A unique academic analysis published today shows that ‘nudge’ interventions are helping governments deliver better outcomes at low cost.
Dr Stefano DellaVigna and Dr Elizabeth Linos from University of California, Berkeley were allowed unprecedented access to data on interventions by two of the best-known nudge units in the world – the Behavioral Insights Team (BIT) and the US Federal Government’s Office of Evaluation Sciences (OES).
‘Nudges’ are small changes to the way choices are presented that make it more likely people will choose the option that brings greater benefits to themselves or the wider community.
Their analysis, published today as a National Bureau of Economic Research Working Paper, found that the projects produced an average improvement of 8.1% (from 17.2% to 18.6%) on key policy outcomes at a low cost. These included encouraging more diverse applications to police departments, increasing engagement with a free primary care service for lower income residents of New Orleans, raising college enrollment among veterans, and reducing overprescribing of antipsychotic medications.
These results translate into real change. Collectively these interventions from BIT and OES reached 24 million people through both local and national government in the United States and covered policy issues such as health, education, tax compliance and uptake of benefits.
Most of these studies were also relatively cheap to implement as many of them involved modifying existing practices such as wording a letter differently or redesigning a form. However, the study finds that the lowest cost interventions did not have a smaller effect on outcomes.
This analysis improves on previous attempts to estimate the impact of nudges in three important ways:
- Full data access. Since the authors had full access to the comprehensive records of each organization, they could look at both published and unpublished studies. This access means that their conclusions are not affected by ‘publication bias’, the issue where positive results are more likely to be published in academic papers than negative ones.
- Many, large trials. The dataset featured 165 trials that tested 349 interventions covering various policies and institutions between 2015 and 2019. These trials involved many more people than most academic studies, and these large sample sizes increase the reliability of the results.
- Randomization. The interventions were tested through carefully-designed Randomized Controlled Trials (RCTs), which are seen as the gold standard of evaluation approaches.
Dr Linos said: “It is very rare to have such rigorous evidence of what works across a host of different policy areas. Our findings provide an optimistic but realistic understanding of the power of a nudge in government services.”
Dr Michael Hallsworth, Managing Director of BIT North America (BITNA), said: “We’re happy to have supported this groundbreaking analysis that confirms the real benefit of nudges for public services. While BIT does use behavioral science in many more wide-ranging ways to achieve fundamental change, this report shows that small changes still produce extremely valuable results.”
Governments around the world have created more than 200 teams dedicated to using behavioral science to improve government services. The trend was started by the launch of BIT by the UK government in 2010 which is now well known for many of its successful nudges and has over the last 10 years expanded beyond these to use behavioral science to tackle major policy issues as well. In the US, the formal launch of the White House’s Social and Behavioral Science Team (SBST) and General Services Administration’s OES followed in 2015.
OES and BIT share common goals: to use behavioral science to improve government services through rigorous and powerful RCTs and to help build the capacity of public agencies to use these approaches themselves.
In 2015, BIT, now independent of the UK government, started its North American operations, BITNA. Today BITNA works out of offices in New York City, Toronto and Washington, DC with local, federal and state agencies, multilateral organizations like the United Nations, philanthropic actors and research institutes. To date BITNA has worked with over 100 US cities and conducted 150 RCTs.
This story of the remarkable growth of behavioral science units across so many areas of public policy in the last decade is set out in the upcoming MIT Press book Behavioral Insights, which will be published this September and is co-authored by BITNA’s Michael Hallsworth.
