Privacy has been at the heart of discussions when it comes to data ethics in politics, but even those using that framing realise the problems arising from the monitoring and analysing of personal data go much wider: issues of siloing, profiling and expensive costs will impact both how effective and ethical data practices are.
The debate can progress with only the slightest reframing: from privacy to private. Whereas privacy is a space free from interference, private data is our property; we own the personal information relating to our habits online, our financial activity, our demographic information and even our dna. Therefore, when the government, a private company or a third sector organisation takes this private data, it must be in exchange for something. When this exchange happens a social contract is formed, whether explicit or not. This social contract takes two dichotomous forms found at the core of any debate surrounding growing internet practices: either centralised or decentralised power.
This dichotomous framing can be found in political communications practices where personal information is gathered to create effective communications between the collector, usually an elite such as a political party or NGO, and the citizen.
Centralised experts in political communications value data for understanding how best to gain support for their own ideas from constituents. By profiling the constituency, trustees can develop personalised persuasive content. The profile can represent an individual or groups demographics, interests, sentiment and behaviours.
They can use data, for example the reach of Facebook posts or click throughs to the website from emails to encourage active support and engagement. They can use the data to test which is most effective rather than relying on gut instincts or pundits. For example, Kreiss (2016) documents how the Obama campaign used testing and resultant data to understand what colour, size and order of information had the most impact on individuals to remain on the website and ultimately to donate. Other than to encourage donation, they can also influence people to mobilise and take action, either by voting, signing a petition or taking to the streets, in support of the expertly chosen policy goals.
The decentralised framing instead focuses on professional facilitators who can use data to listen and empower the citizens to make decisions. For example, Karpf (2017) documents how Avaaz and MoveOn, technology driven civil society organisations, ask their members to respond to surveys and petitions to decide the outcomes for the next year or to prioritise current campaigns. The facilitators can support the leadership of constituents who initiate actions to encourage change. Where data is used to create profiles of constituencies this can be used to record preferred activities of individuals, track their development and help create personalised opportunities for individuals to be directly involved in implementing change. The success of participation can be captured and visualised so citizens can hold the facilitators accountable.
Agents of Data Technologies
While this framing opens up the discussion to broader understandings of what effective practice looks like between citizen and data collectors it also reveals an important problem: this framing doesn’t account for a relationship technology. For example, Lina Dencik talked about the problem of the data double, where decisions are made on a representation of a person in data rather than by the person themselves. The data double and other technology agents, such as programmers, data scientists and algorithms themselves draw power from centralised experts, professional facilitators and citizens.
To advance the discussion of data ethics it is important to consider the allocation of power between elites and citizens, but more importantly to adapt political theory to allow for the role of technology.