When moral reframing backfires at work: New research on diversity, equity, and inclusion messaging
On April 9, Dr. Annie Hoover defended her PhD dissertation at George Mason University. Over the past year, Annie has supported idealis as a contract data analyst. With the dissertation now behind her, she is joining the team full time.
Annie brings a pairing that shows up in her work and is essential to idealis. Deep care for humanity and a steady commitment to rigor and integrity with data. That combination strengthens how idealis leads: advice grounded in evidence, delivered with respect for the people who have to live with the decision.
Annie’s dissertation, When Moral Appeals Miss the Mark: Workplace DEI Persuasion and Backfire Effects, tests a popular persuasion idea: if you want to reach someone, make the case in the moral language they trust. Annie tested that approach in a workplace context where “DEI” has become a charged label, even when the underlying goals are basic. Fair treatment. Belonging. A workplace where people can do their best work.
The result is specific and useful. Moral reframing did not improve persuasion overall. In one key case, it made persuasion worse.
The question Annie tested
Annie first encountered Moral Foundations Theory in graduate school. The theory, pioneered by Jonathan Haidt and Jesse Graham, argues that people draw on a small set of intuitive moral foundations, often summarized as care, fairness, loyalty, authority, and purity. People emphasize these differently, including across political orientation.
Moral reframing builds on that theory. It is a communication technique that translates an argument into the moral values of the audience you are trying to persuade.
Much of the moral reframing research has been done in domains like climate change. Annie saw an open question. Would moral reframing work in workplace messaging about diversity, equity, and inclusion initiatives, or would the workplace context change the effect?
This is not an abstract question for leaders. In many organizations, a senior team may agree on a goal, then lose the room in the rollout. A policy change lands as a lecture, or a training announcement reads as a mandate. The content may be sound, but it isn’t well received
What Annie studied (two studies, about 1,600 participants)
Across two studies with roughly 1,600 participants, Annie tested pro and anti workplace messages about organizational diversity and inclusion efforts. The messages were framed in different moral foundations.
The hypothesis was straightforward. Messages tailored to an audience’s moral values would shift attitudes more than nonmoral arguments.
What Annie found
Overall, moral framing did not outperform nonmoral arguments.
The sharper result showed up among conservative participants. When conservatives read pro inclusion messages framed using traditionally conservative moral foundations, especially authority and purity, they showed less positive attitude change than when they read nonmoral pro inclusion messages.
That is a backfire effect in that the tailored frame reduced persuasion compared to a plain argument.
Annie’s first response was the right one. She rechecked coding and analysis. Study 2 had been designed assuming Study 1 would support the hypothesis, so the result forced a redesign.
Then Annie asked the question that matters. What caused the backfire?
Why it may have happened: reactance
Study 2 pointed to psychological reactance. People resist when they feel their autonomy is being threatened.
Annie describes it simply. People want to act of their own volition, and when they sense someone is trying to control them, they can experience anger and push back.
Study 2 supported that mechanism. Conservatives exposed to conservatively-framed pro inclusion messages reported higher perceived threat to freedom and higher state anger than those who read nonmoral or liberally-framed messages.
One insight that is immediately practical for leaders emerges: When a message uses values that feel personal, the persuasion attempt can read as manipulation, even if the intent is respectful. Reception often controls outcome.
Here is what that can look like in the kind of work idealis does. A leader wants to reduce bias in promotion decisions. They announce a new process with language about “respecting authority” and “upholding standards” and “protecting the culture.” Some employees hear alignment between their values and the organization. Others hear pressure. The process starts to feel like a loyalty test rather than a performance system, and the organization then spends the next 6 months managing distrust instead of improving the decision.
What leaders can take from this
Most leaders hold a reasonable instinct. Meet people where they are. Use language that reflects what employees care about. Annie’s research does not argue against that instinct, rather it shows where it can fail.
Three cautions emerged.
1) Intent does not control impact
A well-meant attempt to speak in someone’s values can still feel controlling if the disagreement is real and the values feel used.
2) The messenger changes the message
Annie’s studies used brief written messages from nonspecified sources. In continuously evolving organizations, trust in the speaker matters, along with history, personal and professional motive. A manager in a one on one conversation may land differently than an anonymous statement.
In practice, this is the difference between an email from “HR” and a leader who has earned credibility saying, plainly: “Here is what we are changing, here is why, here is what stays the same, and here is where you still have choice.”
3) The label can trigger the reaction
In many workplaces, the acronym “DEI” carries enough charge that it can activate resistance before any content gets heard. Leaders can name the work more explicitly at times. For example:
• workplace belonging
• fair access to opportunity
• reducing bias in hiring and promotion
• making performance standards clearer and more consistent
• creating a workplace where people can do their best work
A useful next step for leaders is to pressure-test language before broad rollout. Not “does this sound persuasive,” but “does this sound like control.” If it does, you will get reactance.
A simple example. If you are rolling out a new leader expectation, do not lead with a moral appeal that sounds like “good people agree with this.” Lead with operational clarity: what is changing, what good looks like in observable behavior, and how feedback works. Then back it up with a real conversation format managers can use rather than a slogan.
The research standard behind the result
A PhD is proof of persistence and method. It is a long practice in asking clean questions, testing them with care, and staying honest when the data surprises you.
We are so glad Annie has joined the idealis team full time and we are also glad this research exists. It sharpens how leaders talk about contested topics at work, including diversity, equity, and inclusion. It adds evidence where many organizations currently rely on instinct and reaction.
If your organization is navigating how to communicate about inclusion work in a polarized moment, this research gives a grounded warning. Moral language can help. It can also backfire. The difference often comes down to trust, autonomy, and delivery.
