Exploring Europe through data

Last weekend, Stijn and I visited inkLink13, where I presented a few ideas about using data to cover the European Union. The meet-up was a first encounter between Budapest's tech and journalism communities. As both groups shared some of their ideas for the future of news in Hungary, the discussion was given a special weight by the actions of the Orban government, which has shown little regard for the freedom of the press and routinely rejects requests for information.

I picked the EU as a subject for two reasons: first, because of its undeniable influence on the everyday lives of all of its citizens, whether they live in Budapest, Berlin or Barcelona. Second, because even in 2013, covering the processes and discussions within the Union - especially inside the Brussels bubble - is not routine for many news organizations in the EU. Even Spiegel Online, with a journalistic staff of 150, does not yet have a permanent presence in Brussels.

While this is a real challenge to journalism in the member states, it is also an opportunity: the EU offers a blank slate in terms of government oversight. This creates a chance to re-design the technical and institutional architecture of how we cover government: What role can technology in general (and data in particular) play?

In this context, I think five aspects deserve some thought:

  • Building the right data feeds ("publisher/subscriber accountability"). When I think about the role that data can take in monitoring government activity, static databases increasingly seem like a dead end. To be relevant for news reporting, data must be delivered in feeds, just as in wire services or social media. Information in such services must be current, well-prepared, filtered and easy to use both for human editors and for machines trained to handle those events which can be expressed through sets of rules.

    To some extent this comes back to the question of human-assisted reporting and the use of heuristics to answer key journalistic questions (the biggest, the fastest-growing, etc. items in a dataset). But I think that constructing automated feeds in journalism also makes a great deal of economic sense.

    For most data investigations, the effort required for automation largely overlaps with having a clean and reproducible "chain of evidence" in the first place. The ability to continuously update and perform an analysis is directly related to how well it was done in the first place. A monitoring workflow requires additional investment, but clever use of cloud computing and open source tools can help to turn these projects into cash cows, stories that keep on giving.

    There is already a growing number of such feeds, but they're rarely considered as news sources by journalists. stf's parltrack is one example, OpenTED and LobbyFacts and ItsYourParliament don't offer suitable, continuous streams of data.

    This rift is growing, presumably caused by diverging skill sets and language. Filling it will be beneficial to both sides, the feed providers as well as the journalists.

  • Cooperate across countries and communities. This is a natural result of the complexity of the topics and the methods needed to cover them. FarmSubsidy.org - as well as the recent OffshoreLeaks investigation - are great examples for this, and the argument in their favor within the EU is simple. Most (data-driven) projects require some investment but can then be adapted to each of the 27 member states to provide stories for national media.

    At the same time, cooperations between coders and journalists are necessary, although by no means simple. I'm involved in a few such efforts, and differences in approaches, commitment, resources and interest have made each of them project management nightmares. Making sure that there are sufficient resources and a common approach to managing and planning work seems important in order to hacks/hackers cooperations more successful in the future.

  • Use your rights. Those covering the EU need to make more intense use of the rights to information that have been established for citizens. Probably the most impressive example of this has been Access Info's investigation into the practice of extraordinary rendition flights within the EU. Their team leveraged differences in FoI regimes across member states to gain access to documents and communications even in cases where the country producing the documents had rejected the request.

  • Report actions, not words. The goal of reporting on any government should be to break through political rhetoric. This applies particularly to data-driven reporting, where we need to move away from pre-framed datasets - such as national statistics - towards those data sources which can serve as evidence to actual behaviour: transactional spending, environmental regulations violations, lobbying expenditure.

    Getting the facts is made easier by the Commission's desire to measure it's member's activity in great detail: whether the handling of environmental affairs, consumer protection or state aid to companies and banks: the EU is the country of the ten thousand databases, in which regulation is often based on surprisingly thorough metrics (which, of course, don't always support the eventual decision).

  • Tell a story. Facts can only have an impact when they are used to make a point about the real world. Understanding the EU also means understanding the technocratic language and process of Brussels. Going native in this environment is easy, especially when using a tool as complex and prone to "expertism" as data. So, more than ever, trends and patterns need to be translated into human scale and connected to the political reality of citizens.

    Again, FarmSubsidy.org is a good example: while they were able to make a strong case for the mis-allocation of the agriculture and fisheries funds, their story of big food companies receiving heavy subsidies was not strong enough to significantly influence the political gridlock.

    One of my favorite narrators from the Brussels bubble is Ronny Patz. Together with his colleagues at the BloggingPortal, he tells an updated kind of story in which Europe is considered as a real political place - more than just a trading ground for member state interests. Discussing Europe purely in terms of its members is, in many ways, now inadequate and distorting. In this regard, his blog and his approach are brilliant inspirations for journalists covering the EU.

For news organizations that don't properly cover the EU yet, it's high time to act. Still, this new environment should also be used for experimentation - with technologies, in collaborating with other organizations and in telling different types of stories.

The complexity of decision-making and regulation in the EU makes these approaches necessary, but in many cases they may proove to be useful at home, too - in national politics and even a city government.