- Big Data: A Challenge for Democracy? - March 7, 2023
- Rethinking Agroecological Policies: The ProHuerta Program - January 20, 2023
- The Presidential Image: Do Polls Tell the Truth About Gabriel Boric? - December 14, 2022
In the digital age in which we live in, there is new information available every second. From credit card purchases to Google searches, everything leaves a mark. Big data is that large digital footprint that collects what we do both online and offline. What can all this information be used for?
Big data could have multiple uses. For Cukier and Mayer-Schönberger, it’s a tool that “increases society’s ability to take advantage of the information it generates obtaining useful insights, goods and services of significant value.” Yet, its purpose is distinctive: they predict – and on a large scale. The collection of data, combined with algorithms that search for common patterns or trends, is capable of creating profiles of individuals and predicting their future behaviours.
This forecasting feature attracts the attention of political campaigners. For instance, Barack Obama’s re-election campaign in 2012 pioneered the use of big data to win elections. Similarly, although on a larger scale, Donald Trump did it in 2016. In Argentina, Mauricio Macri used it in the campaign that led him to the presidency and it has been a staple for politicians across parties ever since.
The impact of this type of tech on electoral processes is revolutionary. It proves to be more effective than sample surveys and focus groups. It also creates great competitive advantages for those who use them. Nevertheless, its long-term impact can be negative for democratic institutions.
Big Data’s Role on Electoral Campaigning
The key to understanding how the use of big data affects electoral campaigns is through the triangulation of three methodologies, known as Deep Learning.
For campaigns, two types of data are collected: demographic – gender, location, and age – and psychographic. The latter refers to behavioural patterns of individuals – assessing their openness, precautions and personal traits. By analyzing both sets, personality profiles are created. Once grouped, voters receive different ads.
It is no surprise that in the last decade, it has been a change in advertising techniques. Going from mass communication to personalized campaigns. In this context, big data makes it possible – thanks to the profiles created by algorithms – to reach each voter in the most advantageous way. Usually, everyone receives the same message but with variations in image, colours and wording.
Inevitably, this change in advertising influences the message. In addition to receiving personalized propaganda about the candidates, voters also get persuasive messages outside of the traditional form of election ads.
For example, during the 2016 US presidential election, to convince voters to support the Second Amendment, remaining a controversial issue in the country, republican candidates issued two messages. For voters who had a greater sensitivity or a negative view of carrying guns, they got suggested articles about how having a weapon was a form of protection against thieves. The image showed a thief trying to break into a home. On the hand, for those who support the Second Amendment, the message served as a reinforcement. The overall idea was to influence through emotion; either by the fear of being robbed or as part of a family tradition.
Another impact of big data is on efficiency. Specifically, deep learning techniques allow to plan strategies before the effective start of the campaign by having a better understanding of voters as well as what they want and expect from a candidate. It also speeds up logistics by enabling access to real-time information by geographical area, helping to recognize where it is more effective to hold campaign events. In short, the use of big data makes multiple aspects of the campaing empirically observable and predictable.
Big Data & Representative Democracies
With the end of the Cold War, a new – individual – political subject replaced collective identifications. This individualization of the electorate deepened even more with the digital age. Since people feel less and less identified with traditional collectives, candidates must be more flexible. Therefore, it is possible to believe that the most successful candidates are those who are more permeable to change and adapt their discourse. Big data helps perfectly with this reality, but it may be creating – simultaneously – less critical and less informed voters.
On the one hand, by being continually exposed to information filtered by algorithms, personal views are not challenged, thereby making it more difficult to build critical thoughts. On the other hand, by constantly appealing to the emotions, political discourse avoids giving explanations. Therefore, the voter could end up becoming a passive recipient of an extremely tailored message.
Another impact of big data in democracies is the interference of external actors in national electoral processes. For instance, the presumed Russian interference in the 2016 US presidential election, raised questions about the veracity of the electoral processes in which big data was used. The difference is subtle since it ended up being the electorate who cast the vote. Technically, there is no fraud. Knowing whether the voters made an informed decision or the information made a decision for them is very difficult to prove, almost impossible.
- Up to what extent can these methods be used in electoral processes and the results continue to reflect the expression of citizenship?
- How the use of big data affects voter’s privacy?
- Is it possible to fully persuade the electorate?
Choudhary, L. (2022). “How Predictive Analytics is used to Win Elections.” Analytics India Magazine
Judge, E. Pal, M. (2021). “Voter Privacy and Big-Data Elections.” Osgoode Hall Law Journal Vol 58, N° 1.
Nickerson, D. Rogers, T. (2014). “Political Campaigns and Big Data.” Journal of Economic Perspectives Vol. 28, N°. 2,