Welcome to the first issue of Milkman Delivers of 2026! The last month has been a whirlwind thanks to the usual New Year’s surge of media interest in research on the fresh start effect and behavior change (see below for a roundup of recent press coverage).
One highlight of this year’s media tour was recording an episode of The Mel Robbins Podcast in Boston. The resulting episode, “Change Your Life This Year: How to Get from Where You Are to Where You Want to Be,” was released in the final moments of 2025.
I’ve done hundreds of podcast interviews since publishing How to Change in mid-2021, but working with Mel’s team was a unique experience. Their approach combines curiosity-driven storytelling, an unusually personal tone, and a palpable sense of joy. I had a blast getting to be part of this show, which was the #1 most shared podcast of 2025 and a top-5 most downloaded podcast.
Here’s a short clip from the episode:
A Roundup of New Year’s Press Coverage
Here’s how to make resolutions stick, according to a behavior change scientist The Washington Post
How to make New Year’s resolutions you’ll actually stick to GQ
Want to make your resolution stick this year? Behavioral science has the answers Scientific American
The 5-day brain health challenge, Day 3: Do a brain-boosting workout The New York Times
To make New Year’s resolutions stick, science can work in your favour CBC
How to make New Year`s resolutions that you’ll actually stick to WGN Morning News
Change your life this year: How to get from where you are to where you want to be The Mel Robbins Podcast
Are you ready for a fresh start (update) Freakonomics Radio
Q&A: What is Naïve Realism and How Can Recognizing It Help You Disagree Better?
This month’s Q&A from Choiceology is with Harvard Kennedy School Professor Julia Minson, author of the forthcoming book How to Disagree Better (which I highly recommend!). Here she explains what naïve realism is, how it shapes our judgments, and why it can prevent us from fully incorporating insights from people who see the world differently.
Me: I’m excited to talk about naïve realism. Let’s start at the very beginning: what it is, exactly?
Julia: Naïve realism is basically the belief that most of us have that our perceptions of the world and the reactions to the things we see are reasonable and objective. That “I see the world as it really is.”
Me: Are you telling me that my perceptions of the world are not correct? Because I’m going to have trouble with that.
Julia: It is funny, right? So, we kind of stole the idea of naïve realism from philosophy. We process all of our encounters with the physical world through a set of perceptual organs, so what we experience in our brains is not actually what’s out there. So, when you and I look at a couch of a particular color, we’re actually seeing slightly different colors because we have slightly different eyeballs. And so, in a physical sense, we don’t perceive reality as it really is. We perceive a version that’s funneled to us through our organs. And then that’s even more true when we think about our social reality. We think that we understand what’s just, or fair, or reasonable, and we don’t account for the filtering that happens through our background, our ideology, or our level of education, or whether we’re tired or cranky. We generally don’t stop to consider how all of those factors influenced what we think of as the completely unmediated perception of how the world is.
Me: Where does naïve realism become a real problem?
Julia: Well, first, let me say that naïve realism isn’t necessarily a bad thing. It’s just the way we relate to information in the world. If I’m walking around the world assuming that I’m a rational, reasonable person, that saves me a lot of time. Where naïve realism tends to become more problematic is when we are dealing with more contested opinions and issues. Ironically, the first big thing that naïve realism tends to do is make us not anticipate disagreement. A closely-related bias called the false consensus effect says that because I think I’m a rational, reasonable person and I see things how they really are, I expect other rational, reasonable people to agree with me. So, people tend to overestimate the proportion of people who share their tastes and preferences and attitudes. But the bigger problem is when we see clear evidence of disagreement and rationalize that the other person must be unreasonable. It might lead us to think, “I’m seeing the situation in an objective way, and you disagree with me, ergo, you are not being a reasonable person in this situation.” And then because humans are very good at making up stories to explain the events around them, you can spin up a whole story about what exactly is wrong with the other person that’s making them so unreasonable.
Me: It sounds like anytime I’m interacting with someone who has a different background than I do, and we’re talking about our opinions on, say, which stocks are good buys or how to raise children, these are going to be situations where naïve realism could cause conflict.
Julia: Exactly.
Me: Got it. Could you tell me about some of the research you’ve been involved in proving naïve realism exists?
Julia: There’s a study I love that was conducted by Lee Ross, who came up with the idea of naïve realism in collaboration with Andrew Ward. When I was Lee’s research assistant, he taught an applied social psychology course at Stanford, and so we had this large room of undergrads who were all in this class, and we asked them to fill out a survey about their opinions on a bunch of different policy issues that were at play at the time and you had to say whether you agree or disagree with a bunch of different things. And then we collected all these surveys, basically shuffled them around, and then passed out the surveys back to the students, so you ended up with the survey of some other random student in the class. So you no longer have your own survey. You are looking at somebody else’s responses.
And the question before you is: Consider your own responses that you just filled out and now the responses that you are looking at. How much do you think you overall agree or disagree with this other student’s worldview? And so they respond on this seven-point scale of how much they agree or disagree. Then they answer a bunch of questions about what is driving your own views versus this other person’s views. And we give them sort of this set of considerations about normative facts, normative considerations that could influence your beliefs. So things like a good understanding of American history, engagement in world events, the desire for fairness and justice. Those are things that ideally should influence policy opinions. But we also gave them a list of things that might influence beliefs that we thought of as biases – things like political correctness, the self-interest of the group you belong to, exposure to partisan media and so on.
What we find is that people generally say their own views were more driven by normative considerations than biases, and if their counterpart disagrees with them, this other student whose questionnaire they’re looking at, to the extent that they see that questionnaire as reflecting a different view of the world, they think that worldview is primarily driven by biases rather than normative considerations. And what’s interesting is that it’s not simply: “I think that my views are normative and other people’s views are biased.” It’s that other people’s views are biased if they happen to disagree with mine.
Me: That’s really interesting, and I know you have a great paper looking at how naïve realism affects the way we change our views after exposure to other people’s opinions. Could you describe that work?
Julia: Absolutely. This was work with Lee Ross and Varda Liberman, who is a researcher in Israel, and it involved making quantitative estimates. The great thing about quantitative estimates is that there are lots and lots of them to make in the world, and there are correct answers, so you can actually measure what strategy would be more effective if you’re going for accuracy. Like how many people would support this particular policy or how many inches of snow will fall in Boston next year, and therefore, how many snow plows do we need to prepare?
We make quantitative estimates in the real world constantly. And the thing that’s interesting is that we often make quantitative estimates with the input of other people.
So I have some estimate about the world, and you might have some estimate about the world, and if I have invited you into a conversation about what the right answer is, the question that is interesting from a naïve realism perspective is: How much weight am I going to give to your estimate? If I made an estimate independently and so have you, and we’re equally informed, I should average those two estimates.
What we find instead is that people pretty consistently give more weight to their own judgment than the other person’s judgment, and in fact, they do to such an extent that it’s tantamount to saying that I am twice as knowledgeable about this thing that I really have no idea about as you are. So we have people estimate all kinds of things that they know nothing about, and they very consistently give twice as much weight to their own judgment as they give to the other person’s judgment. So we know mathematically averaging works better. The question is why don’t people do it? An explanation that you would get from naïve realism is that the further away you are from my estimate, the more I think that you’re just wrong.
Me: That’s fascinating. What do you do differently because you know about this research?
Julia: I recognize that in reality, most of us are being naïve realists almost all the time. So, it’s incredibly important to keep in mind that when you disagree with a smart person, you are wrong 50% of the time. You would do well to remember that, especially in either marital arguments or business partnerships.
I would also advise people to actively seek out people who you expect to disagree with you. We have such a strong bias against it, that unless you make it a point to look for those people and get their input, you’ll either avoid it or discount it when you come across it.
Me: That’s a great place to wrap up, Julia, because I think most of us find it uncomfortable to disagree with other people. But recognizing that disagreement can offer a lens into important things we’re missing rather than signaling another person is wrong and biased and we’re right highlights the importance of collecting different perspectives.
This interview has been edited for clarity and length.
To learn more about Julia’s work, listen to the episode of Choiceology where we dig into naïve realism, or pre-order a copy of her excellent book coming out in March, How to Disagree Better.
That’s all for this month’s newsletter. See you in February!
Katy Milkman, PhD
Professor at Wharton, Host of Choiceology, an original podcast from Charles Schwab, and International Bestselling Author of How to Change
P.S. Join my community of ~100,000 followers on social media, where I share ideas, research, and more: LinkedIn / Twitter / Instagram / BlueSky / Threads
The naïve realism framing explains something about fraud that I've never been able to articulate cleanly: the victim's naïve realism is the con artist's primary asset.
The mark doesn't just believe their perception of the investment is accurate — they believe it's so obviously accurate that anyone raising doubts must have a hidden motive. The fraudster engineers this. Every early return, every glossy document, every lunch with a person of apparent substance is designed to make the mark's internal model feel more confirmed, not just correct. By the time the red flags appear, the mark is so epistemically committed that the person raising the alarm seems like the irrational one.
What Professor Minson's research suggests — that naïve realism makes genuine disagreement feel like a character attack rather than an information update — maps directly onto why fraud victims so rarely respond to intervention. Telling someone "I think you've been deceived" doesn't land as new information. It lands as an accusation that their entire model of reality is wrong. The brain treats it as a threat, not a gift.
The structural implication is uncomfortable: the more confident someone feels about a financial decision, the more susceptible they may be, not less. Certainty closes the gap that doubt would otherwise keep open.
Lee's research that found that people tend to attribute their own opinions to normative factors and others' opinions to bias is very interesting. In my own research, we asked open-ended questions about what factors influenced opinions and attitudes. Nearly everyone said that their opinions were completely original ideas that they came to on their own. I remember being very surprised by that!