“It’s a dirty secret that we spend a trillion a year in Britain and we don’t know what works,” David Halpern told CSW in 2023, bemoaning the lack of evaluation in government.
In the same interview, however, Halpern expressed hope about the establishment of the Evaluation Task Force, which was set up in 2021 as a joint unit between the Cabinet Office and the Treasury. The unit, he said, meant “we now have a big chunk of the British government that’s serious about evaluating success”.
One imagines that the founder of the Behavioural Insights Team (which itself was described by Halpern as “a Trojan horse for bringing very hard empirical methods into the heart of government”) would be even more pleased to hear about one of the ETF’s most recent initiatives. Designed and delivered collaboratively with academics from University College London and economists from Frontier Economics, the Government Evaluation Conference 2025 took place in April and brought together more than 400 civil servants, academics and consultants to support and deepen that community of people serious about evaluating success.
Lucie Moore, acting head at the ETF, tells CSW that the key aims for the conference were to explore how academics and businesses can support robust evaluation of policies, and to consider the future of evaluation in government.
On paper, it may have seemed like any other professional conference, with networking sessions, panel discussions and presentations on hot topics such as the safer streets mission and how to evaluate AI interventions.
But the ETF took an unusual approach to designing and developing the event. Rather than creating it entirely themselves, the team built a strong partnership with UCL and Frontier Economics to build the agenda and steer the conference.
Moore explains that this approach followed on naturally from the unit’s aim to bring in external voices to advise on government evaluations. The ETF works closely with a network of academics and businesses, as well as having academics and UK Research and Innovation policy fellows embedded in the team. It has also set up the Evaluation Trials and Advice Panel and is the secretariat for the What Works Networks, which allow it to access a range of expertise to advise on the huge range of evaluations across government.
"It wasn’t just sponsoring a conference and putting our logos on thinking, ‘job done’" –UCL deputy director of public policy Olivia Stevenson
Moore says the ETF wanted to reflect this collaboration in the delivery of this year’s conference, and to use the event as an opportunity to showcase to attendees how working with external stakeholders can benefit the government’s objectives.
To select the event partners, the ETF ran a commercial process, reviewing bids from external organisations and selecting the most appropriate sponsors. UCL and Frontier Economics were chosen, Moore says, because they echo the mission of the Evaluation Task Force and champion the importance of evaluation.
These types of events are “part of UCL’s DNA”, the university’s deputy director of public policy, Olivia Stevenson, tells CSW. “We’re leading ministerial visits, we’re working with government departments on conferences, we’re running round tables, we’re looking at ways to join those dots and ensure that government policy is underpinned by robust evidence.”
But the collaboration for the Government Evaluation Conference 2025 was on another level, Stevenson says. “I have to say this is the first time that I’ve worked with the Cabinet Office in this way for a conference to deliver something at this scale. So it was truly exciting. It wasn’t just sponsoring a conference and putting our logos on thinking, ‘job done’. It actually was a truly co-designed, collaborative effort.”
Marc Stears, pro-provost, policy engagement at UCL, adds: “What was really palpably exciting about the feeling in the room was that there was a mutual respect – co-design in advance of the conference, but then in the conversations as well, people were very eager to learn from the different perspectives that this partnership brought together.”
Theorists and practitioners
One of the standout features of the conference design, according to Andrew Leicester, head of evaluation at Frontier Economics, was the deliberate combination of theory and practice.
Other evaluation conferences tend to be quite focused on “the cutting edge of evaluation methodology”, he says.
While this is “really important”, he continues, “the reality is that on the ground, day-to-day evaluation practitioners in government are having to make very difficult decisions about what to do with limited time, limited resources, limited data, and so can’t always apply the real cutting-edge thinking.
“So it was about thinking it through from the perspective of government evaluators trying to do the best they can to get good evidence.”
As an example of how this played out on the day, Leicester says Frontier developed a session around evaluating AI interventions, having worked with the Cabinet Office to design new guidance for the Magenta Book, “the sort of bible of how to do evaluation within government”.
“It was a really popular session headed by Dr Will Carpenter from Frontier, who was a member of the team that [helped design] that guidance. And working with the taskforce, we were able to also bring in people from the Home Office who had themselves been using the guidance to evaluate some of their own AI interventions, and think about the practice of implementing that guidance, which I think was really useful and really interesting for the audience,” he says.
Learning
One of the key themes that attendees brought up was the importance of using evaluation to drive learning, Leicester says.
“Evaluation is about learning and accountability,” he explains. “I think very often, there’s a focus on the accountability part – have we done a good job and spent money well? But the learning part sometimes gets a bit overlooked. The audience questions emphasised how important it is for them to be able to learn the lessons for policy delivery from evaluation, and to do that almost in real time – being able to evaluate things as they are being delivered.”
This meant there was lots of discussion about how to run effective trials and small-scale tests. “How you do that well is increasingly important for government, which I think is a positive thing,” he says.
“The model of doing a big evaluation, putting the evaluation reports on a shelf, doing the next intervention, clearly isn’t very helpful for anybody.”
The impact
With that attitude in mind, what lessons do the team have for others exploring how to design similar conferences? First, Moore says the evaluation offered by attendee feedback was positive about the partnership model. Attendees reported that the model brought exciting discussions to the day, and said they had continued conversations further with both UCL and Frontier after the conference.
For Moore, the key lesson is that collaboration with sponsors is an amazing opportunity to harness external expertise to enrich your agenda and deliver an exciting event for attendees in a cost-effective way.
There was also a very high demand for the conference. The initial plan was for 250 attendees, based on previous years’ turnout, but sign-ups reached this number in just 24 hours. Within four days, there were more than 500 people registered to attend. For Moore, this appetite shows the clear demand in the evaluation community to come together to learn from and celebrate the work of others and to meet like-minded individuals working in different areas.
Almost everyone who had signed up for the conference attended on the day. Stevenson says this strong turnout “really made a difference in terms of the vibes, the positivity and the cross-conversation that was happening on the day”. She adds that her big takeaway from the collaborative effort is the importance of starting with the purpose, not the format. “Sometimes you can get so stuck in: we’ve got a day and we need to fit this many speakers in. But it’s better actually starting with: why are we doing this? What are we trying to solve? How do we want to get there to solve it? And, therefore, what is the structure? What are the topics? Not losing sight of the purpose.”
“Everybody was very realistic as they left that there’s still a long way to go" – Marc Stears, pro-provost, policy engagement at UCL
One example of this, Leicester explains, was how the team pivoted away from a plan to host surgery-type drop-in sessions at the conference with UCL and Frontier evaluation experts, where civil servants could talk about the challenges they’re facing with evaluation – “a little bit more like an academic conference where people can talk about their work.
“We quickly realised that, particularly once you’ve had a very big sign-up, actually what people wanted to do was talk to each other and connect and that would have felt a little bit awkward… if we’d had that separation, with evaluation experts from outside government on one side. And so we changed that idea, collaboratively deciding what we really needed to do was just have a very good opportunity for people to come together and collaborate and network, with Frontier and UCL people involved in that discussion, rather than it feeling a bit forced and artificial.”
Ultimately, the purpose of the event was to improve how government does evaluations, and Stears echoes Halpern as he reflects on the importance of this work and the collaborative approach to it.
“It really underpins the government’s attempt to rewire the state,” he says. “We need to break down these barriers between different approaches, methods and specialities in order to bring about the change that is beneficial to the whole country.
“I think everybody was very realistic as they left that there’s still a long way to go. There are loads of projects which aren’t evaluated. There are lots of misunderstandings about the methods and approaches that might be best. Working together is a vital first step to improving our own work, but also to helping the country solve some of its biggest problems.”
To learn more about the conference or the innovative delivery model, please contact etf@cabinetoffice.gov.uk
To find out more about the event’s sponsors, please visit www.ucl.ac.uk/public-policy or www.frontier-economics.com/uk/en/expertise/public-policy