Civic Tech at mySociety: How the Imagined Affordances of Data Shape Data Activism


The progressive datafication of social life is often perceived as a threat to democratic publics. Critics warn that filter bubbles would undermine dialogue and consensus, that social discrimination will be reinforced (Barocas and Selbst 2016) or that ‘surveillance capitalism’ is fundamentally anti-democratic (Zuboff 2015). Yet as “democratic power is calculated power” (Rose 1999, 200), datafication is also driven by democratic visions and closely linked to notions of political accountability, fairness, and citizen empowerment. Forms of datafication driven by commercial players or governments have always been accompanied by civil society actors or journalists who have utilized data and related forms of quantification to advance their goals. However, given the new structures of (data) power that shape the workings of governments and businesses today, Milan and Van der Velden (2016, 6) suggest that it is increasingly important to investigate “how activism evolves in relation to big data”. They argue that new forms of ‘data activism’ are “enabled and constrained by data […] and this special relation shapes tactics, identities, and modes of organizing” (Milan and Van der Velden 2016, 3).

In this article, I explore ways in which activism is enabled by datafication by looking at a group of actors who not only reacted to processes of datafication, but pro-actively embraced them: civic technologists. Civic tech is an umbrella term for diverse projects that attempt to make engagement easier for citizens, improve communication and feedback between governments and citizens, and strengthen political accountability. Among other things, civic technologists develop parliamentary monitoring websites, tools to help citizens report local infrastructure problems to local government, or freedom of information (FOI) websites that help users to submit freedom of information requests to public institutions. In its modern incarnation, civic tech is the result of a convergence between “communities of technological and political openness” (Yu and Robinson 2012, 195). Early examples include the British FaxYourMP (2003), which helped citizens to find and contact their representatives in UK parliaments (Townend 2008), or the monitoring website (2004), which made information provided by the US Congress more accessible (Yu and Robinson 2012). From these early volunteer experiments, civic tech has grown substantially in recent years as it has been embraced by governments, corporations and foundations (Baraniuk 2013).

In addition to their success and growth, civic technologists are relevant because they act as pioneers for the use of data to facilitate civic engagement. While there have been predecessors of the tools they developed, the way they utilized data to make them accessible and offer additional services was novel. They took information that was available elsewhere and made it machine-readable, shared it openly, and built services on top of it, e.g. the ability to type in one’s post-code to find one’s representative in the British parliament on FaxYourMP. Through such practices, they combined concepts of legal and technical openness in new ways (cf. Yu and Robinson 2012). Being pioneers of this type of work is also part of the self-conception of at least some civic tech organizations. They possess a sense of mission and make teaching their (data) skills and experiences to other civil society or media organizations an important part of their work.[1] Finally, non-profit civic tech organizations such as the Sunlight Foundation in the US or mySociety in the UK were also among the first to advocate for open data policies, and supported related freedom of information policies (Schrock 2016).

However, despite civic tech’s success and potential influence, it has received little attention in media and communication studies to date. Most of the research that does exist is interested in how the phenomenon might reinforce existing power structures (cf. Gregg 2015), or who is using civic tech applications and in what way (cf. Cantijoch, Galandini, and Gibson 2016). What is missing is a nuanced understanding of the practices, ideas and motivations that guide civic technologists themselves and how those practices and ideas provide orientation for others. Critically examining civic technologists themselves is relevant because the broader impact of civic tech goes beyond the impact of individual civic tech applications. As Hepp (2016, 919) describes, pioneer communities such as civic technologists are influential in the sense that they develop “a horizon of possibility to which the everyday media appropriation of others orients itself, or at least can do so” (Hepp 2016, 919). Accordingly, we also need to be sensitive to how the practices and imaginaries civic technologists develop provide orientation for others, i.e. how other actors adapt and modify them.

This paper therefore critically examines how civic technologists understand and use data to “meet their social ends” (Couldry and Powell 2014, 2). What are the key practices of civic technologists in relation to data and how do they themselves understand what they are doing? Addressing these questions will contribute to our understanding of how activism is enabled or constrained by datafication, as it provides a basis for subsequent studies to examine if and how these practices and imaginaries can be found elsewhere, and how they might have been modified.

I present findings from a case study about the British non-profit organization mySociety. Founded in 2003, mySociety is one of the oldest and most influential civic tech organizations and arguably represents ‘best practice’ in the extremely diverse civic tech sector. Its UK websites have millions of users (mySociety 2015) and the organization had a direct influence on British policy-making (cf. 2014). Some of its more popular projects include FixMyStreet, which lets citizens report local problems like broken streetlights or potholes to local government; its right-to-know website WhatDoTheyKnow which helps users to submit FOI requests to public institutions; or its parliamentary monitoring website TheyWorkForYou which gives detailed information about voting records and makes parliamentary speeches more accessible. mySociety’s projects are also prominent internationally and customized versions of its tools are used in 44 different countries (mySociety 2015), which let mySociety to transition from a UK-centric to an international organization. While it is not representative of the phenomenon as a whole, mySociety’s success and international influence provide a good starting point for studying the values and practices which shape civic technologists’ use of data.

In the following, I shall first discuss my methodological approach. I follow a practice theory approach and use the concept of ‘imagined affordances’ (Nagy and Neff 2015) as a lens which helps amplify how members of mySociety themselves understand data and how they imagine it to advance their agenda. The remainder of the article describes how the data practices of mySociety relate to their broader imaginaries. In the conclusion, I will reflect on the implications for studying civic tech and data activism in general.

Researching the Imagined Affordances of Data

To examine how members of mySociety understand and use data to meet their own ends, this paper relies on the concept of ‘imagined affordances’ (Nagy and Neff 2015). According to Nagy and Neff (2015), whatever actions a particular technology enables or constrains does not solely depend on its features or its material properties, but also on the perception of users and designers. Both may have “expectations about their communication technologies, data, and media that, in effect and practice, shape how they approach them and what actions they think are suggested” (Nagy and Neff 2015, 5). Applied to the subject of this paper, what data enables civic technologists to do does not solely depend on the properties of the data they collect or re-use, or on the applications they are able to develop with it; it also depends on how civic technologists themselves understand and perceive how data can serve their agenda. These perceptions and understandings are the basis for the “horizon of possibility” (Hepp 2016, 919) they develop as a pioneer community, which affects the perceptions of other actors concerning how data can be used to facilitate forms of civic engagement and activism.

To study the imagined affordances of data for members of mySociety, this paper relied on a methodological approach inspired by practice theory. First, a focus on practices aligns well with the concept of imagined affordances because affordances enable or constrain certain actions, and people make sense of affordances “in and through practices” (McVeigh-Schultz and Baym 2015, 2). Second, practices are useful for examining the role of pioneer communities because they act as exemplars. This means that they not only communicate ideas and visions about how a technology can be used, but they also become influential because they demonstrate their own visions and thereby affect the perceptions of others. The practices they develop are expressions of their broader visions and we have to consider them inseparable if we want to understand their influence.

Accordingly, I employ methods that helped exploring what members of mySociety are doing and how they themselves understand and categorize what they are doing in relation to data (cf. Couldry 2004, 2012). I followed a constructivist grounded theory approach (Charmaz 2006). Grounded theory was chosen because of its core principle of theoretical sampling, i.e. an initial data sample is continuously expanded with new data to systematically elaborate and refine the theory. This approach was useful for exploring the open-ended range of practices (Couldry 2012) and to more fully reconstruct the perspectives of the research subjects without applying pre-conceived concepts.

Spread over three rounds of data collection, I conducted five semi-structured interviews with members of mySociety, including its founder and former CEO, two senior developers, a member of the international team, and a member of mySociety’s research team. These interviews had an average length of one-and-a-half hours. A large part of each interview consisted in the reconstruction of a particular project. Each interviewee was asked to pick a project that illustrates her or his work and then walk me through the development process: What was the initial idea behind the project, what were the different steps and phases for implementing the project, what happened after the initial release? Reich (2013, 422) calls this ‘reconstruction interviews’ because it reconstructs “technological ‘biographies’”. Exploring the development of a project in-depth vividly illustrated the practices and routines of my interviewees. If not mentioned by the interviewees, specific questions about the role of data were asked. After one project was explored, I asked whether this was a ‘typical’ project and if there are very different examples. If there were, I explored those as well. Other questions addressed self-understandings (preferred job title, understanding of civic tech) and personal or organizational ambitions and values.

These interviews were complemented by 17 documents found online: mySociety’s homepage (including the use of the Internet Wayback Machine to retrieve older versions), project specific websites, blog posts and forum discussions from Tony Bowden (who has been working for mySociety since 2009), other interviews given by different mySociety members to newspapers or bloggers (e.g. Townend 2008), as well as presentations given by members like Tom Steinberg or Dave Whiteland available online (UsNowFilm 2008; Arcopix 2014; IndigoTrust 2011; mySociety 2014). Some of these documents were included in the initial data sample, others were added later following the theoretical sampling. Moreover, I conducted ethnographic research on two separate conferences which were visited by several mySociety members: The Open Knowledge Festival 2014 in Berlin and the Mozilla Festival 2015 in London. These conferences helped to get an impression of the larger community mySociety is involved with and provided a helpful guidance for both the interviews and the analysis.

How mySociety Members Imagine the Affordances of Structured Data

In the following, I describe the imagined affordances members of mySociety hold around structured data. First, I will explore mySociety’s mission and self-understanding in more detail to give a dense description of the broader ambitions and imaginaries that drive this organization. Then I show how data is used to facilitate this mission by describing four imagined affordances: deep linking into documents to engage citizens with the processes of governments; making the performance of governments legible to affect how they implement laws and public services; affecting its users’ perceptions by demonstrating their impact to them; and scaling technological solutions to support a distributed form of agency.

mySociety’s Mission: Facilitating Engagement

mySociety’s self-proclaimed mission is to “help citizens demand better […] our web tools and apps are breaking down the barriers around governments” (mySociety 2016b). Its tools are supposed to give “greater access for citizens to the work of government and the democratic process”, which essentially means improving how publics can monitor and provide feedback to governments: “We believe that governments tend only to get better at serving the needs of citizens when citizens are capable of demanding better, creating a virtuous circle that leads steadily to better government” (Cridge 2015). While this statement might imply advocacy for specific policy changes, mySociety understands its role as a ‘mere’ facilitator of civic engagement, not as an advocacy organization.

Members generally reject the idea of being a gatekeeper that stands between the citizens and their governments. Instead, they suggest to provide the means by which others are able to take actions more effectively: “What we do is present the facts: This is how your MP [Member of Parliament] voted, this is where the money went, this is what was said. It’s then up to other people to do with that what they will, which might well be using it to promote a cause” (Interview: Research Team). The idealized and simplified scenario that members sketch out rhetorically to describe their role can be outlined as follows: Before mySociety enters the arena, citizens are apathetic and disengaged because engagement is too difficult and time-consuming due to high barriers raised by governments (in most cases unintentionally). mySociety identifies these barriers and then ‘drops’ its tools into the public arena to make engagement easier for citizens, which subsequently facilitates engagement between citizens and their governments and leads to better governance.

Underlying this approach is the assumption that more means for citizens to provide feedback to elected representatives leads to ‘better’ outcomes, i.e. more representative and therefore more democratic outcomes. Given its self-understanding as a ‘facilitator’, mySociety is not advocating for specific outcomes, but is concerned with the processes by which outcomes are generated: “We’re in favour of a vibrant, healthy, lively democracy. That means a rude and obnoxious place. Although we don’t want to do that ourselves, it’s entirely appropriate that we should facilitate other people to” (Steinberg quoted in Krotoski 2010). This approach builds on the principles of open source culture (Kelty 2008; Lewis 2012). “The essence of open source,” Weber (2004, 56) describes, “is not the software. It is the process by which software is created”. As the ideas and practices of the ‘open source process’ were increasingly applied outside of software development (most prominently with Wikipedia) they formed the basis for a larger technological and cultural phenomenon that Jenkins (2006) has described as participatory culture, a culture “which posits that knowledge is richest and most accurate when it reflects the pooled inputs of a distributed population, as opposed to the expertise of a single agent” (Lewis 2012, 847).

MySociety builds on previous forms of participatory culture and has particularly strong connections to technology-driven open data initiatives and rights-based open government or freedom of information initiatives, both of which are interested in applying the ‘logic of open participation’ (Lewis 2012) to institutionalized politics (Schrock 2016; Janssen 2012; Yu and Robinson 2012). mySociety was an early supporter of open data in the UK and its founder was part of a group that articulated the ‘8 principles of open government data’ ( 2007). It also promoted freedom of information laws through its website WhatDoTheyKnow, which helps users to submit FOI requests to public institutions, and advocated for strong FOI legislations. Yet despite this strong connection, advocating for open data and FOI is not mySociety’s main purpose, they are rather perceived as “resources that mySociety needed to function” (Interview: Former CEO). This is because mySociety does not just build on participatory culture, it also extends it in important ways. Participatory culture relied on the connectivity of internet technologies to establish new forms of governance based on collaboration and sharing. As Lewis (2012, 848) describes, participatory culture is based on a forging of technology and culture, in that digitalization “enables greater user participation on a seemingly infinite order, and the socio-cultural context of this technology has encouraged greater participation to achieve normative aims of collective wisdom and well-being”.

MySociety similarly wants to create a more collaborative and participatory process for achieving better outcomes, but it does not solely rely on the connectivity enabled by internet technologies and ways of ‘coordinating collaborations’ (Kelty 2008). Its civic tech applications are not primarily about connecting people, but about facilitating them, i.e. enabling them to engage with governments in ways that go beyond ‘mere’ connectivity or access to information. In other words, mySociety is extending participatory culture by drawing attention to the conditions that would allow and encourage people to participate. Civic tech at mySociety is essentially about feasibility, in that it aims at making engagement more feasible for citizens by removing frictions such as needing to find out who represents them in parliament and how to contact them.

MySociety’s imagined affordances of data are closely tied to this broader mission of creating a participatory culture. To illustrate the fundamental importance of structured data for mySociety, a member uses the analogy of cooking ingredients:

If the useful thing is a cake that people want to eat, you’re interested in the ingredients…But you don’t want raw ingredients like wheat, you need the flour. Some processing has to be done to the ingredients before they are ingredients that you can sensibly make a cake with…Without the structured data, you wouldn’t be able to offer that service and until you can offer that service you couldn’t really prove that the demand for it would be so great. (Interview: International Team)

This suggests that structured data, if ‘served’ correctly, would increase engagement and subsequently alter the relationship between citizens and their governments because it shows that the ‘demand is great’. In the following, I unpack this statement and describe the imagined affordances of data underlying it.

Improving Engagement with Governments: Deep Linking

One of the most popular services on mySociety’s parliamentary monitoring website TheyWorkForYou are email alerts. For example, users can search for keywords in parliamentary discussions and then sign up to regularly receive emails informing them when their keyword comes up in future discussions. This service was significant because British parliaments used to publish transcripts of speeches as PDF files. To monitor keywords or what individual MPs are saying in parliamentary discussions, one needed to download these PDFs regularly and search through them individually. The email alerts at TheyWorkForYou turned this monitoring process into something people are able to do along the way, without investing considerable time and effort.

The key practice behind this service is the idea of ‘deep linking’: “The idea that there is councilor Jones who said ‘This is what we have to do in my home town!’ and you can cite it directly” (Interview: International Team) – similar to Twitter, where every individual Tweet has its own URL which can be shared or embedded on other websites. For deep linking, mySociety (2016c) considers transcripts “made of nicely structured data […] hard to beat”. If documents are in a format that does not allow deep linking, “you can’t cite, you can’t share, you can’t show specific utterances […] that pretty much stops details in documents being called out in debates” (Whiteland in mySociety 2016a). Today, it uses a data standard for modeling parliamentary speeches called Akoma Ntoso. It allows granular filtering (everything this person has said), an analysis of the speakers’ behavior (what was said, how and when?), and the ability to link speakers and what they say to other statements and events: “We would like to see people…build sites like ‘all public statements by the Prime Minister’” (mySociety 2016c). This creates a level of “semantic understanding” with an “awareness of speakers” (Interview: International Team) that is necessary for the services mySociety develops.

The ability to link deeply into a document using structured data is considered “really important in public discourses about documents” (Interview: International Team). More fundamentally, documents are considered important because mySociety is “concerned with the process of government and most government […] is actually the business of making laws, and laws traditionally have been written down” (Interview: International Team).

From the perspective of mySociety members, enabling deep linking into documents and providing services around it improves the public’s awareness of, and engagement with, the businesses of governments (as those businesses are captured in documents). As mySociety explains:

Transcripts are a kind of oil that greases the wheels of well-functioning societies. They let people discover when powerful people have made pronouncements that affect less powerful people. We believe that by making transcripts function better, more people will end up learning about decisions and opinions that affect their lives. (mySociety 2016c)

In the interviews, members also frequently referred to a claim by mySociety’s founder: “everything you can and cannot do in your life has been decided by more powerful people in a meeting” (cf. Whiteland in mySociety 2016a). Deep linking is about improving access to such meetings to help figure out “who was responsible for things” (Interview: Former CEO) by improving the accessibility of those documents which capture what was said by whom. Put simply, deep linking should help to keep track of where and when powerful people make decisions.

Taken together, mySociety is imagining the affordances of structured data in this case as a basis for a document-driven monitoring tool to help engage the public with the businesses and decision-making processes of governments. Through deep linking, structured data is imagined to “even [affect] an apathetic population, it affects the way that they behave and what they know about what’s going on in their own society” (Whiteland in Arcopix 2014). This makes turning documents into structured information a central part of mySociety’s mission of creating a more participatory culture. Importantly, deep linking is more likely to facilitate individuals who are already highly engaged and interested in the processes of governments to begin with, either privately or professionally (e.g. journalists or activist groups). As I will discuss below, mySociety also imagines data to help engaging citizens who are not necessarily interested in politics or feel disengaged and powerless.

Mediating between Governments and Their Publics

On FixMyStreet and WhatDoTheyKnow, mySociety collects all the problem reports or freedom of information requests by its users, tracks the responses by public institutions, and makes both publicly accessible. By making the resulting databases public, mySociety creates new forms of legibility and assessability. The database on FixMyStreet enables the analysis of a city’s infrastructure problems by the public and makes the performance of governments, e.g. how fast they fix problems in specific regions, legible. WhatDoTheyKnow similarly allows an assessment of how different public institutions respond to FOI requests and subsequently how the law is being practiced.

One reason for mySociety to ‘aggressively’ impose transparency on governments is its reliance on their cooperation. Its applications are built on top of services provided by governments and cannot exist independently without them. If public institutions refuse to cooperate, for example by ignoring reports sent via FixMyStreet, their tools would be of little or no use. By imposing transparency and allowing the public to assess their performance, mySociety makes it more difficult for authorities to ignore them. It is a way of pushing institutions to cooperate and to adopt mySociety’s emphasis on user-friendliness and accessibility (see below). This is described as “the one bit of activism that we occasionally engage in” (Interview: Research Team).

A less obvious aspect of this ‘one bit of activism’ is how mySociety’s use of data enables it to mediate between the bureaucratic and legal processes of governments and the users of its applications. The way mySociety’s tools work is usually not a direct reflection of how public institutions or legislations are working:

We think about the aim of the software as not being necessarily to model exactly the processes as they already exist in the world, but to make the software embody a slightly better way of doing things. (Interview: Senior Developer 2)

In this sense, mySociety’s tools are “intended to provoke some friction” (Interview: Senior Developer 2). In the cases of FixMyStreet and WhatDoTheyKnow, the laws and regulations did not prohibit the publication of problem reports or FOI requests, but it also was not something that was suggested. In addition, mySociety reinterprets how services and laws should be implemented in smaller ways, for example by forwarding reports to councils via email, while councils prefer people to use a form on their website; or by allowing users to send FOI requests to institutions that were not subject to the FOI law on WhatDoTheyKnow (mySociety 2011).

MySociety developer Tony Bowden (2010) described this principle as ‘dreaming out loud’: “simply act the way you want the world to be, then wait for reality to catch up” – by finding ways to make institutions cooperate and push ‘reality’ in that direction. This ‘dreaming out loud’ principle fundamentally depends on mySociety’s use of data to make the performance of governments legible and assessable for the public.

However, mySociety cannot simply invent new ways of how governments should work and then impose this vision on institutions. While they are intended to create frictions, the software tools and data structures mySociety develops need to reflect the real-world processes they are supposed to represent to a large enough degree “to actually work” (Interview: Senior Developer 2). Instead of ‘reinventing’ government, mySociety’s tools are compatible with existing services and legislations, but simultaneously utilize data to asses and affect how they are being implemented: “that gap between the way the law works, the way institutions implement the law and the way perhaps it might be ideal from a citizen’s point of view is an important one and one where I think it’s significant to build an artifact that demonstrates that” (Interview: Senior Developer 2). Using data to demonstrate and potentially close this gap is mySociety’s definition of ‘empowerment’: people are empowered by giving them tools which enable them to “see and be able to do what they are legally entitled to as easily as possible” (Interview: Research Team).

This understanding of empowerment encapsulates how mySociety imagines to affect the relationship between citizens and their governments through data. It uses data to facilitate the use of preexisting rights and services, not to fundamentally change them. It envisions its tools to act like a ‘layer’ that translates the bureaucratic and legal procedures followed by public institutions into user-friendly interfaces with accessible language for citizens. They are supposed to both reflect existing processes and visions of how these processes should ‘ideally’ work. The legibility and assessability created by data is imagined by mySociety to affect the implementation of rights and public services in ways intended to make engagement easier for citizens and thus ‘push reality’ closer to its vision of a more participatory culture.

Changing Perceptions: Providing a Sense of Agency

While facilitating already interested users via deep linking is important to mySociety (see above), it ultimately aims at engaging people who are usually not engaged. mySociety aims at a more long-term change in perception about “what is normal rather than what is exceptional” (Interview: Former CEO). Using TripAdvisor as an example, the former CEO explains that it “has caused a massive power shift in the hotel industry from the people who run the hotels towards people who stay in hotels”, even though it is not a “campaign for better hotels”. Services such as TripAdvisor would change people’s expectations about things like going to hotels if they are popular enough, i.e. widely used (cf. Steinberg in IndigoTrust 2011). In a similar vein, mySociety wants its tools to be “popular rather than idealistic” (Interview: Former CEO) and has a strong emphasis on usability, i.e. on making sure its tools are easy to use and provide “the same level of service that the best [commercial] websites we use nowadays would have” (Interview: Senior Developer 1).

In order to simultaneously achieve popularity and promote a more participatory culture, mySociety tries to ‘harness self-interest’ (Bowden 2014b) by generating “public good from private desire” (Steinberg in Nestoria 2008). It aims at developing services that address individual end-users, but use the data gathered from the engagement of these users to create “public value” (Steinberg in UsNowFilm 2008) on top of it. Crowdsourced databases such as those created on FixMyStreet or WhatDoTheyKnow are key to this. FixMyStreet helps individuals to fix their specific problem, but it also collects all the reported problems to create a public database about local problems that can be useful for others, for example local journalists who can sign up to get email alerts for problems reported in a specific region.

By bridging individuals with collectives in this way and by emphasizing ease-of-use, mySociety ultimately wants to affect the perception of users about their own agency. For example, when a problem reported via FixMyStreet has been fixed, it will be removed from the public city map and the user who reported it will receive an email encouraging her to report more, or to try out mySociety’s other projects for contacting her representatives. This way, mySociety is not only solving an individual’s problem, but is also trying to demonstrate the public value created by its action. This is intended to give users a “sense of agency…some ability to change their environment” (Interview: Senior Developer 2) by demonstrating that their actions do have an impact:

[W]hat we often see as apathy is really just learned helplessness. People feel powerless, because they don’t believe they can make a difference. And the best way to change that is not to argue with them […] It’s to simply show them that they do actually have power. That what they do can have an effect, not only for themselves but for people around them. (Bowden 2014b)

Here, the legibility and assessability of crowdsourced data is imagined to create a more abstract, cultural and psychological change in perception.

Taken together, the crowdsourced databases mySociety creates are imagined to affect both governments and the public. On the one hand, they enable mySociety to push governments and influence how laws and public services are being implemented, as they make the performance of governments legible in new ways (see above). On the other hand, it uses the data generated by its users in ways intended to change how they feel about engagement by providing easy-to-use tools and by demonstrating that their actions had an impact. Databases are imagined to affect emotions and perceptions in order to advance a vision of a more participatory culture, which illustrates the importance of emotions and perceptions around data (cf. Kennedy and Hill 2017). As I show in the next section, data is also key to mySociety’s ambition to support the development of similar applications elsewhere.

Scaling Civic Tech: Supporting a Distributed Form of Agency

The bulk of mySociety’s work today is based on collaborations with groups in other countries to create national versions of projects originally created in the UK, especially Alaveteli (an international version of its right-to-know site WhatDoTheyKnow) and tools for parliamentary monitoring. In part, this international orientation is driven by funding opportunities, but also because numerous groups in other countries expressed a desire to have customized versions of mySociety’s UK tools. While partly self-interested, members share the desire to support groups that similarly want to promote a form of participatory culture elsewhere. They think of themselves not merely as ‘tool suppliers’, but as part of an international civic tech community.

MySociety essentially wants to promote its values and practices in many different contexts by supporting local groups with developing customized versions of its tools. To achieve this, it has to accommodate the fact that “people in different places care about different aspects of politics. In some countries what really counts is how politicians vote, in others the crux is campaign finance contributions” (Steinberg 2012). Data is seen as both a key problem and a solution to this approach. One of the main obstacles for transferring an existing application to a new country or for building new civic tech applications from scratch is the design of consistent data models that adequately reflect the structures and legislations of the respective national government. A data model designed to capture the British parliamentary system cannot simply be transferred to another country. If they are not already available (as open data), developing such data models is complex.

Therefore, mySociety is one of the founding members of Poplus, a ‘global civic tech federation’ of organizations similar to mySociety, like Code for America in the US or g0v in Taiwan. The basic idea that drives Poplus is that “pretty much every tool in the civic and democratic space can be broken down into some parts that are universal, with usually only a little bit of local glue holding them together”:

Although pretty much every Parliament has different processes, they’re still dealing with the same raw ingredients – people, parties, speeches, bills, votes, etc. – and if we could create standardised ways of modelling each of those things, it would be a lot less work for people to combine these in the way that makes most sense for their own situation. (Bowden 2014a, emphasize added)

In connection to Poplus, mySociety supports data standards for modelling government structures and develops EveryPolitician, a project that literally collects data about every politician in the world. In this project, mySociety makes “data editorial decision[s]” (Interview: International Team) about how the ‘basic elements’ that can be found in every government are represented in a consistent manner across countries to ensure tools can be easily deployed internationally.

On a technical level, mySociety is trying to reduce problems of scale through data structures and tools that standardize how common ‘ingredients’ of government systems are formalized. However, members perceive this not merely as a technological detail, but as a way to facilitate mySociety’s vision of a more participatory culture elsewhere. By enabling other groups with similar values and social imaginaries to create their own, local versions of civic tech applications that serve their particular needs, mySociety is essentially trying to support a distributed form of agency. By promoting data standards and reusable tools, it tries to create the conditions necessary for supporting its values and approaches in as many different contexts as possible.

Conclusion: The Cultural and Historical Situatedness of Affordances

This article examined the imagined affordances members of mySociety hold around data to gain a deeper understanding of how civic technologists rely on data to meet their own ends. mySociety is trying to facilitate civic engagement and, by extension, create a more participatory culture. Taken together, it is imagining the affordances of structured data in ways that are supposed to expand the agency of publics towards governments: ways intended to enable citizens to better influence and interact with governments or other powerful institutions. Four imagined affordances have been identified.

First, members of mySociety use structured data to make ‘deep linking’ into documents possible, thereby aiming to increase the publics’ awareness of, and engagement with, the processes of governments. Second, crowdsourced databases enable mySociety to mediate between governments and their publics. By collecting data to monitor the performance of governments, and by making this data publicly available, mySociety is trying to push governments to cooperate with its services and thereby affect how existing laws and public services are being implemented in supposedly more user-friendly, i.e. ‘citizen oriented’ ways. Third, members of mySociety are trying to use data to affect the perceptions and self-understandings of their users. Crowdsourced databases are supposed to help connect the individual with the collective by demonstrating to individual users that their actions have an ‘impact’ and create a kind of ‘public value’. Fourth, members of mySociety are scaling technological solutions to support a distributed form of agency that should enable groups in other countries, which similarly want to promote a more participatory culture, to develop customized versions of mySociety tools that serve their particular needs.

The way mySociety is using data to apply notions of participatory culture to politics suggests that we should understand the role of civic technologists in the public arena as facilitators of engagement. Facilitating means that data is used in ways that are supposed to enable others to take actions themselves. mySociety does not advocate specific outcomes (such as particular policy changes), but is concerned with how the processes by which outcomes are generated are designed. It aims to enable users to easily engage with authorities in order to ensure that decision-making “reflects the pooled inputs of a distributed population” (Lewis 2012, 847). In other words, mySociety wants to influence the conditions by which others participate in the public arena without directing the public discourse or influencing policy making. At the same time, it sees its role as complementing actors who do want specific outcomes, like advocacy groups within civil society, or professional journalists who emphasize gatekeeping.

By showing how members of mySociety understand and use data to facilitate civic engagement and realize a more participatory culture, this article draws attention to the fact that “the political and democratic possibilities of data” (Milan and Van der Velden 2016, 8) cannot be determined in an abstract way. What data affords to whom does not only depend on the technological properties of data, but is fundamentally social, and both culturally and historically situated. This has implications for studying data activism in general and civic tech in particular. Related to data activism, the findings presented here invite us to extend Milan and Van der Velden’s (2016) conceptualization of data activism as creating novel epistemic cultures around datafication within civil society. The epistemic cultures and related social imaginaries promoted by mySociety are not novel: the practices and imaginaries described here build on participatory culture, which itself has roots predating computer culture and notions of open source software (Tkacz 2012). mySociety’s practices and epistemologies did not appear out of nowhere and the epistemic cultures they create are not necessarily novel, but they develop imagined affordances around data to implement these ideas in new ways. To understand imagined affordances, historical trajectories are just as important as the new elements added by data activists.

Studying imagined affordances does have a lot of potential for further developing and refining notions of data activism. First, imagined affordances provide a useful angle for moving beyond broad dichotomies of pro-active and re-active forms of data activism (Milan and Van der Velden 2016) and for studying how the epistemic cultures developed by data activists are connected to data in nuanced ways. Especially when combined with a focus on practices, it is a powerful tool for illustrating the “distributions of agency and organising forces” (Tkacz 2012, 404) activists set in motion. Second, the article shows how imagined affordances can be used as an integrative framework for studying how data activists affect the distribution of knowledge and power. I described how mySociety is trying to utilize the legibility and assessability created by data to change the perceptions of users about their own agency. Subsequent studies could examine how the practices and social imaginaries of data activists interrelate or clash with the self-understanding and perception of different groups in the public arena, and how different imagined affordances by various types of users emerge around civic tech applications.

Applied to civic tech, the research approach developed here calls for a more differentiated picture of the civic tech sector. Civic tech organizations and their funders are very much interested in a “coherent and clearly articulated vision and sense of shared identity for civic tech” (Donohue 2016). Yet while the diverse actors in the field of civic tech might align inasmuch as they are all interested in open data, reusability and a vague sense of improving ‘civic life’, their interests might eventually clash. For example, European non-profits like mySociety or the Open Knowledge Foundation Germany advocate for governments to copy civic tech applications. At mySociety, this even goes as far as stating that most mySociety projects “shouldn’t need to exist at all” (Interview: Senior Developer 1). Elsewhere, civic tech is closely aligned with start-up culture and attempts to create new industries. Wanting public institutions to copy civic tech applications in order to change the relationship between citizens and these institutions, or favouring the ‘government as a platform’ paradigm (O’Reilly 2010) and wanting public institutions to ‘step back’ and foster ecosystems of for-profit services, has very different implications for civic life (Baack in 2016).

Pointing out such differences is also important for developing a more differentiated critique of civic tech. Civic tech and open data initiatives have been said to uncritically support a neoliberal agenda driven by commercial interests rather than government accountability or citizen empowerment (cf. Bates 2012; Slee 2012; Gregg 2015). Moreover, their focus on technological solutions would merely reinforce existing power relationships by ‘empowering the empowered’ (Gurstein 2011). While this critique is important, it does not equally apply to every actor in the civic tech sector. We should not discard the agency and intentions of non-profit organizations such as mySociety, who are aware of these discussions and conduct research to better understand the impact of their applications. Members of mySociety constantly experiment and explore ways to support and encourage civic engagement. In this sense, they are “potential ‘laboratories’” (Hepp 2016, 929) that we should not ignore if we want to formulate normative principles for making processes of datafication more democratic and for creating more self-aware and agentic publics (Kennedy and Moss 2015).

The imagined affordances identified and described here can help to map differences among actors in the civic tech sector: Are other civic tech organizations relying on data in the same way, and do they promote similar ends with them? Are actors outside the field of civic tech adopting them, and in what way? Which actor in the civic tech sector influences what field? Given civic tech’s status as a ‘pioneer community’ (Hepp 2016), tackling these questions is one way to conduct empirical research to trace and understand how forms of civic engagement and activism are changing due to their growing reliance on data.