Pushing harder won't change their mind. Here's what will.

You have probably been in this meeting. You make your case. The other person pushes back, so you make it again, more firmly and with more data. They push back harder. By the end of the hour they are further from agreeing than when you started, and you leave wondering why the facts did not win.

Most leaders know that feeling. The research points one way, and the answer is not more facts or more force.

Why pushing backfires

The first problem is the wording. In 2025 researchers combined 35 studies, covering 10,658 people, of health messages such as advice on smoking and drinking. Messages that used commands, absolute claims, exclamation marks or statements that the reader had no choice made people feel their freedom was under threat. That feeling produced resistance, and resistance made the message less persuasive. The effect on persuasion was consistent but modest, and gentler wording such as "could" or "consider" caused less threat.

Moral pressure is the second problem. In two 2026 experiments, more than 600 British meat-eaters imagined posting a photo of a meat dish and receiving criticism from a stranger. When the criticism was moral, saying they did not care about the environment, people felt more offended, thought less of the critic and wanted to avoid them, compared with criticism of their cooking skill. None of the criticism changed how much meat they planned to eat. A private message caused more defensiveness than a public comment, and people read it as a stronger condemnation.

The authors' advice was simple: avoid implying that the other person's choice reflects a moral failing.

The facts are not the problem

A popular idea holds that correcting people's facts makes them cling harder to their beliefs. The evidence says otherwise. In five experiments with more than 10,000 people, covering 52 contested policy issues, researchers found no case where a factual correction made people more wrong. People generally moved towards the facts, even when the facts cut against their politics.

That matters for a leader. Put it beside the first two studies and, as I read it, the facts are not what makes people dig in. The way they are delivered is.

What actually changes minds

Listening changes minds more than arguing does. In 2015 volunteers in Miami knocked on doors to talk with voters about transgender rights. The canvassers came from an LGBT advocacy group, but they did not lecture. In conversations of 10 to 15 minutes they listened, and asked voters to recall a time they had been treated differently themselves. About one voter in ten became more accepting of transgender people, and the change lasted at least three months. The study was published in Science.

Swapping reasons works too. In a 2025 study, adults in the United States and Britain held ten-minute conversations with someone who disagreed with them, on topics from cats and dogs to cancel culture and Joe Biden. Each was asked to share their reasoning and ask about the other's. Both sides moved towards the middle, further than either had predicted, and part of the change was still there a week later. People who were told about these results were more willing to have a difficult conversation themselves.

Being listened to also changes the speaker. In five experiments, people who shared their views with a good listener felt less anxious and became clearer about what they thought. My reading is that a person who feels heard has less to defend.

The strange one: agree, and go further

The oddest finding in this area runs against instinct. Instead of arguing against a view, you agree with it and push it to an extreme, and people step back from it themselves.

In 2013, during an Israeli election, researchers showed 161 Jewish Israelis a series of short clips styled as campaign adverts. Each clip took a belief many viewers held about Israel, such as that it is a moral nation, and argued that Israelis therefore needed the conflict to keep feeling that way. Viewers reconsidered their positions more than a control group, became more willing to compromise, and were still more conciliatory a year later. A 2026 online study on a different issue found mixed but encouraging results across three experiments.

The technique carries a clear risk. In an early pilot, some viewers took the clip literally and became prouder of the view the researchers hoped they would question. One of the 2026 authors put it plainly: there is "both potential and risk". My own view is that this is not a technique to use on your team. It is more useful for testing your own opinions.

Several of these studies dealt with charged political issues, but the methods themselves are neutral. Listening, swapping reasons and avoiding moral pressure work whichever side of an issue you are on, and that is what makes them useful at work.

What this is costing you

None of these studies was run in an office, so applying them to everyday work is my own step. The pattern will be familiar, though. Pressure at work often looks like this:

• the email that says a change is "non-negotiable";
• the second meeting called to make the same case more forcefully;
• feedback that questions someone's commitment rather than their method;
• the decision announced without the reasons behind it.

Each of these may win compliance. My reading of the research is that it often loses agreement, and agreement is what carries a decision through when the leader is not in the room.

The difference can be one paragraph. Here is the same message written two ways (my own example):

• Pushing: "From Monday, all client proposals must go through the new review process. This is non-negotiable. No exceptions."
• Explaining: "From Monday, I'd like all client proposals to go through the new review process. Last quarter two proposals went out with pricing errors, and this is how we stop that happening again. If you can see a problem with it, tell me this week."

Questions worth asking

A few honest questions show how much pushing goes on around you.

• When did you last change your mind in a meeting, and what changed it?
• How often do your messages to the team include "must", "always" or "non-negotiable"?
• When you give critical feedback, is it about the work or about the person's values?
• Before your last disagreement, could you have stated the other person's reasons in a way they would accept?

What you can do this week

Everything in this section is a suggestion of mine, drawn from the research rather than tested by it. Each one takes minutes and costs nothing.

Suggestion

If you lead a team

• Ask for reasons before giving yours. In your next disagreement, ask the other person to explain their reasoning fully, then say back what you heard before you reply.
• Soften the wording, not the message. Replace "you must" and "non-negotiable" with "I'd suggest" or "could we consider", and keep the facts exactly as they are.
• Criticise the method, not the person. Frame a problem as something to fix ("here's what would make this work"), not as a failure of values or care.
• Book ten minutes. When a disagreement has gone round twice, stop arguing and book a ten-minute conversation in which each side only explains its reasons.

Suggestion

For yourself

• Stretch your own view to its extreme. Write down your position on a contested issue in its most extreme form. If you would not sign that version, note where your limit is.
• Expect to move. Before you talk with someone who disagrees, decide that moving a little would be a good result, not a defeat.
• Notice your own push-back. When an instruction makes you bristle, separate the wording from the facts, and judge the facts on their own.

Suggestion

At home

• Offer a choice. With children, "you could do your homework before or after dinner" is likely to meet less resistance than "do your homework now".
• Ask why before saying no. When a child wants something you are about to refuse, ask for their reasons first and say them back.

Sources

• Ma, R., Ma, Z., Kalny, C. S., & Walter, N. (2025). Words that trigger: A meta-analysis of threatening language, reactance, and persuasion in health. Journal of Communication, 75(5). doi:10.1093/joc/jqaf004
• Chung, I., Russell, P. S., & Hepper, E. G. (2026). Maybe I can't cook, but it doesn't mean I don't care: Moral framing and private criticism of meat consumption leads to psychological defensiveness. Personality and Social Psychology Bulletin. doi:10.1177/01461672261472608
• Wood, T., & Porter, E. (2019). The elusive backfire effect: Mass attitudes' steadfast factual adherence. Political Behavior, 41(1), 135 to 163. doi:10.1007/s11109-018-9443-y
• Broockman, D., & Kalla, J. (2016). Durably reducing transphobia: A field experiment on door-to-door canvassing. Science, 352(6282), 220 to 224. doi:10.1126/science.aad9713
• Kardas, M., Nordgren, L. F., & Rucker, D. D. (2026). Unnecessarily divided: Civil conversations reduce attitude polarization more than people expect. Journal of Personality and Social Psychology, 130(2), 187 to 214. doi:10.1037/pspa0000469
• Itzchakov, G., DeMarree, K. G., Kluger, A. N., & Turjeman-Levi, Y. (2018). The listener sets the tone: High-quality listening increases attitude clarity and behavior-intention consequences. Personality and Social Psychology Bulletin, 44(5), 762 to 778.
• Hameiri, B., Porat, R., Bar-Tal, D., Bieler, A., & Halperin, E. (2014). Paradoxical thinking as a new avenue of intervention to promote peace. Proceedings of the National Academy of Sciences, 111(30), 10996 to 11001.
• Yaffe, D., Kochav, O., & Lifshin, U. (2026). Paradoxical thinking and attitudes toward prostitution: Preliminary experimental findings. The Journal of Social Psychology. doi:10.1080/00224545.2026.2667805