Skip to main content

This blog post has not been approved

Posted by: , Posted on: - Categories: Defra digital

I’m a government romantic. I think there is a Nirvana out there where all I need is trust, not approval, to deliver. But my own default position is to seek approval. I feel that approval is so built-in to everything we do, it’s hard to imagine another version of the Civil Service where trust is our default position.

I’m not an advocate of total anarchy – though a little bit is lovely, particularly if you're having fun and delivering. If something goes wrong – if I step over a line; if I break the trust – then I expect to be told, and told very early – not years after my work has finished. I expect that I should learn from my mistakes and the mistakes of others, as well as their successes. Again, I’m not sure this is my default position. I’m guessing that, if I take a long hard look back at my career , I make the same types of mistakes more than once , so again this is something I need to change.

The problem with approval is that it can be used to deflect responsibility. Trust is more empowering and more personal. The problem I wrestle with when it comes to trust, is that it’s often implied rather than being explicit; approval can only be explicit. I keep flipping from the need to tell everyone, individually, that they are trusted, to the idea of just putting it on a poster. Now, not everyone has such an understanding line management chain as me – I get that. That’s why it's important that I write this down and share it, without them reviewing the content (trust me, it will be fine).

I think that, as Defra group becomes a data-driven organisation, trust and a continual reaffirmation of this trust will become more important than approval. Ultimately because, in the long-run, we will be able to ever more readily automate approval. This really does create big questions for the Civil Service and Defra, and I feel in somewhat of a privileged position to be considering them, though of course anyone can – and I would encourage you to share your thinking. Am I worried that as a result I may lose my job? Nope. The current role I’m doing? Well maybe. I just saw an advert for a tool to automatically write blog posts, and given there’s so much content out there, I wouldn’t want to use it right now (I don’t trust it). I can see a future where this is perfectly feasible, I said, there are big questions to consider.

But as with everything we do in the Civil Service (at least for me), it’s about teamwork. Yes – we need to sometimes artificially break that down into the ‘I’, and what I’m focused on doing, but all the best work I’ve been involved in has been delivered by a team. Some of the opportunities that being data-driven provides, and the questions it throws up as for Defra group, I think we are a little bit too early in our journey to consider – we are still quite busy, as a team, laying the foundations, or as James would say ‘strapping wings on a rocket so we can fly’. Getting some of the basics right is important.

I’m really looking forward to the coming year, and delivering stuff… as well as overcomplicating things a bit less, and yes – having fun.


Sharing and comments

Share this page


  1. Comment by Stefan Janusz posted on

    Thinking about approval and whether it's always explicit...what about situations where approval of a general type of thing is given (which is more implicit); but approval of a specific case of that general thing hasn't been given? (i.e. I can imagine a manager saying 'yes, I agreed to something around X but the type of X you've done required further approval?')

    • Replies to Stefan Janusz>

      Comment by Elspeth Body posted on

      I think that as with all communication, the onus would be on both parties to ensure clarity of understanding. If you think the thing you want to do might need a bit more approval - ask. If you're the manager and you did not make clear what approval was required, it's on your shoulders

  2. Comment by Alan Rider posted on

    I think it is possible to combine trust and approval by having clear standards and delegating delivery within those boundaries e.g. "providing you meet the standards, then I will leave it to you to decide how to do this" . The tricky bit is how to encourage innovation, which brings with it an element of risk and an acceptance that rules may need to be stretched or broken at times to make new innovative stuff happen. Providing serious harm doesn't result, should we be more relaxed about rules and protocols?

  3. Comment by Owen Boswarva posted on

    Thought-provoking ... but also problematic. If you had said something inaccurate or controversial in the above post, who should the public hold to account? Are you speaking as the government or as yourself?

    Building more trust into the approvals process is a fine objective. But individual civil servants publishing *without* approval or authority effectively creates a class of "deniable" public sector information. Those of us who aren't government romantics (most of the general public I suspect ...) are likely to view that as bad policy.

  4. Comment by Elspeth Body posted on

    Thanks for writing this David, it's a great post, and so valuable for these thoughts to be aired. I think you're right that we're too early in the journey for some things to happen, and different parts of Defra are at different stages of embracing a data-driven and trust-driven way of working. It's an interesting and occasionally frustrating time, but we are going in a good direction. Keep at it with all the good work you do (both plural and singular)!

  5. Comment by Travis Andrew posted on

    Hi David, very interesting read. I think trust is something any organisation finds difficult when looking to make decisions yet at the same time engage the entire team and ensure their opinion is taken into consideration and trusted. I agree this can be tiresome and such a time-consuming process....I'd be pulling my hair out if I had any left!

    The key question I think should come down to; how do you currently ensure visibility is spread across the entire team as so to give them visibility into key questions and allow for better collaboration across the board? What is the current thinking about 'how do we currently collaborate and are we willing to challenge the status quo?' I'd imagine it would be tough to do this, but thinking about the time is takes to communicate decisions and get responses is something that can allow for thinking into challenging the status quo with different tools and methods.

    Interesting to get your thoughts on that....

  6. Comment by ChemistryPoet posted on

    Ideally the envelope of the area of agreement is defined, and the details of the implementation can be agreed by peer review rather than hierarchical management structures.

  7. Comment by Emily posted on

    Thanks for this excellent blog. Really thought provoking. I was visiting Wilkie & Son in Essex today and got talking to one of their directors about the culture of trust there. He said that they have few staff, mainly because they trust each other and don't have to keep records or produce big reports to argue for new product lines - because they don't need to go back later and prove that, given what they knew, they made a perfectly reasonable decision. If it turns out a product doesn't work - which some of them don't (apparently an ice cream range, for an example) -they bank the learning, are grateful for it, and strike out again. It's definitely the world of prototyping, failing fast. He talked about 'effectuation' - It sounded similar to user-centred designing, and agile approaches.

    I think doing this in Government inevitably has its limits - we need to be accountable for the taxpayers' money we spend, which in turn means we need to track what we did and why. Parliament needs to know, and that means we need to report on what we did, and the National Audit Office gets to look at the choices we made. That takes us in to the realm of business cases. And the business case process can make decision making slow and laborious, not agile. But we should improve on our language of prototyping, and work that in to the process of developing an approach if we can; and more explicitly talk about what risks we are prepared to take (ie what money we are prepared to put on trials) in order to learn - because of course we would waste much more if we got it wrong. That way we can make more space for trust, and less for approval.

    Thanks for getting me thinking.