Understanding Flows of Information and Power in Open Data

Berlin.
I’m chairing the next workshop at the Berlin Symposium, which features a paper by Viktor Mayer-Schönberger and Zarino Zappia. Zarino starts us off by highlighting the Obama administration’s statement that government should be transparent, participative, and collaborative – and a number of open data sites by governments and non-government have now been set up.

But where is the research into how this material has been used, by whom, why, and with what results? Will such re-routing of information flows bring about a democratic renaissance, or will we see the rise of intermediaries who wield new forms of power? To address some of these issues, Zarino and Viktor have begun to map the new field.

Viktor notes that this mapping does not work from specific ideological pretexts, but aims to investigate the sociopolitical contexts of the re-routing of information flows. This can also be compared to other instances of information re-routing in the context of government: first, electronic rulemaking, which aims to engage a broader range of stakeholders in governmental policymaking (but on the terms of the government); this worked fine in the first place for engaging small numbers of lobbyists in the U.S., but not necessarily for more comprehensive public participation, in spite of high hopes for mass participation. In terms of democratic participation, e-rulemaking was a failure – not least also because the system was gamed by lobbyist organisations encouraging their audiences to mass-submit ready-made boilerplate contributions.

At the same time, e-rulemaking did improve the outcomes of policymaking processes (once the boilerplate mass submissions were removed); a handful of valuable new contributions from non-standard stakeholders could be identified and could influence the rulemaking process. So, this does redirect information flows; it also resulted in a shift in information intermediaries.

Second, open science and open source: these generate a substantial amount of participatory input, but this input is concentrated amongst a specific group of participants; software development processes are changed, but not democratised. At the same time, the outcome of these processes is usually of high quality – mass participation did not happen, but was also not necessary to improve outcomes. Again, the landscape of intermediaries in the software development process was reconfigured.

Third, the (political) blogosphere, whose aim could be descried as opening up journalism to new voices. Again, this did not necessarily result in a mass movement of citizen journalists and citizen engagement – rather, there are certain levels of concentration, homogeneity, and groupthink in this environment; at the same time, new voices have emerged, information flows have been redirected, and new information intermediaries have emerged. Has this improved journalism? That remains a disputed question.

Zarino explored similar issues in open data by studying some 175 open data Web applications, from local authorities through national government data to worldwide datasets, as well as interviewing a cross-section of open data participants; he also generated a network diagram showing the interactions between these actors. Almost three quarters of the 215 developers Zarino identified were ‘lone individuals’, as it turned out; only 29% of the apps were collaboratively developed.

Zarino also found that there was no more than a loose alliance between parties interested in open data issues; people followed their own agendas rather than being driven by a strong focus on open data. Notably, there were communities around the areas which deal with open data, rather than around the idea of open data as such. Only some 10% of actors used more than one open data source, incidentally.

Some obvious intermediaries also emerged: the World Bank was a main source of information for one group of developers and apps, for example; some of these intermediaries also take the role of gatekeepers between specific thematic communities, or inside specific communities. Developers tend to be well aware of intermediaries, both good and bad; overall, however, the landscape of intermediaries is still emerging and developing at this point, too. Whether dominant intermediaries will form remains an open question. It’s too early to tell – but it’s always too early to tell.

Intermediaries per se aren’t a problem – they do perform important roles of nurturing and supporting development, for example, and provide collaborative platforms for developers. Past experience suggests that dominant intermediaries will form; this may help nurture involvement. There’s no need for full representative involvement, of course; a simplistic notion of mass participation is unhelpful, and a more heterogeneous structure of open data ecosystems may in fact be productive. At the same time, the role of developers, in particular, must be observed closely, because of their significant power – the future of our democracy may well be in their hands.