David Halpern

Trusted policy advisor bringing formal experimentation and behavioural insights into governments, first in the United Kingdom and now in many countries

Figuring out what does and doesn’t work and getting that evidence into the hands of policymakers and practitioners is a public good. There are financial, and sometimes political, costs in creating such knowledge. But it is of enormous benefit to all. The Evidence Commission drives home that it’s not just a national public good, but an international one. Whether you live in Birmingham, Boston or Beirut, we’re all asking what’s a better way of teaching our kids maths, reducing crime, or boosting sustainable growth. It’s mad not to learn from each other.

Previously, David Halpern shared the following comment as part of the report launch.

For me the key take-aways are: 1) the sheer scale of the ‘catch-up’ needed for other sectors if they are to ever get to where the health sector is in all aspects of the production, sharing and use of evidence; 2) the need for a global mechanism for governments to jointly commission evidence syntheses – not least to avoid duplication – and for a set of global public-good producers to respond with high-quality and timely evidence products; and 3) the need to build ‘absorptive capacity’ in governments and professional bodies. I’m both passionate and impatient on these points.

On the first point, we need to lay bare the fragility of our evidence base in so many areas, but more positively what’s possible when we do build it. COVID-19 illustrates both sides of this – incredible and rapid advance in some domains, but also some serious lacuna. This sets up our recommendation 2 – all of us should pay attention when a claim is being made and ask about the quality and applicability of the evidence on which the claim is based. Demand better!

Turning to the second point, we need to ‘flush out’ the questions that government departments should know the answers to but don’t – or said another way, we need to identify the areas of policy and practice that are ‘built on sand.’ We’ve had some success with this in the UK with what we call ‘areas of research interest.’ These questions posed by government departments now help shape the research funding agenda of UK Research and Innovation (£8 billion per annum). This connects to our recommendation 5 about making government evidence-support systems more fit-for-purpose. We also need a global coordination mechanism to respond to these questions by generating, synthesizing and sharing evidence. We would call them a global network of What Work Centres (extending what we have already in the UK), but other countries may want to use a different name for the network. The global network can help to address the uneven coverage and quality of the available evidence, and the unnecessary duplication that we see now with each country doing its own thing (or free riding on the investments of others). This connects to our recommendation 24 directed at funders.

The last point brings me to the weakness of the institutions that people think of as offering definitive policy advice. The shocking truth is that, across large swathes of policy and practice, we’re stumbling in the dark. Robust evaluations are rare. At the same time, policymakers are prone to over-confidence. Technical guides such as the UK’s Magenta Book on designing evaluations and the Green Book on how to appraise and evaluate policies, programs and projects are a good starting point. We need more fit-for-purpose evidence-support staff and partnerships, science advisors, and advisory bodies in government (recommendations 6-8), and corresponding improvements in professional bodies (recommendation 12). Building evaluation capacity, such as the UK’s new Evaluation Task Force, is especially important as pump-primes for evidence building alongside the capacity to utilize it. One day I’d like to see us select, periodically test and internationally compare senior policy advisors on their ability to understand and use evidence. The Evidence Commission report brings such ideas together, along with a lot of ‘how to’ guidance.