It would be discouraging if humans’ choice of mates and romantic partners operated just like a market. That would violate our idealistic notions of romance, and it would imply a deep source of inequality. Above the level of basic subsistence and safety, many people care about nothing more than their partners. And if the pairing process works like a market, then some people have much more market power than others. That would be a form of inequality that is very hard to address, since we must have the freedom to choose whom we love.

Indeed, much of the previous literature suggests that romantic pairing does work like a market. Everyone has a perceived value. You try to snag the person with the highest value, and what you offer in return is your own value. Thus more highly valued individuals can expect higher ranked partners. As Eastwick and Hunt summarize previous research:

The classic perspective on mate value suggests that people possess romantically desirable qualities to different degrees; that is, some people are more attractive, more intelligent, or more popular than others. …

The most consistent finding in the mate value literature is that people with higher self-reported mate value (i.e., higher self-esteem in the mating domain) report higher standards for the qualities that they desire in romantic partners … . This finding is consistent with both the social exchange and evolutionary predictions that people should pursue the best partner that they can realistically obtain …—that is, people with wonderful traits should expect that their partners will have wonderful traits.

It’s also true that college students converge in rating the same students as most romantically desirable–evidence that they each have a market value:

participants tended to agree on which of their opposite-sex classmates did and did not possess these desirable traits. They also achieved consensus regarding which of their classmates were popular with members of the opposite sex—another classic measure of mate value.

But Eastwick and Hunt offer news to warm the hearts of romantics and idealists. Once college students know a group better, their estimations of who would make the best partner diverge dramatically. After they have interacted, everyone is not drawn to the same target; they diverge substantially in their valuations of the pool:

Participants exhibited considerable uniqueness in their judgments of who was attractive, intelligent, and popular, and they strongly disagreed about who was likely to be a good relationship partner. When the Study 3 participants reported on opposite-sex individuals whom they had known for a considerable period, they reached very little consensus about these qualities and exhibited huge amounts of relationship variance.

That finding could mean at least three different things. First, people could be “settling,” once they can see who is a realistic partner for them:

Participants could be using their self-ratings as a guide to settle for the best partner they could realistically obtain. … For example, participants could rate a target as especially high in vitality/attractiveness to the extent that the participant’s self-assessment and the target’s self-assessment on vitality/attractiveness are similar.

But the study finds little empirical support for this first explanation–in fact, it is empirically refuted.

Second, people could have very diverse tastes. Some like brie; others prefer Velveeta. There is nevertheless a market for cheese, and each brand has a price. It just happens to be a very segmented market. Once you have tried both kinds of cheese, you will realize which fits your tastes.

If the same were true for relationships, it would mean that after we get to know people better, we go beyond superficiality and form more accurate perceptions of them; and once we have that data, we vary more in our assessments. It would be like seeing two pieces of cheese and rating them the same, but deciding after you chew them both that you much prefer one. This would be modestly good news because more people could be fully satisfied by their romantic choices on account of their varied tastes. But brie still costs more than Velveeta because more people (with more money) want to buy the former. In the same way, Molly might have a higher market value than Sally even if some actually prefer Sally. Inequality and disappointment would persist, just not as badly as would happen if everyone liked Sally better.

The third explanation is most idealistic–yet still consistent with the data. Perhaps what matters is not what the other person brings to the pair but what you build together. As Eastwick and Hunt put it, “Given that two people can uniquely inspire the expression of traits and the experience of positive affect in each other, much of the variance in mate value judgments may be a function of the dyad.” We imply that romantic relationships are functions of two inputs when we talk about “chemistry” or “compatibility.” The result would still be beyond the control of the parties if the function were automatic: if Sally plus Barry (or Molly) equals a good result. But it could rather be that Sally and Barry make the relationship, and they can do that either well or badly. They are not consumers of each other but co-producers of a new thing. People change their estimates of the romantic potential of others while they are getting to know one another because they are already starting to construct relationships, and it’s the relationship that matters most.

Although consensus emerges on desirable qualities in initial impression settings, this consensus is weaker than the tendency for participants to see one another as uniquely desirable or undesirable, and over time, relationship variance grows while consensus declines.

This all sounds very dry and clinical, but it’s conceptually interesting because it suggests that the market metaphor is not useful for understanding relationships. That’s actually the cheeriest scientific finding I have heard in a while–unless you prefer to know that mice love to run on wheels and will choose to do so even if they’re free.

(See also: Dickens and the right to be loved.)

Source: Eastwick, Paul W., and Lucy L. Hunt. “Relational Mate Value: Consensus and Uniqueness in Romantic Evaluations.” Journal of personality and social psychology 106.5 (2014): 728-51. ProQuest. Web. 21 May 2014.