Panel Paper: Preferences Between English-Speaking and German-Speaking Buyers on eBay Kleinanzeigen

Friday, March 9, 2018
Room 16 (Burkle Family Building at Claremont Graduate University)

*Names in bold indicate Presenter

Janet Li, University of Southern California

Motivated by recent sentiments of a high-profile German politician, Jens Spahn, against English-speakers in Berlin, our experiment examines differences in the behavior of German sellers on the used goods platform eBay Kleinanzeigen when responding to messages sent in German versus English. When online market platforms like eBay were first released, many predicted that they would be better than traditional face-to-face markets at preventing discrimination against users, as a user’s identity can be better anonymized. However, as platforms are allowing for the input of more user data, this theory comes into question. Research on this topic through randomized controlled trials has been conducted across many platforms, including Airbnb and eBay. Our experiment is focused on the iPhone market in the four largest German cities. We sent roughly 600 messages over the course of two weeks, randomizing the source between two different eBay accounts with the same name, with the same generic message asking for a 10 percent discount on an iPhone listing. Our hypotheses were that English messages, as compared to German messages, would receive fewer responses, more negative responses, slower responses, and less cordial responses from German sellers. Our study found that German sellers of iPhones on eBay Kleinanzeigen respond less to buyers who send messages in English compared to those who send messages in German. A message in English is 30 percent less likely to get a response than a German message. These results show a demonstrable difference in German sellers’ preferences between users using different languages, a factor which could also be a flag of nationality. Significant differences between responses to English vs. German messages for other factors, like the type of response (positive, negative, or neutral), the length of the response, and the politeness of the response, were not found. Future research can expand the sample size, market type, and market locale, and also analyze response times. This research has implications on policies addressing the inclusivity of transactions on online platforms, which are increasingly prevalent in our current information age.