A tale of two datasets

Controversially, Gavin Freeguard, head of data and transparency at the Institute for Government, was allowed a PowerPoint presentation at Open Data Camp 4. However, it was in a good cause.

His slides enabled him to give some concrete examples of the data in the Whitehall Monitoring Project, which he runs. The project monitors the shape and size of government, the morale of civil servants, and other factors.


Freeguard picked out two data sets for discussion. First, picking up on his Dickensian session theme, “the best of times” – the annual Civil Service Employment Survey. “This has been going on for decades, and is part of an official release from the ONS, so it’s pretty good.”

Even better, it enables interesting analysis: showing the grade at which civil servants are employed and enabling mapping of how seniority changes over time. This can be shown as shapes – many departments look like a “kite” – and “unfortunately for the Cabinet Office a coffin.”

The project’s analysis is being used by government. Freeguard said: “We know at least one government department asked why their department looked like it did when another department looked different” and then used it for workforce planning.

The worst of times (or, perhaps, a time like too many others)

But then there’s the “worst of times” – departmental organograms. Former prime minister David Cameron asked departments to publish these, and departments are supposed to publish them every six months.

However, IfG analysis showed some are not doing that.  Also, there’s little evidence that departments are using these stats. These “two tales” raise questions: such as ‘how do we persuade government to publish more data’, ‘how do we get departments to engage with it’ and ‘how can outside organisations help’?

Or, perhaps, how do we persuade official bodies that it’s a far, far better thing to publish data, let it be scrutinised, and then react to the scrutiny, than leave data closed up and neglected?

Politics is essential – so we need to apply political pressure

Participants asked to discuss these questions argued the answers were likely to depend on factors like complexity and value. “The real question is why would we do that analysis, why would we do that work? And why would we publish it?”

But some people working in government noted that political commitment could make a big difference. For example, departments have been told to draw up an annual plan to say what data they will release: and while some are struggling, others have made significant progress.

Freeguard agreed that ministerial interest was vital. Departments with ministers committed to this agenda have made appreciably more progress. But that just raised the question of how to get ministers interested. After all, he pointed out: “This is not an obvious agenda for politicians: it is not something that comes up on the doorstep.”

Naturally, this focused the attention of the group on how to make data a doorstep issue. Reports from think-tanks making use of the data might help in itself; but so might creating communities to do more with them. Freeguard said the IfG is already thinking this way: “A hack-day sounds like a very 2010 solution to a problem, but we have been wondering whether to run something like that.”

A participant coming from a local government perspective argued that requiring councils and departments to release data to access grants or take part in particular programmes might help. Another argued that official bodies should be required to give – and publish – proper reasons for not releasing data. Another suggested the media needed to be brought into the debate: and encouraged to make more use of the data available.