How 'digital gerrymandering' can swing the American election
Are you having a crushingly tedious day at work? Try playing Google autocomplete. Type the start of a question into the world's most trusted search engine, and marvel as the aggregated curiosity of the global crowd is marshalled to anticipate your inquiry
When I searched "who", "what" and "where", the inappropriately punctuated responses were: "who is donald trump"; "what does trump mean"; "where do i vote". It seems that the US presidential election is the most pressing global issue among Google users; and that many are still unfamiliar with the attention-seeking, wall-building, women-groping, thrice-married septuagenarian Republican nominee, with a history of six corporate bankruptcies.
Could Google influence the presidential election?
Or is it possible that search results and the way they are ordered reflect the interests of Google rather than its users? The company is under no obligation to divulge details of the proprietary algorithm it uses to deal with search queries and rank web pages. So we can only guess at whether there is any intended or unintended bias in the information returned and possibly acted on by voters. It has led the journal, Science, to ask a pertinent, though better punctuated, question: "Could Google influence the presidential election?"
A survey of the research suggests yes - in theory. In one study, Robert Epstein, a psychologist at the American Institute for Behavioural Research in California, built a rigged search engine, enrolled about 300 typical American voters, and asked them to choose their preferred candidate in a two-horse race (it happened to be the 2015 contest between Julia Gillard and Tony Abbott for Australian prime minister, about which the subjects knew little).
Participants whose search queries resulted in one candidate's web pages near the top were more likely to choose that candidate; those who saw a more balanced mix of links were more evenly divided. Mr Epstein also tested his theories in real elections in India, with similar outcomes.
The ability to manipulate search engines, he concluded, is "a means of mind control on a massive scale". There is no evidence, incidentally, that Google is manipulating search engines in this way, although the company does use "relevance" and "credibility" criteria to rank search results.
In 2010, Facebook decided to encourage voting in midterm congressional elections by including reminders in some users' feeds, a button to click once they had voted and information about whether their friends had already done so. The social media giant calculated that its intervention accounted for 60,000 votes overall. That sounds paltry - until you remember that in 2000 George W Bush secured the presidency with 537 votes in Florida.
Facebook calculated that its intervention accounted for 60,000 votes overall
Professor Jonathan Zittrain of Harvard Law School has since described the potential for "digital gerrymandering" - tech corporations influencing the outcome of elections by, say, sending voting reminders to some users and not others. Facebook and Twitter already know your political leanings. Digital gerrymandering is possible whenever personalised information is served up by an intermediary, and it can tilt the game when margins are narrow.
Most insidiously, the tilting can be done surreptitiously. After all, with personalised news feeds nothing is false; rather, different users are fed variants of the same world, each of us gazing out on to a bespoke reality. Facebook now owns messaging service WhatsApp and Instagram, giving it more data and more channels of influence, especially among the young.
That is why a public conversation about the role of tech companies as brokers of unbiased information is so urgently needed. Because, as Mr Trump didn't quite say: "There might be something going on."
The writer is a science commentator
For more articles from the Financial Times/ FT.com please register here
Source: Anjana Ahuja. 2016. 'How 'digital gerrymandering' can swing the American election
'. Financial Times / FT.com. November 3, 2016 Used under licence from the Financial Times. © The Financial Times Limited 2016.
All Rights Reserved. Articles sourced from the Financial Times have been referenced and are used under licence from The Financial Times Limited. These articles remain the copyright of The Financial Times Limited and were originally published in 2016. All rights reserved. “FT” and “Financial Times” are trade marks of The Financial Times Limited. The Financial Times Limited has not endorsed, verified or been involved in the creation of the information provided from other sources in this publication, and is not responsible or liable for its accuracy, completeness or content.