The speed at which the DEI reversal occurred is not what makes it peculiar. It’s how softly. The language began to vanish from business websites at some point in late 2023, much like a shoreline recedes before a wave. A diversity report here, a sentence there. The change was already well under way by the time Forbes began publishing weekly lists of companies pulling out. It was based more on subtle changes made after hours by people who most likely didn’t enjoy making them than on big announcements.
Its shape is indicated by the numbers. DEI was mentioned in over 29% more job postings between November 2020 and November 2021. They had decreased by 23% after two years. According to one Aura Intelligence analysis, between early 2022 and the end of 2024, DEI-focused positions declined by 40% overall. A similar curve was followed by sentiment. By February 2025, the percentage of positive social media mentions of DEI had decreased from 87% in late 2021 to 38%. It is not a drift in policy. It’s a collapse.
| DEI Backlash — Key Data & Context | Details |
|---|---|
| DEI Job Posting Decline (Jan 2022 – End 2024) | 40%, per Aura Intelligence |
| Social Media Positive Sentiment Toward DEI (Nov 2021) | 87% |
| Social Media Positive Sentiment Toward DEI (Feb 2025) | 38% |
| DEI Posting Surge (Nov 2020 – Nov 2021) | +29% |
| DEI Posting Drop (Nov 2022 – Nov 2023) | –23% |
| Companies Publicly Rolling Back DEI in 2024–2025 | Includes IBM, Target, Constellation Brands, Meta |
| Key Academic Framework | Social Identity Theory, signaling theory |
| Notable Case Study | Stanford GSB study on Glassdoor reviews |
| Trigger Events | Supreme Court affirmative action ruling, 2023 state-level anti-DEI legislation |
The fact that the retreat’s economics haven’t really supported the choice is intriguing—and perhaps unsettling. Target did not face backlash from the critics it was attempting to placate when it withdrew from its publicly declared DEI commitments. It came from the clients who had faith in the initial stance. Sales fell. In areas where the brand had developed a certain level of loyalty, foot traffic declined. Costco’s customer loyalty scores remained stable after it refused to back down from an anti-DEI shareholder proposal. It’s not a subtle contrast. One retailer paid for it, while the other did not blink.
Last August, research from Stanford’s Graduate School of Business examined what transpired within businesses following DEI disputes. Morale declined. Glassdoor ratings decreased. In the weeks that followed reports that executives were, in the words of one CEO, “reprioritizing,” leadership approval scores fell. Those declines might have occurred for other reasons. However, the pattern is sufficiently consistent across businesses to read more like a signal than noise. Workers are listening to what their employers say and what they subtly stop saying, especially those from the groups that DEI programs were intended to keep.
According to a conceptual model released this year by Pitt researchers, the rollback itself is not the true harm caused by the DEI backlash. It’s the discrepancy it causes between a company’s stated values and what it is now prepared to defend. That gap was interpreted by the workers as a signal. Belonging diminishes. Those who are perceived begin to feel like advertising copy.

As this develops, there’s a sense that corporate America hasn’t given it enough thought. After the legal uncertainty has subsided and the political climate has changed once more, some of these companies are likely to reverse the reversal within five years. More slowly, others will learn that the employees they lost don’t return for a press release.
It’s still unclear whether the retreat’s cost will be reflected in consumer trust surveys, hiring pipelines, or the more difficult-to-measure area where a company’s reputation truly resides. However, the bill is currently being drafted. Additionally, when the invoice arrives, the businesses that claim to have saved money by leaving DEI may discover that the savings were borrowed from a source they neglected to investigate.
