As Twitter bans all forms of political advertisement and Facebook takes the opposite stand, regulators struggle to make sense of the complex issue of targeted ads and their impact on the democratic process.
With the 2019 U.K. general election underway and the 2020 U.S. election just around the corner, the shadows of the Cambridge Analytica scandal and of Russian interference in the 2016 U.S. elections have reemerged. Political advertisement through social media is once again under the spotlight for its potential disruptive effects on the democratic process. Part of the debate has revolved around micro-targeting, which enables advertisers to deliver ads, including potentially misleading ones, to specific groups. Pressure is on tech companies to take appropriate measures to avoid past mistakes For the first time, those concerns have been voiced by a social media giant itself: in an unexpected move, Twitter’s CEO Jack Patrick Dorsey has announced a decision to stop all political advertising on Twitter. Facebook has instead refused to even fact-check political ads, and insisted on increased advertisement transparency as an efficient strategy to tackle misinformation.
According to The Wall Street Journal, however, Facebook has also been discussing restricting micro-targeted political ads – specifically by raising the minimum number of people who can be singled out from a hundred to a few thousands. For its part, Google has announced that it will no longer allow political advertisers to target voters based on their political affiliation and interests.
But are these policies enough to protect democratic processes from digital political advertisement? And does democracy need to be protected from it in the first place?
According to political and advocacy advertising consultant Joe Fuld, digital ads are not problematic per se: “From what we saw in the last U.S. election, it’s clear that the root of the problem was not legitimate ads, but fake news. Google and Twitter, with their policies, are not really dealing with the core issue.” All those policies do, claims Fuld, is aggravate an already uneven playfield: “In the context of increasing media consolidation, an ad ban creates more difficulties for legitimate candidates with less resources, advocacy groups and NGOs. These policies are only going to make the political discourse worse.”
However, in Fuld’s opinion, Twitter’s announcement is not particularly worrying: “Twitter is such a small part of the problem. No one was using it for political advertisement in the first place. Google’s policy is a bigger issue. If Facebook made that kind of change, it would be even more troublesome.”
As stated by Amber Macintyre, researcher at Tactical Tech, a Berlin-based NGO that investigates the impact of digital technologies on society, one of the problematic aspects of these regulations is that they are reactive: “All these platforms are responding to negative press by adopting quick fixes”. A broader debate is needed.
Broadening the picture
A wider conversation must include the topic of intermediate actors such as political campaign firms and consultants, whose importance has been overlooked. As Macintyre explains, “There has been much talk about the Cambridge Analytica scandal, but it is important to realize that Cambridge Analytica was just one of the companies that make up a larger political influence industry. In our research, we found at least 250 such companies, but I am sure there are hundreds more.”
These actors use a wide variety of tools, not just digital ads, to target and influence voters. They base their targeting on personal data they acquire through various sources. These sources include, just to list a few examples, voter registration records, supporter databases, or information acquired through “social listening”, that is finding out what a potential voter is interested in through how they interact with social media posts.
Another valuable source is consumer data, such as information about financial transactions or any loyalty cards a person might have. This type of information is bought from data brokers – large companies that possess in-depth data about millions of people. According to Macintyre, the use of these sources causes even the most mundane actions to become political ones: which supermarket you do your shopping at says something political about you.
What is the big issue?
With this material in hand, political actors are able to craft different messages to appeal to different groups. According to Elliot Jones, researcher at Demos, a British think-tank, these micro-targeted messages are not inherently dangerous for democracy: they might even be beneficial, in the sense that they allow people to be reached by information on topics they care about. They also don’t have such a worrying role in the spread of fake news since, as Jones explains, misinformation in the United Kingdom spreads mainly organically. However, problems arise when a political actor uses targeted ads to spread conflicting messages: it creates a breach of trust.
In Macintyre’s opinion, even more worrying is how this modus operandi shapes and limits the conversations that people around the world are having: “Quantitative metrics are substituting political discussions. This leaves not much room for nuance, which is so important in the democratic discourse.” She argues that politicians are now able to identify and target those who might vote for them, and leave out of the conversation those who they think would not cast a vote in their favor. This really limits the extent of the knowledge some groups of people have of certain political parties.
Political actors are making decisions about which messages potential voters should receive based not on who they are, but on what political consultants think their personal data says about them. “Instead of talking to you, they’re talking to a ‘data double’ of you, which becomes their constituent”, Macintyre says. The notion that ads simply provide citizens with relevant information is, according to her, an easy justification to dismiss the most important question: at what cost?
Is greater transparency the solution?
Legislators have not yet been able to regulate this complex and rapidly changing phenomenon. One of the possible solutions that is often discussed is increasing the transparency of political ads. This is the option that Facebook, among other actors, such as Snapchat or Google, has decided to adopt: all political ads are freely accessible in the Ad Library, an archive which includes information about who paid for the adverts and the targeting, reach and amount spent.
According to Elliot Jones, however, “Facebook’s Ad Library is better than anyone else’s, but still quite bad. The Application Programming Interface through which you access the data is buggy, and the data is less granular than it should be”. But the main problem – Jones argues – is that Facebook is left free to decide which ads should be classified as political, and which information to show. Ultimately, it is impossible to completely trust Facebook’s transparency. According to Jones, a solution might be regulating what data the governments expect ads libraries to provide. Additionally, all political actors should be required to disclose information about digital adverts to an electoral commission, thereby creating a central repository.
Macintyre, while agreeing that Facebook’s Library is a step in the right direction, points to a different problem: “Those who are going to look at the library will be mainly researchers and journalist. Their analysis will take months, whereas as human beings, we start making decisions the moment we are presented with information. The library does not really help with the real-time sharing of information.”
She agrees with Jones that the decision on what information is disclosed shouldn’t be left to the platforms: “Tactical Tech suggests that it’s time to make more long-term, strategic decisions about regulations. The process should involve legislators, citizens, businesses and even future stakeholders, such as WhatsApp. Political actors should also take more responsibility, instead of handing over important decisions to campaign consultants.”
But in the meantime, Macintyre adds, there are actions that individuals who want to detoxify from targeted political advertisement can take: “The complexity of the phenomenon makes people feel overwhelmed. But citizens can take a number of steps, such as keeping location data in check, having a look at ads libraries, using incognito browsing modes or changing settings on social media.”