Diversity Trumps Experience in Problem Solving

Many other articles on this site have reported evidence that diverse groups will figure out better ways to run a firm or other organisation than will a bunch of middle-aged upper-middle-class white men. But critics could argue that maybe many of those old-boys leaderships are simply incompetent or the diverse groups particularly brighter. No longer.

A study by University of Michigan and Loyola University researchers found that a completely random and thus more varied group of competent individuals will prove better at solving problems than individual experts or panels composed of people who are among the most qualified and competent at figuring out dilemmas.

How can this be? “This result relies on the intuition that, as the initial pool of problem solvers becomes large, the best-performing agents necessarily become similar in the space of problem solvers. Their relatively greater ability is more than offset by their lack of problem-solving diversity.”

In other words, at least in problem-solving mere competence combined with broad diversity will work out better than a smaller but more expert leadership. Broader thinking and more ideas can do better at selling more widgets or running a country than relying on a less creative but previously more successful group. And in some cases the best way to get that diversity and use it most effectively is to have a fully random set of competent leaders.

The authors are careful to make the obvious point that a truly random group of decision makers is not necessarily gender (or in any other manner) diverse; it is diverse because of the randomness. In fact, while they note comparative advantages of gender diverse groups, they also say such groups can be less effective (period and, presumably, compared with random groups) simply because of tensions that arise from the deliberate fostering of gender diversity.

So when does this reliance on random diversity really pay off? It depends on what form the diversity takes. The researchers distinguish between the approaches they call perspective (internally conceptualizing the problem and then trying to develop a solution) and heuristics (using logical and methodical processes). Many algorithms and formulae later, the researchers show that the random group does better the larger it is; but after a certain size, the group of experts also starts to improve, as they become, through expansion of their number, more diverse.

“Firms then may want to hire people with similar perspectives yet maintain a diversity of heuristics,” the authors conclude. “In this way, the firm can exploit diversity while minimizing communication costs.”