Abstract:

This paper briefly describes the set-up and results of an experiment conducted among undergraduate students of human resource management at the University of Hamburg. The research focuses on the effects that social capital and migratory background may have on the students’ performance – individually and in teams of two – and on their choice of team partners. The results are intended both to contribute to a range of related literature and to generate stimuli for improvements to teaching practices.

The project is motivated by recent evidence suggesting a positive connection between social capital and student performance. This paper will revisit the relationship between social capital and student performance, though with special attention devoted to the students’ migratory status. Firstly, we expect that as the students choose a team partner, they prefer persons who are similar to them in terms of their socio-demographic characteristics, including social capital and migratory status. In other words, we are looking for evidence of in-group bias. Secondly, we expect that the students’ socio-demographic characteristics, in particular again their social capital and migratory background, have a measurable impact on their performance in class, both individually and in teams.

We find significant evidence of in-group bias with respect to gender, migratory background, and the frequency of meeting colleagues outside of work or university. Regarding individual performance, we do find a significant negative impact for migratory background, which confirms the authors’ prior results obtained in similar settings. If we forego individual performance and instead try to explain a team’s performance by the personal characteristics of its members, it becomes much harder to achieve a reasonable goodness of fit. We first try to capture social capital by an index, which we also introduce in squared form to achieve at least marginal significance of the coefficient estimates. We thus see weak evidence of an inverted-U shape impact of the social capital index on team performance. Therefore, social capital is beneficial only up to a certain point, beyond which performance tends to decline again because time spent building up social capital (‘socialising’) is time spent not studying. In other words, more social capital need not be better.

Much clearer results on social capital can be obtained if we drop the index and instead examine the individual components. We discover a strongly negative coefficient on the frequency of meeting friends. By contrast, meeting colleagues more often is associated with higher team performance. The fact that two seemingly quite similar indicators of social capital can have radically different effects on performance suggests that the phenomenon of social capital should be studied very carefully.

Keywords:

Social capital, migratory background, team formation, team performance, diversity, peer instruction.