I’ve worked in government for over ten years, moving between the Home Office, Ministry of Justice, Department for Energy & Climate Change, then back to the MoJ.
My early career was focused on using perhaps more traditional analytical tools, forecasting, economic appraisal, a bit of performance management. As an operational researcher, there was also lots of modelling; discrete event simulations of the justice system; optimisation models of the UK energy system; micro-simulation of the UK housing stock; and, sadly, plenty of simple stock flow models knocked-up in Excel. This was almost entirely focused on providing support to policy and strategy decisions and done with limited access to data.
Since I came back to the MoJ in 2015, I’ve focused heavily on making sure analysts can get access to high quality open source tools and data, lots of operational data, as close to live as is feasible.
We’ve managed to close the cycle from around six weeks from data being entered into a system to the first insights being available, to overnight (and in some cases almost live). This has totally changed analysts’ relationship with the business and the sort of decisions where data and analytics can directly support our staff.
Probably two early career highlights, where I was much more hands on with the analysis, were working on significantly changing the Sentencing Framework in 2012 and then the following year the UK’s input into the EU 2030 renewables targets.
These are both examples where the data and analysis that my teams provided had a direct, and fundamental, impact on the shape of the policy. Hard, intense work where it was difficult to change minds when the evidence was perhaps counterintuitive but also a joy to work alongside some of the brightest analysts and policy professions I’ve come across in my time in government. Fun times.
I’m not really one for looking to role models for inspiration unless you count Jonathan Davies, Stephi Graff or Viv Richards, but sadly I didn’t make it as a top-class sportsman - or even an average one.
I do though genuinely take inspiration from my team, who are almost without exception passionate, dedicated, creative and continually driving to find innovative ways to use data and evidence to improve what we do in MoJ. Trying to be a good leader for them and making sure their work has the maximum impact to improve outcomes for the public is what keeps me motivated.
I think the relationship between our digital community and our data and analysis community has fundamentally changed and we’ve worked out how to harness the value of combining digital and data transformation. Exactly what many of us were trying to achieve over the year; but in many ways it turned out far better than expected. This is really exciting and means that together we’ll be able to drive real positive change for MoJ and the public.
My team has also continued to build great analytical products for our front-line staff and an especially nice surprise was the Prison Data Science team winning the OR Societies President’s medals for their work on the Prison Safety Diagnostic tool. I no longer feel that working in government our tools are a decade behind the private sector, and we can produce some genuinely innovative work.
I guess there’ll be the usual hype around certain technologies. But what I want to see happen is continued improvement in our ability to explain what’s driving decisions in the most complex algorithm. Transparency (including explainability) is one of the four fundamental tenets for everything we do within the MoJ Data Science Community.
The biggest opportunities are in the personalisation of services, and the public sector’s ability to be much more precise in our ability to know when and how to intervene to improve outcomes for the public. For example, using better data and advanced analytics to support probation officers in choosing the most effective intervention for an offender.
Of course, the flip side of this is ensuring we get the ethics right ahead of our ability to deliver this. Ensuring that the public trusts us with their data, and that any applications of AI or machine learning have ethical principles embedded at their heart.
Stating the obvious, especially for a big, complex government department, I guess it’s always legacy systems. But leaving the “fix the plumbing” challenges to my digital colleagues, perhaps a more subtle challenge is ensuring that the data we need to deliver data driven insights is seen as a strategic asset.
While our digital teams are rightly absolutely focused on user centred design and delivering digital products that meet the needs of the end user; it can be a challenge to get those teams to appreciate the wider strategic value of data to support analytics and decision making. That said, I feel very positive about this year as we have now moved from thinking about purely digital transforming into thinking very clearly about digital and data transformation.