Disinformation campaigns are being used everywhere to undermine democracy. Yet, most accounts of online disinformation swaying elections come out of the US and the UK. The media’s disproportionate coverage of these two countries is partially to blame for this - but are there similarities between the US and UK political systems that make them more vulnerable to manipulation attempts?
Both the US and the UK effectively have a two-party system in which the winner takes it all. This means whoever gets most votes in the constituency will represent the entire region - indifferent of how close the runner-up came to beating the eventual winner. In the UK, this led to around 68% of 'wasted votes' in 2017. These are votes that were not cast for the winning candidates in their respective constituencies. In the US, this is illustrated by the difference between the popular vote and the electoral vote. Hillary Clinton won around 2.9 million (or 2.1%) more votes than Trump. Yet, Trump won the election based on the share of states in which he came first.
How does this affect the role disinformation campaigns play in swaying elections? There are two types of constituencies in winner-takes-it-all (or first-past-the-post) electoral systems. Those that are safely in the hands of one party (e.g. red or blue 'states'), and those that are disputed (e.g 'swing states'). The latter are the constituencies that determine the outcome of an election. Many disinformation campaigns, and particularly those that deploy micro-targeting, only impact a marginal share of the entire electorate. Yet, it is this marginal share that is most valuable in winner-takes-it-all electoral systems. In a disputed constituency, a tiny share of the electorate can have sweeping consequences for the entire election. Famously, only 537 votes (or 0.00009%) out of around 6 million votes eventually lead to George W. Bush winning the state of Florida in 2000. He subsequently won the presidential election. If disinformation campaigns, supported by techniques such as micro-targeting, can individually identify and mobilise these key swing voters, the potential for bad-faith actors to successfully interfere in democratic elections is enormous.
The effectiveness of different types of disinformation campaigns
Researchers are still trying to understand the effect of online disinformation campaigns on political preference formation. Yet, the relative value of a tiny share of votes in winner-takes-it-all systems compared to more proportionally representative systems means that disinformation campaigns, aided by micro-targeting, will be more relevant in places like the US or the UK.
However, representative systems are not inherently more resilient to disinformation attempts. Many disinformation campaigns rely less on micro-targeting but on large-scale inauthentic and coordinated activity such as astroturfing or hashtag-trending. These efforts seek to artificially inflate the perceived popularity of a certain political agenda or actor. As the recent (representative) European Parliament Elections have shown, the goal here is usually to boost turn-out for insurgent parties on either side of the political extremes. In representative systems, this will inevitably lead to these parties gaining a share of seats in parliament that is proportional to their turn-out. In comparison, the electoral thresholds of winner-takes-it-all systems keep insurgent parties out of parliament - at least in the short term.
Indifferent of the electoral system, the availability of personal data and the reach of social media both afford bad-faith political actors with a diverse disinformation toolkit. The type of electoral system, however, determines which tools are most effective for bad-faith actors to wield influence.
How can we respond?
In the long term, electoral reform may be inevitable - not only with an eye to electoral interference, but also representational issues arising from the decline in popular support for catch-all parties. In the short term, governments must seek to improve the transparency surrounding elections. This may be done by increasing reporting requirements for party donations, campaign financing and political (online) advertising. Online imprint rules are a first step in the right direction. They have been introduced by the Australian government last year and will likely soon be introduced in the UK (PDF). Yet, solely relying on transparency will not be enough to combat online disinformation. Data protection authorities must be vigilant of personal data (ab)use by political campaigns. The UK's Information Commissioner has already shown initiative in this area.
Unfortunately, government regulation is a long and tedious process - authorities will always play catch-up with the ever-changing manipulation tactics of bad-faith actors. This is why potential victims of such campaigns cannot solely rely on government action to protect them. They must also have the capacity to identify small-scale contextual anomalies on social media that may indicate an emerging disinformation attempt. That's why granular monitoring of social network activity is a key prerequisite in the fight against online disinformation. Once vulnerable actors have that understanding and the processes in place, they can then start tackling the spread of disinformation through pre-emptive strategic communication.
Sparked your interest? Want to stay informed amidst the daily onslaught of online disinformation? Subscribe to our weekly newsletter!