Why opposites don't always attract: A global study of 41,606 people across 74 countries found shared values matter most, while believing your partner is kinder and more attractive predicts happier relationships

If you’ve ever wondered whether your relationship needs more in common with your partner, or just needs you to see them through slightly rosier glasses, science has an answer, and it’s a bit of both. A new study, ‘Partner idealization and perceived partner similarity predict relationship quality across 74 countries,’ published in the Journal of Research in Personality and led by psychologist Marta Kowal of the University of Wrocław, surveyed 41,606 people in relationships across 74 countries and found “opposites attract” to be mostly a myth. But exactly what you’re being similar or dissimilar about is a question of compatibility, and for American millennials who probably came of age swiping through dating apps, and then saw politics seep into their first dates, this might land at a strangely personal moment.

The plot twist: not all "sameness" is created equal
In the study, participants rated themselves and their partners on nine traits: health, kindness, physical attractiveness, religiousness, resources, social class, education, political orientation, and age, as well as their overall love and relationship satisfaction.

The results presented two clear patterns. For values, especially political orientation, being on the same page mattered a lot. The study found that differences in political views predicted a lower quality of relationship. This is true in the US, where dating surveys have repeatedly shown political identity shaping who Americans are willing to date.


Kindness and looks: it's not about matching, it's about admiring
Now here is where it gets interesting. For traits like kindness and physical attractiveness, similarity barely mattered. Kowal said the findings suggest what matters most is not being evenly matched, but perceiving a partner as very kind and attractive. Of the nine traits, kindness was the single strongest predictor of relationship quality, followed closely by physical attractiveness. The study also found an “additive” effect: relationships were happiest when both people rated themselves and their partner highly, a base of mutual admiration.

This is not a new concept. According to a landmark 1996 study in the Journal of Personality and Social Psychology, by Sandra Murray and colleagues on "positive illusions" in relationships, people in happier partnerships tend to see their partners in a more positive light than their partners see themselves, and the study suggests similarity and idealization are associated with relationship quality. This isn’t a Western thing, though how much it happens depends on where you are: Kowal’s global data shows.

It’s worth being precise: this is about perception, not objective reality. It’s not that your partner is objectively nicer than you on some universal scale; it’s just that really believing that tends to go along with a happier relationship.

Where you live changes the rules
Culture turned out to be a major factor. A country’s Human Development Index, individualism, gender equality, and “relational mobility” (how freely people can pick or leave a partner) affected which traits mattered most, the study found. The strongest effects were for kindness and attractiveness in highly modernized, individualistic countries with high relational mobility. In less modernized countries, the focus shifted to status and matching on social class, and education was a better predictor of stability.

This is not just Kowal’s study; earlier research points in the same direction. In a meta-analysis of 313 separate studies, titled ‘Is actual similarity necessary for attraction? A meta-analysis of actual and perceived similarity,’ R. Matthew Montoya and colleagues find that actual similarity predicts attraction early on in relationships, but perceived similarity continues to predict closeness and satisfaction in established relationships.

A necessary reality check
Before you start using your partner’s politics as a compatibility test, a few caveats matter. This study relied on self-reported data from one partner, not both, so it’s about perception, not verifiable fact. This is also just a snapshot, not a years-long tracking study, so it can't prove that idealizing your partner causes happiness, rather than happiness leading to idealization; people who are already happy might just start seeing their partners more favorably. Kowal herself acknowledged this limitation, suggesting future research could look at couples longitudinally and seek the views of both partners.

The takeaway for your dating life
So what do you do with this? Shared values seem to really matter for long-term compatibility, especially when it comes to politics. But for subjective traits like kindness or attractiveness, worrying about whether you and your partner are “evenly matched” might be the wrong instinct. This research shows that the healthier pattern is mutual admiration, where both people see the best in each other and are seen in the same way in return.

Opposites might not always attract, but perfect matching isn't the secret either. What seems to work is getting on the same page about what is really important to you, and really admiring the person you have chosen.