The UK Nudge Unit is the most trendy thing in behavioural economics these days, and quite possibly one of the most exciting things in government. To my mind, their basic message is twofold. First, that people should pay a little more attention to the details of how policies and programs are implemented; little things, like how a notice of overdue taxes is phrased, can make a big difference to how effective it is. Second, that we should be testing our policies: that experiments and randomized control trials can play a key role in making sure we know what works and how governments can best serve their citizens.

The idea behind behavioural insights is that small, even superficial changes to things can make a big difference. If you tell people that everyone else pays their taxes, they’re more likely to do so; if you offer to clean people’s lofts if they agree to insulate it, even if you charge them for the cleaning, they’re more likely to insulate their lofts (which reduces carbon emissions); texting people that they still owed fines to the UK Court service increased payment rates, saving money on bailiffs and courts’ time. Small changes, but a few percentage points increase in taxes paid can save millions of dollars.

The team started with an admirable focus: if they didn’t manage to transform at least two major areas of policy, spread understanding of behavioural approaches across the UK government, and achieve a tenfold return on the cost of the unit, it would be shut down on its second anniversary. More government programs could do with such a motivator. The fact that they’re still around testifies to the enormous impact they’ve had.

David Halpern, as head of the Nudge Unit, is clearly a true believer, and he glosses over some things that I suspect his critics would not. He doesn’t mention, for example, one of the Nudge Unit’s most controversial policies, attempting to convince immigrants to leave the UK: while he discusses some potential ethical objections to nudging generally, he rapidly concludes that none of them hold water and that nudging is perfectly acceptable, which I’m not sure all readers will agree with. He also spends a lot of time explaining behavioural economics more generally, but to be honest if you want a general primer I’d start with Nudge or Thinking Fast and Slow – the value added of this book is the actual story of the Nudge Unit, and it is unfortunate he doesn’t spend more time on that. Still, if you’re interested in behavioural economics, well worth the read, though I wouldn’t start with it.