A resignation letter, left on a desk. What leads there?
People accumulate evidence, over time, about whether they can count on the people and places around them. Everyday follow-through, easy to overlook, is one of the ways that evidence builds. Can the person or place you depend on be counted on?
Now turn that same question around: it runs on you too, whenever someone depends on you for something only you can give them. Say you tell someone something specific: I'll get back to you with an answer by the end of the day. I'll let you know the moment I hear anything, even if it's a no. I'll tell you myself, before you hear it from anyone else. The moment you say it, they're holding something that wasn't there before: a genuine expectation, with your name on it. Then the moment arrives, and you either do what you said or you don't. Whatever happens, they now have something concrete to go on.
September is full of these moments. New routines are starting. New relationships are forming. People are making commitments to each other, many of them for the first time, and learning whether those commitments hold. Those ordinary early exchanges are already answering a version of that question, person to person. Researchers have studied a related question, in a large study of schools: not whether one person can be counted on, but whether the conditions people work in relate to whether they stay.
In 2011, researchers Donald Boyd, Pam Grossman, Marsha Ing, Hamilton Lankford, Susanna Loeb, and James Wyckoff published a study in the American Educational Research Journal. They followed 4,360 first-year teachers in New York City and asked them to rate several things about their working conditions: the school's administration, student behavior, and facilities. Then the researchers tracked what teachers actually did next: stay, transfer schools, or leave teaching in the city altogether.
The design solves a genuine problem. If you ask an unhappy teacher both how they feel about their school and whether they might leave, their unhappiness pushes both answers down at the same time. So a link between the two answers might not mean much. It might just mean the same person was in a bad mood when they answered both questions. Boyd and his colleagues checked for this separately. They used one group of teachers' ratings of their school's administration to predict what happened to a different group of teachers, in the same school, afterward. Because it's two different sets of people, one teacher's bad mood couldn't be secretly inflating both numbers in that check. Administration came out as the strongest predictor there too.
In the study's main analysis, of every working condition they measured, one stood out. Researchers can test several possible reasons at once and see which one still matters after the others are ruled out. When Boyd's team ran that test, on teachers' own ratings and their own later outcomes, one answer stood alone. Teachers' rating of the school's administration was the only factor that still predicted retention. A big, meaningful improvement in how teachers rated their administration made a real difference. Teachers were 44 percent less likely to transfer schools, and 28 percent less likely to leave the city's teaching workforce entirely. In a separate part of the study, researchers also asked teachers who had left teaching in the city to name, in their own words, the single biggest reason. More than 40 percent cited the administration. About 15 percent named student behavior, the next most common answer.
What exactly were teachers rating, when they rated "the administration"? The study broke it into several specific behaviors: whether the administration supported and encouraged them, and whether it consulted staff before decisions that affected them. It asked whether the administration enforced discipline policy consistently, whether it evaluated performance fairly, and whether teachers got the resources the job required, among others.
Look closer at three of those behaviors. Being consulted before a decision that affects you. Being evaluated fairly. Being supported and encouraged. Those are three of the behaviors Boyd's team asked about, part of the broader set they measured for the first-year New York City teachers in the study. In everyday terms, behaviors like these are part of how people find out whether someone can be counted on.
A 2022 meta-analysis by Ruoxuan Li and Meilin Yao, published in Educational Research Review, pulled together ninety-four studies covering nearly 40,000 teachers. It measured something narrower than Boyd's study: not whether a teacher actually left, but whether they wanted to. Across all of them, trust was one of the three factors most strongly linked with teachers being less likely to want to leave. The other two: feeling a strong sense of identity as a teacher, and feeling committed to their school.
A separate line of research makes one part of trustworthiness especially concrete: honesty. For almost forty years, James Kouzes and Barry Posner have studied what people look for in someone they would willingly follow. They've published their research across multiple editions of The Leadership Challenge, the first in 1987. It surveys people around the world and asks a simple question: which characteristics do you most look for and admire in someone you'd willingly follow? One quality has ranked first every single time: honest. In that same body of research, Kouzes and Posner point to a related test for it: do what you say you will do.
Picture yourself telling someone, on a Tuesday, that you'll have an answer for them by the end of the week. There are two ways that Friday can go. Friday arrives, and you don't follow up. Friday arrives, and you do, exactly as you said you would, on time, whatever the answer turns out to be. That second version is the kind of track record Kouzes and Posner's related test points to: someone whose word keeps turning out to be good. Either way, the other person now has real evidence to weigh, whether they realize it or not.
The difference between those two outcomes is small enough to build into a regular habit. Tell someone what they can expect, whether it's an answer, a decision, or a plan. Then follow through on it, on the timeline you gave them. Put it on the calendar the way you would any other recurring task, so it happens on schedule even on the days that are already full.
Honoring a commitment, deciding something fairly, answering honestly: each one becomes evidence about whether someone can be counted on. Separately, the research points to two related findings: the conditions people work in relate to whether they actually stay, and trust relates to whether they want to leave. Once that evidence turns against the person you depend on, what does it take to turn it back?