By Sarah Cliffe
There’s sometimes a disconnect between how we talk about leadership qualities (we tend to use words like authority, power, and emotional intelligence) and what we actually require from the people leading teams and other working groups (arguably, competence and a deep knowledge of the specific work that needs to get done).
In a forthcoming Journal of Applied Psychology article, researchers from Stanford and Erasmus University explore which set of qualities matters most to team performance. The paper also looks at when power differences contribute to team success, and when they damage it.
I spoke with Stanford’s Lindred Greer about the research; an edited version of our conversation appears below. The other authors on the article are Murat Taraki (lead author) and Patrick Groenen, both at the Rotterdam School of Management.
HBR: What did you hope to learn from this research?
Greer: First, we wanted to understand when it’s ideal to have a strong hierarchy, and when it’s better to let groups manage themselves. People talk a lot about “holocracies” and self-management right now, but from a research point of view they’re largely untested.
Second, we were interested in investigating how good people actually are at recognizing good leadership. We teach our students about things like power poses – how to appear to be someone with authority – and how to fake it ‘til you make it. Those things are based on great research and they have real value, but are we losing sight of whether people actually have the goods? (David Dunning and Justin Kruger at Cornell have great research showing that the least competent people often end up in charge because they’re overconfident about their own abilities.)
In the first study, you simulated how well three different types of teams performed. The teams were searching collaboratively for the best solution to a complex problem. How’d they do?
In one group of teams, influence was aligned with competence: the person who knew the most about the task to be done led the team. These groups performed best.
A second group of teams shared power – they were relatively non-hierarchical. This group did not perform as well as the first, but they did outperform our third group of teams — hierarchical teams with a randomly chosen leader.
We replicated these findings in the field, by the way. We studied 49 teams at a publicly held Dutch company; the teams were auditing finances in search of tax evasion and fraud. If the team leader didn’t have a deep, technical understanding of tax fraud, he or she led the team badly astray.
The last study also looked in depth at how leaders get chosen. Tell us how that went.
Actually, this is a well-known exercise we do every year with students at Stanford. A team is given a list of items they can use to survive after a plane crash has left them in the desert. First they decide whether to stay or wait to be rescued, and then they rank order the importance of the items. We then compare their answers to that of a wilderness survival expert.
When doing this exercise in the context of our current study, one group of teams solved the problem collaboratively, without a leader. We compared these teams to a second group of teams whose members were asked to select a leader to manage group discussions, make final decisions if disagreements exist, and hand in the final rankings.
After 10 minutes of work, we took a break and the teams with leaders were informed publicly of how well each individual was doing, and how well the group was doing. Then they were asked to reconvene and rerank the items – but they could choose a new leader, as well.
This blows my mind: Only 55% of the teams chose the most expert person. Forty-five percent did not choose the most expert person, even though they knew who that was by now. Instead, they chose people who were, for example, taller, louder, or more confident.
Once again, the self-managing groups did better on the task than the teams who chose the wrong leaders, but less well than the teams with the most competent person in charge.
We’ve done this exercise for years, with similar outcomes. This means that I have watched people make poor choices about who to give a leadership role to, year after year. This is powerfully persuasive.
What lessons should managers take away from these studies?
One, we need to pay closer attention to how we choose leaders. Select for competence. Don’t get snowed by political connections or persuasiveness or the appearance of authority. Some of those things can be useful, but competence comes first. It’s essential to use objective measures of performance in hiring and promotion.
Two, pay attention to the dangers of formal hierarchy. For any important decision, you want to bring the best possible information to bear – which means you need leaders who are able to value the expertise of other team members and to share power when they’re not the person best suited to make a decision.
Three, It’s important to know who knows what. Often in a growth setting, like the ones here in Silicon Valley, you lose track. Every couple of months, a team should take stock of what challenges are most pressing, and who has deep knowledge that’s relevant to those challenges. That issue comes up in consulting engagements, too. The person who brought in the client isn’t necessarily the most knowledgeable about the work to be done: leadership roles are better assigned once you understand who knows what, and they may need to shift in the course of the project as new issues arise.
When do self-managed teams, or “holocracies,” work best?
I’ve got some interesting new research on that, but I’m not quite ready to share it. From this set of studies, here’s the lesson I see: If power can’t be aligned with the right set of competencies, an egalitarian team may be a good idea.
I want to go back to your emphasis on “competence.” By that you mean deep knowledge of the technical work being done, is that right?
Exactly. As a result of that depth, competent leaders can enable their teams to seek out new ideas and propose better solutions.
Are top-tier business schools overly focused on developing “leaders” who are ready to take charge of anything? It sounds as if what growth-oriented companies in the STEM fields need most is people who are best-in-class at the technical work, plus good at assessing other people’s strengths.
Both skill sets are important, of course, but we may be over-emphasizing generalized leadership qualities and under-emphasizing task competence. That’s a real risk.