Being More Educated and Earning More Increases Romantic Interest: Data from 1.8 M Online Daters from 24 Nations

Being More Educated and Earning More Increases Romantic Interest: Data from 1.8 M Online Daters from 24 Nations. Peter K. Jonason & Andrew G. Thomas. Human Nature, Apr 5 2022. https://link.springer.com/article/10.1007/s12110-022-09422-2

Abstract: How humans choose their mates is a central feature of adult life and an area of considerable disagreement among relationship researchers. However, few studies have examined mate choice (instead of mate preferences) around the world, and fewer still have considered data from online dating services. Using data from more than 1.8 million online daters from 24 countries, we examined the role of sex and resource-acquisition ability (as indicated by level of education and income) in mate choice using multilevel modeling. We then attempted to understand country-level variance by examining factors such as gender equality and the operational sex ratio. In every nation, a person’s resource-acquisition ability was positively associated with the amount of attention they received from other site members. There was a marked sex difference in this effect; resource-acquisition ability improved the attention received by men almost 2.5 times that of women. This sex difference was in every country, admittedly with some variance between nations. Several country-level traits moderated the effects of resource-acquisition ability, and in the case of unemployment this moderating role differed by sex. Overall, country-level effects were more consistent with evolutionary explanations than sociocultural ones. The results suggest a robust effect of resource-acquisition ability on real-life mate choice that transcends international boundaries and is reliably stronger for men than women. Cross-cultural variance in the role of resource-acquisition ability appears sensitive to local competition and gender equality at the country level.

Discussion

Using 1.8 M online dating profiles, we found that resource-acquisition ability and sex had a small, but robust influence on the amount of interest a dating profile received. Specifically, being a woman or having higher resource-acquisition ability led to increased numbers of messages, “winks,” and “likes” from other members (i.e., IOI). These patterns showed considerable cross-cultural consistency: resource-acquisition ability generally increased IOI in all countries, and, except for the USA, profiles of women generally received more IOI than those of men. Even in the USA, this sex difference reversed only at high levels of resource-acquisition ability (more than 2 SD above the mean). There was some variability between nations in the enhancing effect of resource-acquisition ability and its differential effect on the sexes, but this was simply a matter of degree. That is, resource-acquisition ability enhanced dating profile attention broadly, and for men more than women specifically, in all countries, though some more than others. These national sex differences accounted for only a slither of the total variance in attention received among the population. Nonetheless, we were still able to associate this variance with some country-level traits: gross national income, sex ratio, unemployment, and gender development.

Besides informing on the roles of culture, sex, and resource-acquisition ability on mate choice, we examined how differences in social, political, and economic differences in the various nations accounted for mate choice and sex differences therein. Although we cannot claim our results are definitive (given, e.g., sampling homogeneity in the countries), our results reveal that (1) sex differences persist in all countries sampled and (2) they appear relatively insensitive to the nation-level variables we considered. This may be more in line with evolutionary models of sex differences in mate choice than sociocultural ones because the latter treat sex differences as artifacts of culture (e.g., Eagly & Wood, 1999). Our results showed considerable agreement with studies from labs, smaller datasets, and mate preference research suggesting that even when people are actively choosing mates from the comfort of their couches, regardless of their country of origin, evolved mate selection tendencies are expressed.

Our research draws attention to the unique challenges of working with data of this magnitude (e.g., everything was significant). To cope with these challenges, we relied on confidence intervals to understand country-level patterns and descriptive differences (i.e., percent increase) as prima facie evidence of ostensible population-level effects. This process revealed just how small some of these effects might be. This may, in part, be the result of being unable to account for the primary feature that predicts romantic interest—physical attraction (Jonason & Antoon, 2019; Jonason et al., 2019; Kenrick et al., 1993; Li et al., 2002). At the same time, as both epidemiologists and evolutionary theorists have appreciated for some time, small effects over large populations and periods of time are not bereft of impact (Dawkins, 1996; Rose et al., 2008). In the online dating space, one additional message received might, for some people, change their mating trajectory entirely, with real consequences for their happiness and their reproductive success. One fruitful approach to deal with these kinds of data in the future may be to adopt Bayesian analyses instead of null-hypothesis-testing procedures.

Limitations and Conclusions

Despite the size and scope of our study, it still had several limitations. First, we focused only on two predictors of romantic interest even though our data have several more. We did so because of the exponentially increasing complexity afforded by including more variables, and instead of focusing on describing who gets more IOI, we focused on theory-testing of microscopic issues. Of all the variables we have, we felt that resource-acquisition ability was the timeliest (e.g., the rise of the topic of sapiosexuality), the most useful for considering mate choice in relation to two theoretical paradigms, and one that has applied implications for mate searching and child mortality (Egebark et al., 2021; Hopcroft, 2021). The magnitude of these data is simply too much to conscientiously allow for exploratory tests when everything is likely to be “significant” but unlikely to be meaningful. Subsequent studies will examine the effects of height, marital status, number of children, and more. We presented here the first of a series of studies relying on “real” and “really big” data to understand cross-cultural patterns in mate choice using those seeking mates and people’s bone fide interest in them, not some hypothetical interest.

Second, resource-acquisition ability, as a factor influencing mate choice, is likely to have several related indicators, such as ambition, social status/level, and earning capacity (Buss, 1989; Li et al., 2002). We were only able to examine two of them—treated as a single index—given the limitations of what was collected. Although the two may not fully represent the larger construct of competence or resource-acquisition ability as we envision them, we think the results are more than defensible given their alignment with theory and having, themselves, been used as indicators of research in the past (Egebark et al., 2021; Hopcroft, 2021; Jonason & March, 2021). Indeed, the moderate correlation between the two may be reflective of the fact that we have only two indicators on a much larger mate-choice determinant (along with potential error in that data).

Third, despite the cross-national nature of this data, our sample was still WEIRD. (Henrich et al., 2010). Although countries such as Chile and Mexico might not traditionally be considered “Western,” they are educated, industrialized, and rich enough to have online dating services and Internet access. This may have created some range restriction and limit our results to just the countries where the dating service operates. It remains to be seen whether these patterns would hold up in African, South American, and Asian nations. Nevertheless, if we take an evolutionary perspective, differences in countries are a matter of degree for local calibration; Homo sapiens are humans everywhere (Buss, 1989; Thomas et al., 2020).

A final limitation involves our ability to account for country-level variance. Members in our dataset came from 24 countries, which allowed us to develop an understanding of how consistently resource-acquisition ability affects dating profile attention and how much this varies from country to country. However, even with two dozen nations, this aspect of our analysis was underpowered. Thus, we took a conservative approach when adding them to the base model. In contrast, other studies on the idea of “evoked culture” tend to examine country-level effects while controlling for variables such as longitude and latitude (Gangestad et al., 2006), though arguably, even then, such analyses are underpowered. Including more countries, particularly from non-WEIRD nations, would help us draw firmer conclusions.

In conclusion, we have provided the most definitive answers yet to the questions of the role of resource-acquisition ability in mate choice, whether there are sex differences in that role, and what nation-level factors might account for national patterns overall and in the sexes. We showed that greater resource-acquisition ability leads to more dating profile interest in data from more than 1.8 million people living in 24 nations who use the services of an international, online dating company. While both sexes received a boost in interest when they had more resource-acquisition ability, the increase was almost 2.5 times stronger in men than in women. And last, resource-acquisition ability tended to be slightly less important in richer countries with more women of reproductive age than men, and slightly more important in cultures with greater gender equality. Higher levels of unemployment also seemed to make resource-acquisition ability more important, but this effect was restricted to the amount of attention women’s profiles received. The relative primacy and robustness of sex differences suggest evolutionary models of mate choice may be more powerful than sociocultural ones when it comes to resource-acquisition ability.