On Wednesday Gartner’s Andrea Dimaio wrote an interesting blog post entitled Open Data and Application Contests: Government 2.0 at the Peak of Inflated Expectations which Peter Smith nicely linked to the Gartner’s Hype Cycle graph from Wikipedia. I want to break his post down into three components. Two – the bad and the good, I’m going to talk about today – the third, which I’ll tackle on Monday involves some mapping and fun.
The Bad
As someone whose been thinking about and working on Open Data and Gov 2.0 for several years now three things struck me as problematic about Andrea’s post. Firstly, he misunderstands the point of open data. While many people – self-included- talk about how it can empower citizens, citizens will not be its primary beneficiary. The biggest user of open data portals is going to be government employees. Indeed, Tim Wilson reminded me the other day of our conversation with Jason Birch, the thought leader who made much of Nanaimo’s geo-data public, where he talked about how he wasn’t actually tasked with sharing data publicly – he was tasked with making the data available to other Nanaimo city employees. Sharing it with citizens was a (relatively) cost free addition. These projects aren’t about serving some techo-literati, it is about getting a city to first and foremost talk to itself – having it talk to its citizens is an important (and democracy expanding) benefit.
Second, was this unfortunate anecdote:
Yesterday I was discussing with a British client over lunch and he told me how the publication of data may lead to requests for more data (through the Freedom of Information Act), in a never-ending cycle of information gathering which is likely to cost a lot to both government and taxpayers. Another client observed (as I said in a previous post) that there is no way people will be able to tell to what extent a mash up on an application actually uses official, trusted government data.
Could government become swamped with data requests? Who knows, but in theory… it shouldn’t. Making data available should reduce the amount of time public servants spend responding to requests by diverting requests to open data portals. But let’s say Andrea’s concerns are valid and that, as a result of open data, citizens become more actively concerned and interested in how government works and thus Freedom of Information Act requests increase. The horror… citizens are interested in government! Citizens want to know how decisions are made! Remind me again… why is this a problem?
The real problem here isn’t access to data, it’s that the Freedom of Information Act process is itself broken. If open data creates a further demand for more transparent government and pushes us to foster better mechanisms for sharing government information, this is a good consequence. As for concerns that people might misrepresent public data, well a) people can already do this and we haven’t had a rash of bad applications, but even if they tried… people will stop using their service pretty quick.
Finally, another nice thing about public data is that it tends to get very clean, very quickly. My concern isn’t that government data will be misrepresented… I’m concerned that government data is already wrong and isn’t being verified. Knowing that someone might actually look at a data set is one of the most powerful incentives for organization to improve its collection. (Something Clay Shirky noted in a talk he made the other day at a Bioinformatics conference I’m at).
(There is of course, one group who may not see these a good consequences as it will change how they work: British public servant like Andrea’s client’s who raised the objections… but then they pay Gartner’s bills, not you.)
The Good
The end of Andrea Dimiao’s piece is where we find common ground. I agree that the Apps for Democracy competitions run the risk of limiting the definition of “the public” to citizen coders. We want broader participation – particularly once more complex data sets like budgets, procurement and crime data are released – from academics, citizens groups and NGOs. Here in Vancouver we’ve talked about focusing any Apps competition on the themes of homelessness, housing and the environment, since these have been the dominant concerns of citizens in recent years.
More importantly, I agree (and love) Dimiao’s concept of employee-centric government. Indeed, my chapter for Tim O’Reilly’s upcoming book on Open Government makes a parallel argument, that namely we should stop trying to teach an analogue government to talk to a digital public and instead focus on making government digital (ie. getting it “open,” networked and using web 2.0 internally) first.
And perhaps most importantly, I agree that government 2.0 risks being over-hyped. I still believe in the potential, but know that getting there is going to be a painful process (mind the gap!). Government 2.0 advocates should expect lots of resistance and adoption problems ahead – but then change is painful.
eaves.ca 
For more info about HIM, here's a link to Andrea DiMaio's bio:http://www.gartner.com/AnalystBiography?authorI…And I'm glad to see that people like you and Andrea see that we need to step back from the Gov2.0 hype and remember that, in the end, it is the citizen's (customer's) satisfaction that shows us which tools worked well.But, because those tools are provided (or approved) by Management to the government employee, then it is the employee who tells the Manager how well those tools worked in satisfying the citizen.As such, the government employee is the (internal) customer of Management (i.e., Management helps *them* in helping the citizen). And that is what is meant by “employee-centric”, i.e., focusing on what *employees* need (from Management) in order to do a good job.vr,Stephen Buckleyhttp://www.UStransparency.com
My comment on her post: I think you [Andrea] miss a point that Gavin alludes to (though I agree with you re: the international situation), that regardless of who develops the applications, if they are open to use by the public then it doesn’t matter what side of the divide you are on. I’m pretty sure I’m not a member of the digirati, and all I had to do to use an app in Vancouver was click a few boxes and now I get an email reminder that garbage day is tomorrow. Perhaps I fall into the digirati, as I have access to the internet, but then so do 90+% of Canadians (last I heard it was 95%, could be way off as I didn’t check).
This is exactly right. Open data is good for the public service and _very_ good for data quality. Public engagement is an excellent side benefit.
Awkward! Stephen, thank you so much for the correction. I've changed the gender references to Andrea. Also thank you for sending his bio link (which I've also included) I looked all their website and couldn't find it! I should have just googled his name…cheers,dave
First of all, thanks a lot for your thoughful comments to my post. About the gender, well by now I'm used to this for people who do not know me yet. But I hope the more I post controversial blogs, the more people find out :-)Coming to your observations, I am glad to see that you agree with the term “employee-centric government”: I have been getting some pushback about that, as people say that government has always been “employee-centric” and now should become “citizen-centric”. My point is that government has always been SELF-centric, has tried to become “citizen-centric” (mostly failing) and now needs to become “employee-centric” to really be citizen-centric (or citizen-driven, as I prefer to say).About your first criticism, your example is a great one but I'm not sure it is the norm. There are plenty of “open data” initiatives now mushrooming around the world and the focus (assuming there is one) inevitably is on citizens. If the employee-centric view that you agree with was brought straight into these initiatives, I'm sure they would be far more impactful (although – probably – less hyped and press-worthy).Your second criticism is about FOIA requests. On surface, I would tend to agree with you. The more data, the more requests, the better it is. But the problem that I believe our British client was pointing to is that unless processes are able to handle a likely increase of requests, the impression of greater openness and transparency is going to be challenged by slow reaction times. We've seen that already at the dawn of e-government, when the lack of integration between on-line services and back-office processes put off many potential users. All our client was saying was that one needs to think through all possible consequences of opening data and make sure issues and limitations are clearly understood to set expectations right.Finally, I love your point about data quality: the more eyes look at that data, the better its quality. I do agree that the whole exercise of publishing data will expose inconsistencies that – for a good part – will be hopefully resolved as soon as or immediately before data is published. Even when they are exposed outside and their resolution is “crowdsource”, this is positive. What is slightly more obscure is what happens when incorrect data are published and mashed up in all possible ways by social media, corporations, NGOs, individual citizens: the impact of that incorrect data is going to be multiplied several times. Of course this is an incentive to invest more in information assurance, but I would argue that this should call for slowing down and pondering more about “opening data”.
I'm sorry if I am missing something, but how would open data projects increase the number of requests under the records act there? Would love to know more about the 'whys' of that. In my city in Washington state, we have one person who is costing us $40,000 a year in staff time for requests, and he's doing that without significant open data initiatives at the local level yet. It's more whether a person wants to ask, isn't it?
I'm sorry if I am missing something, but how would open data projects increase the number of requests under the records act there? Would love to know more about the 'whys' of that. In my city in Washington state, we have one person who is costing us $40,000 a year in staff time for requests, and he's doing that without significant open data initiatives at the local level yet. It's more whether a person wants to ask, isn't it?