Choosing Your Own Team?

Why self-selected teams strongly differ from those composed randomly

Rainer Michael Rilke - April 18, 2023

Tips for practitioners

It’s in our nature to prefer working with colleagues who share our sensibilities. Yet, within homogenous teams, there is often a pressure to follow majority opinion in order to maintain harmony and to avoid appearing as a disrupter or a “spoke in the wheels.” This can result in more suitable problem-solving approaches going unheard, which, in turn, leads the team to make suboptimal decisions. Recent research in this area has made it increasingly clearer that people in more diverse teams are more motivated, productive, creative, and aware of risk. So, it’s obvious how employers should compose their teams, right? Perhaps not!

Agile working methods lead to homogeneous teams

Many companies follow the trend of employing agile working methods to render their working processes more flexible. One essential aspect of these methods is allowing employees to compose their teams on their own or to determine of their own accord the colleagues with whom they would like to collaborate on any given task. This process is known as “self-selection” and is intended to ensure that employees more strongly identify with the team and their tasks, theoretically leading to increased productivity and better end results.

Composing diverse teams is in direct conflict with the human tendency to surround oneself with people that have interests, personality traits, and abilities similar to their own. As the old saying goes, “birds of a feather flock together,” a concept that social psychology calls “homophily.” On the one hand, members of self-selected teams better identify with their tasks and the group as a whole; on the other hand, they lose sight of diversity and often overemphasize their homogeneity.

Who is more successful in their endeavors—self-selected or randomly selected teams?

In which cases will self-selection have a positive (or negative) effect on the team’s productivity? In this experiment, participants had to work in pairs for two months, completing different tasks throughout. Half of the participants were allowed to choose the person with whom they wanted to work, i.e., through self-selection. The other half were assigned their partners at random. Intriguingly, the earliest results of the experiment came during this process, before the teams had even assumed their tasks. Reflecting our homophilic nature as humans, the self-selected teams often comprised members of the same sex and with similar cognitive abilities; their counterpart teams were more varied in both regards.

Depending on how the teams were assembled, and on the task at hand, there were clear differences in the results of their work. In the experiment, each duo had to complete two different tasks in different formats. For the first, they had to complete and hand in a worksheet with written solutions; for the second, they had to film a short video presentation on their findings. The results were astounding: While the randomly composed teams produced better results for the written task, the self-selected teams soared when it came to the video task.

It was revealed that, on average, self-selected teams perform better in tasks that require a high level of coordination and a strong sense of teamwork. For an assignment such as the video task, for example, homogeneity works to a team’s benefit, as the two members understand each other with ease and are quicker to agree with each other. By comparison, randomly assigned teams produced stronger results for tasks that required little coordination or teamwork and for which the individual abilities of the partners were of more importance. When such is the case, the randomness of their team composition most often ensures that there is at least one highly capable person on board. When it comes to self-selected teams, this is not always the case, as it is just as likely that two weaker participants could find each other as could two stronger ones.

Tips for practitioners

  • Remember that birds of a feather flock together. Employees choosing their own teams will inevitably lead to a lower degree of diversity overall. That being said, in being less diverse, they will also be similar in their cognitive abilities.
  • Take note that the degree to which either a self-selected or a randomly selected team will be productive is highly contingent on the task at hand. In practice, you should evaluate what the task demands: If it requires more coordination and alignment (and less cognitive ability), employees should be allowed to choose their own teams. If, by comparison, the opposite is true, then the manager should build the teams randomly.

Literature reference and methodology

  • Fischer, M., Rilke, R.M., & Yurtoglu, B. (2023): When, and Why, Do Teams Benefit from Self-Selection? Experimental Economics 26, 749–774 (2023). doi.org/10.1007/s10683-023-09800-2.

Co-Autor of the study

Assistant Professor Rainer Michael Rilke

Rainer Michael Rilke is an Associate Professor of Business Economics at the IHK-Chair of Small and Medium-Sized Enterprises. He researches and teaches in the field of experimental economics, focusing on the analysis of human behavior in social contexts. In his work, he looks at topics relevant to loyalty, lying, and corruption as they relate to team incentive, leadership, and proposals made by artificial intelligence. With his research, he seeks to offer insights into the factors that shape human behavior within social and economic circles and how to incentivize people such that they make ethical and efficient decisions.