This is the first election for which the UK has a single government domain. Here's how we're making it clear what was published under previous governments.
Most visitors to departmental and agency content want to see what the current position is, so there is a need to manage historical content. Users expect this to be available.
In the past there have been different approaches to this. Having a single government website gives us an opportunity to make this consistent, manageable and effective for users and publishers.
The approach to content from previous governments
Content from previous governments will remain on GOV.UK and receive ‘history mode’. There's guidance here on what this means and how it works.
Our research suggests users see the record of the ambitions and evolution of previous governments as providing important insight and context. They need a record of government activity, not a particular government’s activity.
The National Archives play an important role in maintaining a record of every website run by government. However, when people access those periodic snapshots we’re unable to send them back to the most current information on GOV.UK.
Introducing ‘history mode’
‘History mode’ on GOV.UK will make sure any new government has a platform from which to communicate its policies while meeting users’ need to keep information for posterity. There's an example here of what this looks like. The design of this labelling may evolve. If needed we’ll iterate the design and search implementation to make sure the distinction between new and old is clear.
There are over 900 organisations whose content is on GOV.UK (including historical bodies). After discussions with the Public Bodies team in Cabinet Office we’ve established that it wouldn’t be appropriate for operational organisations like the Planning Inspectorate to be identified with the ministerial politics of their parent department.
We have consulted with each department to confirm which bodies publish material which relates to a current government’s policies and therefore need their content tagging, and which don’t. Please let us know through your department or agency's single point of contact if there's anything to discuss.
An API of British governments
To implement ‘history mode’ we’ve built a new API (application programme interface). For every government since 1801 this API details the dates it started, and ended, as well as its political makeup. It starts from 1801 as this is the beginning of the modern era of British governments since the Acts of Union 1800.
It's currently unsupported but given that it’s helping us identify content from historic governments in order to preserve it (with caveats) it’s unlikely to disappear without any warning. If you do something with it then let us know.
9 comments
Comment by Stephen Edwards posted on
Are there any plans to implement any of these features other than when a government changes? There are big user needs for seeing content that relates to before/after a change in law or guidance, and this happens every day, not just when a government changes.
Focusing on archiving only when governments change is missing the bigger point. Specialists outside the Whitehall bubble are dealing with changes every single day. It is specialist needs like this that were poorly understood and not delivered by the GOV.UK team during transition. There was plenty of user research conducted and evidence collected on this. One small example that was blogged about: https://hmrctransition.blog.gov.uk/2013/10/21/tracking-changes-to-manuals/
I really hope this can be addressed soon and I would have thought this would have had a higher priority than an API for which government was in office in 1801... Following transition, users in many professions are still struggling to find important (and current) content from just a couple of years ago that has ended up buried deep in the national archive.
I'm also concerned about how you will manage the list of content types and departments to determine which content should receive these "history" notices. The limited content types on GOV.UK mean that content is often published in formats/types that don't really properly describe the content. Policy papers are a good example: we often see content published in a policy paper that doesn't relate to 'policy' in the way you think of in this blog post. It would be interesting to see the operational public bodies that will be excluded from this system to see if this might help this problem.
Comment by Simon Howard posted on
Both of these are great points, Stephen.
From my perspective, when public health guidelines are updated on gov.uk, it can be near impossible to work out what has changed between iterations. I've spent too much of my life going through documents line-by-line by hand, which is an absurd use of time. And there have been important occasions - which I've reported separately to the web team - when documents have been published with a list of changes which is simply incorrect, which is a very high risk situation indeed.
Your second point about the limited content types struck a chord, too: in several roles I've had in recent years one of the most frequent refrains I've heard about gov.uk is "this isn't a news story, but that's the only way we can publish it". How anyone can now separate the complex web of stuff published as news stories that is actually news vs stuff that is continually relevant information is beyond me; and, again, it's a very high risk situation to be in if current clinical guidance, for example, is badged as historical (as happened for quite some time during the PHE transition).
I wonder, too, whether reports which are independent of government but hosted on gov.uk will be badged with the "historical" label, given that it implies a level of government control or endorsement.
Comment by Alan Maddrell posted on
Hi Simon
On the formats point this is often tricky. We're always on hand to help with these questions if that's useful - ask the content team to raise a Zendesk ticket.
As regards independent reports, the guidance says they don't receive history mode:
https://www.gov.uk/guidance/how-to-publish-on-gov-uk/creating-and-updating-pages#what-content-doesnt-get-history-mode
Comment by Stephen posted on
Glad to see that this has been considered, although I do wonder how the process as outlined above would deal with things like this from last week (Current government publishes research reports commissioned by the previous government): https://www.gov.uk/government/speeches/unpublished-research-reports-commissioned-by-the-last-administration
Would those get flagged as historic content from the 2005 to 2010 Labour Government even though the papers were only first published under the current government?
Comment by Alan Maddrell posted on
Hi Stephen
Well that's a case we could never have predicted... History mode is derived from the 'first published' date. You can set this to be whenever makes most sense. Hope that helps.
Comment by Alan Maddrell posted on
Hi Stephen
History mode addresses the particular issue we face at this moment but we have also done some work looking at how the current functionality might evolve. It's only exploratory at the moment, however.
If significant numbers of people aren't finding current useful information on GOV.UK that indicates an unmet need. Those items should be remapped to current content - that seems pretty evident.
It's true that this work shines a new light on how formats are being used. It remains to be seen what the implications of this will be, whether there is a need for more formats or fewer. It's likely there's some misunderstanding about the proper use of some formats.
Comment by Mark posted on
I hope that you can consider some way of continuing to make statistics available after a change of government - whether through National Archives or gov.uk. Statistics under previous governments will be extremely popular for all sorts of users, and it will look extremely odd - to the point, I would think, of being a reputational issue for gov.uk - if they all suddenly vanish when a new government is elected.
Comment by Alan Maddrell posted on
Hi Mark
There's no intention to change anything about how statistics are presented, they'll still be available exactly as before.
Comment by Dain Morritt posted on
I'm glad to see that thought has been given to the evolution of the site, though I am not sure whether the 'history' solution solves more problems than it raises. As a user (both academically and at work inside government) I can see two issues with the proposed approach. My starting assumption is one which may challenge the .gov model. To my mind, when the government changes, the only thing that can change in the short term is the Policy layer, the 227 policies listed. Each policy is linked to its Policy, Detail and Latest entries and then on to action from thousands of people working to deliver this policy from across the globe, through a link to ministry business plans and Perm Sec objectives - an amazing achievement. The incoming government has three options, to adopt, evolve or reject (and maybe replace) the policy. The way these policies were set out (until early Mar 15) meant that most policy aims would not change. ie the Ends that each government seeks are pretty much unarguable - eg "Stimulating economic growth in rural areas", rather it is the Ways that change ('we will do this by...').
From this starting point, my concerns are:
a. It takes time to cascade new policy 'ways' into objectives, actions, policy, command papers and so on, in this time the UK's civil and public servants, economists, military and diplomats are considering what possible changes mean to them, and then seek to re-issue or update key documents. The 'history' annotation, whilst factual, does not enable a user to see whether the particular document (one of the 227 or indeed wider 6,349 policy paper documents) is still live or has been superseded. It is also not clear to me what happens to show that a previously issued document, unaffected by a change of government, is indeed still valid and has been adopted by the current government as policy. Is the document simply re-issued? By way of example, consider the government's approach to increasing exports in country x. Much work has gone in to translating 'increase' in to activity. How is the link made by the new government and the action that follows on their watch, together with links to business plans and personal objectives?
b. My second concern, again on the top level policies, is how to track changes in the overall picture. The figure 227 has stayed constant for over a month (possibly just an oversight) whilst in that period over 10 policies have been issued. Under the new government it is possible to delete whole policies (Policy, Detail, Latest) and there to be no simple way, bar trawling the 226 line by line, for users to see which one is missing. At this point the old entry would not be visible as there is no place in the gov.uk API that allows users to compare what is there now to what the previous Policy was. By way of example, consider 'Reducing barriers to international free trade' (Aug 13), if this is not a policy that the new government wishes to pursue, it may simply disappear. Would here be a link under collections to its previous incarnation, or would it be up to researchers to use the National Archive to find it and make the link?
Overall, my view is that flagging of documents as being issued under a previous government, whilst helpful, is of little value - the key here is the date of publication and the signature under which it is published. Readers will be trying to place this in to the context of 'is this document still valid? Do I still follow it?' What is needed, and is called for by the PASC in its search for Strategy, Grand Strategy and National Interests, is a clear statement of what the UK's policy is on all its issues. To my mind, the 227 policies in the format used in Feb 15 was a great start, but there is a huge risk of the use of smoke and mirrors in governmental transition. In a perfect world, it would be nice to think that one of the first tasks for any new government would be to review the 227 policies and to see if their Ends are comprehensive and still sensible, their Ways match party policy and that the Means to deliver them are then guaranteed or adjusted to match the resources allocated.