The Week 19 January 2024
Senior Researcher
As policymaking ramps back up in the New Year, this week saw a major speech by NAO Director, Gareth Davies, and further detail on Labour’s mission to ‘Get Britain building’.
First, Davies’ speech on ‘Getting the most from every public pound’ — a topic close to Re:State’s heart and one we published on in October last year — focused on five areas of opportunity which could potentially save government “tens of billions of pounds a year”.
Many of these will be familiar to readers of The Week. For example, Davies rightly highlights that government is poor at managing its biggest, most costly projects — such as HS2 or the New Hospitals Programme — the majority of which regularly receive amber or red ratings in the IPA’s annual reports on “delivery confidence”. Meaning they are deemed undeliverable or are at risk of becoming undeliverable.
He also points to the public sector’s “patchy” ability to buy goods and services efficiently, with issues around how data is used, government manages contracts (especially temporary contracts), and engages with key markets. Re:State has a lengthy back catalogue of reports on how government can procure more responsibly, and save billions a year by doing so.
Common to each of these is how government approaches productivity and the emphasis it places on continuous improvement. And it’s certainly true that cutting spending is an easier, but much less serious, task than taking a “focused, cross-government approach over several years” to “fully realise the potential” of high impact spending.
Something an incoming government would be wise to prioritise from day one…
Labour struck a similar note on Monday with Rachel Reeves, the Shadow Chancellor, announcing they would “examine every ongoing major infrastructure project” if they were to win the next general election, and drawing attention to the fact that too many major capital projects are “over time, over budget and are in danger of going undelivered”.
As part of this drive, Labour would create a “central unit” to oversee major projects, reporting directly to the Prime Minister and Chancellor. So … the Infrastructure and Projects Authority? We wonder how much of an improvement this will be over the already well-established and resourced body committed to overseeing major projects.
The speech did, however, contain very sensible ambitions to boost public investment (we currently lag far behind our international peers) whilst trying to ‘crowd in’ additional private investment, and to fast-track planning applications for critical infrastructure, including “battery factories, laboratories and 5G infrastructure”.
Onto the read…
This week we’ve been reading the mission statement of the Government’s recently created AI Safety Institute, aimed at developing the “infrastructure needed to understand the risks of advanced AI”. While there is a growing body of research on AI safety, the new Institute claims as its USP the ability to run “evaluations of issues related to national security”, uphold common standards in the quality and consistency of this research, and consider AI models “with various forms of access control”.
Importantly, the AI white paper last year did not create a new, AI-specialist regulator (instead setting out general principles existing regulators should adhere to), and this mission statement puts beyond doubt that the Institute is intended to function as one. So its success will very much depend on how well it can work with relevant regulators, take advantage of cutting-edge AI research occurring elsewhere, and bring in the talent needed to really understand “the most advanced AI models”.