June 19, 2024 - 1:00pm

This week, the Washington Post and market research group Ipsos jointly released a new poll finding that 61% of Americans think DEI (Diversity, Equity, and Inclusion) efforts in businesses are “a good thing”, and that support rises to 69% when the concept of DEI is explained in a full sentence-long definition. Meanwhile, businesses around the country are gutting DEI programmes adopted in 2020 and 2021, citing their unpopularity. Can these new figures really be true?

This example demonstrates the degree to which the framing of an opinion poll can influence the results. What, we might ask, is the fuller description of DEI that respondents were offered? It is this: “programs to hire more employees from groups that are underrepresented in their workforce, such as racial and ethnic minorities and people with disabilities and to promote equity in the workplace”. The first thing we might notice here is that the most important word to define in the DEI trinity and the most prone to misinterpretation — equity — goes undefined.

Why? Because for poll respondents to respond positively to DEI, the pollster must avoid disrupting the popular misimpression that equity means the same thing as equality (the core American value of offering equal opportunity to all) even if that means offering us a circular definition. The words are easy to confuse, but equity, unlike equality, means that our measuring stick of success should be equal outcomes across different groups, not equal opportunity. It is out of this concept of “equity” that many of the more extreme policies of the DEI era have arisen.

It also colours respondents’ responses on DEI that many of these extreme (but terribly common) corporate policies are not listed in the subsequent section of the poll that asks their opinion on specific policies. Those that are listed are presented in the most charitable possible light. The policies listed (all but the last achieving well over 60% approval) were “mentorship opportunities for underrepresented groups”, “anti-bias trainings”, “internships for underrepresented groups”, “efforts to recruit underrepresented groups”, “employee resource or affinity groups for underrepresented groups”, and “paying executives more if diversity targets are met”.

Absent from this list are policies such as setting racial or gender targets for hiring and demanding explanations from executives or committees if job searches fail to yield a candidate of the “right” group, requiring diversity statements from job applicants, new company initiatives on racial impact before other business considerations are made, and setting racial or gender quotas for board representation. That’s before we get to donating corporate funds to political charities based on social justice priorities, or requiring employee attendance at trainings based around texts such as Robin DiAngelo’s White Fragility (none of which could be fairly described as “anti-bias trainings”).

Another vital (and rarely remarked upon) nuance is this pollster’s choice of the word “underrepresented” when describing each policy. This word is designed to convince the reader that the disparity in question is a problem that must be addressed rather than simply a gap due to any number of factors (many of which don’t indicate any underlying unfairness). To call a group of people “underrepresented”, much like calling a tyre “underinflated”, is to tell us that the underrepresentation must be fixed, and by extension that any initiative designed to fix it is good. It’s a subtle way of inserting circular reasoning into the phrasing of the question.

Considering the source of the poll, these biases should not surprise us. Over the last several years, Ipsos has been a frequent cheerleader for DEI initiatives. As recently as last year, it released a report entitled “Diversity, Equity and Inclusion are Central to Good Research: Here’s How Ipsos is Delivering on That.” Among other recommendations, the report committed to altering Ipsos’s survey methods to redress the historical marginalisation of some social groups in its polling. Just a few days ago, the company released “a selection of Ipsos’s latest and greatest insights and data on DEI for business leaders, policymakers, advertisers, and insights professionals”.

We would be unlikely to take the results very seriously if a major newspaper partnered with the National Rifle Association to evaluate public opinion on gun control. Why should we take one seriously when it hires a company with a vested financial stake in DEI to evaluate what today’s public thinks of diversity initiatives? Perhaps we should believe the evidence we see before us instead.


John Masko is a journalist based in Boston, specialising in business and international politics.