This post originally appeared on RUSI.org on 14 November 2018
Social media is often seen as having a negative impact on politics and society. However, the evidence shows that many individuals are sceptical about the information they consume online.
That much is certain: you are not yourself when you are online. Whether you browse the web or explore cyberspace in other ways, however you get square eyed, while you are doing it, you are somebody else. However, the impact of social media on our politics and our democracy remains a controversial subject. Abstractly, the belief that social media is damaging our politics is now accepted as almost axiomatic, although studies suggest the answer may be more complex.
It is worth recalling that early in the digital revolution there was quite a debate on the supposed damage the practice of staring at phones might inflict on young people. That debate continues as 0–8-year olds are now spending almost an hour a day on their phones, although much evidence for one side or another in the argument remains scant. Still, digital addiction is being seen as a more serious issue: political extremism; radicalisation; recourse to violence; and the broader trend of political polarisation are all thought to be linked to our digital environment.
The usual argument is that social media has helped move the so-called Overton Window, the range of ideas now tolerated in public discourse by fundamentally moving the goalposts and by changing what is acceptable to say in public. As more extreme content and ideas are shared on social media, what is considered normal and moderate shrinks, pushing users to the extremes. But is this truly the case?
An instinct shared by both those who are and are not addicted to social media is to blame the user, to claim that it is ultimately a lack of self-control that leads people to spend on average almost two and half hours a day on social media.
Of course, those who produce content also believe in this concept even though they understand that the reasons are more complex. Advertisers, for instance, have always sought to make media addictive and more manipulative of the user. In the 1950s Vance Packard studied the effect of subliminal stimulation, which is ‘the presentation of short messages that tell us what to do but that are flashed so briefly we aren’t aware we have seen them’. Subliminal stimulation did not work, and its measurable effects were negligible, but it started a pattern of seeking new ways to exert influence on users that continues to this day. Media companies want to have as much influence on their users as possible, and if possible, without their users even realising what effect is being had on them.
And today’s social media companies have specialised teams, using increasingly personal data, to keep users hooked. The design of the sites we use is targeted at creating habits and addictions which in turn they profit from. There is now a ‘zero sum race for finite attention’, using design tricks ranging from red notification bubbles designed to grab your attention, swiping down to refresh, which mimics the action of a slot machine, or infinite scrolling, which without pages to give the user a chance to opt out and stop using an app keeps users locked in a digital cycle.
There is little doubt that these refining tactics are open to manipulation by bad actors. Giving people what they want in order to keep their attention is a problem, because what people want are highly emotive stories, which often turn out to be false. These false stories have an enormous impact on both the inter-social lives of people and on the way that they relate to state apparatus.
Social media’s impact on elections goes deeper, with ‘39 % of adults using social media solely to discuss politics’, and darker, false stories travel further and quicker than the truth. Still, the conclusions drawn from this narrative of social media having a polarising effect may turn out to be empirically wrong.
Levi Boxell, Matthew Gentzkow and Jesse M Shapiro, US-based economists working with the US’s National Bureau of Economic Research, argue in their recently published paper that, while political polarisation among US adults is on the rise, it is rising in demographic groups that use the internet and social media the least. It is not the youth who are becoming radicalised by social media; it is the elderly. And those aged over 75 are the one age group most polarised by their online experience.
Younger, more technology savvy audiences are exposed to more digital content from younger ages, and as result do not take what they see and read online at face value. Older users may be more susceptible to fake news as they are more likely to hold a belief that what is online is generally true. For them the internet has replaced newspapers but in a way that retains their integrity.
So young people, who otherwise may be deemed vulnerable to radicalisation and indoctrination into violent ideologies online, simply by being on social media, may not be affected as much as we fear, and the background chatter is not polarising their beliefs, although admittedly young people may be singled out for indoctrination and recruitment by violent organisations.
This counterintuitive conclusion reached by the National Bureau of Economic Research paper is not an isolated case: Michael Beam from Kent State University in the US arrives at a similar place. He found that ‘Facebook news use was related to a modest over-time spiral of depolarization’. ‘Depolarization’ is the effect of users moving towards more politically moderate views. What Beam seems to be arguing is that, despite the persistent argument that social media creates ‘echo chambers’ where beliefs a user has are repeated and amplified, the truth appears to be that Facebook does provide counter-attitudinal news feeds which have a depolarising effect.
And that conclusion may be entirely sustainable. For it is easy to think that Facebook creates ‘opinion silos—deep but narrow, socially homogenous echo chambers, held together by shared political assumptions’ but 2.23 billion people are on Facebook, and only a minority of them are becoming polarised. A majority can use the platform as it was intended, and by seeing more of the world and the lives of others their worldview is broadened.
In this debate, there is also confusion about how these social sites work, especially since the algorithms that decide what one does or does not see on their timelines are secret. Knowing what you can get posted is also difficult. To be safe many companies will produce exhaustive terms and conditions, leading to many that are too long for users to read.
In the 2016 US presidential election, the advertisements that Donald Trump’s campaign ran were more controversial and resulted in a higher engagement rate. This higher engagement (measured in clicks, comments, and ‘likes’ and ‘dislikes’) resulted in adverts costing less to post compared with the less inflammatory adverts posted by the opposing Hillary Clinton campaign. So, the Trump campaign may have had an edge by just being more controversial, rather than by resorting to any other underhand practices.
Still, this lack of understanding about how these sites can be used is amplified as they can appear to be out of bounds to many legal jurisdictions. Facebook has its own rulebook on what content is and is not acceptable, but as the site becomes more ingrained in all aspects of its users’ lives, the problems become more intractable. How it finds unacceptable content is itself problematic.
And, while algorithms that can identify images are being developed, it is left to human moderators to review at all the content that is flagged, and many are suffering from PTSD from all the extreme content they are being exposed to. Facebook is not looking to national laws to help solve this issue either, with Mark Zuckerberg, its boss, instead hinting that will need its own ‘Supreme Court’ to decide what is and is not allowed.
A reason that social media takes the brunt of the blame for all manner of current social ills could simply be that journalists spend a disproportionate amount of time there, and the time they spend there is not pleasant. Not only do people swear at each other at least twice as much online as in reality, but journalists get the brunt of that as it is ‘in vogue to attack the messenger for the message’. It should be unsurprising then that, when pushed to find answers to current political trends, both journalists and politicians see persistent problems online.
It is not that these journalists are wrong; social media certainly has some part to play in the process of radicalisation. But at a time when real wages are stagnating, life expectancy is dropping, homelessness is growing, and the richest 1,000 families have more than doubled their wealth since 2009, it is perhaps reductive to point the finger at just, say, Twitter.