티스토리 뷰
Studies in America suggest that the social networking site has the tools to influence voter turnout
There are two things about 2015 of which one can be reasonably certain: there will be a general election in May and it's unlikely to produce an overall majority for either of the two big parties. In those circumstances, small, localise events might have big implications: a Ukip candidate shoots his mouth off about, er, non-white people; a Labour candidate turns out to have an embarrassing past; a Tory garagiste cannot differentiate between sexual harassment and bum pinching. The kind of stuff, in other words, that could affect the outcome in a finely balanced constituency
which brings us to social media and the question of whether the 2015 general election could be the first one in which the outcome is affected by what goes on there. Could Facebook, for example, be a factor in determining the outcome of some local constituency battles?
Fat-fetched? Maybe. But the question is worth asking because in the 2010 US congressional elections, Facebook conducted an interesting experiment in social engineering, which made some of us sit up. The company collaborated with some political scientists to see if a social network could persuade apathetic American voters to get off their butt and vote. And the answer was yes.
The methodology used was simple enough. Sixth-one million Facebook users were shown an icon containing a link for looking up polling stations, an "I voted" button to click to announce they had voted, and the profile pictures of up to six of their Facebook friends who had indicated they'd already done the same. The icon and button were inserted in the newsfeeds of tens of millions of users, while others were shown either a generic get-out-the-vote exhortation or no message at all. Then the researchers cross-referenced their subjects' names with the day's actual voting records from precincts across the country to measure how much the Facebook voting prompt actually increased turnout.
The Harvard law professor Jonathan Zittrain summarised the findings thus: "Overall, users notified of their friends' voting were 0.39% more likely to vote than those in the control group, and any resulting decisions to cast a ballot also appeared to ripple to the behavior of close Facebook friends, even if those people hadn't received the original message. That small increase in turnout rates amounted to a lot of new votes. The researchers concluded that their Facebook graphic directly mobilised 60,000 voters, and, thanks to the ripple effect, ultimately caused an additional 340,000 votes to be cast that day. As they point out, [in 2000] George W Bush won Florid, and thus the presidency, by 537 votes - fewer than 0.01% of the votes cast in that state."
In itself, the experiment was innocuous: after all, in a democracy encouraging people to vote can only be a good thing. But Facebook is a big data company and what such companies do is experiment on their users all the time. Most Facebook users probably have no idea that what appears on their newsfeeds is determined by algorithms, which are constantly making guesses about what they might want to see and determining what Facebook wants them to see
So far, so unremarkable: that's the manipulative reality of social networking services. What's more interesting is that some of these ongoing user "experiments" may have emotional or political dimensions. In one such study, for example, an experiment involving 660,000, Facebook users showed that "emotional states can be transferred to others via emotional contagion, leading people to experience the same emotions without their awareness". It provided "experimental evidence that emotional contagion occurs without direct interaction between people (exposure to a friend expressing an emotion is sufficient) and in the complete absence of non-verbal cues"
So Facebook can influence the emotions of its users. Could it also influence their interest in politics? Micah : Sifry, the co-founder of Personal Democracy Media, reports that in the months leading up to election day in 2012, Facebook made a change to the newsfeeds of 1.9 million users in order to see whether it could influence those users to become more interested in political activity: it did this by increasing the number of hard news items that appeared at the top of a user's newsfeed. The results were a "statistically significant" increase in the amount of attention users paid to government-related news
None of this amounts to any kind of smoking gun. But, given that social media clearly influence behavior in many other areas of life, it seems implausible to imagine that when it comes to politics, they don't have any impact. Which means they now wield power of an unaccountable kind. In an election period, we fiercely regulate broadcasters' coverage of the campaign to ensure "balance" and "fairness". Should we now do the same for Facebook? More importantly, could we?
to ensure balance and fairness
fiercely regulate broadcasters' coverage of the campaign
they now wield power of an unaccountable kind
don't have any impact
implausible to imagine that when it comes to politics
given that social media clearly influence behavior in many other areas of life
smoking gun
amounts to
None of this amounts to any kind of smoking gun
in the amount of attention
the results were a statistically significant increase
appeared at the top of a user's newsfeed
by increasing the number of hard news items
to becom more interested in political activity
in order to see whether it could influence those users
made a change to the newsfeeds
reports that in the months leading up to election day in 2012
the co-founder of Personal Democracy Media
also influence their interest in politics
can influence the emotions of its users
the complete absence of non-verbal cues
an emotion
exposure to a friend expressing an emotion is sufficient
emotional contagion occurs without direct interaction between people
leading people to experience the same emotions without their awareness
emotional states can be transferred to others via emotional contagion
an experiment involving 660,000
some of these ongoing user experiments may have emotional or political dimensions
the manipulative reality of social networking services
so unremarkable
determining what Facebook wants them to see
which are constantly making guesses
is determined by algorithms
have no idea that what appears on their newsfeed
encouraging people to vote
the experiment was innocuous
cast in that state
thus the presidency
ultimately caused an additional 340,000 votes to be cast that day
ripple effect
concluded that their Facebook graphic directly mobilized 60,000 voters
amounted to a lot of new votes
in turnout rates
hadn't received the original message
any resulting decisions to cast a ballot also appeared to ripple to the behavior of close Facebook friends
more likely to vote than those in the control group
overall, users notified of their friends' voting
summarised the findings
to measure how much the Facebook voting prompt actually increased turnout
from precints across the country
cross-referenced their subjects' names with the day's actual voting records
a generic exhortation
while others were shown either a generic get-out-the-vote exhortation
of tens of millions of users
were inserted in the newsfeeds
who had indicated they'd already done the same
of up to six of their Facebook friends
to click to announce they had voted
polling station
were shown an icon containing a link for looking up polling stations
The methodology used was simple enough
to get off their butt and vote
to see if a social network could persuade apathetic American voters
collaborated with some political scientists
which made some of us sit up
conducted an interesting experiment in social engineering
congressional elections
is worth asking
fat-fetched
in determining the outcome of some local constituency battles
in which the outcome is affected by what goes on there
brings us to social media
could affect the outcome in a finely balanced constituency
cannot differentiate between sexual harassment and bum pinching
turns out to have an embarrassing past
shoots his mouth off about non-white people
localize events might have big implications
in those circumstances
is unlikely to produce an overall majority for either of the two big parties
a general election
of which one can be reasonably certain
has the tools to influence vote turnout