Key takeaways from the 2022 Meeting of Minds

Most business leaders and investors agree that diversity is important. It’s near the top of the agenda not just as a moral imperative, but also because mounting evidence suggests that it creates lasting value for firms. However, the mechanisms by which diversity creates value are poorly understood, and, perhaps as a result, progress on diversity is slow.

To break the impasse, we set out to develop a better understanding of the nature and value of diversity, and the conditions under which its benefits manifest themselves. For this, we went beyond the realm of business, assembling thinkers from academia and public institutions alongside corporate leaders for a one-day, in-person discussion at our 2022 Meeting of Minds:

• Delia Baldassarri, Professor in the Department of Sociology at New York University
• John Chisholm, CEO and founder of John Chisholm Ventures
• Susan Fitzpatrick, President of the James S. McDonnell Foundation
• Cathy Kling, Professor of Environmental, Energy, and Resource Economics at Cornell Dyson School of Applied Economics and Management
• Richard Lenski, Professor of Microbial Ecology at Michigan State University
• Mathieu Lefevre, CEO and co-founder of More in Common
• Simon Levin, Professor of Ecology and Evolutionary Biology at Princeton
• Daniel Levinthal, Professor of Corporate Strategy at Wharton School of the University of Pennsylvania
• Frida Polli, CEO and co-founder of pymetrics, Chief Data Science Officer at Haver
• Martin Reeves, Chairman of BCG Henderson Institute, Managing Director and Senior Partner at BCG
• Corina Tarnita, Professor of Ecology and Evolutionary Biology at Princeton

The group set out to tackle the following key questions:

• What is diversity and how can it be measured?
• How can diversity create value?
• What are the costs of diversity?
• Which conditions need to be created for the net benefits of diversity to be maximized?

Below, we summarize the emerging insights that our group of thinkers discussed, as well as the potential implications for businesses.

What is diversity and how can it be measured?

In business and academia alike, diversity is often conceived of narrowly—as differences in gender, race, and social background—but it is useful to think of it more broadly. There can be diversity of inputs (this is the traditional view, for example, a student body diverse in terms of gender or race), but also of outputs (the career paths of alumni). There can be diversity within institutions (a college offering many different majors) and between them (one school providing a different offering than others). Further, there can be diversity of roles (a college with highly specialized majors) and diversity of selection or metrics for success (individual departments measuring results differently).

Listing these kinds of diversity is not just about semantics. Rather, clarity on the kind of diversity an organization is aiming for, and which others they are willing to deprioritize, is important because the different kinds of diversity do not necessarily coincide. On the one hand, homogenous input can lead to diverse output. In both workplaces and academia, individuals who look alike can be very diverse cognitively and intellectually: They may be abstract vs. concrete thinkers, risk-averse vs. risk-neutral, or have long vs. short time horizons. Even in nature, different specializations, from forager to nurse, emerge in small groups of genetically and demographically identical ants through task allocation. On the other hand, diverse input can lead to homogenous output. For example, intentionally diverse foundations end up funding similar research due to standardized grant application and selection processes. Finally, interlinkages between types of diversity also need to be considered. Groups diverse along one dimension may still be homogenous across others.

Further, while diversity can be managed, it can also emerge (arise naturally) or be inhibited by the structures and environment of an organization. For example, a diverse set of employees could be pigeon-holed or locked into their identities or backgrounds, precluding diversity of career destinations.

Business leaders need to be clear about which kinds of diversity they want to achieve. Towards this end, they must be aware of the axes on which their firms can meaningfully diversify and the interactions between those axes.

How can diversity create value?

Many studies link diversity within organizations to positive outcomes. Some also offer hypotheses for the mechanism underlying this effect. Our thinkers discussed how observations from their own fields supported hypotheses from business:

Diversity can foster innovation. In biology, the mechanisms for innovation are the same as those for diversity: Genetic mutation and recombination create diversity in species that allow them, over many generations, to adapt to new habitats, overcome antibiotics or immunity, or find new food sources.

In organizations, humans are the innovators, and evidence suggests that groups of problem solvers with diverse perspectives are more effective than individuals or homogenous groups. The latent potential of individuals may only emerge once they are in a group: Scholars have looked at genomics to figure out the origins of multicellularity and realized that 80% of what we thought was multicellular innovation was already present in single-celled organisms—we just did not see it until they were in a group.

BCG has found similar trends for businesses: Companies with above-average total diversity, measured across six dimensions (national origin, industry, career path, gender, education, and age) achieve 19% points higher innovation revenues and 9% points higher EBIT margins, on average.

Diversity can enhance resilience. Intellectual/cognitive diversity makes committees, boards, and workgroups better decision makers and innovators. That innovativeness enables organizations to better adapt to and thrive in changed circumstances—to be resilient to shocks.

In biology, diversity enables individuals and species to adapt to stress. In humans, an extremely steady heartbeat is a warning sign for serious health problems; continual variation in rhythm is healthy and allows the body to rapidly adjust to stress or change. Diversity also enables species to find a successful niche after a shock: A fire in Yellowstone National Park in 1988 caused aspen trees in the region to reproduce sexually for the first time in many generations.

For companies, resilience is critical not only to survival but also to competitive advantage: The ability to outperform in unfavorable periods accounts for nearly 30% of long-term outperformance.

Diversity can foster trust and prevent polarization. Sociology traditionally teaches us that homogenous communities are more trusting because they have stronger relationships, based on more commonalities. Now, our society is less homogenous and more polarized than ever. And new insights suggest that, from a starting point of some heterogeneity, more diversity can bring us closer together—under certain conditions.

Specifically, as long as the axes along which groups are diverse are independent of one another (rather than co-varying), diversity creates more opportunities for individuals to cross-cut and discover common ties with others, thus enhancing trust and altruistic behavior beyond in-group boundaries. Think, “We don’t bowl together anymore, but we still work together.”

Diversity can also contribute to limiting polarization by closing perception gaps—the gaps between what we imagine opposing groups believe, and what they actually believe. By engaging with a wider range of news sources, with people who think differently, and with people who have different life experiences, diversity helps us understand others and thereby decreases polarization.

What are the costs of diversity?

There are also very real costs of diversity, which decision-makers must be aware of to have a chance at mitigating them.

Diversity increases coordination costs. In organizations, diversity of people, processes, structures, and strategies increases complexity, and thus the costs of coordination.

Diversity reduces economies of scale and experience. In businesses, variation reduces short-run efficiency by diluting economies of scale (the decline of marginal costs with volume) and economies of experience (the decline of marginal costs with cumulative volume).

Diversity can create polarization or fracture. When the heterogeneity of beliefs across groups increases beyond a critical threshold, social polarization or even fracture can result. This is a particularly big risk if interactions among groups are limited, and if they have little else in common. The canonical business example is high-income, urban executives ensconced in corporate headquarters and low-income, rural workers in the same company’s less-attractive facilities.

Diversity increases near-term uncertainty. By nature, diversity is unpredictable and may require relinquishing some control, which is uncomfortable for many leaders. Further, we know that many of the benefits of diversity come from mitigating long-term uncertainty through innovation and resilience, so it can be extremely hard to predict how and when those benefits will manifest.

Which conditions need to be created for the benefits of diversity to be maximized?

We now know some of the mechanisms by which diversity can create and destroy value. But neither of these effects is guaranteed. Rather, the extent of value creation depends on the context.

For example, diversity helps organizations adapt to and succeed in changing environments. During periods of significant change, innovation, resilience, and trust are all critical, and diversity becomes a more important driver of each. Think again about the aspen trees in Yellowstone. It is in part because of their diverse “skill sets”—their ability to reproduce either sexually or by cloning—that they were able to find new habitats in the park after the fire.

While the degree of external variation is hard to control, our group of experts discussed a variety of other conditions that can be managed to optimize the benefits of diversity.

Create robust variation and selection mechanisms 

In nature, the rate at which evolution takes place is a product of the amount of variation (through mutation and recombination) and the strength of the selection mechanism. Across businesses, these pressures naturally emerge. Each firm develops with some level of uniqueness relative to the market, and then market and competitive pressures decide whether it survives.

Unlike organisms in nature, firms can deliberately foster variation and selection, not only in response to market and competitive pressures but also in anticipation of them. Fostering variation can take the form, for example, of periodically reshuffling teams or increasing variation rates when the need for innovation increases. The landscape explored from a random walk with smaller and larger step sizes provides a good analogy for serendipitous discoveries from innovation. Fostering selection could take the form of budgetary and performance-review incentives for outcomes that benefit the company. Engineering these mechanisms is a balancing act—leaders must make sure they are strict enough to benefit the company, while still preserving room for creativity and serendipity. In order to finetune this trade-off, leaders can take inspiration from biological principles:

Avoid winner-take-all selection. In biology, gene combinations that best fit a particular environment will proliferate and eventually dominate. COVID-19 has an incredible ability to evolve exactly because it mutates so much, but each wave brings a new dominant strain, which wipes out the previous viral winner. Similar phenomena can happen in companies, where diversity allows a firm to find success in adversity. But pursuing the most successful idea exclusively is fraught: The firm loses any further diversity, putting its ability to further adapt and survive at risk. One way of countering these tendencies is to periodically reshuffle groups, creating opportunities to work in small, generalist teams.

Leave room for bad or neutral moves. Just like leaving room for diversity itself is crucial, so is refraining from shutting down past innovations that at first seem unsuccessful. Instead, leaders should recognize that these perceived failures may create optionality in the future. For example, researchers found in a 75,000-generation experiment with bacteria that only one of the twelve populations studied evolved the ability to use a secondary source of energy that was present throughout the experiment. Initially, this ability was so inefficient that it provided no measurable fitness advantage. However, many generations later, descendants acquired another mutation, which refined the new ability and conferred an enormous benefit.

Embrace uncertainty and serendipity. Evolution is a process without foresight—there is no reason or goal in gene mutation or recombination, only a random change and an opportunity to test it in the environment. In fact, most genetic mutations in any species have no effect on a species’ ability to thrive, and most of the rest are harmful; only a select few benefit the species. Similarly, the benefits of resilience become apparent only when the environment changes and old strategies of survival become obsolete. For this reason, short-term perspectives that reward the best fit with the current environment will always undervalue diversity. Instead, businesses should embrace uncertainty, even at the expense of efficiency. Firms should consider going as far as having multiple groups working on the same important problem to diversify approaches.

Develop robust evaluation approaches 

Because of the long-term nature of many costs and benefits of diversity, as well as a tendency for impacts to unfold under uncertain circumstances, methods for evaluating success are required that go beyond our standard short-term metrics of productivity:

Real options valuation. Instead of deciding whether to pursue a project by only looking at the immediate gains, assess its value based on what other projects will become possible upon its completion (for example, an R&D project that, if successful, will let you launch a new line of products) and how valuable they may be. The framework for such an evaluation is known in corporate finance as real options valuation.

Co-optimized multi-objective functions. The field of environmental economics has faced similar challenges in quantifying the trade-offs between the more-immediate costs of getting an environmental policy passed and implemented, and the longer-term benefits once it is in place. Here, one can evaluate options by optimizing multi-objective functions within a defined set of constraints.

Empirical observation. More radically, instead of evaluating and selecting projects, people, or options ex ante, develop robust empirical observation methods to evaluate success post hoc. This is, in effect, what natural selection does. There is no team selection before the game starts, but survivors rise to the top nevertheless. AI is making this option more realistic: Already, anti-bias algorithms, including the one pioneered by one of our participants, allow for companies to use unbiased evaluation metrics to provide a comprehensive profile, against which firms can track employee success and eventually select new hires. Likewise, another participant advocates for the use of AI in evaluating post-collegiate success, and then using the insights to have AI evaluate future applicants.

With its “Micro-masters” programs in such areas as supply chain management and business analytics, MIT is using empirical observation in admitting master’s degree candidates. Thousands of learners around the world take the one-year program’s first semester online for free and can then opt to take the first-semester final exam at modest cost. The learners who perform the best on that exam are admitted to the master’s degree program and invited to come to MIT for the final semester. In effect, applicants must perform exceptionally on half of the one-year degree program’s course work before they are accepted into the program.

Embrace modular org design 

A modular organization of loosely linked parts can enhance diversity by allowing specialists to hone their skills within modules, while generalists maintain connective tissue by talking to everyone, facilitating the spread of information and consensus-building.

Modularity also promotes polyarchy, whereby power is distributed across modules, allowing for more diversity in structures, methodologies, and perspectives across teams. Further, it allows for recombination with high variability across teams while decentralizing intelligence, which reduces risk from rapid information spread.

At universities, for instance, this is implemented by maintaining specialized departments that are connected via interdisciplinary groups or institutes. This design also implements the principle of polyarchy, as there is no clear hierarchy among the departments.

Diversity is important and impactful, and leaders across business and society are striving to accelerate progress. In so doing, they will benefit from a better understanding of the nature of diversity, as well as how and under which conditions it creates or destroys value.

The perspectives from leading experts across different disciplines brought forward in the Meeting of Minds can serve as a crucial unlock by broadening perception and deepening understanding of diversity.

Further reading

Participants prepared written provocations related to their unique expertise in advance of the Meeting of Minds discussion. These were used to seed the conversation and enable focus on points of intersection, commonality, and debate across fields. The provocations can be read here.

About the Meeting of Minds

The Meeting of Minds is a multi-disciplinary meeting of leaders in business and science discussing major issues in society and business, hosted by the BCG Henderson Institute. Previous meetings discussed what we learned from the COVID crisis, strategizing and managing on multiple timescales, commitment and flexibility in strategy, learning and forgetting in the age of AI, collective imagination and innovation, and the “meta-problem” of climate change.