Pradel, Franziska (2020). Biased Representation of Politicians in Google and Wikipedia Search? The Joint Effect of Party Identity, Gender Identity and Elections. Political Communication, 1–32. https://doi.org/10.1080/10584609.2020.1793846

Abstract

Web search engines have become an important and trusted source when people seek political information. Even though previous research suggests that information about politicians in traditional and new media can provide content that makes stereotypes based on gender and party, little is known about the presence of such bias in search engines, which function as information gatekeepers in the digital age. Using quantitative text analysis and human coding techniques on a novel data set of members of the German parliament, this study examines whether search engine suggestions, i.e. search predictions, for politicians differ with respect to personal and role-oriented information based on the gender and party of the politician. It also explores whether the search engine representation of politicians changes around elections. The study further compares gender and party differences in search engine results with corresponding Wikipedia articles of the same politicians, as users are most often redirected to Wikipedia from Google. The results suggest that politicians’ representation in search engines and Wikipedia are structured by a joint effect of their gender and party identity. While Google suggestions provide less personal information about female politicians belonging to a right-wing party compared to their male counterparts, this relationship is not observable for left-wing parties. Moreover, there are changes in gender biases around the election. In Wikipedia articles, politicians belonging to right-wing parties are represented with more personal information compared to politicians belonging to left ones, an effect which is even stronger for females.