A review of the two standard cases for diversity

It has come to my attention that I should have images at the top of my articles.

We’ve all heard that diversity is important, and most of us know the two main arguments for it: The social justice case and the business case. In my next article, I’m going to share my very own third case, which I thrillingly call the “managerial case.” But first, I want to make sure we are all on the same page (that’s an amazing pun, as you’ll soon see) with these two prominent, established reasons for diversity.

The social justice case says that having a more diverse workforce is the right thing to do. This is because the lack of diversity in organizations in the US (and, you know, elsewhere) is a reflection of a long history of structural, systemic, and systematic disadvantaging, marginalizing, oppressing, and enslaving certain demographic groups. Some examples of these groups include: black people, Asian people, Latinx people, people with disabilities, working class people, gay people, anyone who identifies as anything other than a cisgender man, and many more.

Furthermore, the profound homogeneity we currently observe at the top of many organizations in the US is a reflection not just of this history, but of the ongoing manifestations of it in the form of inequality of opportunity and the corresponding lack of social and economic mobility that both stems from and feeds it.

You know, easy stuff.

The business case says that hiring, retaining, and promoting diverse talent isn’t just the right thing to do, but also it’s good for business. (In the corporate world they might say it’s good for the “bottom line,” but my academic elitism prevents me from using that phrase without embedding it in this strange, if not downright obnoxious, aside.)

Many smart people have written about the business case for diversity. One of them is my mentor Scott E. Page. I hereby acknowledge my bias in favor of his ideas, but I will share them anyway because I am nothing if not a fan of the person who shaped how I think about the world (hint: models!). Also, I think you’ll like them.

Scott developed something called the Diversity Prediction Theorem. This shows mathematically that groups made of up people with diverse perspectives outperform homogeneous groups, even of experts, when it comes to finding solutions to difficult problems in complex environments.

Because you all avidly read my previous article, you are already asking: What do you mean by “diverse perspectives”? Great question! We will talk more about cognitive diversity later, but for now let’s summarize it as bringing together people with different areas of expertise, lenses through which they view the world, and lived experiences to work on a particular problem.

Your next question is probably: What am I doing with my life? Would my childhood self be proud of me? Is it too early to have a drink? These are good questions, but I don’t have the answers to those. What I do have an answer to is: What do you mean by difficult problems and complex environments?

I’m so glad you asked! (Please, have a drink.)

A difficult problem is one that is multidimensional and where there is no obvious right answer. Concretely, it means a problem where there are a bunch of features you could optimize over, but it’s not obvious which one(s) they should be. For example, suppose you are designing a smartphone. Should you optimize over battery life, display quality, camera functionality, durability, speed, size (and even then, bigger or smaller?), something else, or some combination? Personally, given the number and scale of cracks on my phone, I recommend durability, and urge Fisher-Price to get in on the smartphone game (the real one, that is).

In contrast, to use an example Scott gives a lot, chopping down trees to form logs is hard (Exhausting! Requires skills I don’t have! Might die from an axe!), but it’s not difficult in this strict sense. I shall now brace myself for hate mail from the logging community.

All of this means that if you’re running an assembly line, diversity may not help, and could even make things worse (I don’t want a bunch of poets on my factory floor, either). But, if you’re thinking about how to optimize the setup of your assembly line, or what to assemble in the first place — then bring in the poets. And, heck, maybe even a modern dancer (maybe).

A complex environment is one that is dynamic and adaptive. The way to think about this is that the difficult problem your company is solving usually exists on what Scott calls a “dancing landscape,” or a changing context. For example, you might spend years designing the world’s biggest, highest resolution new smartphone, only to discover that by the time it’s ready for release, the market winds have changed, and everyone wants slim, foldable phones that never break no matter what. (Fisher-Price, I’m begging you.)

Part of what drives this complexity is something called endogeneity. In addition to being one of my favorite concepts in the world, endogeneity means one thing causes another thing, which causes the first thing. For example, the features of a smartphone might influence consumer preferences, which in turn influence the next generation of smartphones.

It’s a chicken-and-egg problem: What came first, better cameras on phones, or our own desperation for permanence manifested through the otherwise inexplicable act of voluntarily putting photos of our own faces on the Internet?

The business case for diversity says that diverse teams are what you want when you’re dealing with these kinds of problems because the range of perspectives means you’re more likely to spot and address the various dimensionalities of your problem and changing landscapes of your environment.

Again, easy stuff.

Now, I must call out three quick issues with the business case.

1. A cynical take on the business case for diversity is that it’s gross companies even need to hear about it, because doing the right thing should be reason enough.* But, until we figure out how to make “doing the right thing” aligned with “maximizing shareholder value,” I think it’s not bad to have some tangible reasons to spur companies to prioritize diversity.**
2. The push for cognitive diversity is something companies sometimes misuse as justification for a lack of diversity on other dimensions. I shall address this in a future article, but the short version is: Don’t do it. The medium version is: Identity diversity is part of cognitive diversity.
3. The business case has been studied empirically and the results are mixed. It turns out the benefits only pan out in real life if the cognitively diverse people are willing and able to actually work together. It also turns out this “if” is not trivial.***

This brings me to my third, “managerial” case for diversity, which is that the lack of diversity in a company is a symptom not just of socioeconomic problems in the country, but also of other processes going wrong inside that company that are allowing biases and second-order policy consequences to give rise to an imbalance of types of people at the top of companies. (Say it with me now: E-A-S-Y!) And, correcting the inputs that are giving rise to that homogeneity will also just so happen to make for more efficient and effective practices and policies along the way.

And that is where we will pick up next week.

Thanks, as always, for reading!

This is Scott Page’s Diversity Prediction Theorem. The math checks out!

This article is the third in a ten-part series on social science and diversity. Previous articles include:

1. A new series about diversity and social science
2. The importance of defining diversity (and how to do it!)

Questions? (Positive) Feedback? Comment below or contact me on social media. Remember, I’m insecure, so unless you bathe me in flattery (gross?) alongside your constructive comments, I might not be able to take it.

* Scott has also emphasized over the years that the two cases need not be at odds with each other — increasing diversity for social justice reasons can just so happen to yield material benefits to companies, too.

** I credit Seisei Tatebe-Goddu for opening my eyes to this language.

*** To be fair, a variant on this third one also plagues the social justice case: it’s not a great step forward if you hire a bunch of people different from you and then treat them like crap.