In the heat of battle, sometimes it can be hard to tell that you have won.

The battle for government data has been hard fought and bloody. For a long time the outcome was unclear. The battle ebbed and flowed, with clouds of smoke and seas of churned up mud making it hard to see who was really in the ascendant. But now we can say with some confidence that the battle has been won. There are still pockets of resistance. There is still danger of counter attack and much greater danger of guerilla insurgency behind the lines. But the battle has been won.

The veterans of that battle deserve great honour. They have been persistent and persuasive campaigners and they have been demonstrators of the most powerful kind: not demonstrating by marching with placards (though there has been some of that too), but demonstrating by doing, by building, by showing what is possible.

Winning a battle is not the same as winning a war. Sometimes the result of one battle is to set the opening positions for the next one

The availability of data is a critical enabler, but what matters in the end is what it enables. The recent release of the COINS database of public expenditure illustrates the point: it is useless to any normal person without the skills and tools to manipulate 4Gb datasets – but the point is more subtly true for most of the rest of what is now available. The existence of is without question a good thing, but its audience is necessarily a specialised one.

It is hard to believe that it was not much more than a year ago that Rewired State so clearly represented both challenge and achievement. The solutions presented there were brilliant and the ingenuity involved in acquiring the data (or in diverting to a different problem when data acquisition proved insuperable) was deeply impressive. But that ingenuity is already less necessary and will become less necessary still. With all the precision of hindsight, it was possible to see at least the beginning of the end at Young Rewired State last August. I recorded three reflections at the time. The first and third – that we needed to find ways of not repeatedly starting from scratch, and that we needed to find ways of harnessing the same kind of energy and inventiveness to deeper transactional problems – are less directly relevant here. But in between them was an idea which is, I think, at the heart of this turnng point:

The second – for which doing the first would create some space and opportunity – would be to bring in users and customers more explicitly. These projects can get off to a great start using their originators as their own use case, but they won’t become sustainable on that basis. Government has painfully learned – or, rather, is painfully learning – that starting off with the assumption that you know what is best for people doesn’t deliver the greatest results. I am not quite sure where the tipping point comes between creator-evangelists and customer-centred design, but I am sure it has to come somewhere.

My sense is that we are fast approaching that tipping point, and indeed in many areas are already beyond it. The next battle, the battle which is moving from tactical skirmishes on the flanks to the core of government transparency and service, is the battle of design. Transparency is vital, but it is not sufficient. As danah boyd recently argued:

The issues with transparency are similar to the issues with Internet access and the digital divide. In focusing on the first step – transparency or access – it’s easy to forget the bigger picture. Internet access does not automagically create an informed citizenry. Likewise, transparent data doesn’t make an informed citizenry. Transparency is only the first step. And when we treat transparency as an end in itself, we can create all sorts of unintended consequences. For this reason, I think that we need to critically think through not just transparency, but the information landscape around transparency.

So the first key question is how do we achieve transparency rather than just moving the boundary of opacity?

Part of boyd’s answer is to focus on information literacy:

Information literacy includes the skills necessary to interpret information in a context. Information literacy isn’t something that people develop just because information is available. So assuming that they will emerge once we unlock information is naive. Furthermore, skills aren’t distributed randomly across the population. Eszter Hargittai has consistently shown that those who are most privileged in our society are more likely to have information literacy skills. What this means is that those who are most privileged are more equipped to make sense of and use the information that they have access to. If you want information access because you want a better informed citizenry and a fairer society, you must start embracing the importance of information literacy and the need to provide infrastructure to help people build these skills. Providing broadband access is wonderful, but without the skills to make sense of what the Internet provides, access does nothing. The same is true for information transparency. And we can’t wait until we get transparency to start creating a citizenry who has the skills to interpret the data that will be made available.

That’s clearly an argument with some pretty big implications for approaches to digital inclusion, which I am not even going to touch on here beyond the obvious point that information literacy is not created by the more open availability of data. Instead, I want to pose as the second key question, how do we use new sources of data to create services which are compelling, useful and attractive?

The answer, I think is to look in slightly different directions, not because people whose initial focus is on pummelling the data can’t do design and usability – a moment looking at They Work for You puts paid to that idea – but because the ecosystem of players and providers and of the kinds of problems which people attempt to solve are different ones. Some of that is to be found in well established (by the standards of such things) companies such as FutureGov, some newly emerging from a pure design background such as Snook. But that’s not the only source of such thinking, as a glance at the online car tax renewal service shows, the one government online service which consistently gets unprompted praise.

One implication of all that is that the time for a one size fits all model may already have passed.  There were some good arguments for what has become the web convergence project, to bring govenment content together in a small number of so-called super sites, but of necessity each of them embodies a single architecture and a single navigation model.  If there has to be a single solution, they may be the best single solution there is,  but convergence is looking increasingly like an inflection point rather than a destination:  the question of whether they are in fact the best solution there is becomes much less relevant if solutions don’t have to be single.  I wrote a couple of days ago about an example of the Guardian being re-presented in a radically different way from that used by the Guardian’s own site.  It may be valued and used only by a tiny minority, but the needs of that minority are met vastly better than the necessarily lowest common deniminator single site can possibly achieve.

The Guardian has got that external benefit as a result of opening its data.  Directgov will – we must hope – get some of the same benefit from doing the same thing. The point here though is not to argue about whether Directgov can be improved on, but to underline the fact that the nature of the problem has changed:  it is no longer fundamentally a data extraction and manipulation problem, it is a customer experience design problem.  So the third question this prompts is, how do we embrace the power of fragmentation?

There is another Young Rewired State coming in a few weeks.  Eager young developers will be mentored by ninja data geeks and will produce views of public sector data which will be new and exciting.  That much is predictable with considerable confidence.  But will they produce embryonic services which make a real difference to the lives of users of public services and members of the public polity?  I think that is less certain, but no less important.  The next battle will need a newly constituted army, not rejecting the winners of the last battle, but augmenting their forces to fight on new fronts.

Update:  I have added a couple of additional references in a supplementary note here.  Well worth reading the articles it links to.


  1. The lesson of the Rewired State hack days and most of the independent development around government data is that you can do good work by doing what would conventionally be thought of as cheating. That is, working with a different set of constraints and objectives than are generally prevalent in official government work.

    The web is all about fragmentation, as you call it. Programmers talk about code that has high cohesion, low coupling — each part does things well and is connected to other parts at its natural seams in a way that each component can easily be replaced with another. The web has mechanisms for tying disparate parts together — hyperlinks, search engines, feeds and APIs. Rather than constraining designers by producing mega-projects with strict design and technical consistency across a range of activities and tight internal coupling, break things up.

    Usefulness is more important than consistency. Flexibility is more important than slavishly following conventions. Everything on the web is just a single click away, so free up designers to work in a way that makes sense for every fragment of the web using the designs and tools that make sense to them rather than expecting that one size will fit all.

    Things move so quickly now that agility counts above all else. Digital services need to be designed and built quickly, easy to adapt and small and cheap enough that they can be mothballed the second they no longer serve their purpose.

    The title of David Weinberger’s book sums it up nicely: Small Pieces Loosely Joined.

  2. Getting the data is the easy bit and presenting this to those who need the intelligence. What is more difficult – and this requires commitment at the highest levels of Whitehall and Downing Street – is ensuring that those who require the spend data have it.

    A good example is with COINS. There are companies such as Rosslyn Analytics that have turned this data around within 24 hours (, making it available to the general public in easy-to-use analytical reports from the web at RA.Pid Gateway (

    We applaud the government’s move toward more transparency; and for companies/organizations to make the data available in a more user friendly format. But the general public aren’t the ones who need this intelligence. From ministerial departments to councils, there are thousands of public sector employees (the front line troops) who desperately need empowerment as they are the ones who have been asked to reduce the deficit through intelligent cost reductions – and if they don’t find the savings then they will be held to account.

    Until our ministers focus on the needs of public sector employees, George Osborne’s efforts to balance the books will be in vain.

Comments are closed.