Sun 8 Feb, 2009
The role of diversity in group decision-making
Comments (7) Filed under: Remedies, Thinking about race, privilege and inequalityTags: Class, Diversity, Gender, Race
Columnist Nicholas Kristof of the New York Times writes today about the overwhelming domination of Wall Street by male executives.
I’m not sure that I agree with Kristof’s conclusion that what the global banking industry needed in order to avoid its current woes was “women, women and women.” However, he devotes most of the column to highlighting important research showing that in areas such as race, gender, and class, diversity improves the quality of group decision-making.
This research offers a distinct rationale for diversity in education and in the workplace, beyond questions of fairness to the individuals involved or other arguments about diversity which may not garner universal agreement.
This particular justification for diversity is also more palatable to many of those who are skeptical of affirmative action or multiculturalism, being focused on generating measurably superior outcomes for the entire institution or for society as a whole. Moreover, this approach defines diversity in a subversive manner: it assumes that diversity today means having different experiences and perspectives, while giving no credence to beliefs that there are fundamental differences between people on account of race, ethnicity, gender, or other superficial traits.
Kristof points to several studies which highlight the role of gender in specific decision-making contexts. One set of researchers, for instance, found that men (but not women) were more likely to make high-risk decisions when their audience was of equal social status, which the authors suggest may be an evolutionary response to the need to compete for females.
He also reviews an article by Lu Hong and Scott Page from 2000, in which they created a formal model to analyze the superior decision-making abilities of diverse groups.
Kristof could have mentioned other important work on diversity by Hong and Page, including a 2007 paper on decision-making by agents using generated and interpreted signals and a notable 2004 article in the Proceedings of the National Academy of Sciences which argued that when choosing a group of decision-makers, randomly-chosen but diverse participants are likely to out-perform elite participants with similar backgrounds.
Scott Page is also the author of The Difference: How the Power of Diversity Creates Better Groups, Firms, Schools, and Societies (Princeton University Press, 2007). His argument is that group decision-making is best when the group reflects unique perspectives and ways of thinking. The key, according to Page, lies in bringing together individuals with different backgrounds and life experiences; the result will be a messy process, but one which out-performs groups of like-minded individuals chosen on the basis of extensive experience or high IQ.
bobbo says:
I hate articles like this one: nothing is defined. Maybe we all know what a woman and a man is, but what defines all the other terms used?
What is a "better performing Bank?"
"The problem" with most integrated acitivities is that most often the women come to act just like the men. If everyone is responding to the same motivations and penalties wouldn't that be expected. It should be the "rare" case where the sex differences actually play out.
Take the US Military. Have women made it a kinder/gentler/"better performing" military==and how do we define this better military?
So—I'm thinking yes, anything with kiddies would show a difference and better or not depending on what you think of kiddies. The sex scandal of the Catholic Church likely would have avoided if women were priests and/or priests married. Those folks aren't going to put up with child abuse. Single Males do.
Likewise with voting. Women more likely to vote in favor of child oriented policies. Thats only good if you define it that way.
James says:
It should be the “rare” case where the sex differences actually play out.
And yet that's not what studies show, nor what the formal models predict should happen.
The reason is that the role of diversity in group decision-making isn't that members of the group will face different "motivations and penalties." The professional incentives are assumed to be identical. The difference is that that diverse members will have different backgrounds and perspectives, which will cause them to evaluate issues differently. The assumption is that this diversity will help the group to avoid such decision-making pitfalls as "groupthink."
bobbo says:
James–I don't see any example of sex/diversity making a difference. I don't see any difference in the military which is about the "best" example of an all male establishment being taken over.
Same with banks. The motives and incentives rewarded are high risk/high profit. The women will act just like the men or get weeded out.
The notion that one can create an artificial study group and be given an artificial challenge and the higher diversity group will do better than any single sex group is not surprising==the test was designed for that outcome.
I'll restate my main concern: What is a better performing bank? Military? Congress? Its all definitional, and no definitions are given.
James says:
I don’t see any difference in the military which is about the “best” example of an all male establishment being taken over.
Bobbo, there's a difference between asking whether group decision-making is improved by diversity, and looking into whether the U.S. military performs better with women than without. In any event, there's been a lot written about the inclusion of women in the military, and the links in the post will take you to references to the literature on diversity and decision-making.
The motives and incentives rewarded are high risk/high profit. The women will act just like the men or get weeded out.
I think you need to focus on the fact that this entry is entirely about the quality of group decision-making. The question is whether groups, such as bank executives, will make better joint decisions with diversity than without, not whether they will maintain an incentive structure that will cause women to be weeded out.
In fact, though, there is fascinating research which shows that male traders will engage in riskier behavior in the presence of their equals. This suggests that women will actually perform better under the reward structures of banking and finance, because the degree of risk in their decision-making won't vary with an irrelevant factor like the social standing of those around them.
the test was designed for that outcome
What artificial study group are you referring to, bobbo? I'm not familiar with any studies that created artificial tests weighted toward this outcome.
What is a better performing bank? Military? Congress? Its all definitional, and no definitions are given.
Where are you looking for these definitions, bobbo? I think it's fairly clear what Kristof meant by a better-performing bank, even if his notion wasn't terribly precise or capable of accurate measurement. Certainly the research in this area has been careful to employ precise definitions.
bobbo says:
In fact, though, there is fascinating research which shows that male traders will engage in riskier behavior in the presence of their equals. This suggests that women will actually perform better under the reward structures of banking and finance, because the degree of risk in their decision-making won’t vary with an irrelevant factor like the social standing of those around them. /// I can believe that. Could it also mean that any other group will perform better than macho jock straps? (smile)
I didn't find any links in the links above. The PDF document was overly academic, indecipherable, and therefore to painful to read to see if it ever had anything interesting to say.
I have participated in group studies where different teams are put together to solve certain problems. EG–you are on a moon base under a plastic dome that has cracked and you have 24 hours to get to the next dome that is safe. You are then given a bunch of equipment, maps and other role playing information. ((Yes, just like a GAME!)) The point of the exercise is always to include every member of the team as everyone has valuable input to make. Everyone has a good laugh when top management is always bested by the janitor team.
Maybe not all studies are like that?
I think we both agree there are some situations where diversity creates more options, and others where it does not?===or maybe some situations where options are not sought?
James says:
I didn’t find any links in the links above.
Most academic papers, bobbo, contain traditional citations to the literature, not hyperlinks to online content.
As for the difficulty in reading academic working papers, I'm afraid that's simply an inherent challenge in reading specialized professional materials.
Maybe not all studies are like that?
Oh, definitely not, bobbo. Although it's not uncommon to find researchers designing carefully-controlled experiments with test subjects.
Those sound like fun exercises. Were they research studies, or team-building exercises? They sound like the latter, especially since your impression is that they were designed to teach participants a lesson, rather than as neutral experiments to learn how people behave or what the outcomes of different behaviors are.
This may be why you suggested, above, that studies are designed to produce particular outcomes. You may be thinking of team-building exercises and similar activities.
I think we both agree there are some situations where diversity creates more options, and others where it does not?
I think that seems quite plausible, bobbo. I think the key is to have rigorous research to learn under what conditions diversity does, and does not, improve group decision-making. Common sense can be misleading, as shown by the research above which indicates that choosing a group based on merit (just for example, those with perceived high IQs) can be less effective than seeking diversity.
bobbo says:
James==again, you are quite right. It was team building–not a study. Confused by something "close" but not the same thing.
I wonder how often that happens? Embarassing, really.