I am less biased than thou

Behavioral New World

March 1, 2026

I am less biased than thou

What types of biases do we as humans have? In what circumstances are these biases important? How strong are these biases? These are some of the fundamental questions sitting at the very heart of Behavioral Economics.

A related but not as frequently explored question is, “Are MY biases stronger than other people’s?” Research shows that the tendency is for individuals to think that their biases are not as strong as others’, implying that they are fairer and more objective than other people. A study (link) shows that people consistently rank themselves as being less biased than the average person. This result recalls studies that show that a substantial majority of Americans rate themselves as better-than-average drivers, which of course they can’t be.

What explains this bias about biases? Ali Almossawi provides three explanations (link). First, the belief that our internal reasoning is reliable. It makes us uncomfortable to think otherwise. Contemplating that our internal reasoning is fallible requires cognitive effort and being open to “epistemic humility,” which I wrote about here.

Second, if an idea “feels true,” then (we often believe) it must be true. Hence the view that our belief is unbiased, or perhaps weakly biased at worst. How often have we heard in our lives, from say friends and family members, “Stop thinking so much and go with your gut on this one,” or some similar admonition? While that might have served us well at times, I believe there is a high failure rate.

An antidote to this notion comes in the form of a comment a therapist friend of mine once offered: “Feelings are information, but they are not the truth.” That puts feelings in their proper place. Truth requires something beyond feelings.

Third, fundamental attribution error may be in play. In short, we attribute good outcomes to our genius and bad outcomes to forces beyond our control. Although this view makes us feel better about ourselves, it imparts a bias to our understanding of the world.

The belief that “I’m not biased, but you are” is critically important when people have different viewpoints. One doesn’t have to look beyond the current political scene here in the U.S. to see substantial differences of opinion. Surely the people with viewpoints other than mine are deeply biased, right?

Or could it be that I’m more biased than I think I am? Perhaps I’m even completely wrong about some issue. The thought of an open-minded thinker! As someone who strives to gain an accurate understanding of how the world works, I might seek out data and facts to either support or refute my beliefs.[1] But especially the latter.

An article in The New Yorker in 2017 provides some bad news in its title: “Why Facts Don’t Change Our Minds.” The link is here, but it is behind a paywall. The article reports, “Even after the evidence ‘for their beliefs has been totally refuted, people fail to make appropriate revisions in those beliefs,’ the researchers noted.” (emphasis added)

So what does change people’s minds? My friend Scott Christianson has written about the effectiveness of AI in influencing people’s beliefs (link). Reporting on one study, he writes, “…people weren’t being persuaded by balanced facts. They were being persuaded by confidence, repetition, and volume.” He adds, “Now, to be clear, humans have always persuaded other humans with bias, emotion, and misinformation. That’s not new. What is new is scale.”[2]

And that scale is what makes it harder than ever, in my estimation, to battle the forces of bias, especially our own.

So what can we do to mitigate our biases? As with all biases, the first step is awareness that they exist and that we are susceptible to them. (There is an ongoing conversation about reducing the biases in AI models. We focus here on debiasing ourselves.)

Some research shows specific actions we can take to mitigate bias. For example, one study shows that a one-time “training intervention” reduced confirmation bias, fundamental attribution error, and anchoring. Other research shows five strategies that reduce implicit bias.

But as is often the case, the research results are mixed. For example, one study reports that “Of the hundreds of studies we examine, a small fraction speak convincingly to the questions of whether, why, and under what conditions a given type of intervention works” (in this study, to reduce prejudice).

I wish the evidence were more clear-cut, but that’s just not the case. Perhaps our best hope at personal bias mitigation comes from awareness, open-mindedness, and epistemic humility. Meditation may also help, as it seems to reduce confirmation bias. (link)

Did the facts presented here change your mind about anything? 😊 At a minimum, I hope that they were informative. Comments welcome.

[1] Note that confirmation bias (link) can color our search. For example, if we ask AI for three reasons that our belief is correct, we will feel comfortable with the response, but it won’t illustrate alternative viewpoints.

[2] I encourage you to read the article. If you are interested in the role that technology plays in our world, Scott’s Substack is a valuable resource. https://substack.com/@profc