The college or university a person attends can be a key factor in determining who they will marry, according to a new analysis of Norwegian higher education data, a phenomenon that is not explained by the self-selection of individuals into particular institutions or fields of study via preexisting traits.

The study, published as a National Bureau of Economic Research working paper April 19, found that one in four college-educated couples in the dataset both graduated from the same institution. Moreover, the academic field an individual chooses to enroll in increases the chances of marrying someone within that field, as long as both individuals attend the same institution; enrolling in a particular field of study doesn't make it more likely to marry someone in the same field if the other person attended a different institution.

Assortative mating — who marries whom — is something that "fundamentally shapes our society, as it determines the joint attributes of married couples," according to the study, and the effect of education and occupation on assortative mating has been intensively studied. 

For example, a 2020 analysis showed that U.S. couples are more likely to match in their occupation if they share the same vocational-specific field of study.  While research has observed that these associations exist, though, it's unclear why. 

As an intuitive proposition, researchers have theorized that individuals within institutions might match on traits that are themselves correlated with the choice of college field or institution. In other words, individuals may be self-selecting into a particular college, university or field of study based on pre-determined traits, which in turn leads these individuals to marry others with similar traits. This should theoretically lead to high rates of homogamy — marriage between people from similar sociological or educational backgrounds — within institutions. 

"Basically, we have preferences for partners who are similar to ourselves," Edwin Leuven, a co-author of the Norwegian study and professor at the University of Oslo, told The Academic Times.

A second theory hypothesizes that the choice of college education causally impacts whom one marries. In other words, being at a particular place and time with open opportunities, rather than self-selection based on pre-existing traits, might causally produce high rates of homogamy among college-educated individuals. 

Leuven and his colleagues set out to untangle these two theories using Norway's higher education system. Norway is a particularly good case study for this research for several reasons, as Norwegian register data allows researchers to observe people's college education choices, including field and institution, but also their workplace after they graduate and whom they marry or cohabitate with. 

Another strength of using this data is that Norway's centralized admissions process effectively randomizes applicants, with nearly unpredictable admissions cutoffs, into different institutions and fields of study. This allows researchers to naturally observe how outcomes differ when a student is placed in one institution and field of study versus another. Since a student may not make the cutoff for a particular field based on their competition, researchers are able to isolate students who get pushed into their second choice.

"This is what allows us to pin down the causal effect — it's that there are these application thresholds in the system that people cannot control exactly on which side they find themselves," Leuven said. "So that means that we can compare people who are around these thresholds."

The researchers used records that provided information on nearly all applications to higher education in Norway from 1998 to 2004. They merged this information with administrative registers provided by Statistics Norway that cover information on each resident from 1967 to 2017. The final sample size was 110,345 students who applied for a least two programs, where the most preferred program had an admission cutoff and the next-best alternative had a lower cutoff. 

The researchers found that a student enrolling in their preferred institution increased the homogamy rate within that institution by 9 percentage points by the 10th year of the study. Likewise, a student enrolling in the preferred field of study had a smaller effect, increasing homogamy with respect to that field by about 4 percentage points. 

After adjusting for the differences in the distribution of men and women, since they differ more across fields and programs than they do institutions, the researchers found that the homogamy rate increased by 15 percentage points for institutions and about 7 percentage points for fields of study. 

The researchers also noted that that one in eight college-educated couples have degrees in the same field, and while there are some fields of study that have higher rates of homogamy than others — such as law and medicine, where enrollment increases homogamy by 8 and 13 percentage points, respectively — the researchers ultimately found that enrolling in a particular field makes it no more likely to marry someone from other institutions with the same field. What really mattered was that individuals in the same field attended the same institution.

Since the admissions cutoffs randomize applicants into particular institutions and fields of studies, the researchers also eliminated any correlation between educational choices and predetermined preferences and abilities, implying that the large impacts on homogamy and assortativity reflect the educational choices that individuals make rather than predetermined traits.

This paper contributes to academic literature in this field of study by managing to pin down the causal effect of education choices on mating markets, Leuven said, bringing academia one step closer toward more fully understanding mating assortativity and the driving factors behind high levels of homogamy among college-educated individuals.

"What we show is that institutions are extremely important in explaining this," Leuven said. "And that the nature of this matching is very local within institutions, which suggests that people are making these matches in very, very narrow local environments." 

The study "College as a marriage market," published as a working paper on April 19 by the National Bureau of Economic Research, was co-authored by Lars Kirkebøen, Statistics Norway; Edwin Leuven, University of Oslo and Statistics Norway; and Magne Mogstad, University of Chicago and National Bureau of Economic Research.