A general view of the Radcliffe Camera building, part of the Bodleian Library, in Oxford city center in England. Oxford University has a student population in excess of 20,000, taken from over 140 countries around the world. (Oli Scarff/Getty Images)

Since Kant, liberal scholars of international relations have hypothesized that greater cross-border contact can be a powerful force for good. The idea is that such contact encourages a sense of shared international community, breaking down artificial barriers separating people into different nations and inhibiting their natural human affinities for one another. This intuitively appealing hypothesis has inspired several famous student exchange programs, which among other potential positive effects are expected to quell nationalist fervors and ward off international conflict.

Despite its ubiquity, this hypothesis has rarely been tested in a rigorous way. Does cross-border contact really foster a feeling of community? In a recent study, I used a natural experiment across a sample of American “study abroad” students at 11 colleges in New England, the Midwest and the South to carry out a unique test. The institutional structure of study abroad makes it well-suited for a natural experiment. Students are typically placed in foreign settings for either the fall or spring semester, with the winter break providing a valuable window during which a treatment group of students just returning from a semester abroad can be compared with a control group of students who are about to embark. Since all subjects are predisposed to participate, the design controls for self-selection, and the choice of which semester is a logistical one with no obvious implications. These are significant design improvements over earlier studies that did not control for self-selection or lacked a strong control group.

More than 500 students were surveyed on their feelings of international community, perceptions of foreign threat, and levels of nationalism and patriotism, as well as demographics and study abroad program characteristics. As expected, those returning from a semester abroad (the treatment group) were not significantly different either demographically or in terms of program choices from those about to take their semester abroad (the control group). For instance, they selected the same host countries in which to study abroad, especially Spain, France, Italy and the United Kingdom, and females outnumbered males in both groups. All this mirrors the general population of American study abroad students, who are majority female and tend to study abroad most in Western Europe.

First, I tested the core liberal hypothesis that cross-border contact promotes a sense of shared international community, or what political scientist Karl Deutsch called a “we-feeling” across cultural divides. Theorists define this in terms of warmth, shared understandings and values, and trust. Surprisingly, the hypothesis was not supported: None of the indicators for international community was higher on average for students returning from study abroad than for those yet to travel. In fact, those who had just returned from a semester abroad felt they had significantly fewer values in common and were more likely to say their understandings of key concepts were different from the people of their host country. None of this was sensitive to potential moderators like whether or not students opted to live with a host family. Given the intuitive plausibility of the liberal hypothesis, these results are striking.

Jones_Fig1Blue

How about threat perceptions? I asked students to rate how threatening they would consider their study abroad host country if it were to surpass the United States in terms of material power, such as economic growth or military expansion. In theory, cross-border contact should mitigate perceptions of foreign threat and foster expectations of peaceful change and cooperation, despite uncertainty and shifts in the distribution of power. And indeed, given identical scenarios, those just returned from a semester abroad rated their host countries as less threatening than did students about to leave. So the liberal hypothesis that cross-border contact mitigates threat perceptions was supported, even though the hypothesis that it fosters “community” was not.

Finally, I tested a variant on the “clash of civilizations” hypothesis—that cross-border contact, rather than encouraging a sense of shared international community, promotes nationalism. Perhaps troubling for some, the results strongly supported that hypothesis.

Jones_Fig2Red

Students returning from their study abroad experience were considerably prouder of America along a range of dimensions, including its literature, achievements in the arts, armed forces, athletic accomplishments and political influence. They were also prouder to be American, warmer toward American culture and more patriotic. Importantly, however, they did not display a heightened belief in America’s superiority; there was no difference in that attitude across the two groups. So while cross-border contact heightened nationalism, it did not appear to promote a virulent or chauvinistic form of it.

There are several key take-aways. The first is that efforts to influence group identity in positive ways — such as programs that engineer cross-border contact — may not work the way we think they do. The second is that “community” at the international level may not be so important after all. Although treated students did not come back with a stronger “we-feeling,” they did show a lessened tendency to view “the other” as threatening. Cross-border contact may still be a strong force for peace, even if community is not the underlying mechanism.

We are used to thinking about nationalism and internationalism as mutually exclusive; people who are highly nationalistic are often assumed to lack the cosmopolitan mindset of a “global citizen.” Yet study abroad returnees were both more nationalistic and less prone to seeing other nations as threatening. Rather than fostering a sense of shared international community and warm realizations of “we are the same,” cross-border contact may instead encourage a form of “enlightened nationalism”—a sharper sense of national difference, and pride in that difference, tempered by tolerance and the realization that such differences need not be threatening. In a globalizing world where cross-border contact continues to grow, it is perhaps enlightened nationalism rather than utopian notions of international community that should be encouraged. Indeed, if nationalism were capable of evolving from the bogeyman of the 20th century into a peace-promoting norm of the 21st, we would do well to help it get there.

Calvert Jones is an assistant professor of political science at the University of Maryland.