So it’s a Saturday afternoon, it’s 20+ degrees outside – you know what you should do?
That’s right, read a thread about GOVERNMENT ORGANOGRAMS, a year on from our @instituteforgov #IfGhackday #opendata
Organograms aren’t most obvious data thing to get pulses racing. Hardly AI, blockchain (#buzzword), Citymapper etc.
Part of Cameron’s May 2010 pledges (bit.ly/2LXhdqE) to open up data, they include names, salaries, professions, jobs of civil servants & units w/in depts
Maude called them ‘the most comprehensive organisation charts of the UK civil service ever released online’ bit.ly/2uVd28k. Not the flashiest or sexiest datasets, but mgmt info vital both for effectiveness (understanding your workforce) and accountability.
In Whitehall Monitor 2017, we found a lot of them hadn’t been published bit.ly/2vkUQEH.
Patchy publication, and data quality, suggested they weren’t being used.
After a session @ODcamp, bit.ly/2NO6C1R, & a Govt Data Publishers meeting (thks @nickmhalliday), a few reasons emerged:
- Political mandate unclear/faded
- Difficult (although bit.ly/2dHnglv)
- But the biggest one: people didn’t see the value in them
We thought we could do something about that. So – after showing what we could do with the data – we decided to organise a hackday, a practical solution to see what value people could get out of the data bit.ly/2tgbntG
People did some great things – we tried visualising the internal structures of DfE, NAO did some great #dataviz, GDS some useful data science to work out impact of managers on teams, all here: bit.ly/2uO8Veo
As well as showing the value of using the data, there were great conversations throughout the day about how organograms could be used and improved.
Here’s a flipchart of what people thought of about the data – pink = things that could be improved, yellow = good things…
So what’s happened since?
- People still talk about the hackday as a model that could be used to show the value of other datasets
- Refreshed Cabinet Office guidance (Dec 2017) made depts’ responsibilities for transparency data clearer bit.ly/2jWzQzd
However, despite some improvement, some depts still aren’t publishing organograms as mandated. (Chart flatters CO btw – some recent files blank.)
Also not convinced they’re being used – still seem like a tick-box transparency exercise, rather than info used to run depts
So what does all this say about govt data more widely?
Prompts questions: what problems are you trying to solve with the data? What questions are you trying to answer?
There’s a sense of organograms being a transparency pledge without thinking these through. Even though…
There are a lot of questions to which organograms are an answer. People *inside* govt complain of not being able to find the right people – real consequences for institutional memory, knowledge mgmt.
Useful for us on the outside.
Useful mgmt info. Etc.
Seems to be a lot of data duplication. Depts produce organograms twice a year. Different system from what they give to the ONS. Probably different system again from any HR system – or systems – which depts are actually using. Lots of duplicated effort. A waste.
Hackday useful for showing value of using data, but question remains – how do you encourage depts to use their data?
How do you move from dogfooding (bit.ly/1nDaUfH) to foie gras?
That is, from idea depts should use own data to getting them to do so?
And from the outside – still real issues of data discoverability. Even where organogram data is published, it’s more difficult to find than it should be.
As we said last summer:
If only there were some, I don’t know, national data strategy on the horizon (🤔 bit.ly/2HORfTy) that could look at big qs like:
- What data do you need to run a govt?
- How do you make it as easy as possible for people to record, find and use government data?
Thanks for indulging me. You can find everything that came out of our hackday, and some further background and resources, here: bit.ly/2uO8Veo