I have watched the DEI debate collapse into two camps that both feel more like tribal loyalty tests than serious policy positions. One side treats any criticism of a diversity office as proof of bad faith. The other side treats the mere existence of a diversity office as proof of ideological capture. Neither posture actually engages with the evidence, and having spent years working inside institutions where these programs lived or died based on budget politics rather than results, I think the honest answer is more complicated and more useful than either camp wants to hear.
What the Backlash Gets Right
Start with the legitimate part of the backlash, because I think it gets dismissed too quickly by DEI's defenders. A meaningful body of research on workplace diversity training has found that mandatory, one-off training sessions, the kind where employees are herded into a conference room once a year to click through slides about unconscious bias, often produce little measurable change in behavior and occasionally produce backlash effects, where participants become more resistant to the message rather than less. Some diversity offices did expand during a period of enormous institutional pressure without much thought given to what specific problem they were solving or how success would be measured, and I do not think it is unreasonable for a board or a legislature to ask pointed questions about a program that cannot articulate its own metrics.
What Gets Lost in Wholesale Elimination
But here is where I part ways with the elimination-first crowd. The same broader body of research that is skeptical of mandatory bias training tends to find real, measurable value in a different set of practices that often lived under the DEI umbrella: structured, standardized hiring processes that reduce the influence of individual reviewer bias, mentorship and sponsorship programs that measurably improve retention and promotion rates for employees who lack informal institutional networks, and accessibility accommodations that expand who can even apply for a role in the first place. When a state or a company eliminates its entire diversity function in one legislative stroke or one press release, it does not surgically remove the ineffective training modules and keep the structured hiring reforms. It usually kills the whole thing, evidence-based practices included, because the political fight is being fought at the level of the word diversity rather than at the level of individual program effectiveness.
An All-or-Nothing Failure on Both Sides
I think that all-or-nothing dynamic is the actual failure here, on both sides. Advocates for these programs often resisted the kind of rigorous, skeptical evaluation that would have let them prune the ineffective pieces and defend the effective ones with real data, because evaluation felt like giving ammunition to bad-faith critics. Critics, for their part, frequently were not actually interested in a program-by-program evaluation either, treating elimination as the goal regardless of what the data showed about any specific practice. Both postures avoided the harder, less satisfying work of asking which specific interventions produce which specific outcomes.
Judging Programs by Results, Not Labels
What I would actually want to see, and what I think a serious institution should demand, is treating these programs the way we would treat any other institutional investment: with clear goals, defined metrics, and a willingness to cut what does not work while protecting what does. If a mentorship program measurably improves retention among employees who would otherwise leave within two years, that is worth keeping regardless of what political label gets attached to it, because retention costs organizations real money and losing capable people is a genuine loss. If a mandatory training module produces no measurable change or a backlash effect, cut it, and do not treat that as a betrayal of the underlying goal of a fair workplace.
I recognize this position will frustrate people on both sides, and I have made peace with that. It denies the elimination camp the clean political victory of ending DEI as a category, and it denies the defense camp the comfort of treating every program under that umbrella as equally beyond scrutiny. But institutions exist to produce results, not to signal loyalty to a side of a culture war, and the evidence we already have is good enough to tell the difference between what is working and what is not, if anyone is actually willing to look at it rather than simply pick a side.
