The Market of Disinformation

The 2016 US elections continue to cast a long shadow over democratic processes around the world. More than 40 countries are pondering legislative responses (Bradshaw, Neudert, & Howard, 2018). Meanwhile, the tech platforms have made more than 125 announcements describing how, through self-regulation, they will solve the manipulation of their platforms by bad actors (Taylor, Walsh, & Bradshaw, 2018). Among the more frequently referenced self-regulatory measures are changes to algorithms and the use of artificial intelligence (AI) to demote disinformation and junk news. We ask whether these changes took place, and if so, have they had the intended impact of reducing the spread of disinformation on social media platforms? To date, much of the policy debate has focused on paid-for advertising on the platforms, but what about the viral spread of unpaid, organic content? The ‘black box’ nature of today’s most widely used platforms makes it difficult for researchers and journalists to understand how algorithmic changes might be affecting both legitimate political campaigning and disinformation. It is essential that any reform of electoral regulation or oversight in the UK is informed by an understanding of the techniques used in both the paid and the unpaid markets of disinformation. The digital marketing industry can offer insights, albeit incomplete and heuristic in nature, into the impact of algorithmic changes. Social media marketing and search engine optimization (SEO) – that is, the practice of guessing, testing, and experimenting with algorithms so that searches for particular words appear higher in search results – are part of a multi-billion-dollar industry built upon understanding how these obscure technical systems rank, order, sort, and prioritize information. By interviewing professionals and reviewing reports from the digital marketing industry, we can gain insight into the impact that algorithmic changes might have had on the distribution of content online. The findings provide an additional evidence base that can inform the Oxford Technology and Elections Commission’s project to identify potential regulatory reform of elections.