Psychologist and researcher Robert Epstein has published research (embedded below) that he contends shows how search rankings “have the potential to profoundly influence voters without them noticing the impact.” He calls this capacity the “Search Engine Manipulation Effect” or “SEME.”
The study tested how search rankings affected the perceptions and candidate preferences of “2,000 undecided voters throughout India.” India just concluded national elections.
The research found that rankings could affect undecided voter opinion significantly: “votes in India can easily be pushed toward one candidate or another by about 12 percent . . . enough to determine the outcomes of many close races.”
The research was all hypothetical. There’s no claim that search rankings on Google or elsewhere did in fact bias or impact the Indian election. The idea is that “manipulated rankings” could affect election outcomes.
The study also argues that certain groups may be more susceptible to “SEME.” For example the voting preferences of women over 35 and “the unemployed” were shifted more dramatically that the group as a whole. Epstein also argues that in the real world front-runners tend to rank higher in search results thus giving them a kind of built-in advantage over challengers. (This is an empirical question not tested in the study.)
Epstein is apparently a vocal Google critic. It’s unclear whether his anti-Google bias entered into the design of the study or the presentation of results. If that were true, however, it would be highly unSEMEly.