Motivated Reasoning: Why Facts Do Not Change Minds

Show two people the same study and they can walk away with opposite conclusions, each one more convinced than before they read it. This is not a failure of intelligence. Social psychologist Ziva Kunda first laid out the mechanism in 1990: people process information as if a goal, not just accuracy, were on trial. When the goal is to reach a particular conclusion, the brain searches memory for supporting evidence, applies looser standards to friendly claims, and applies much stricter standards to unfriendly ones. The result is that facts alone are a weak tool for changing minds, and understanding why is more useful than being frustrated by it.

What motivated reasoning actually is

Motivated reasoning is the tendency to process information in a way that serves an existing goal rather than in a way that reaches the most accurate conclusion. Kunda's original paper, "The Case for Motivated Reasoning," published in Psychological Bulletin, distinguished two kinds of goals: accuracy goals, where a person genuinely wants the correct answer regardless of what it is, and directional goals, where a person wants to arrive at a particular, preferred answer (Kunda, 1990). Both goals use the same cognitive machinery, memory retrieval, evidence weighting, and inference rules. The difference is which memories get retrieved and which standards get applied.

Kunda's key insight was that directional motivated reasoning is not simply wishful thinking or self-deception in the crude sense. People do not consciously decide "I will believe what I want." Instead, the mind constructs a justification that feels, from the inside, like ordinary reasoning. A person motivated to reject a claim about, say, a medication's side effects does not think "I refuse to believe this." They think "this study had a small sample size," or "the researchers have a conflict of interest," and those objections may even be partly true. The bias is in which objections get generated and pursued, not in a blank refusal to think.

This matters because it explains why motivated reasoning is so hard to catch in yourself. A person using directional reasoning still feels like they are being a careful, skeptical thinker. They are just being selectively careful and selectively skeptical, and the selection criterion is which conclusion they want to reach.

In a classic 1979 Stanford study, researchers gave both supporters and opponents of capital punishment the exact same two fabricated studies, one showing capital punishment deters crime, one showing it does not. Both groups rated the study that agreed with their existing view as more scientifically rigorous, and both groups left the experiment more confident in their original position than when they walked in, even though they had just read identical, balanced evidence ([Lord, Ross, and Lepper, 1979, Journal of Personality and Social Psychology](https://fbaum.unc.edu/teaching/articles/jpsp-1979-Lord-Ross-Lepper.pdf)).

How the mechanism works

Motivated reasoning is not one bias. It is a bundle of smaller, well-documented biases that all point in the same direction when a person has a stake in the outcome. Three of them do most of the work.

Confirmation bias is the tendency to search for, notice, and remember information that supports what you already believe, while underweighting or forgetting information that contradicts it. It shows up before you even evaluate an argument, at the stage of what information you choose to look at.

Disconfirmation bias is what happens once you actually encounter a counter-argument. Political scientists Charles Taber and Milton Lodge ran experiments in which participants evaluated arguments for and against issues like gun control and affirmative action. People spent significantly more time and mental effort generating counterarguments against positions they disagreed with than against positions they agreed with, a pattern the researchers called disconfirmation bias, paired with an uncritical acceptance of agreeable arguments (Taber and Lodge, 2006, American Journal of Political Science). In the same study, when participants were free to choose which sources to read, they actively sought out sources that matched their prior views, a self-selected echo chamber built one click at a time.

Identity-protective cognition is the version of motivated reasoning that runs on group belonging rather than a single opinion. Yale law professor Dan Kahan and colleagues found that people do not process risk information as isolated individuals weighing evidence. They process it as members of cultural groups whose standing depends on holding the "right" position. In a widely cited 2012 study published in Nature Climate Change, Kahan's team found that people with higher scientific literacy and numeracy were not more accurate about climate change risk. They were more polarized. Higher scientific ability gave people better tools to construct sophisticated arguments for whatever conclusion their cultural group already held (Kahan et al., 2012, Nature Climate Change). This single finding upended the assumption that public disagreement on contested science is mainly a knowledge gap that more facts will close.

Why does cognitive dissonance make this worse?

Cognitive dissonance is the psychological discomfort of holding two contradictory ideas at once, first described by Leon Festinger in 1957. When new evidence contradicts a belief a person has staked their identity, social relationships, or past decisions on, that discomfort has to go somewhere. Rejecting or reinterpreting the new evidence is usually cheaper than rejecting the old belief, because the old belief is load-bearing for a person's sense of self and social standing.

Does having more information make people less biased?

No. Across the studies above, higher education, science literacy, and numeracy consistently predicted more sophisticated motivated reasoning, not less bias. More cognitive horsepower gives a person more tools to construct persuasive-sounding counterarguments against unwelcome evidence and to generate confirming interpretations of welcome evidence. Intelligence changes the quality of the justification, not the presence of the motive.

The backfire effect: a case study in getting the science wrong, then right

No discussion of facts failing to change minds is complete without the "backfire effect," and it is also a useful lesson in how science self-corrects. In 2010, political scientists Brendan Nyhan and Jason Reifler published an experiment in which they corrected false political claims, for example about weapons of mass destruction in Iraq, and found that for some conservative participants, the correction actually increased belief in the original false claim (Nyhan and Reifler, 2010, Political Behavior). The finding spread widely, in journalism and in psychology writing, as evidence that correcting misinformation was not just useless but actively counterproductive.

That version of the story does not hold up. In 2019, political scientists Thomas Wood and Ethan Porter ran five much larger experiments, more than 10,000 participants across 52 contested political claims, specifically designed to detect a backfire effect if one existed. They found corrections nudged nearly every group toward the accurate information, across the political spectrum, and could not reliably produce a backfire on a single one of the 52 items (Wood and Porter, 2019, Political Behavior). Notably, Nyhan and Reifler did not dig in. They collaborated with Wood and Porter on follow-up work and accepted that the original backfire effect, while real in that one experiment, was not the reliable, general phenomenon it had been treated as.

The corrected picture, laid out in a 2017 review by Flynn, Nyhan, and Reifler, is more precise and more useful: corrections generally do reduce false beliefs on average, but the effect is often small, temporary, and weaker among people for whom the false belief is identity-relevant (Flynn, Nyhan, and Reifler, 2017, Political Psychology). Facts are not powerless. They are just far weaker than the strength of the identity commitment attached to the belief they are trying to correct.

The most surprising fact in this literature is not that facts fail. It is that the specific "backfire effect" study everyone cites to explain why facts fail did not replicate at scale, and even its original authors now say so.

Why simple exposure to the other side often backfires

If motivated reasoning is the disease, more exposure to opposing views sounds like the obvious cure. It usually is not, at least not when delivered the way social media delivers it. Sociologist Christopher Bail and colleagues at Duke ran a field experiment in 2018: they paid Republican and Democratic Twitter users to follow a bot that retweeted messages from the opposing political side for one month. If exposure alone reduced polarization, both groups should have moderated. Instead, Republican participants became measurably more conservative after the exposure, and Democrats showed a small, not statistically significant shift in the opposite direction (Bail et al., 2018, PNAS).

Here is the analogy that makes sense of this: showing someone hostile arguments from an out-group on a platform built for public point-scoring is like handing them a script to defend their team in front of an audience, not a quiet space to reconsider. The content of the opposing argument is often less important than the social frame it arrives in. Content delivered as an attack on your group triggers identity-protective cognition before the argument itself is even evaluated.

Compare that to Broockman and Kalla's 2016 door-to-door canvassing study on reducing anti-transgender prejudice. A single 10-minute conversation, built around asking canvassed voters to recall a time they were judged unfairly and then connect that memory to the experience of transgender people, produced attitude shifts that persisted for at least three months, larger than the average national shift in anti-gay attitudes measured between 1998 and 2012 (Broockman and Kalla, 2016, Science). The difference between the Twitter experiment and the doorstep conversation was not the strength of the facts. It was whether the delivery made the listener feel attacked or heard.

Where motivated reasoning shows up outside politics

Politics gets the most research funding for this topic because polling makes it measurable, but the mechanism is not political. It shows up anywhere a conclusion is tied to identity, money, or a past decision a person has already committed to.

Domain	What the directional motive protects	Typical evidence-resistant claim
Health decisions	Feeling like a competent, in-control decision-maker	Dismissing a diagnosis or downplaying a risk factor tied to a lifestyle choice already made
Personal finance	The belief that a past investment or purchase was smart	Holding a losing stock or asset longer than the evidence supports, to avoid admitting the initial call was wrong
Workplace decisions	A manager's authority and track record	Reinterpreting a failed project's data to protect the reputation of the person who approved it
Relationships	Trust already placed in a partner, friend, or leader	Explaining away a pattern of red-flag behavior as a one-time exception
Consumer brand loyalty	Identity built around a product, team, or brand	Rejecting a negative product review as biased or unrepresentative
Scientific and political belief	Group belonging and cultural worldview	Rejecting a well-supported finding, such as vaccine safety data or climate trend data, that implies your group's stance is wrong

Table 1: Motivated reasoning across everyday domains. The pattern is structurally identical in each row: the strength of the resistance to evidence tracks how much identity, money, or self-image is riding on the existing belief, not how strong the new evidence is.

Behavioral economists have documented a closely related pattern in investing called the disposition effect, the tendency to sell winning investments too early and hold losing ones too long, which researchers link to loss aversion and the reluctance to convert a "paper loss" into a confirmed mistake (Shefrin and Statman, 1985, Journal of Finance). It is the financial cousin of biased assimilation: the data says sell, but selling means admitting the original decision was wrong, so the data gets reinterpreted instead.

What actually moves people, and what does not

If facts alone rarely work, the practical question is what does. The research points to a short list of levers, and none of them involve simply repeating the correct information louder.

Self-affirmation before the threatening information. Psychologist Geoffrey Cohen and colleagues found that having people write briefly about a personal value unrelated to the contested topic, before being shown threatening evidence, made them significantly more open to belief change. The logic is that identity threat is the fuel for motivated reasoning, and if a person's sense of self-worth is topped up from an unrelated source, the threatening fact does not have to be fought off as hard (Cohen, Aronson, and Steele, 2000, Personality and Social Psychology Bulletin).

Messenger trust matters more than message quality. A correction delivered by someone the listener already trusts, a co-partisan, a respected local figure, a messenger with no perceived stake in the outcome, lands differently than the identical correction from a distrusted or adversarial source. This is why public health messaging research increasingly focuses on "trusted messenger" programs rather than mass broadcast of statistics alone.

Perspective-taking outperforms fact-dumping. The Broockman and Kalla canvassing result above is the strongest field evidence for this: asking someone to connect a memory of being judged to another group's experience changed attitudes more durably than any volume of statistics would have.

Slowing down the decision, not just correcting the belief. Adding friction, a delay, a second look, a written prediction made in advance, gives the accuracy motive more room to operate before the directional motive locks in. This is closer to a design fix than a persuasion technique: it changes the conditions under which the reasoning happens rather than trying to out-argue the bias in the moment.

What reliably fails: simply repeating the correct fact louder or more often, mocking the false belief, or delivering the correction as an attack on the person's group. All three tend to increase the sense of identity threat that motivated reasoning exists to manage, which strengthens the very defense mechanism you are trying to dismantle.

Common misunderstandings about motivated reasoning

"It only happens to the other side." Every study cited here found the pattern on both ends of the political spectrum, and in domains with no political content at all, such as personal finance and workplace decisions. Motivated reasoning is a property of how human cognition handles identity-relevant information, not a trait that only affects people you disagree with.

"It means people are lying." Directional motivated reasoning is not conscious deception. The person doing it typically believes, sincerely, that they are reasoning carefully. That is precisely what makes it durable: there is no dishonesty to catch them in, because from the inside it feels like ordinary critical thinking.

"Smarter people are less susceptible." The Kahan climate literacy study and the Taber and Lodge findings both point the other way. Analytical skill and education give a person more resources to construct sophisticated justifications for a preferred conclusion. Being smart changes what the bias looks like, not whether it exists.

"The backfire effect proves correcting misinformation is pointless." As detailed above, the specific 2010 backfire effect did not hold up under large-scale replication, and the researchers who found it originally now say the same. Corrections do work, on average, just less powerfully than plain evidence would predict, and much less powerfully when the false belief is tied to identity.

The takeaway

Motivated reasoning is not a character flaw or a sign of low intelligence. It is what happens when the brain's evidence-evaluation machinery gets pointed at a goal other than accuracy, usually the goal of protecting an identity, a past decision, or a group membership. The fix is not a better fact or a louder correction. It is changing the conditions around the fact: who delivers it, whether it threatens identity on arrival, and whether the listener has room to reconsider without losing face. Facts still matter. They just are not sufficient on their own, and treating them as if they were is the biggest mistake in most attempts at persuasion.

---

Read next:

• Cognitive Biases Explained: Why Your Brain Shortcuts Cost You
• How Memory Can Be False: The Science of Reconstructive Memory
• The Asch Conformity Experiment: Why People Deny What They Can See

Frequently Asked Questions

Q: What is motivated reasoning in simple terms? A: Motivated reasoning is processing information to reach a conclusion you already want, rather than the most accurate conclusion. The brain uses the same skepticism and evidence-weighing tools either way; it just applies them selectively, going easy on friendly claims and hard on unfriendly ones, according to Ziva Kunda's foundational 1990 review in Psychological Bulletin.

Q: Is the backfire effect real? A: A specific 2010 study found it once, but a much larger 2019 replication with over 10,000 participants across 52 issues could not reproduce it, and the original researchers accepted the result. Corrections generally do reduce false beliefs on average; they are just weaker and slower than people assume, especially when the belief is tied to identity.

Q: Why do smart, educated people fall for motivated reasoning? A: Because analytical skill is a tool, not a filter. Dan Kahan's 2012 Nature Climate Change study found that higher scientific literacy and numeracy predicted more political polarization on climate risk, not less, because smarter people are better at constructing sophisticated justifications for the conclusion their group already holds.

Q: Can showing someone the opposing viewpoint reduce their bias? A: Not reliably, and it can backfire. A 2018 Duke University field experiment found that Republican Twitter users became more conservative, not less, after a month of exposure to opposing political content via a bot, likely because the exposure felt like a social attack rather than a neutral information source.

Q: What actually works to change someone's mind? A: Field research points to trusted messengers, self-affirmation before presenting threatening evidence, and structured perspective-taking conversations. A 2016 Science study found a single 10-minute door-to-door conversation using perspective-taking reduced anti-transgender prejudice for at least three months, more than mass information campaigns typically achieve.

Q: Is motivated reasoning the same thing as confirmation bias? A: They overlap but are not identical. Confirmation bias is the narrower tendency to seek and remember evidence that fits an existing belief. Motivated reasoning is the broader umbrella that includes confirmation bias, disconfirmation bias (unfairly scrutinizing unwelcome evidence), and identity-protective cognition (defending group standing) as different mechanisms serving the same directional goal.

Q: Does motivated reasoning affect decisions outside of politics? A: Yes. Behavioral finance research documents the "disposition effect," where investors hold losing stocks too long to avoid admitting a bad initial decision, which follows the same logic as biased political reasoning: the data says one thing, but accepting it means admitting a past choice was wrong.

Q: Why does cognitive dissonance make people double down instead of changing their mind? A: Cognitive dissonance, first described by Leon Festinger in 1957, is the discomfort of holding two contradictory beliefs. When a belief is tied to identity or a past decision, rejecting the new contradictory evidence is often psychologically cheaper than rejecting the old belief, so the mind reinterprets or dismisses the evidence instead.

Q: Does correcting misinformation ever make things worse? A: Rarely, according to the current evidence. The narrow 2010 finding that suggested this was largely not replicated at scale. The bigger risk is not that corrections backfire, but that they are delivered in a way that feels like a social attack, which activates the same identity-protective defenses that made the original false belief sticky.

Q: How can I reduce motivated reasoning in my own thinking? A: Research suggests writing down a prediction before seeing the outcome, actively asking "what would change my mind" before evaluating evidence, and getting feedback from people without a stake in your conclusion. These add friction between the identity-protective impulse and the final judgment, giving the accuracy motive more room to operate.

Sources

• Kunda, Z. (1990). "The Case for Motivated Reasoning." Psychological Bulletin, 108(3), 480-498. Full text (PDF)
• Lord, C. G., Ross, L., & Lepper, M. R. (1979). "Biased Assimilation and Attitude Polarization." Journal of Personality and Social Psychology, 37(11), 2098-2109. Full text (PDF)
• Taber, C. S., & Lodge, M. (2006). "Motivated Skepticism in the Evaluation of Political Beliefs." American Journal of Political Science, 50(3), 755-769. Wiley Online Library
• Kahan, D. M., Peters, E., Wittlin, M., et al. (2012). "The Polarizing Impact of Science Literacy and Numeracy on Perceived Climate Change Risks." Nature Climate Change, 2, 732-735. Nature
• Nyhan, B., & Reifler, J. (2010). "When Corrections Fail: The Persistence of Political Misperceptions." Political Behavior, 32, 303-330. Springer
• Wood, T., & Porter, E. (2019). "The Elusive Backfire Effect: Mass Attitudes' Steadfast Factual Adherence." Political Behavior, 41, 135-163. Springer
• Flynn, D. J., Nyhan, B., & Reifler, J. (2017). "The Nature and Origins of Misperceptions: Understanding False and Unsupported Beliefs About Politics." Political Psychology, 38(S1), 127-150. Full text (PDF)
• Bail, C. A., Argyle, L. P., Brown, T. W., et al. (2018). "Exposure to Opposing Views on Social Media Can Increase Political Polarization." Proceedings of the National Academy of Sciences, 115(37), 9216-9221. PNAS
• Broockman, D., & Kalla, J. (2016). "Durably Reducing Transphobia: A Field Experiment on Door-to-Door Canvassing." Science, 352(6282), 220-224. Science
• Cohen, G. L., Aronson, J., & Steele, C. M. (2000). "When Beliefs Yield to Evidence: Reducing Biased Evaluation by Affirming the Self." Personality and Social Psychology Bulletin, 26(9), 1151-1164. SAGE Journals
• Shefrin, H., & Statman, M. (1985). "The Disposition to Sell Winners Too Early and Ride Losers Too Long: Theory and Evidence." Journal of Finance, 40(3), 777-790. JSTOR