|Originally published in Inside the Influence Industry's Personal Data: Political Persuasion - How it works by Tactical Tech's Data & Politics team.|
Apps aren't just for music, shopping and ride-sharing anymore; increasingly, they are becoming tools for politicians and campaigns to gain support and win votes.
Political campaign apps generally fall into one or more of three categories:
↘ mobile apps designed to support specific political candidates or particular causes.
↘ enhanced canvassing apps developed to combine information gathered from door-to-door canvassing with data from the campaign, commercial sources and public records.
↘ games or gamified apps created to mobilise an existing base of supporters and attract new voters.
↘ Mobile apps give politically like-minded people an exclusive space to interact and share ideas outside of larger social media platforms, where views can be more mixed. Some apps encourage participation by gamifying the experience, such as letting users accrue points and unlock badges for completing certain tasks like watching campaign advertisements, tweeting pre-written political messages, sharing their contacts with the campaign or calling their representatives to discuss preset talking points.
↘ Enhanced canvassing apps ostensibly allow campaign volunteers to visit homes door-to-door more efficiently. They give canvassers detailed information about the homes in their area, including who lives there, what party they are registered with, when they have voted in the past and what issues they care about. Apps can also supply customised scripts and survey questions for canvassers to ask residents based on their profiles. As canvassers visit homes, they upload the information they collect via the app and it is immediately recorded in the campaign's central database. Behind the scenes these apps gather and match voter registration, bankruptcy, criminal offence and other public-record data. Canvassers using NGP VAN's enhanced canvassing app, for instance, can present a candidate's views on issues affecting former military members if the app indicates that a veteran lives at the home. These apps may also employ elements of gamification, such as leaderboards for the most productive canvassers.
↘ Games or gamified apps: Although they do not appear to use personal data and are the least common among these three groups, online political games like CorbynRun (created to support the Labour Party in the UK) and Super Klaver (supporting the centre-left De Groenen party in the Netherlands) are worth mentioning for their novelty. These games, which tend to use 8-bit graphics and lo-fi audio, are easy to understand and can help build communities around political goals. The introduction to CorbynRun, for instance, reads: 'we're in a race against time time to defeat a rigged system... Together we can win!' Fiscal Kombat, built for 2017 French presidential candidate Jean-Luc Mélenchon, shows the candidate fighting political opponents for money to pay for his policies while the rich try to defeat him. Over the course of the game, Mélenchon encounters the chairwoman of the International Monetary Fund, Christine Lagarde; French politician Jérôme Cahuzac, who was prosecuted for tax evasion; and former French president Nicolas Sarkozy. By combining creative elements with topical political matters, games like these introduce new ways of engaging with political ideas while also perhaps blurring the line between reality and fiction.
Campaign apps capture various types of data that can benefit a campaign as well as the app creator, who can adapt the user experience to solicit even more information from the user. Signing up is often free, and according to makers of such apps, the cost of creating and maintaining an app is 'set off by the data that can be gathered'.
Campaign apps typically collect four types of data:
↘ Data explicitly supplied by the user (such as name, email address, phone number, postal code, gender and age).
↘ Information about the user's social networks. Some apps reward users with points for sharing their address book contacts, which the campaign can cross-reference with its list of target voters. If a potential swing voter is found in the address book of an app user, the user will be prompted to invite the voter to the app with a preset, personalised message. Thomas Peters, CEO of the political app service uCampaign, explained how their Ted Cruz 2016 app reached out to potential swing voters they'd already identified based on voter files: 'if we identify that you have 10 friends in Iowa who are potential Cruz supporters, then we'll ask you to reach out to those people'.
↘ Surveys or quizzes within apps can also supply personal data to campaigns. According to Peters, 'app supporters have completed over 20,000 political ID surveys about themselves, their friends and their neighbors, generating valuable cross-section data on the supporters' political views, activism affinities and personal network, essential information for a modern, data-driven campaign'. While Cambridge Analytica's final product was not an app, the data for its psychometric profiles originated from a Facebook app called 'This is Your Digital Life', which was one of several personality quizzes available on the site.
↘ Behavioural data from interactions on the app. If a user responds to video instead of text, for example, this information may be logged and used to inform future versions of the app.
In India: Prime Minister Narendra Modi's official campaign app, NaMo, launched in June 2015, promised to 'bring [users] the latest information' and important updates about Modi's government. In March 2018, it was discovered that the app on Android requested access to 22 different features of users' data, including access to their camera, microphone, contacts, photographs and location. (In contrast, Amazon's app in India requests access to 17 features.) NaMO also seems to have collected personal data from 1.3 million members of the National Cadet Corps, a branch of the Indian military, to facilitate personal interaction between the Prime Minister and cadets.
In the Dominican Republic: In 2012, Danilo Madina of the Dominican Liberation Party was narrowly elected President of the Dominican Republic. Four years later, after investing in both his country's and his campaign's technological base, Madina was reelected by a much wider margin. His app, Danilo 2016, was also built by uCampaign and had been downloaded nearly 14,000 times in the country of 10.65 million. About 65% of the app's users shared their address book contacts with the campaign, and nearly all agreed to receive push notifications. Through the app, users checked into events, shared content on social media, watched videos, posted selfies with President Madina, looked up GPS directions to polling stations, shared their votes on election day and invited friends to join. In total, uCampaign claims that voters completed over 360,000 actions in support of President Madina's reelection. uCampaign boasts over 600,000 app downloads for political candidates and causes in nine languages across 12 countries.
In France: In the lead-up to the 2017 French Presidential elections, Nicolas Sarkozy's campaign developed an app called Knockin, which mapped the campaign's database of contacts for door-to-door canvassers. The map marked each contact's address with a red dot, along with the resident's name. Canvassers approached the app's contacts at their homes and addressed them by name, leading to a public outcry over its invasiveness and an investigation by the French data protection authority, which ruled the app legal.
Knowing exactly what information you share with apps requires reading their privacy policies, which vary app-by-app and are cumbersome to read. An alternative is to identify all the explicitly political apps on your device and conduct a quick internet search on each of them.
Even official campaign apps are not guaranteed to be secure. Ted Cruz's mobile app, for example, leaked users' IMSI number, a unique number that identifies mobile phone users and can potentially be used to track or eavesdrop on users.
↘ Overall, campaigns claim that apps help them operate more efficiently.
↘ Apps can enhance the efficiency of canvassing efforts, which may increase participation in elections.
↘ Campaign apps collect a lot of data, much of it without the user's clear consent or awareness.
↘ Voters who may not want to receive messages from a certain campaign may have their names, email addresses, places of work, websites and other contact data shared without their permission or knowledge via an address book contact.
↘ By profiling voters' homes and tailoring canvassers' talking points, enhanced canvassing apps can make interactions more personalised, but they can also feel like an invasion of privacy.
↘ To the extent that canvassing can encourage voters to vote, selectively knocking on doors believed to support one candidate and skipping those thought to support the opposition is problematic. Companies that focus on knocking on the 'right doors' and making the 'right calls' implicitly attempt to avoid spending resources on those assumed to be politically misaligned or unengaged.
↘ While gamified campaign apps can help high scorers accrue social capital, they also risk publicly shaming or pressuring some voters to be more politically engaged as measured by the app's scoring system.
↘ Because apps attract people who think similarly and circumvent the ideological diversity of larger social networks, they risk creating filter bubbles and perpetuating confirmation biases.
↘ Because these digital products often survive in some form post-election, they set a precedent for a never-ending campaign season.
Author: Varoon Bashyakarla