Picture a company that ranks its employees against one another every few months and fires the bottom performers. The people who survive get to keep their jobs, but what happens to how they treat each other afterward? Do the winners become sharper collaborators, or does the pressure of that competition quietly reshape how much they trust, help, and cooperate with the colleagues who made it through alongside them?
A study published in the Journal of Economic Psychology set out to answer that question. Using two controlled experiments, the researchers found that while competitive firing systems can push individual productivity up in the short term, they also appear to leave surviving employees less generous, less cooperative, and more willing to undercut one another in later interactions.
The system under the microscope
The practice at the center of this research goes by several names: forced distribution rating systems, stack ranking, or, more bluntly, “rank-and-yank.” Companies evaluate a group of workers, rank them against each other, and terminate a set fraction of the lowest performers. Some estimates suggest that between 20 and 30 percent of U.S. companies have used such systems, and firms including Amazon, Microsoft, Ford, and General Electric have been associated with the approach over the years.
The appeal is straightforward. Managers hope that by weeding out the weakest performers and keeping only the strongest, they will build a high-performing team. But critics have long argued that these systems breed a “dog-eat-dog” culture that erodes teamwork over time. Until now, much of that criticism has rested on anecdotes and theory rather than direct evidence.
Eric Cardella of Texas Tech University, along with Tamar Kugler of the University of Arizona and Kyle Mao of Texas State University, wanted to test the claim rigorously. The problem with studying this in real workplaces is that companies choose to adopt these systems for reasons that are hard to untangle, and the employees who join or stay at such firms may already differ from others. To get around this, the researchers built an experiment where they could randomly assign people to different conditions and observe their actual behavior.
Building a lab version of the workplace
In the first study, 194 university students took part in a two-stage exercise. In the first stage, everyone completed a real work task: they were given lists of book titles and their matching ISBN numbers, then asked to type in as many correct ISBNs as they could within five minutes. Performance was ranked against others in the session.
Participants were randomly sorted into three groups. In the Baseline group, everyone moved on to the second stage regardless of how they ranked. In the main Rank-and-Yank group, only the top half of performers were retained; the bottom half were removed from the study. A third group, called Yank-Random, also cut half the participants, but did so at random rather than based on performance. That third group let the researchers separate the effect of being fired based on rank from the effect of simply being fired.
To keep the comparison fair, both the Baseline and Rank-and-Yank groups worked under the same competitive bonus during the first stage. The difference was the looming threat of termination in the Rank-and-Yank condition.
In the second stage, the surviving participants played a series of well-established economic games designed to measure how people treat one another. In the dictator game, one person decides how much of a $12 pot to share with a partner, a common measure of altruism. In the public goods game, participants choose how much to contribute to a shared account, a proxy for cooperation. In the trust game, one player decides how much money to send to a partner, which gets tripled, and the partner decides how much to return, measuring trust and trustworthiness. Finally, a sabotage game let one player take money from a partner to boost their own payout, capturing the willingness to harm others for personal gain.
Higher output, lower generosity
On the productivity front, the results matched the conventional argument for these systems. Participants in the Rank-and-Yank group completed about 21.5 percent more entries than those in the Baseline, and this gain did not come with a drop in accuracy. The threat of being cut appeared to motivate harder work.
But the second stage told a different story. To make a clean comparison, the researchers matched the survivors of Rank-and-Yank against the top-performing half of the Baseline group, since both represented high performers. Across every measure, the survivors behaved in a less other-regarding way. They gave less in the dictator game, contributed less to the shared pot, sent less money and returned less in the trust game, and took more from their partners in the sabotage game.
The Yank-Random group landed in between. Its behavior was closer to the Baseline than to Rank-and-Yank, which the authors interpret as evidence that the competitive ranking itself, not just the experience of being fired, drives the effect. In their reading, it is the fight to avoid the bottom rung that changes how people treat one another.
Why survivors pull back
The second study tried to explain the mechanism behind this shift. Here, 564 working adults recruited online read a scenario describing a company that had adopted a new evaluation system. Depending on their assigned group, they were told either that low performers faced no consequences, that the bottom 50 percent would be fired, or that the bottom 20 percent would be fired. Participants were then asked to imagine they had ranked in the top half and to rate their beliefs about the coworkers who survived alongside them.
People who imagined a rank-and-yank workplace consistently viewed their surviving peers as less cooperative, less altruistic, less trusting, and more inclined to sabotage. They also rated those survivors as more likely to have cheated to earn their rank, and as feeling more entitled to company resources because they had “earned” their spot.
The authors suggest these beliefs help explain the behavior seen in the first study. If you expect the people around you to be self-serving, or suspect they cut corners to get ahead, you may respond by protecting yourself and giving less. A sense of having earned one’s position may also feed a feeling of entitlement that makes people less generous.
One finding stood out: the 50 percent and 20 percent termination groups produced nearly identical results. This suggests to the researchers that the mere presence of the system, rather than how many people get cut, is what shifts attitudes. Participants in both firing scenarios also reported that such a system would hurt morale, satisfaction, and the company’s reputation, and said they would be less likely to work for a firm that used it.
What it means, and what to keep in mind
The authors argue that the short-term productivity gains from rank-and-yank may be offset over time by the erosion of trust and cooperation that hold teams together. They suggest such systems may fit best in settings where results depend mostly on individual output rather than collaboration. They also point to possible ways to soften the downsides, such as being more transparent about how performance is judged, which might ease suspicions of cheating, and building a stronger shared identity among the survivors.
A few caveats are worth holding onto. The participants were students and online workers in short, anonymous interactions, not employees who build relationships over years, so the findings may not transfer neatly to real workplaces. The first study also used a steeper cutoff, firing half the group, than most companies apply. And because people knew the rules before they started, the survivors of the Rank-and-Yank condition were not a perfectly matched group. The researchers frame their work as a first step in producing causal evidence on a practice that has been debated largely through anecdote.
