When People Change their Beliefs about Change

From proverbs to Pinterest, the world seems to be brimming with messages that convey a simple truth about humanity: that people can’t truly change. Oh wait, no. That people can change. Because we all know that leopards can’t change their spots. Well, unless they’re turning over a new leaf. This basic tension between notions of change and stability is even reflected in a song from the Disney Movie “Frozen,” claiming in one line “We aren’t saying you can change him, ‘Cause people don’t really change” yet in another line affirming that “Everyone’s a bit of a fixer upper.”

In light of all of these contradictory cultural messages about change, how do we make sense of the information we encounter about ourselves – and others – over time?

Stanford psychologist Carol Dweck and her colleagues have compellingly demonstrated that people tend to have a favorite go-to belief or implicit theory about change. Can hard work make you smarter, or do you have a certain level of braininess no matter what you try? Is your basic morality – or personality for that matter – set in stone, or can it be altered? Entity theorists believe these traits can’t truly be changed; incremental theorists are confident that attributes are malleable. Even though these implicit beliefs are really just “all in our head,” they affect how we interpret the world, and have dramatic effects on learningmotivation, and judgments.  

Even though it makes sense that people develop these habits of thinking about change or stability, intuitively I can think of cases where my own theories flip-flop. I’m an incremental theorist much of the time, but even though I firmly believe that people can change, I’ve heard myself give the exact opposite advice to a friend who was reconsidering a not-so-deserving ex. Suddenly, I was all about leopards not changing their spots – maybe because this alternate viewpoint better supported my argument.

The idea that people selectively appeal to evidence to support their position isn’t a new one. Ziva Kunda theorized that we often engage in motivated reasoning – that we’ll preferentially seek evidence, search memory, or appeal to causal theories that stack the deck in favor of the conclusion we wanted to come to all along. My colleagues and I wondered if people’s allegiance to implicit theories of change might be subject to change themselves, especially when shifting them could help support a goal.

Beliefs about change and stability matter for how we interpret information about people over time: to what extent is the past a good predictor of future behavior? There’s not always a right answer to that question – but sometimes we definitely have a preferred answer.

For example, imagine you were advising the owner of a bike shop. A likable fellow, Jack applies with the right qualifications. However, you learn that Jack served time for theft in his early 20s. He has been out of jail for five years. Should you hire him? If Jack is a stranger, your habitual belief in the old adage “once a thief, always a thief” may determine your reaction. Now, imagine that the person in Jack’s position was someone you care about– your brother perhaps, or your son. Then, even if you normally think of moral character as set in stone, you may decide – at least for a while – that people can change and deserve second chances.

In a recent paper (Leith, Ward, Giacomin, Landau, Ehrlinger & Wilson, 2014), we tested the idea that people might temporarily alter their beliefs about change and stability when a different perspective served their motives. First, we know people are often motivated to protect their egos in the face of failure. We threatened people’s feelings of competence by giving them (false) feedback on a test, telling them they’d failed. A comparison group was told they’d done especially well on the test. We then asked people how much they agreed with statements like “Your intelligence is something about you that you can’t change very much.” Although people in the two testing conditions didn’t differ from one another in their beliefs about intelligence at the start of the session, after getting their score, those who failed were much more likely to start leaning toward the view that intelligence is, in fact, changeable. It’s easier to handle failure when it’s temporary and alterable rather than enduring.

Although people are notoriously motivated to protect and defend the self, it’s not the only time they show their biases. People’s reasoning also gets lopsided when they start talking politics. For instance, voters have to evaluate political candidates by taking into account information about their current, recent, and often distant past. When voters confront the dirt dug up from their favorite candidate’s past, do they activate different beliefs about change than they do when they contemplate the skeletons in the opponent’s closet? To test this we first approached voters shortly before the last Canadian Federal election, when political hackles were naturally raised. Inspired by the content of the attack ads circulating at the time, we compiled a set of unflattering statements that each of the two leading candidates had uttered in the distant past – an average of 10 years earlier. Liberal and Conservative voters read either questionable quotations from the Liberal or the Conservative candidate. Once again, people showed a lot of flexibility in their views of change: the past missteps of their favored politician were forgiven by appealing to a belief in people’s essential malleability. On the other hand, people ensured that past mud continued to “stick” to disliked candidates by highlighting how people’s core characteristics really cannot change.

In real life, people’s beliefs about the nature of change inform their views on crime and punishment, rehabilitation, and recidivism. Once again, though, we wondered how stable those beliefs really were. In a final study, we asked American adults – about half of whom were parents – to review the case of a previously convicted child sex offender who was paroled after showing evidence of rehabilitation. We asked half of our respondents to imagine that the offender would soon be moving to a community 200 miles away; we asked the other half to imagine the offender was moving into their own neighborhood. We thought that parents who contemplated a child sex offender moving nearby would be driven by the urge to protect their family. Regardless of their habitual beliefs, this group of highly threatened parents shifted to endorse the belief that people, at the core, really don’t change. Importantly, the more they came to believe that people can’t change, the less they accepted evidence that this former offender had been rehabilitated. We recognize that this topic is fraught with legitimate ethical complexities about recidivism, public safety, and rights of the former offender, especially in light of the concrete entity assumptions inherent in the National Sex Offender registry. This research can’t directly comment on the wisdom of the registry; however, it can demonstrate a process by which emotional threat can influence policy decisions. Rather than relying on empirical evidence of rehabilitation or actual recidivism risks, people under threat may reframe a debate by shifting their underlying beliefs about whether rehabilitation is even possible.

As Dan Sperber and Hugo Mercier have argued, “Reasoning was not designed to pursue the truth. Reasoning was designed by evolution to help us win arguments.” Our evidence suggests that one way people may win arguments (with themselves or others) is by selectively appealing to culturally-available beliefs about change and stability in ways that allow them to either disregard the past or affirm its enduring nature. How much does this matter? Short-term at least, it seems to matter a lot. People who even temporarily believed that attributes can change were more likely to agree to retake a test, to forgive a political candidate, and to acknowledge the possibility of criminal rehabilitation. Even temporary fluctuations in people’s beliefs, then, could alter in-the-moment learning, voting, and policy decisions. And conceivably, if these motivated shifts happen often enough, they could form the basis of some of the enduring theories that we come to hold about the nature of change.


Leith, S., Ward, C., Giacomin, M., Landau, E., Ehrlinger, J., & Wilson, A. E.  (2014). Changing theories of change: Strategic shifting in implicit theory endorsement. Journal of Personality and Social Psychology,107, 597-620doi: 10.1037/a0037699


Anne Wilson is a social psychologist at Wilfrid Laurier University in Waterloo, Ontario, Canada. Her research focuses on self and identity, psychological time, and motivated social cognition.