Last week the government released spending data for departments since it arrived in May. It’s part of a movement it hopes will make the UK into the world leaders for open data – and unleash an ‘army of armchair auditors’ that will pore over every infinitesimal detail of data and thus improve the way we deliver public services.
There’ll be an army, but probably no the sort we’d be expecting. This picture is taken from Flickr. It is by Angus Kirk and is of an English Civil War re-enactment in Andwell.
That’s a laudable ambition, but there is an awful lot that needs to happen before that takes place.
In my own rather inept way, I’ve been trying out this ‘armchair auditor’ role with my own website about swimming pools. It’s helped me begin to think about how a local authority (or any form of government) can identify and support people who really care about services and want to make them better.
This will probably end up with a series of posts, but I thought I’d kick off with one that looks specifically at tips I’d give anyone who is thinking about filling that armchair auditor role. I think that by doing this it might also help government to understand what this army might be doing and how it will need lots of help to happen.
1). Find data you can work with (it’ll be small chunks).
The Guardian’s tool allows you to pick small chunks of the departmental spending release to look at and analyse yourself.
You’ll never as an individual be able to cope with the massive amounts of information that are being thrown at you. That can be odd and off-putting, because the most press attention the open data movement gets is when a big database is released. So you might be tempted to try to download them, but they’re often too big. Instead, if you want to interrogate this stuff, look to see who’s made the data digestible or find other, smaller releases and then look to pick off bits you can actually start to work with. Don’t worry, if there is to be an army, each soldier will have their fair share.
2). Don’t bother chasing big stories. Look for stuff that actually matters.
The devil, they say, is in the detail. While spending data is likely to throw up a few scandals (which journalists will look for), there is more worthy work to be done in the long term. That means picking a particular furrow and ploughing it diligently. Choose a nice, narrow bit of data that you feel passionately about and find out everything you can.
3). It isn’t just spending that’s important.
There’s been a huge amount of emphasis on spending data, but on its own it can’t tell you whether that money is being spent well – only offer clues. You need to find out what the money’s being spent on and look at performance data before coming to any conclusions. We’re not quite as good, yet, at linking spending data to other information so you’ll have to work hard, but your work will be invaluable.
4). Remember that there’s a story behind the data that you can’t tell on your own.
Just as data isn’t always linked to other forms of information, it’s also important to ask the authorities and other people about what you’re finding. Data releases don’t tell stories without context and much of this context can only be found out by asking questions – for example: are my conclusions correct? Am I missing something? What does this mean? It’s important to remember that it can be all too easy to jump to conclusions about data that are missing key elements.
5). Take advantage of the online tools available to make your life easier.
I’ll cover all this in more detail, but you should take every advantage of free tools that you can. There are so many, I’m not going to list them, but there are clever applications that will help you in crunching, analysing and visualising data that usually don’t ask for a penny of your hard-earned cash. Most will also help you to publish and share your details, which reminds me…
6). Work together: Try to take care not to repeat work.
From the Library of Congress c 1903 – Flickr user: Thiophene_Guy
If there really is to be an ‘army of armchair auditors’ out there then you must make sure you don’t end up all working on the same pieces of data and coming to the same conclusions. That’d be a silly waste of resources. That’s one of the reasons why you shouldn’t just publish your results, but also…
7). Share your workings.
I may have made this rule number seven, but it could well be one of the golden rules. If this is all going to work then people need not to work in isolation, but together. If you share how you’ve come to a conclusion then it can be peer-reviewed and, if necessary, improved. You might also be able to pick up and share tips by looking in real detail at what other people are doing. So get yourself a blog and make sure you share what you’re doing.
8). Be reasonable.
This might seem rather facetious, but it’s early days. We’re not yet brilliant at this stuff and we have systems, like Freedom of Information requests, that are a little bit, well, bureaucratic. As I’ve learned, data can take ages to come your way and sometimes it’s in a horrible mess. Don’t be put off and keep smiling. But…
9). Ask for better data.
Thousands of data releases mean thousands of inconsistencies. These are things you can tell government about. If something is badly presented, poorly annotated or just doesn’t make sense, you should make sure you can find the right person to tell.
