Lots of things happened this week. Many of them have been extremely unpleasant, so instead of all of that I want to talk about the moments I’ve been able to focus on something interesting and – dare I say – fun.
This week I have been reflecting a lot on system health indicators. I’ve had some really energising conversations about how we know what we are doing is working and what we are optimising for.
Much of the output of government, observably, is framed through performance indicators.
It often stems from the centre. The Treasury’s focus on managing the government’s budget creates a momentum behind a certain way of monitoring progress. GDP per capita. The size of the deficit. The share of national debt.
It flows all the way down to every programme we deliver, procurement we run and business case we write. No matter how small, we regularly end up shoehorning often abstract outcomes into crude financial measures. Cost per transaction. Value for money. You get the idea.
As civil servants it often feels like we encourage each other to only think in this way. The result is that if you can’t measure it in pounds, it’s unlikely anyone will take it seriously.
Most of the time this is not a problem. If you’re buying widgets, you can optimise cost and quality to an equilibrium you’re content with most of the time. It takes years of experience watching things go a bit wonky to realise how to do this well; but nonetheless once you have a broad enough toolkit and set of experiences, you figure out how to work with the system to give you the outcome that works.
But what if your economic indicators aren’t the only thing that are important to your goals? Like – say – when you’re trying to deliver an outcome that is privacy-enhancing, trustworthy and trusted alongside that.
If you optimise for efficiency, you definitely get growth. Case in point; before the introduction of digital identity services, an employer could waste weeks trying to conclude a Right to Work check for a new employee. This process can now take minutes, using a passport and an app on your phone.
You can measure that. You can calculate the time savings. You can track the cost. You can roll that all together and ta-da; growth!
Maybe efficiency isn’t the only metric. Maybe we also care about usage. Maybe what we want is simply number goes up. Or percent of X achieving Y.
Maybe it’s also the number of firms, employees, Gross Valued Added or tax revenues.
All of that can be added up too. Growth!
Where that leads might not be where you want to be though. If you only optimise for counting all the things, you don’t necessarily create a better thing. You end up measuring the outputs but not the outcome.
This way of thinking means we get the faster horses we asked for, but not the cars we needed.
Perhaps simply faster horses is enough. Maybe a faster horse delivers on the vision; but maybe it doesn’t. Maybe this all leads to ending up in a tepid bath.
My sense is that we should try to look wider; to optimise for more than any one set of financial or economic indicators.
How can we optimise for things that are more intangible like trust, or privacy, or good governance, or interoperability in a way that is worthy of a Treasury business case?
How do we use all of those measures, as well the economic ones, to optimise our work?
Maybe the former are pre-conditions of the latter. Maybe our system health indicators aren’t and shouldn’t be growth indicators. Maybe they are the other things. And if we can crack those, then will the economics flow from that?
Most of the discussions I’ve had about this have been thought experiments. We already have a plan, and a business case, and a set of performance indicators. There is no burning need to change how we think about this in our team. Nonetheless, it’s been an interesting exercise in challenging myself and has provided a useful way to take my mind – and my colleagues minds – out of the day-to-day for a bit to chat about it.