The Center for Data Innovation spoke with Rebecca Rumbul, head of research at mySociety, a not-for-profit social enterprise based in the UK which builds technologies to engage communities online. Dr. Rumbul discussed how, by connecting networks of people, civic technologies can empower individuals and improve their experience as citizens.
This interview has been edited.
Eline Chivot: What is the “civic tech” movement, and what problems does it address?
Rebecca Rumbul: The civic tech movement is a loosely connected sphere of digital activity, concentrating on developing digital tools and capacity to empower citizens in some way. Civic tech is normally borne out of simple frustration with the system, and tends to focus on issues of democracy, open government, international development, information rights, and public administration. At its core it is designed to reduce the distance between individuals and their public institutions, enabling citizens to exert their rights and assisting institutions to administer those obligations efficiently.
Civic technologies can address a whole spectrum of issues where the political will and digital capacity exists to do so, and where it doesn’t, civic tech has the potential to circumvent those barriers. Civic tech covers everything from small scale local issue reporting such as FixMyStreet, to parliamentary monitoring such as TheyWorkForYou, to information rights services such as AskTheEU.
Chivot: “TheyWorkForYou” is an application that uses open government data to increase democratic accountability and transparency. How does the system work, and how have people used it?
Rumbul: TheyWorkForYou (TWFY) is run on a system we designed called Pombola. The system is based on two databases. The first one is about “people and places.” At the core of many parliamentary monitoring sites is a structured database of politicians that links elected representatives to the places they represent, the parties of which they are a member, and the roles they perform. We helped these sites display a page for each politician. The data could additionally be cut in various ways to populate other areas of each site. The debates they publish from parliament, for example, have the politicians’ names and photographs beside their contributions. Users can use filters to display all of the members of a particular house of parliament, all of the elected representatives of particular areas, members of a specific party or even all of the politicians who attended a specific college.
The second database is about “the words they speak.” It stores parliamentary transcripts, with speeches ascribed to the individual speakers. This is, of course, only possible in countries where the parliaments provide debate text, ideally as open, regularly released structured data, although in some cases we have parsed PDFs. Once the content is on the parliamentary monitoring sites, users can search the entire database for a specific keyword or phrase, or browse to read recent debates or ones going back into the past. This database also allows the use of code that does nifty things like display on a politician’s page how often they have spoken in parliament and what they have said.
Boundary information allows users to input their location, and discover the elected politicians who represent them. This feature is a very powerful way to ensure that content is relevant, even for people who do not know much about Parliament or its workings. Boundary information is stored on our mapping product MapIt, which makes it easy for platforms to link geographic location data with political or boundary data.
Finally, we helped deploy integration with major social media channels, with tools that allow users to comment on and socially share individual pages, Twitter streams on the home page, and a blog for news items and infographics.
We use Google Analytics on all of our sites in line with our privacy policy. We have certain tracking functions turned off as we don’t want to intrude on user privacy too much, but do collect the basic analytical data and use this to glean insights into our user base. TWFY is used by the majority of people in and around the British Parliament—indeed, most members of parliament and their staff tend to use TWFY rather than the parliament’s own website—and is used by individuals working professionally in policy, news media and campaigning roles. We always see spikes in usage around election times, when citizens are looking up their local candidates voting records, and we see spikes in usage that relate to items in the news. We conducted some research a few years back into the demographic usage, and found that the user base tends to fit primarily the ABC1 social category (middle class), with a slight bias towards white and male users. Because of this research, we have tried much harder to make the site better known and more accessible to a wider user base.
Chivot: “FixMyStreet” provides a platform to improve security and quality of life in people’s neighborhoods. What are the challenges you’ve faced in using data-driven applications? What do you wish to do and develop but may not be able to (yet)?
Rumbul: FixMyStreet is like a “first step” into using civic tech for most people, and possibly the first time an individual might have considered exercising their right to request the local authority to execute their legal obligation to maintain public areas. It is much less daunting than sending an email to a faceless local authority email address, and easier and quicker to do.
The way the app works is to enable anyone to report any nuisance without necessarily knowing which public authority is responsible for them. Users click on a map, enter the nature of the problem, and can add a picture of it as well.
With the development of this site and its application, we were thinking about how we, as normal individuals, would want to be able to report issues to the council, rather than how the council wants us to do it. The platform itself is based on MapIt. While this technology is actually relatively easy to use in the UK, there are obviously challenges in developing countries, where good boundary data can be difficult to source.
The way we convert documents into structured information happens very differently across our different codebases, all of which are available free online. We have multiple functions on the back end converting information from multiple sources. We have multiple scrapers in play, and code to convert open data into more user-friendly formats.
One of the biggest challenges we face is what we would like to do with our tools, but which we don’t feel we can at present. The tension between open data and privacy is something that will not be resolved any time soon, and we deliberately collect as little data as possible on our users. This inevitably means that we don’t really have a great idea about how people are using our sites, why they do, or any other peripheral data that could tell us a lot, but could be considered an intrusion to privacy. As a researcher this make my job particularly difficult!
Chivot: Why is the role of governments so important in using and supporting civic technologies?
Rumbul: Government support for civic tech enables it to create more useful and efficient platforms for citizens, with related features that can plug into government systems, easing workloads and reducing bureaucracy. Government support also brings two worlds together, introducing more adaptive and user-centered practices within often archaic government IT departments, and opening up the black box of necessary administration to enable civic tech producers to build more appropriate tools. Partnership between government and civic tech is vital for it to move forward and be useful to all.
Governments and their departments, for the most part, produce the bulk of information that is needed to create civic tech tools. Data is the oxygen by which civic tech lives and dies, and so even where government and civic tech are very distant from each other, they are meaningfully close in data production and re-use. Governments that do not produce data, or that deliberately withhold data, frustrate the development of civic tech, and such actions catalyze democracy activists to find alternative routes to data production. This often takes the form of crowdsourcing, automated scraping, or manual data collection, which are less perfect and accurate methods than using properly produced open data in a computer-readable format.
Chivot: Some of your recent research looked at how digital technologies are shaping democratic information flows. What have you discovered?
Rumbul: The rise of the Internet means that we are no longer a society which receives information via a hierarchical system of broadcast. Whereas information traditionally flowed from positions of authority and trickled downwards through the media or organizational hierarchy, information is now produced, published, and popularized in a web format, where anyone with an Internet connection can be a broadcaster or a consumer. This has significant consequences for how citizens consume information about their democratic institutions, and is the reason why the fake news phenomenon has mushroomed over the past decade. In many places in Africa, governments are lagging behind in understanding how to harness and regulate digital tools effectively. Many choose not to engage with the digitalization of the world, and some choose to overregulate, predominantly to keep control of the narrative of their choice concerning their leadership, and to quash any disquiet. This manifests in reducing the production and publication of government information, and the suppression of the media, reducing the data necessary to create good civic tech. This frustrates legitimate information flow, and reduces the ability of individuals to differentiate genuine government information from propaganda.
