I spent a day last week at The Guardian’s Managing Public Sector Information conference in London (hashtag #psi2011), and came away with head buzzing about transparency and open data.
It’s probably fair to say that the mood of the conference was a combination of vague approval – “transparency” is a little like apple pie, you can’t really argue with it as an idea – mingled with worry. Actually, perhaps that was only my mood. But it’s prompted me to think more about the worries.
Mike Martin led a session which I think probably had most attendees a little baffled. Well OK, it left me a little baffled. But what I took from it was one key message (and my apologies to Mike if I misrepresent him at all here).
Technology changes in the promulgation of ideas – from the invention of moveable type to typesetting to web publication – have involved constant renegotiation of the roles that we describe as “author”, “editor” and “publisher”. This negotiation has taken place within the marketplace, the law, and popular understanding, and each time leads to different views of the value those roles contribute (financial, social, etc), of their liabilities and responsibilities, and of their their role in giving authority to content.
So what does it mean when we talk about public bodies “publishing” data? What responsibilities and liabilities are involved? What authority does the name of the body as “publisher” bestow on the data? To whom does the data belong once it is published?
Do these questions matter? Well, I suspect they do, partly because of two thoughts that came up during the day.
The first related to that idea of “authority”. Does the emphasis on publishing data as an end in itself reflect a rather naive belief in the objectivity and authority of data against other kinds of information? Do we really believe the figures for exam passes at a college represent a better way of selecting a college, as against the anecdotal and experience-based views we come across in conversations at the school gates (or on FaceBook)? Do we really think that crime maps give us a better picture of the experience of crime in our area than local gossip and maybe a local news-sheet does? I realise that these questions don’t imply that we shouldn’t have access to*both the data and the anecdotal, of course. But will the weight carried by each of the two words “government data” overpower other voices?
The second point came in a presentation from Pat Ellison of Barnardo’s. Barnardo’s work is frequently paid for under contract to local authorities. Pat explained how voluntary sector organisations like hers work with people who are frequently vulnerable and almost always facing hugely difficult situations. So there needs to be a high level of trust by clients of the voluntary organisation involved, trust which must sometimes be built on precisely the fact that they are not state agents. Pat explained that Barnardo’s and others have no issue with the principle that they must account to the authority for the work which they have funded: this is not completely unproblematic, but can be managed. It’s when that data is to be made public that things change somewhat. It’s not that there’s a threat to individual privacy – at least, one hopes and assumes not. The question is, is this a case where transparency, far from encouraging trust, actually erodes it? Will people have quite the same trustful approach to sharing very sensitive personal information with an organisation when they know that their data (however well anonymised) is to be put in the public domain? Do we have the right to re-use the personal information of people who, by the nature of things, are unlikely themselves to be able to access or take advantage of the neatly-packaged CSV files which document their misery? (Please note that I am paraphrasing Pat’s presentation wildly; don’t blame her if I’m exaggerating for effect.)
Of course, these are edge-cases. I’m not suggesting that they throw the whole transparency project into doubt. But they do suggest to me that we need some kind of classification scheme by which to distinguish different kinds of “public data” and the differing degrees of authority they carry – and that we in the public sector need to recognise our responsibilities as a publisher as we push the data out there.
Which is the idea I’ll explore, I hope, in another blog.