Over the past several years, data has been utilised to facilitate election campaigns in various countries, but the multifaceted notion of Data-Driven Campaigns (DDC) has become more prevalent in recent years. With elections scheduled for more than 60 countries in 2024 (with 44 countries having already voted, including the European Union Parliament), voter data has played a major role in how political parties prepared and contested their national elections. The advent of big data and enhanced analytical techniques have given political parties in-depth insights into voter behaviour and attitudes. These data-driven insights enable electoral campaigns to tailor their messages and amplify their impact.
The United States (US) is set to go to the polls in November. In the US, successful use of data and statistics in the 2008 and 2012 presidential campaigns of Barack Obama and Donald Trump set the tone for employing data-related techniques in presidential elections. Political parties collect data through various sources ranging from available voter records, online and offline events, social media sources, as well as data brokers, etc. Data brokers combine voter lists with their data and create a behavioural profile of voters based on the inferences regarding demographics, affiliations, interests, hobbies, etc. These voter profiles are further used to micro-target different segments of voters. In 2020, various US political groups spent approximately USD 23 million to obtain data and services from 37 different data brokers. This expenditure highlights the significant role that data brokers play in political campaigns, offering services ranging from voter information to targeted advertising strategies. The heavy investment in data underscores the importance of data-driven decision-making in modern political campaigns.
US political ad spending is expected to reach USD 12.32 billion in 2024, marking a 30% increase from 2020. Notably, there will be a substantial 156% rise in spending on digital platforms. On these platforms, candidates are opting for innovative strategies and moving towards contextual AI target advertisements, which process vast data from different sources to identify trends. For instance, by analysing what people read online, AI can aid in crafting messages related to pressing issues like abortion and immigration, which are the anticipated topics likely to dominate the US election cycle.
Meanwhile, in the United Kingdom, the winning Labour Party also employed data-driven strategies and digital infrastructure to enhance targeted messaging. Their campaign hired digital experts from Google and established a digital trainee scheme to stimulate regional digital campaigning.
Elections in India, the world’s most populated country, were also characterised by micro-targeting. In India, the BJP government is the biggest holder of voters’ personal data and has been involved in selling it to private companies. Moreover, for Lok Sabha 2024, the Saral app (Sangathan Reporting and Analysis) was developed by the BJP to engage with its followers. The app collected detailed voter data to the extent that many data and electoral experts raised concerns about how it might have helped the BJP in targeting voters at a granular level for political leverage.
This shows that political parties now increasingly mobilise their core supporters through DDC and engage disengaged and unsure voters. In this regard, the wealth of social media data is a pivotal source (along with polling and voter registration data) of tailored decision-making for many election campaigns worldwide.
Ironically, though, in the recent French election, data showed that it was anticipated the right-wing National Rally party would win a majority in the National Assembly. Bettors on Polymarket, a popular crypto-based prediction market platform, indicated a 97% chance of this outcome, which also aligned with the polls. However, neither prediction markets nor polls data predicted the actual result: a hung parliament.
Many scholars have raised serious transparency and accountability concerns regarding DDC’s implications on democratic norms. The 2018 Cambridge Analytica scandal highlighted how data could be used and abused to attain political ends. The political consultancy firm’s unauthorised harvesting of users’ data from Facebook garnered criticism about how it undermined the democratic process and citizen’s right to privacy.
Data-Driven Campaigns have also been connected to voter surveillance and profiling, leading to electoral manipulation. As discussed earlier, by analysing various data sets, political parties can attain a detailed understanding of attitudes and opinions to the extent that they can predict people’s beliefs even before they have fully formed. People’s beliefs are subtly steered through expensive market research and psychological manipulation, making them think they were the ones who made the decision. This raises serious questions about voter privacy and ability to exercise their choice freely.
Advancements in voter data analysis are making the field of Digital Data Campaigning increasingly ambiguous, often pushing ethical and legal boundaries. Political parties, in their quest to micro-target voters for electoral gains, are likely to start treating voters like commercial commodities, potentially bypassing national privacy regulations.
This situation necessitates a concerted effort at the individual and national level.
Voters need to be critically informed and educated regarding their online data consumption and how it can be used. Hence, they should verify information from various sources and be mindful of sharing their private data online. Moreover, at the national level, political parties should comply with regulations related to usage of voter data in election campaigns and where no regulations exist, they should be drafted e.g., like UK’s General Data Protection Regulations and Data Protection Act 2018. States also need to formulate more concrete national legal regimes regarding the use of voter data and the extent of micro-targeting per their local norms and contexts because the nature of data usage is likely to vary in each state. Measures like these would ensure that, in the Age of Data, the integrity of a voter’s personal data remains intact and safe from manipulative electoral targeting.
Ajwa Hijazi is a Research Assistant at the Centre for Aerospace & Security Studies (CASS), Islamabad, Pakistan. She can be reached at cass.thinkers@casstt.com.
