I was asked to act as scribe for the last session of a broad and quite interesting conference last week, whose express goal was a quantitative exploration of the value, role, and meaning of “diversity” in many settings: Not just the race-themed version familiar to those of us at Michigan since the recent Supreme Court case, but also biodiversity and its practical (and ineffable) meaning; the measurable advantage of diverse problem-solvers in business teams and stacked machine learning solutions; the role of social networks and diversity in the dynamics of ethnic violence and racism; the economics of reparations, and the difficulties of identifying who, let alone how much; what it might mean to try to capture concepts of “social power” and “justice” in a complex systems model; the points at which diversity becomes a problem; the places in interpersonal relationships where diversity is hidden by prejudice, not played upon by it.
I’ll write more about the conference after I finish some homework. But in the meantime, some of my homework involves polishing the notes I took, which I present here without much annotation.
These notes are extensively paraphrased from notes, not attempts to quote the speaker, except as noted by, well… quote marks.
Friday 11.11.05, 2:30-3:30 Session: “Interacting Across our Differences”
Speakers:
- Cindy Barton Rabe, Innovation Strategist, Intel
- Jeff Polzer, Organizational Behavior, Harvard Business School
- Katherine Williams Phillips, Kellogg School of Management, Northwestern
Intro [Jerry Davis]: Contrary to many folks’ preconceptions, business is in fact at the vanguard of multiculturalism and diversity. Not because of altruism or a sense of do-gooderism, but because there’s money in it.
Large manufacturers (e.g., Intel) make scads of money outside the US. They cannot afford to be ethnocentric or biased. Indeed, who are “they” that make up the corporation, if they’re multinational?
So: businesses have learned institutional patterns and structures to make themselves more diverse.
Speaker #1: Cynthia Rabe“Innovation and an Expanded View of Diversity: Connecting the dots for business”
Observed in a lunchtime discussion: Diversity tends to gain traction when it’s in the best interest of those in power. One powerful approach to making diversity in the best interest of those in business is linking it to innovation.
Business needs to expand its definition of the “diverse workgroup” to include multiple expertise: Especially smart non-experts.
For example, consider the case of disposable diapers: Experts at many consumer products manufacturing firms, by not listening to the inventor of disposable diapers (a “nonexpert” mother), turned down her offers of licenses and kept the invention from the public for a decade. The idea was patented by a mom in the 1950s. She took it to a number of manufacturers and asked them to market it, but they ignored her. The idea sat on the shelf for ten years. Finally, Victor Mills bought the patent, and introduced Pampers.
Point: Smart nonexperts can see things experts can’t.
Defining “innovation”: Application of an idea that results in a valuable improvement. Might be advertising, inventions, approaches — not limited to designs, inventions or patents.
According to a recent McKinsey poll: The single most important factor in growth was considered to be innovation.
According to Bruce Lehman, former US patent commissioner: In 20 years, India & China will be responsible for producing more patents than US.
There is an ongoing leveling of the playing field. The knowledge economy is here, and starting to slow. The next great resource will be creativity, and the next mode for business the “Creativity Economy”.
Yet US “creativity” is not keeping pace. The number of named inventors per patent has increased from 1.7 in 1975, to 2.3 in 1999. Larger teams need to be brought to bear, for some reason. Why?
Many businesses already get the idea and apply “diverse” teams, with people from manufacturing, finance, marketing, and purchasing. But these teams tend to act by splitting up the problem, so that everybody works only on their own field of expertise. There are fewer smart nonexperts working on any one component of the process, so in fact much less effective diversity is being brought to bear on the greater problem.
It’s clear that optimal teams for innovation are those that include varying degrees of expertise and ability regarding the problem at hand.
Examples
- Ford’s Model T — Ford brought in Flanders, an engineer from the Singer sewing machine company, where they used the idea of interchangeable parts. Ford until that time had not; cars were hand-crafted, essentially. This transformed everything at Ford.
- Bluetooth — When Kardash, an engineer, brought the name “Bluetooth” in to handle engineering side of SIG, he was invoking the historical symbolism of Edward Bluetooth, the Danish king who united the Scandinavian countries. Marketing however thought the name connoted “poor dental hygiene”; then again, they were used to technology names like “PCMCIA”. Kardash’s idea held sway by subversion, effectively in an end-run around marketing and into the techie public. Here again, a smart nonexpert added value, in this case invaluable marketing insight.
- The Gossamer Condor — The famous human-powered aircraft, winner of a substantial cash engineering prize. Paul MacCready, team lead, when asked why his team won: Everybody else on all the other teams was very sophisticated. Every other team had a specialist for each necessary discipline. But our wings were developed from first principles, not from specialist knowledge, and from insights from building model airplanes when we were kids. The other teams were so experienced that they never considered an easier approach like ours.
Most steps towards fostering innovation through diversity in expertise are intuitive. Others are more purposeful, e.g., Brighthouse; IDEO’s insistence on “T-shaped people” (who are very deep in one discipline, but very broadly experienced in others); Intel’s “Embedded alien” approach for introducing smart non-experts onto design teams; Microsoft’s “Smart Dummies” approach.
But how do we value the non-expert? How does one instill a culture in which non-experts are allowed to participate. Groups of experts are subject to “expertthink”: a lot like groupthink on steroids. Worse, there is a perception that including non-experts can compromise the “speed of business”, since it can be slower to work with people who have different training.
What’s needed is research to validate the link between innovation and diversity. We need to educate business on that link, and help them identify ways to foster innovation by adapting social structures. We need to close the theory-application gap. There are clear advantages, anecdotally, but the study of expertise diversity has a unique set of issues to address and overcome.
[end of talk]
Question: Question on Google and their employees 80%/20% rule for scheduling time: 80% on the assigned project, 20% on any useful thing at all. “Thinking time”, in other words.
CR: There’s a cultural problem when you allow calendaring time to “think about off-topic stuff”. That time will be seen as “unused”, and will rapidly fill with meetings and projects. Even if it’s not 20%, time not spent on-topic is seen as wasted. Do you think you can do it in an organization in which time-filling entrenched? We’ll see.
Question: One of the great unanswered questions we’ve seen at this conference is how to address power. Any non-expert in a group of experts doesn’t have power (or credibility) among the experts. WHat might be done?
CR: From an applications standpoint, there needs to be not only a basic language for discussion between experts and non-experts, but also explicit full support and buy-in of the team’s leaders and all team members. This is clearly a responsibility that rests on the leaders’ shoulders. Perhaps they will need to point out that the new person is being added to the team not a punishment for not being creative enough, but rather as a new tool to help them solve problems.
Speaker #2: Jeff Polzer, on the social psychology of diversity, collaboration and conflicts in small groups.
We’re talking here about small group interactions, where everyone knows everyone else. Nonetheless there can exist conflict, in terms of barriers and stress caused by differing ideas and values, or in terms of the self-censoring of ideas, even though some members might think they’re good, in order to avoid expected conflict.
The desired outcomes of these team processes are things like: joint decisions, creativity, productivity. The processes studied include the level of conflict and debate, the degree of information sharing, the type and level of behavioral integration, the capacity for joint decision-making, mode and degree of cooperation.
Some of the more important diversity axes for small groups (in academia and business) are tenure and cohort; there are of course many others, including race, sex, nationality, organizational, functional area of specialization, personality type.
One struggle we face: How can we capture the combinatorial aspects of diversity — say of a small group of six people, differing in eight dimensions from one another — when assuming that conflicts arise due to, e.g., different genders on this team? Is it true that one of many axes of difference inevitably trumps the others? [and Cf. next talk]
We can make some progress if we think in terms of social identity and self-categorization theory. Which groups do the team members identify with, and how do these groups become salient during the social dynamics of discussion?
Problems caused by diversity are rather well-studied. How can a team capture the value in diversity? What are the conditions under which the informational benefits of having multiple viewpoints will take hold, and the potential detriments not manifest?
In the studies described here, we’re looking at social norms as a moderator or ameliorator of diversity. This framing places commonality of opinion as a goal of individual members, but it should be pointed out that this may be in tension with the expected value the team as a whole derives from its diversity.
Business school students (at the University of Texas) are obliged to work in assigned teams throughout their entire tenure. It was noted that some were able to utilize one another, and as a result were able to be creative and bond tightly. Other teams just fell apart, and quickly.
It’s useful to think of these dynamics in light of William Swann’s self-verification theory. Essentially, people come into any social situation with a self-concept, and as they interact they are appraising one another and assigning roles to one another. A very salient question about the basis of team success is: Is a team member’s self-concept congruent with others’ assessment of them?
Study surveyed 82 teams, collecting data on the students’ self-views and their appraisals of teammates. The match of these was used to predict long-term survey measures on conflict, outcomes, grades.
Results in praecis: When there are low levels of congruence between self-concept and that assigned by others, diversity hurts. There is more conflict, and worse team performance. When there’s high congruence, diversity helps, especially in creative projects. Even very early measures — taken within fifteen minutes — were able to predict team performance outcomes four months later. These impressions were formed of one another very very quickly.
Ongoing followup work: Video analysis and interviews of team members throughout the process, to explore the dynamics of first impressions and how congruence may shift over time.
Some prescriptive advice from this: It’s wise to design a first meeting in order to foster understanding and respect for others’ expertise and identities, and their differing goals for time spent on campus. For example: Some degree of obligatory self-disclosure can help, such as every team member explicitly stating one of their own shortcomings. This can open up each member to the idea of taking risks with interpersonal images.
Three directions for future work:
- Fault lines and binding seams. There are potentially lots of dimensions of diversity in even a small group. One big question is not “what’s the most diverse group?”, but rather how do these many dimensions align [correlate]? Do they tend to break the group into two strongly-aligned sub-groups, or are they such that each members differs from the majority of the others? Strong alignment can lead to the strongest forms of inter-group conflict.
- What’s up with virtual teams? Organizations are putting people into teams these days, and in many cases the people are far apart and rarely (if ever) meet face to face. To date there’s but little research on these remote social dynamics, especially in terms of how other kinds of diversity interact with the big, physical kind caused by distance.
- Status and power matter. People are not all equal. In what cases can this be bad? In what cases might it be beneficial?
Speaker #3: Katherine Williams-Phillips “Value in Diversity? Joint Impact of Surface- and Deep-level Diversity”
She recently undertook a review of forty years of diversity research. This focused on how diversity impacts teams, and included work in the fields of psychology, sociology, and organizational behavior. Many aspects of diversity were examined, from tenure to race. The results: Cross-functional teams were the only place where there was a clear value from adding diversity to business teams. In all other cases, the benefits were unclear.
This leads to an obvious question: What impact will diversity have on the bottom line?
These studies focused on “surface level” diversity: difference in team members’ group, office location, race, sex, &c.
[quote from John Stuart Mill, 1848, on the obvious benefit of diversity in society]
Let’s look at the critical assumption in much of the Organizational Behavior literature on this subject: different people hold different perspectives; different perspectives lead to performance benefits.
This, in light of the studies’ focus on surface diversity, implies:
- Socially different individuals are the only individuals who can bring different perspectives to the table. People who look different are the ones who fight; people who look the same are supposed to agree with one another. But if you examine the dynamics carefully, you’ll realize that team members who are new, or less powerful, or who lack status tend to stay quiet, tacitly being more agreeable. Is it the case that where there’s conflict, it’s the different people who are fighting?
- The social majority will share the same perspective…. You think?
As it happens, different perspectives do not always come from where they’re expected to come from.
Consider the process by which we assume and maintain congruence:
- We categorize to simplify the world: things and people are like me, or not
- We prefer information consistent with these simple categories, which tends to reinforce our assumptions.
- Consider Heider’s Balance theory (1940s): We would rather agree with people we like. We would rather change ourselves (or others) to maintain congruence & social relationships, than change our conceptions.
We tend to underestimate the importance of dynamics on in-group members, and instead focus on inter-subgroup dynamics. We ignore the differences of opinion between people who look alike (in terms of tenure, job description, race, whatever), and emphasize differences of opinion between those who don’t.
As it happens, congruent groups tend to outperform incongruent groups. But also incongruence is better than homogeneous groups, where there are no surface-level differences between members. When people see surface-level diversity, they are prepared to believe there will be deep differences. Whereas where the group is homogeneous, the surface similarities mask deep differences; these are self-suppressed by group members because they are assumed to be illegitimate.
Interestingly, the existence of surface differences acts as a cue that makes it more likely that anybody in the group will be able to voice their differences of opinion, even when these occur in-group.
GROUP DISCUSSION (JP and KWP only)
Jerry Davis’s prefatory summary: There are clearly a number of interesting social mechanisms by which diversity, and the value that it might bring, can be suppressed.
Audience Q: It may be interesting to look at results on cross-race mentoring among faculty. Problems can sometimes be predicted by how much the pair think race should be on the table. If there’s a mismatch between how much the individuals think race plays a role (regardless of their individual races), it has been shown to cause problems in the social dynamics.
KWP: There was a related interesting discussion on social identity and diversity at a recent conference in Amsterdam. In brief, it seems that each individual has their own perspective on the value of diversity. To the extent there’s a match in a group, congruence of views can have a huge impact.
JP: There have been a number of cases of interventions in companies, under the assumption that it’s better to talk about cross-race mentoring and related issues. Unexpectedly, such open discussions can sometimes be helpful, but other times very destructive. It seems that the outcomes where there are deep differences in what people expect others to contribute tend to be worst.
Audience Q: What about team diversity when there’s already power differentials in the organization, or barriers to entry? Can there then be social disruption caused by enforced introduction of diversity?
JP: That’s not my research, as such. A lot of companies, especially global ones, do have a lot of experience with very diverse teams. it’s clear that they’re institutionally very willing to invest, but admittedly they have to figure out what to spend their money on.
KWP: In some cases, this idea of “bringing in” diversity can have unforeseen problems. For example, firms far outside big cities can have issues, for one reason because in essence they need to import people (for example, from outside the country, or from big city universities). While the firms understand this can convey benefits, the effort can substantial.
Conference wrap-up, kept cunningly and intentionally brief:
Scott Page: This is not a closing of the conversation. It’s an opening.
Subscribe to posts feed