Leadership selection methods: why random selection outperforms the "best" approach

The experiment: four leadership selection methods compared

The researchers designed a series of experiments where small groups had to solve a survival task together. Think: your plane crashed in a frozen wilderness, and you need to rank a set of items in order of importance for survival. Rankings were then compared to those of actual survival experts.

But the real experiment wasn't about survival skills. It was about how each group's leader was selected.

Across multiple experiments, the researchers tested four different leadership selection methods:

Formal selection. Participants filled out a leadership skills inventory, scores were tallied, and the highest scorer was appointed leader. This is the closest to what most traditional organizations do: assess, rank, appoint. It's also the approach most susceptible to the Peter Principle. Research from Mercuri Urval confirms this is the dominant model, while also noting that roughly 50% of leader appointments made this way end up failing.

Informal selection. The group was simply told to decide among themselves who should lead, by whatever means they saw fit.

No leader. These groups just received the task and got on with it.

Random selection. The person whose last name came first in alphabetical order was appointed leader. No assessment. No discussion. Just the alphabet.

What everyone predicted vs. what actually happened

In a separate study, the researchers asked people to predict which leadership selection method would produce the best results. The predictions were near-unanimous: people expected formally selected leaders to perform best and randomly selected leaders to perform worst.

Makes sense, right? Leadership matters. And picking the most capable person should give your team the best shot.

Here's what actually happened.

In the first experiment, random selection came out on top. Informal selection landed in the middle. And formal selection came in last. The difference between random and the other two was statistically significant.

The researchers then ran a second experiment, replacing informal selection with a no-leader condition. Again, random selection produced the best group decisions. And groups with formally selected leaders performed about the same as groups with no leader at all.

All that effort to identify the best leader? It delivered the exact same result as not bothering to pick one.

Why formal leadership selection methods backfire

The researchers offer an explanation rooted in social identity theory. When you formally select a leader based on individual competence, the selection process itself draws attention to differences between group members. It signals that one person is above the rest. Research on hierarchy in teams confirms this dynamic often does more harm than good.

That can erode the group's sense of shared identity and togetherness. The selected leader may start focusing on proving they deserve the role. The rest of the group may disengage.

This aligns with what leadership psychologist Tomas Chamorro-Premuzic has long argued: organizations tend to confuse confidence with competence and mistake extroversion for vision. Formal selection processes are particularly vulnerable to what Mercuri Urval describes as three fundamental flaws: over-reliance on superficial impressions, stereotyping based on generic leadership models, and the subjectivity of decision-makers.

When a leader is randomly assigned, none of that happens. There's no implied superiority. The group stays a group. Members feel more involved and contribute more actively.

Perhaps most painfully: in the second experiment, formally selected leaders reported enjoying the task more, believing they were more effective, and claiming they put in more effort. The leaders most convinced of their own effectiveness were the least effective ones. A classic case of the HiPPO effect in action.

What this means for leadership selection in practice

The researchers are careful to note that random selection isn't a universal fix. They suggest it's most beneficial when a group has a clearly defined shared goal and is disposed to work in a democratic, collaborative way. In other words: environments where distributed decision-making is already the norm.

That said, the findings challenge the deeply held assumption that more rigorous leadership selection methods automatically lead to better outcomes. If anything, the research suggests that the process of elevating one person can be the very thing that undermines team performance.

One practical alternative is distributing leadership across multiple roles rather than concentrating it in a single person. For example, some organizations split coordination into three distinct functions: operational decisions, people and culture topics, and strategic direction. Team members then step into whichever coordinating role fits their specific talents. This is one of the core principles behind self-managing teams.

This approach shares something with both informal and random selection. The team decides, but because leadership is split into multiple roles, it becomes less likely that one person gets elevated above the rest, which is exactly the dynamic the researchers identified as harmful.

Rethinking how we select leaders

The century-old science of leadership assessment has given us increasingly precise tools for predicting who will succeed in leadership roles. Psychometric assessments, structured interviews, and data-driven evaluation methods all have their place.

But this research adds an important nuance: the selection process itself shapes group dynamics. Even when the "right" person is chosen, the act of formally appointing them can undermine the very collaboration that makes teams effective.

So the next time your organization is deliberating over leadership selection methods, consider this: maybe the most sophisticated approach isn't always the most effective one. Sometimes, the alphabet works just fine.

Ready to move beyond traditional leadership?

This research makes one thing clear: the way most organizations select and structure leadership is working against them. But knowing that is only step one. The real question is: what do you replace it with?

In our 6-week Masterclass: Progressive Organizational Design, you'll learn exactly how pioneering organizations have redesigned leadership, from distributed roles and team-elected coordinators to the structures that make it all work at scale. Through 5 real-world case studies and live sessions with the pioneers behind them, you'll walk away with a complete framework you can apply in your own organization.

Join 1,500+ leaders and changemakers who already did