I see a lot of misuse of statistics, some deliberately dishonest, on all sides of arguments. As a Statistics graduate I think this should be a criminal offence, of course!
In particular now, we need fair and accurate narratives to avoid encouraging polarisation within society and within organisations’ working populations.
The large consultancy firm McKinsey position themselves as data-driven thought leaders on many topics. As a result many organisations follow their conclusions. But ordinary readers should never be unthinking; look at the workings carefully, before accepting conclusions. Learning basic economics and statistics is extremely helpful in maintaining an accurate view and avoiding being led astray with false or exaggerated narratives.
I&D (as McKinsey call Inclusion and Diversity) is one of McKinsey’s topics, with their longitudinal study of gender and ethnic diversity from 2014, 2017 and 2019 (published May 2020). A link to the Report is below. After glowing narratives about the compelling business case for gender and ethnic diversity:
“The business case for diversity, equity, and inclusion (DE&I) is stronger than ever. Taking a closer look at diversity winners reveals what can drive real progress.”
The facts provided by McKinsey on gender diversity are:
“Our 2019 analysis finds that companies in the top quartile for gender diversity on executive teams were 25 percent more likely to have above-average profitability than companies in the fourth quartile—up from 21 percent in 2017 and 15 percent in 2014 (Exhibit 1).”
Now, there MAY be something here to be said for a correlation between gender diversity and profits (see *footnote for exception conditions), but if so, it is a small effect. We see (turning the 25% on its head), that the least diverse organisations have a 75% chance of not having below-average profitability than the most diverse due to their lack of diversity. So roughly, three-quarters of the time it probably makes no difference, or may work slightly the other way.
So McKinsey are not exactly compelling, though certainly enough of an effect to be worth thinking about, especially as:
the greater the representation, the higher the likelihood of outperformance
– again subject to the possible ‘cause-and-effect’ fallacy.
But with this rare or small an effect, it’s clear each organisation would need to tailor its considerations very carefully in context, to make gender diversity approaches work. And that is perhaps the main point we can learn from looking at this report. Each organisation should get the right facts, and consider carefully its potential approaches by having representative conversations before embarking on a potentially wasteful or counter-productive action (see Lily Zheng’s Medium article on the polarisation caused by many initiatives).
McKinsey describe this result as:
Our latest analysis reaffirms the strong business case for both gender diversity and ethnic and cultural diversity in corporate leadership—and shows that this business case continues to strengthen. The most diverse companies are now more likely than ever to outperform less diverse peers on profitability.
That’s a bit strong. With 2014 results being cited as a strong business case, when they showed an 85% chance that (roughly speaking) gender diversity makes no difference, the ‘stronger than ever’ narrative is built on low foundations. It’s not that there is no effect; it’s the fatigue from hearing such hyperbolic rhetoric that creates push-back and stalls otherwise-good intentions; or creates sub-optimal policies and approaches based on the rhetoric, not the facts.
And that’s sad for progress towards what we all want – good function and fairness.
Another (worse) example in the McKinsey Report is the rhetoric on ethnicity in Executive Teams and Board. McKinsey say:
This lack of material progress is evident across all industries and in most countries. Similarly, the representation of ethnic-minorities on UK and US executive teams stood at only 13 percent in 2019.
Er…what’s the proportion of ethnic minorities in the UK population? Just under 13%. Come on, McKinsey, now your rhetoric is straining reality and therefore being counter-productive in winning over ‘traditional majorities’ to ally in creating equality of opportunity.
But that’s the problem with Applied Critical Theory – by taking only the worst of ‘old systems’ and turning it around, it wants unbounded power, not fairness. And hence makes push-back to itself inevitable. We can all do better.
The McKinsey report is not terrible, but their story is magnified unhelpfully. This develops polarisation of opinion – those who want to use the rhetoric to make change; and those who oppose the inaccuracy and unfairness of the rhetoric and so push back. This is commonplace in 2022 in the UK and McKinsey have a responsibility to address this in its own work.
In general, as a ‘Stat’ myself, I say to all sides to beware exaggerated narratives based on pin-thin (and often misleading) statistical arguments. Especially when the advocates are making a living from the narrative they are choosing to concoct. This is true of many types of I&D advocates, whether they be large consultancies like McKinsey, or individual traders. The incentive structure currently rewards the I&D advocate for promoting an exaggerated or concocted narrative. This is an unhealthy part of the new I&D system.
We will get ‘Good I&D’ – and particularly ‘Good Results from I&D’ when the narratives are credible on all sides, and based on strong stats.
What I do like is that McKinsey are looking at the business case for I&D, and not just a moral case (which also exists but that’s another story). By providing data on diversity and profitability, organisations can see from such Reports what is in their self-interest, and the problem will therefore weed itself out due to natural market forces. Firms don’t need moral coercion if they see that their own discrimination or laziness costs them in the pocket and puts them eventually out of business in the competitive market-place. But tell them the true story, not a distorted one, or they won’t believe you and they won’t see the truth and the opportunity. It’s often simply about encouraging enlightened self-interest.
*Footnote for the McKinsey study: Are organisations who increase gender diversity making more profits as a result? Or are organisations who are already in the top half of profitability increasing their gender diversity without losing their current position as ‘top-half’ companies? In the latter case, McKinsey may be misrepresenting a political shift as a ‘cause-and-effect’. For the sake of this article, I’ve assumed McKinsey have proved the ‘cause-and-effect’, though they provide no evidence of this in the Report.