This article is written by Pamela Dow, the COO of Civic Future, a new organisation created with the intention of opening up politics and public life to a broader range of talent. Prior to this, Pamela was a Senior Civil Servant, most recently at the Cabinet Office, where she founded and led the Government Skills Campus.

The issues we deal with in government are hard. There are no easy, quick and cheap answers. Most reforms take a long time if they happen at all. That's why it's so beautiful when the planets do align.

In my last role, establishing the Government Skills Campus and Leadership College for Government, I knew we had to “raise the bar, and the ceiling” for data and digital skills. We needed a higher, universal,level of competence and confidence, and we also had to increase specialist knowledge and expertise.

There was collective appreciation of this, from ministers and senior leaders alike, but there wasn’t much money around and there were a lot of legacy systems: training frameworks and clunky platforms. It was a huge challenge.

Luckily, I had great colleagues in Nick Walker, Programme Director of the Skills Campus and Curriculum, Robyn Scott of Apolitical, Stephan Chambers of the LSE and Josie Cluer of EY. Where there was a will, there was a way!

💬 Let the author know what you think about upskilling in government by leaving a comment below

The Skills Campus recognises that 450,000 people have very different starting points in terms of digital ability, experience and needs in their roles. It recognises what modern life and work is like. It recognises how we learn, where we learn and with whom. It is built around the shared, defined, five-strand curriculum framework, allowing people to build up from the foundations to the advanced, from the generalist to the specialist.

A few years ago, I took part in a panel on the importance of data competence in public administration. The gist of it was that we needed to move beyond the slightly performative cries of “be more data-driven!”

We needed to equip ourselves with the knowledge and skills that being data-driven entails. Improved basic numeracy, for a start. Confidence with pivot tables. Not necessarily being the expert in the room, but knowing how to get the expert into the room. Knowing how to distinguish between actual expertise and misplaced confidence. Understanding the difference between descriptive, diagnostic, prescriptive and predictive, data analytics. Recognising that ‘Policy’, ‘Economics’, 'Digital' ‘Data’ and ‘Analysis’ weren’t separate disciplines, never to meet, and that good advice to ministers is a seamless combination of all.

Those trying to sell digital products to the public sector need to “show not tell”.

I used the slightly whimsical device of ‘Seven Sins to Avoid’ to describe the most common mistakes in using data in policy. This seemed to be quite well received, so Apolitical asked me to do a companion piece on the Seven Sins to avoid when trying to build a digital culture in policymaking.

And here they are.

1. Believing the hype.

Hundreds of companies and entrepreneurs will try to tell you they've found the silver bullet, the holy grail, the great new digital platform, software, or app. Their pitch may be excellent. You may feel a bit out of your depth because of the words they use and a superb presentation. But where's their proof? Can they demonstrate any promising impact evidence at all, even if it's from a tiny cohort test? Any emerging signs of impact in real contexts? If it sounds too good to be true, it might be.

You may not be an expert but you are not an idiot. If the person trying to sell you something — an idea, or a technology — can’t explain how it works, in terms you understand, then they don’t understand it. Alarm bells should be ringing. There are no stupid questions. The best questions are very simple and so are the best answers.

Those trying to sell digital products to the public sector need to “show not tell”. Those buying need to be confident asking simple questions and requesting some impact evidence, however incomplete or early stage.

2. Blaming the people, not the product.

This is self-explanatory, I hope. If human beings don't or can’t use the thing it's almost always the thing's fault,_ not_ the human being's fault. Too often leaders will blame employees for lack of digital process compliance. Or public services blame the public for low take-up of a digital service. Get back to the designing stage, and understand why! Now we have a 20-year case of horrifying proportions to learn from, where postmasters were blamed, not the flawed Horizon software.

3. Not trying it yourself.

Whatever the digital product innovation is, you should mystery shop properly at an early design stage, however senior you are. Try to sign up, call the help desk and download the app. Why should others comply or find it easy if you don't? Even better, have a go in the persona of a few others – your dad, your cousin, your neighbour. Obviously, at a later stage you need to test new digital technologies with the people who are going to be using them, but by ‘beta testing’ time you’re quite a long way in. You’re fighting optimism bias and confirmation bias. You can save time and money by testing informally, a lot earlier.

4. Underestimating or underinvesting in workforce training.

If you have been involved in any kind of digital project you will know this well. Everyone in the digital product roll-out process – policy strategists and leaders, frontline workforce, citizens and customers – needs to be supported to change their behaviour, to adopt a new way of doing things. This is as much an underinvestment of time as of money. Consistent, tailored, communication is vital: “repetition doesn't spoil the prayer”. (This Atul Gawande essay is brilliant in making the case for this – a long read, but a great one.)

5. Scrimping generally.

Especially up-front. Explain in your business case why money now saves money later. I know the Treasury Green and Magenta Books are not always the friends of doing novel things, but if you build a very robust argument, with numbers, and a supportive coalition of voices, you can win them over (at the right point in a spending cycle, and with the wind behind you…).

6. Fixing your behavioural insight, or your product design/delivery, in aspic.

When we all got a bit starry-eyed over nudging, merrily segmenting customers and their behaviour, we were all very pleased with ourselves. To an extent rightly so. (Finally public policy had caught up with what the advertising industry had been doing for 50 years...) But the next sophisticated step to make was understanding that behaviour is dynamic – it changes! Especially digital behaviour. Early adopters become rejectors, saturation affects trends, etc. etc. This is particularly a risk with middle-aged comms people observing the platforms that the kids use: e.g. “we must have a TikTok campaign!” When all the kids have moved onto something else...

7. Not seeing the bigger picture.

Speaking of TikTok… Is your provider or product in a resilient supply chain? Who might have access to data, now and in the future? Does that cheap supplier represent a liberal democracy and ally, or a totalitarian state? Domestic policy is no longer distinct from national security, if it ever was. All public services have to be considered in the context of national interest and economic prosperity. Digital technologies don't recognise borders, so you will have to.

Done reading? Make sure to share your own tips on building a digital culture in policymaking by leaving a comment below ⬇️

(Image credit: Unsplash)