Beyond the Bubble: Assessing the Diversity of Political Search Results

Abstract

Search engines are among the most popular online services used in a range of contexts, including to find information on political issues. As such, they increasingly act as powerful mediators between news organizations and their audiences. The claim that search results are politically biased, while hardly new, has also recently received fresh support from President Donald Trump, who has blamed Google for unfairly prioritizing news outlets critical of his policies. In the context of elections, search engines may serve to inform citizens and have been argued to sway the choices of undecided voters. I examine two related issues: How political parties and candidates are represented in Google Search results and how strongly results in both Google Search and Google News are personalized in the run up to the 2017 German general elections. My results suggest that some parties and candidates are able to exert greater influence over how they are represented in search results than others, through a combination of local branch websites and social media presences. I furthermore find only a small share of results which differ from the mainstream, while controlling for time, language, and location, calling into question the validity of the filter bubble concept.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 119 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.