“Competitive,” “decisive,” “action-oriented,” even “intimidating”: many people invoke these words to describe good leaders. Indeed, several studies suggest extraverted, dominant individuals are perceived as competent, influential leaders in industry and politics. Think of the late former General Electric CEO Jack Welch, Amazon founder Jeff Bezos or Tesla CEO Elon Musk. Many people find these leaders appealing and inspiring.

But such individuals have shortcomings as well. Dominant leaders sometimes seek to influence co-workers by fiat or force—insisting on their own way or intimidating others—rather than taking steps to discuss, debate or consult with colleagues. And that has serious downsides for the companies, organizations and nations that they lead.

In our recent research, we examined some of the unintentional negative consequences of a dominant leadership style. Across eight studies, we explored how such leaders can inadvertently reduce cooperation among their employees by fostering a competitive climate. Past research shows that societies and organizations flourish when members help one another, share information and engage in collective problem-solving. Dominant leadership can stifle those activities, however. We argue that’s because a leader’s hyperindividualist approach can foster a widespread zero-sum mindset, in which people believe they can only progress at the expense of others.

In our first foray into this investigation, we looked at political leadership, specifically comparing democracies and dictatorships. Though some democratic leaders are aggressive and competitive, dictators exhibit extremely dominant behavior. They subjugate others to serve their own best interests. Given our hypothesis that dominance may foster a highly competitive culture, we wondered whether citizens in dictatorships engage in more zero-sum thinking than those in democracies. To test that idea, we examined data from 70 countries surveyed between 1981 and 2014 through the World Values Survey, which seeks to understand peoples’ social, political and cultural beliefs. We specifically attended to how much residents reported their agreement with such statements as “people can only get rich at the expense of others.” We also looked at their inclination toward helping behaviors, including how highly they rated the importance of caring for their neighbors. We found that citizens of countries governed by dictators reported greater zero-sum mindsets and were less likely to help others when compared with residents of democracies.

For our second study, we designed an experiment to directly test whether dominance influences how people think about cooperation and competition in a work context. We recruited male and female professional actors and then filmed them in a series of videos. The performers introduced themselves at the start of each video and described their leadership approach to newly onboarded workplace subordinates. One of these approaches was dominance: in it, the leader described their tendency to be authoritative and decide what is best for the team. The other approach was what we call the prestige style. In it, the leader emphasized how much they valued others’ input and an egalitarian approach.

We then recruited about 600 participants who watched one of these videos (either a male or female leader in the dominance or prestige condition). Afterward, they rated how much they agreed with statements related to zero-sum thinking and how likely they would be to engage in helping behaviors—such as listening to a co-worker’s problems—if they worked for the boss whose video they had just seen. We found that participants who had watched a dominant leader were more prone to express a zero-sum mindset and less likely to help others, compared with participants who had just watched a prestige leader.

Additional questions and analysis allowed us to rule out other factors that could influence these findings. Gender had no effect: dominant men and women in the videos both reduced helpfulness and increased zero-sum thinking among participants. In a follow-up study, we asked people questions to assess how much autonomy they possessed and whether they considered assisting others to be an important part of their work. After all, if people feel they lack control over their tasks or that their job simply doesn’t involve cooperation, it’s no surprise they might hesitate to help others, regardless of their leader’s style. But these factors, our analyses revealed, had minimal effect on thinking and behavior in comparison with leader dominance and zero-sum thinking.

We also assessed actual helping behavior rather than relying solely on people’s reported inclinations. We gave participants a written description of a leader. Then we put them on teams for an online task and measured the degree to which they volunteered to transcribe text for their fellow group members. Our hypothesis held. People who had read descriptions of their leader’s dominant styles were significantly less willing to help their team out during these exercises.

Finally, we tested whether this finding could be replicated with actual working groups. We surveyed 249 employees in 50 teams, along with their supervisors, at companies in India. We began by asking employees about their leader’s tendency to influence based on dominance and about their own zero-sum mindset. Six weeks later, supervisors rated their employees’ helping behaviors. When we looked at our combined data, we found a familiar pattern: Employees supervised by a dominant leader reported greater zero-sum thinking. And as their supervisors subsequently revealed, these employees displayed fewer helping behaviors. Importantly, this effect remained robust even when employees had a positive relationship with their supervisor and saw this leader as highly ethical, two factors that might otherwise explain variation in their thinking and behavior.

Although a number of leadership books and popular coaching manuals celebrate the effectiveness of a confident, decisive leader, our work underscores how this approach may breed an “each to their own” culture. Managers need to be aware that an assertive or forceful approach could reduce cohesiveness and collaboration. Organizations, meanwhile, should be careful about whom they promote. If a leader cannot rein in their dominant style, management should incentivize helping others. Companies can, for example, stress that employees understand how supporting one another’s career is part of their job. And some academics have suggested that job crafting—in which organizations help employees expand and define their role to build skills—should include activities that involve helping others. Researchers have found that such structural arrangements promote employee cooperation.

Many real-world examples bear out our conclusions. For instance, changes at Microsoft in the past few years illustrate both the repercussions of dominance and the positive power of changing leadership. Steve Ballmer, former CEO of Microsoft, was known for his domineering approach. Under his management, the company lost a lot of ground to its competitors and suffered from a culture of fear and internal conflict. But company culture changed in 2014 with the arrival of its current CEO Satya Nadella, a leader known for his expertise and empathetic approach. His main focus has been to channel employees’ attention away from zero-sum thinking to a growth and learning mindset, which encourages people to accept both successes and failures as opportunities to gain insight that can benefit all involved. Microsoft has since seen record revenues and stock share prices.

Are you a scientist who specializes in neuroscience, cognitive science or psychology? And have you read a recent peer-reviewed paper that you would like to write about for Mind Matters? Please send suggestions to Scientific American’s Mind Matters editor Daisy Yuhas at pitchmindmatters@gmail.com.