Seeing government, being seen by government.

5 min readJun 14, 2020

I’m writing new sections of PhD stuff and in the spirit of peer review, I’m going to try and blog as much of it as I can and as much of it as is interesting to a general (but slightly knowledgeable) audience. This is a section I’ve written out differently several times. I am excluding complete and detailed references and some linking material to other sections. It is unfinished.

What, in relation to civic technology, is government?

As stated elsewhere, this project is not a history of government’s relationship with the web. Those stories are told elsewhere, and have a contested history around who did what, when, for whom and the erasure of work done by people who weren’t white and male.

For the purposes of this project we need to assess how participants in the field of civic technology.


Scott’s concept of legibility, the process in government of making its humans understandable is an incredibly powerful framework for understanding aspects of civic technology. The idea develops Foucault’s work on governmentality, stemming from his earlier iteration of the relationship of power-knowledge: the idea that power is consolidated by knowledge which is created by power. Within this, we can understand projects such as the 2005–10 UK Labour government’s attempt to institute ID cards as a project in legibility. In the same way as Facebook’s data model is an ontology (pace Henare et al 2007) of human life, so is an ID card. It attempts to make a messy individual into something machine readable. The comparison between ID cards and Facebook is partially also satisfying because of the evolution in the data model in the years between the comparison (assuming we take Facebook at the point where it hit the mainstream rather than being a timewaster for undergraduates). Facebook’s process of reduction of human characteristics into boxes (Zuboff 2019) that stratify political stance, religious belief, favourite music, purchase history and relationship status served a similar process to that of a government ID card scheme: legibility, although in their case, to advertisers. For government, the subject volunteers less data so government can only construct an image from the data to hand, which more or less forms a raft of tax, education, health and police records.

I raise the ID card issue as it often seems foundational to the creation of the UK’s civic technology scene in the early 2010s. The issue of ID cards had marked a shift it what was regarded as a consensus on the state’s right to surveil citizens (we’ll come back to this) and a coalition of lobby groups, unions and citizens were able to generate enough noise and objections that the plan was shelved.

It’s worth pointing out at this stage that the plan was not shelved. Progressive layers of increased legal need for proof of ID were introduced, to start a new job, to rent a room in a flat, but what was lacking was the infrastructure to do that simply. Government still wished to see. The “will to knowledge” as Foucault put it was still alive and well.

But what is most interesting about this iteration of civic technology in this perspective is that it is a rejoinder to that will, to the concept of governmentality. It tries to counter-surveil government through making government itself legible.

What is important about this is how many people involved were former civil servants. That is important because of the way that they constructed the government they were critiquing. It wasn’t a government of Kafka-esque bureaucracy or of Blairite soundbites, it was neither Yes Minister nor The Thick Of It. It was a government of fiefdoms, empires and power — something that was quite probably [quote needed] borne of experience. The prescription was open government, radical transparency. A space for armchair auditors to review the operation of government. Access to the outputs of government without a mediator, censor or

This is where pre-existing literature on civic tech exists: pitching it as part of a wider right-to-information movement. And that is great, as far as it goes, but in noting that many people involved in civic technology have come from government, we need to problematise the boundary between civic technology and “straight ahead” government (for want of a better term).

The tragedy of the armchair auditor

Implicit in the narratives of open government is a pivotal assumption: “if we make government easier to understand/the data easy to access/the code open to others can redeploy it, then things will be better”. This isn’t always wrong, but it does not have a perfect track record. It leaves two questions that this project will attempt to cover. The first is whether these approaches are successful by any standard. Open data hasn’t changed the world, open code and iterations leads to misunderstandings and press bollockings about things that don’t work first time, and concepts like open government became so linked to the Obama administration that they have struggled to survive after his term of office. In short, these movements were not the end of a teleological route, but a goal in a cyclical, attritional set of actions between civil society and government.

The other question is what civic technology users are asking for and whether it is about something other than oversight or more about the creation of a category of “expertise”.

The Institute for Government is able to tweet out each time a minister has to direct (an order-in-writing) the civil service to do something that the civil servant believes is unusual, unwise or likely to lead to consequences. This is because the directions are published as open data. It allows those who are able to consume APIs to add that feed of data to their pre-existing schemas and infrastructure for handling and understanding it. In short, it priveleges those already in the know and is presented with so little wrapping information that no one new is likely to be able to induct themselves.

So, in this instance then, the open is both globally possible (anyone can access), but also guarded by implicit gatekeeping mechanisms (it would help if you understood the policy area/knew the layout of the building/have a copy of Erskine May).

There is going to be more of this. I had been very anxious about talking about UK civic tech and GDS in the same project and I’m still hoping to avoid that, but the start of a discussion this week about the UK’s digital government model and its place in modern colonialism got me thinking a lot about how we construct the categories in civic tech and they are absolutely either borrowing from or in direct reaction to moves that the government makes. So, I’m trying to incorporate a bit more of that perspective and see them both as the same people who just happen to do different jobs in their lives. So many things that are implicit need to be made explicit.




Public sector specialist. Anthropologist on the internet.