If the European Data Protection Supervisor gets his way, a ban on the use of targeted political advertisements, aka political microtargeting, is just around the corner.
In 2018, the British-American data company Cambridge Analytica came under heavy fire for allegedly having been involved in manipulating Facebook users in support of the Trump campaign during the build-up to the 2016 U.S. presidential elections.
Political microtargeting, however, is by no means an exclusively American phenomenon, being common practice in many European countries as well. Dutch citizens, for instance, may have seen political ads popping up on YouTube and Facebook during the recent municipal elections where, as shown by an Argos study, Dutch parties collectively spent over half a million Euros on Facebook advertising.
In November 2021, the European Commission presented a legislative proposal on political advertisements containing guidelines related to transparency and targeting. The measures outlined include labelling ads as political content where appropriate and subjecting the use of sensitive personal data for targeting techniques to the requirement of data subject consent.
The proposed measures, however, do not offer sufficient safeguards in the eyes of the European Data Protection Supervisor who, instead, argues for an all-out ban on microtargeting for political purposes.
In this blog, we will be asking the following questions. What exactly is political microtargeting and what are the dangers? We will also touch on a number of highlights from ‘Opinion 2/2022 on the Proposal for Regulation on the transparency and targeting of political advertising’ as published by the EDPS.
Microtargeting allows advertisers to specifically target pre-defined audiences. In order to do so, large amounts of data have to be collected first, which is done through internet platforms or by data merchants. The next step is an analysis of these data, the results of which then serve as the basis for the definition of target groups consisting of people who share common traits, characteristics or interests. Finally, it is up to the advertiser to create group-specific content aligned with the preferences of a particular audience segment.
One of the frequently made objections to political applications of microtargeting is the inherent threat to democracy because of the possibility the technique offers of selectively communicating tailor-made messages to very specific target groups. As a result, the people seeing the ad may be presented with a far from complete picture of what a particular political party stands for.
So what, you may say. In real life, candidates for public office are commonly just as selective in the presentation of their political agenda. That hardly means they should not be allowed to campaign in public places or directly interact with members of the electorate, does it? Correct. The difference, however, is that the digital world makes it uniquely possible, based on the analysis of online behaviour and personal data, to create detailed user profiles which can then be used to manipulate large audiences by showing each particular segment the message that perfectly fits specific sentiments, beliefs and needs. Moreover, there is the ever-accelerating pace of technological development. Advertising techniques that we consider to be cutting edge today, may very well be used with even more target precision in just a few years’ time. Already, the existing algorithms are pretty effective in estimating which political message, communicated in which way, will appeal to which demographics. And they will only get better at it in the future.
As mentioned above, the EDPS is not convinced by the proposal put forward by the European Commission, finding it lacking in scope and, as a result, unable to successfully tackle all varying types of microtargeting techniques. One of the concerns voiced by the EDPS, for instance, is that possibly significant electoral impact may not be exclusive to microtargeting based on sensitive or special categories of personal data. ‘Normal’ personal data may be equally effective in fuelling the technique.
In short, the EDPS finds the proposed bill to be lacking in protective power and suggests an all-out ban on political microtargeting instead. Among the examples of microtargeting listed by the EDPS are the selection of messages and/or target groups based on specific character traits, interests or preferences. Also, the EDPS suggests further reduction of the categories of personal data admissible for processing in support of political advertising, with the intention, in particular, of preventing targeted advertising based on intrusive tracking, as in the case of processing information related to online behaviour.
On the one hand, a total ban on online political advertising would be a remarkable initiative. A significant part of our modern lives is taking place in the virtual world, which therefore presents the perfect platform for reaching audiences less inclined to watch traditional televised content like talk shows and political debates in times of elections. On the other hand, the extreme pace of technological development is making it increasingly easy to manipulate large parts of the electorate through online advertising.
Weighing the pros and cons against each other, a ban on political microtargeting seems to be the preferable long-term solution in view of the highly biased perception it may create in the voter’s mind, particularly when considering the ever more advanced capabilities of the technique. There is a real and significant danger of voters making their choices based on seriously selective information, whereas the very idea of a democratic society is rooted in the notion of citizens being able to make truly, properly and literally “informed” decisions. The manipulation of choice by means of advanced political microtargeting techniques will, in the long run, be counterproductive to the cause of democracy.