What is in a surname? The role of ethnicity in economic decision making

Abstract

This article reports results from two experiments that investigate possible incidence of discrimination against people with foreign backgrounds in Sweden. In the first experiment, participants played the trust game and the dictator game with co-players of different ethnic affiliation. The family name of the players was exposed to their co-players. Results for the trust game showed no significant discrimination against co-players with foreign backgrounds. On the other hand, the results for the dictator game showed a statistically significant discriminatory behaviour by men against co-players with non-European backgrounds. The discriminatory behaviour was solely a male phenomenon. In the second experiment, the dictator game was replicated to check the stability of the results in the first experiment. The second experiment also examined whether people with foreign backgrounds discriminate against other people with foreign backgrounds; that is, the purpose was to discover whether discrimination is systematic. The observations in the second experiment underlined the results found in the first experiment: foreign co-players are discriminated against by Swedish players. However, we did not find that people with foreign backgrounds discriminated against other people with foreign backgrounds.

Acknowledgements

The author would like to thank Karl-Markus Modén, Ebba Hallsenius and Johanna Sturesson for their help during the experiment. The author would also like to thank James Albrecht, Fredrik Andersson, Lars Behrenz, Fredrik Carlsson, Jan Ekberg, Fredrik Hansen, Håkan Holm, Susanna Holzer, Håkan Locking, Jonas Månsson, Marcus Selart, and seminar participants at Växjö University, Karolinska Institutet, Göteborg University and the AMER and Swedish Institute for Emigration Symposium on Research in International Migration for valuable comments.

Notes

¹ The same debate could probably be found in many other countries as well. But we discuss the Swedish case since we conducted our experiments in Sweden. Probably, the same argument could be applied to other countries. There are many studies based on registered data that indicate discrimination; for example, see Hammarstedt (2001) and Rooth and Ekberg (2003) for Sweden and Darity et al. (2002) and Queneau (2005) for the US.

² Source: Swedish Patent and Registration Office. In 1999, 1159 individuals in Sweden changed their surnames, 114 of whom changed from a foreign-sounding surname to a Swedish-sounding surname. In 2004, the corresponding figures were 1252 and 276, respectively. It is reasonable to suspect that individuals who changed their foreign-sounding surname were overrepresented in 2004, since roughly 10% of the Swedish populations had a foreign background during these years.

³ It should also be mentioned that it is presumably a bigger step to change a foreign-sounding surname to a Swedish-sounding surname than it is to change a Swedish-sounding surname to another Swedish-sounding surname.

⁴ Bouckaert and Dhaene (2004) is another study that is related to the present study. They conducted a trust game experiment, but in the field instead of the laboratory. They did not find any discriminatory behaviour.

⁵ See, for example, Holm (2000), Saad and Gill (2001a, b), Solnick (2001) and Eckel and Grossman (2001). Eckel and Grossman (2008) gave an excellent overview of experiments with gender in focus.

⁶ Individuals with foreign backgrounds were individuals who immigrated to Sweden or individuals born in Sweden with parents who immigrated to Sweden.

⁷ Includes individuals with non-European backgrounds, except individuals with North American and Australian backgrounds.

⁸ Includes individuals with European (except Swedes), North American and Australian backgrounds.

⁹ The written directions and decision sheets are available upon request.

¹⁰ Which was equivalent to €11 or $12.50, at that time.

¹¹ This method was first proposed by Selten (1967). Rather than observing the participant's behaviour in only those nodes a player reaches, the strategy method asks for a decision of the participant in every node of the game. There are both advantages and disadvantages of the strategy method; see Roth (1995). Most importantly, the advantage was the possibility of observing participants’ entire strategies, while a drawback is that participants were forced to think about behaviour in each node, which could lead to different outcomes than if the game had been played in an ordinary manner. To avoid potential misunderstandings, extensive written and oral information was provided.

¹² See Siegel and Castellan Jr (1988) for an overview of these tests.

¹³ In the post-experimental questionnaire, trustors where asked to state the amount of money they expected their counterpart to send back. There were no significant differences in expectations between different groups. However, it is important to point out that this analysis was done on individuals who had sent positive amounts. Of course, those players who did not send anything did not expect anything in return.

¹⁴ Trustees were also asked to state the amount of money they expected their counterpart to send them. There were no significant differences in expectations among different groups.