wanted to modernize their biology instruction. Based on our observations of teaching practice, reports from participants, and the progress shown in interim student assessments, these two initiatives were successful.
Conversely, the changes brought on by a third initiative were less so. The school engaged in a data-informed reform effort to help teacher teams use assessments, student work, and observations to improve reading and math. But teachers didn’t join voluntarily; instead, the district mandated their participation. As you might expect, that effort started behind the eight ball with a number of teachers.
Respect is also crucial in how we treat our teachers. The NGSP was a joint effort among a university center, a school district, and a team of educators, and the teachers deeply appreciated being treated as equals by the university folks. Simple things, like responding to teacher emails within 24 hours, were remarkably important for building trust. Similarly, in the Literacy Collaborative, because coaches were in teachers’ rooms at least once a week, they understood the classroom context and weren’t just offering generic recommendations.
Leaders also showed respect by using teacher feedback to refine the work. For example, the NGSP initiative began over the summer with a long retreat at the university campus; participants described this as a wonderful chance to get away and reconceive the curriculum. But the plan then called for teachers to work with the researchers to redevelop the curriculum as the year went on. Teachers found it overwhelming to add this to their already full plates, so the leaders adjusted. They put more collaborative time back in the summer and gave researchers more of the curriculum writing responsibility during the year.
In her book Respect, Sara Lawrence-Lightfoot (2000) differentiates between two kinds of respect: the respect that those in power demand (“You owe me some respect, young man!”) and the respect that those with more formal authority extend to those with less. She argues that the people on the receiving end of this second kind of respect are deeply appreciative because they feel seen and valued.
For example, in one conversation Jal was having with a school principal, she cut him off mid-sentence: “If we want to see that 8th-grade presentation,” she said, “we need to go now. The teacher and student said we could come, but we can’t come in mid-presentation.” Compare this respectful attitude to the blithe way many administrators conduct unannounced walkthroughs, and you can see why this school hadn’t had an opening for a teaching position in seven years. Who would leave a school with a respectful principal like that?
Jal’s mom, a longtime school administrator who is wise in all things, once heard Jal extolling the power of apprenticeship, where newer teachers are under the wings of more experienced ones. “Don’t forget,” she said, “that younger people bring a different something to the table, too. They know more non-Western history, they’re better with technology, and they can coach teams.”
In a similar vein, when our colleague Sarah Fine was running the San Diego Teacher Residency program, which was strong in project-based learning, she had the mentee teachers pick the topic of the project. Then the mentor teachers coached them through how to build high-quality project-based learning—based on the project the mentees chose. She found that by making the mentees the experts on the subjects, they gained confidence more quickly, and the mentor teachers often learned something, too.
Similarly, in the science initiative, some of the teachers initially were in awe of the lead professor, who worked with them on a biology curriculum, because he was one of the foremost experts on science instruction in the United States. But he humanized himself by inviting people to his house for dinner, and he built an environment that welcomed all views. One of the teachers remembers calling home after the first day and saying, “This is amazing!” She appreciated being treated like a professional with a valued perspective among the several PhDs who participated. “They know a lot,” she said, “then I suggest something, and they say ‘Oh, that’s a great idea. Let’s do that!’ Like, whoa, I’m a professional now?”
In a time-starved environment like schools, building a successful collaboration requires drawing on each group’s unique skills and not asking people to do what isn’t central to their work. This has been crucial to the Freshman on Track work, perhaps the most well-known example of a continuous improvement effort resulting in dramatic and sustained benefits for students. Under the program, graduation rates in participating schools have increased by close to 15 percentage points over 15 years. Ironically, this initiative stems from Chicago (not a system known for its trusting culture), but it has built a working relational container within the larger district.
Everyone had a part to play that aligned with their skill sets. The University of Chicago Consortium on School Research first prepared the data and provided lists of students who needed intervention. Then 9th grade teachers who had strong relationships with those students reached out to them to bring them back into the fold. Finally, key district administrators developed the credit recovery program and ran political interference.
Crucial to the success of these processes was building cultures where people could be open and honest and center learning.
In the Literacy Collaborative, educators noted that their coaches, who worked for the Carnegie Foundation for the Advancement of Teaching, were highly approachable and willing to poke fun at themselves for the jargon they used. As one educator said, “It’s the first time I’ve had a training where I didn’t feel like I was being talked to or talked at. I was being talked with.” Another recalled, “I could feel free to use my own words and not . . . all the buzz words. I think that also hooked me into wanting to participate.”
In contrast, in the less successful data-informed reform effort, principals said that once district administrators were in the room, it became a “dog and pony show” where you were expected to show progress—or else. As one principal explained, “We have a problem with the game of trust, and we have a problem with the game of telephone in [this district]. And if those things are being recorded . . . it’s not really safe to say, ‘I messed up’ because then you’re judged, like, ‘You should have known!’ ”
So how do leaders build psychological safety? We’ve learned that trust is built on deep relationships that connect both the personal and the professional. For example, the Deeper Learning Dozen, a collaboration for district change that Jal helped lead, opened with an exercise called Tree of Life. In this exercise, participants draw a tree that includes their roots (ancestors, cultural traditions, birthplaces, religious influences); their soil (what nourishes them); their trunk (their values); and their leaves or blossoms (what they contribute to the world). Participants are invited to share their trees in trios, and then the trees are posted around the room during the remainder of the program. This exercise, and others like it, invites people to more deeply share who they are, including aspects of their identity that they may not often bring into professional spaces. This enables participants to get beyond the superficial and move from “performing” (all too common in gatherings of any kind) to real learning.
Finally, these collaborations protected their participants. In the NGSP, in particular, the constructivist assumptions underlying the biology curriculum were not aligned with the district’s more traditional teacher evaluation system. The university leaders developed a crosswalk tool that aligned evaluations with the goals of the NGSP. Without this effort, the teachers would have been caught between their enthusiasm for the new effort and the need to align themselves with the expectations of the system in which they worked.
Despite district-level challenges, several schools that participated in the data-informed reform built a school-level relational container that enabled their teams to engage in real improvement work. In these schools, the leaders were less focused on implementing a process with fidelity and more focused on building team members’ comfort in discussing their practice and data with one another. To foster continued sharing, leaders were also careful to respond to moments of vulnerability or transparency with enthusiastic support.
For example, despite the school’s lower scores this year, one teacher explained, “At first, I was like, this is going to be embarrassing to discuss our grade-level [data], but then when we did, it made me feel more comfortable because people were positive about it. So, I think as the year has progressed, I’m feeling more and more comfortable in those meetings, whether they’re positive or negative. I’m OK discussing that.”
Respect. Collaboration. Humanity. Boundaries. We set out to study methods, but we discovered that the methods were just an occasion to create space for humans to work together. The methods are technical, but it’s the very human act of building and holding a space for collaborative learning that’s crucial to their success.
Successful schools are doing work that is consistent with what Yong Zhao (2026) has recently called the “courageous minority” or what Ron Berger has called “the rebel alliance” (Maxwell, 2026). Along with nonprofits, these teachers, school leaders, and district and state leaders are moving away from our inherited industrial, transactional, and bureaucratic ways of working to replace them with something that is more meaningful, relational, and deeply human—and that works.
References
•
Lawrence-Lightfoot, S. (2000). Respect: An exploration. Basic Books.
•
Maxwell, R. (Host). (2026, May 4). Ron Berger, a leader in the Rebel Alliance: Why not give people important work to do now? [Audio podcast episode]. In Exceptionally Good: Leaders for a Better World.
•
Mehta, J., Yurkofsky, M., Frumin, K., Peterson, A., Horwitz-Willis, R., & Jack, J. (2026). A chance to solve their own problems: Inside the new methods seeking to remake public schools. Harvard University Press.
•
Zhao, Y. (2026). Obsolete schooling and the courageous minority: Rethinking educational change in the age of artificial intelligence. ECNU Review of Education, 9(3).
