Many of us like to think our moral values are stable. We imagine that we’ve each got a moral compass inside us, with the needle reliably pointing us north — toward our sense of right and wrong.

But according to a new study published in Proceedings of the National Academy of Sciences, our deepest moral values are not some fixed point we’re magnetically drawn to. They fluctuate depending on whether it’s spring, summer, fall, or winter.

We already know that seasonal changes affect our mood (I’m looking at you, seasonal affective disorder). And research also shows they influence everything from our attention to our memory, to color preferences to crime rates. But our morality?

“It was definitely surprising to me,” lead author Ian Hohm of the University of British Columbia told me in an interview.

Hohm said he didn’t expect morality to shift “because it’s not a behavior, and it’s not a kind of transient emotion. It’s a fundamental part of how people make decisions about what’s good and bad and right and wrong.”

Yet when he and his coauthors analyzed survey responses from more than 230,000 Americans over the span of a decade, they found that people embraced certain values — purity, loyalty, and respect for authority — more strongly in spring and fall, compared to summer and winter.

Psychologists call those “binding values” because they tend to promote social cohesion. They’re a double-edged sword, however: Their upside is that they increase cooperation among members of an in-group, but they can also increase prejudice against out-groups. (The study authors also investigated care and fairness, sometimes known as “individualizing values,” but found no consistent seasonal pattern there.)

The findings raise some big questions. First, why do people embrace binding values more in the spring and fall? And second, given that binding values are strongly associated with political conservatism, what are the implications for how we run a society — including major court cases and elections?

There’s something else that peaks in spring and fall: self-reported anxiety.

The study authors noticed this trend when they analyzed more than 90,000 questionnaire responses. They hypothesize that the higher anxiety they found in spring and fall drives people to put a higher premium on binding values and their manifestations, like close-knit social groups and long-running traditions. That sort of stuff makes us feel safe and comforted — exactly what you want when you’re feeling anxious.

The authors also analyzed survey data from Canada and Australia, and found that there, too, people embrace binding values more strongly in spring and fall. Endorsement of those values dips in the summer — and by looking at respondents’ zip codes, the authors found that the dip is extra sharp in areas where the weather changes dramatically from one season to the next. But, interestingly, the winter dip wasn’t related to the weather itself.

“So the summer effect might be some kind of sunshine effect — you know, your anxiety goes down because you’ve got nice weather outside,” Hohm said. Whether it’s because we’re getting more daylight, more time in nature, more exercise, or more socializing, lots of us notice an improved mood in summertime.

“But the winter effect, we think, is about some sort of cultural practice, like the changes in people’s activities and motivations around the holiday season.” In other words, you might be less anxious in December because you’ve got time off work or school and you’re hanging out with family and friends.

Of course, this isn’t true for every single person. And individuals who do see a change aren’t experiencing massive shifts in their moral values every time the seasons turn.

“The effects on people’s moral values at the individual level are quite small,” Hohm told me. “If you’re looking at individuals, this effect probably isn’t going to explain much. But when it comes to human populations in the aggregate, when everybody is endorsing these values a little bit more, that can have large consequences.”

The study argues that these findings could have wide-ranging social implications.

We know that conservatives endorse binding values more than liberals do. So, one possible implication is that if we hold elections when people are endorsing binding values more, that could make on-the-fence voters slightly more inclined to opt for a conservative candidate than they would’ve been at other times of year. “The timing of political elections (whether they are scheduled to occur in the summer or the autumn, for instance) might plausibly have some subtle effect on election results,” the study argues.

Research also shows that people who more strongly endorse binding values tend to be more punitive toward norm violators. So, conceivably, there could be some seasonal variation in legal decision-making.

Finally, since binding moral values emphasize group cohesion, there might also be seasonal variation in prejudices against people — like immigrants, racial minorities, or LGBTQ people — who are viewed as outsiders. (The study authors have begun looking into this, and although their research is still in progress, Hohm told me prejudice “does seem to increase at the same time of the year when people are endorsing binding values more.”)

There are important limitations to the study. For one, it doesn’t analyze nationally representative samples of people, but a more self-selecting crew that chose to answer survey questions online. Plus, it’s not a longitudinal study — different people responded to the survey at different times. We’d get a fuller picture by assessing the values endorsed by the same people across different seasons.

One more wrinkle: the findings are not generalizable to every country. Although the authors found a similar seasonal cycle in moral values in the US, Canada, and Australia, they did not find it in the United Kingdom, and they don’t know why.

Still, the study challenges the popular view that each person’s values are stable.

“The idea that people’s moral values can change with the seasons is consistent with a growing body of research highlighting just how much moral values — which are typically thought of as fixed and immutable — can change based on subtle or seemingly mundane features of people’s environment,” said Daniel Yudkin, a social psychologist not involved in the new study.

With apologies to your moral compass, the takeaway here is: There is no moral North Pole. Your sense of right and wrong might not be a static point, but something that’s subject to seasons of change — literally.

This story originally appeared in Today, Explained, Vox’s flagship daily newsletter. Sign up here for future editions.