Beyond Nudge: Behavioral Science, Policy and Knowing What Works

By espousing behavioral tests in policy making and policy outcomes, the UK Cabinet Office Behavioral Insights Unit Team is imposing itself as the world leader among the first governmental efforts to integrate behavioral research into policy making


I had the unique chance to be invited by the UK Cabinet Office Behavioral Insights Unit Team (BIT) to attend “Nudge and Beyond: Behavioral Science, Policy and Knowing What Works” that took place in London last week. As suggested by its title, the aim of the meeting was to take stock of BIT’s first two years of existence and to boldly address the questions recently posed by PM David Cameron to Professor Richard Thaler, the initiator of Nudge thinking and one of the fathers of behavioral economics:

What’s should be next in the Nudge agenda? How to depoliticize the BIT’s pioneering work in integrating behavioral research into policy-making?

Richard Thaler, who has been advising the BIT since its inception two years ago, delivered a key note speech that nicely set the scene for the subsequent discussion. His main advice to smart regulators across the world goes back to the RECAP idea originally formulated in his best-selling book and UK-government bible for policy-making, Nudge, that he co-authored with Cass Sunstein, Obama’s regulatory tsar. The idea behind such a mysterious acronym, which has in the meantime been relabeled Smart Disclosure, is that both governments and private companies be encouraged to release the information and data sets they own (e.g. cost and fee paid, individual consumption rates, preferences, patterns, etc) because ‘if they do an App may come’. In particular, by disclosing data in machine readable format, both governments and companies will enable not only an easy access to valuable, yet today largely unavailable, information but will also lead third-parties to repackage  that information by thus developing ‘choice-engines’ similar to those that we use to buy flight tickets and book hotels. Why shouldn’t be able to subscribe to a credit card, sign up for a mortgage or conclude an electricity supply contract as we buy a flight ticket? There is apparently no reason why those markets should continue to be less transparent and systematically more complex to the final consumer. As a result of the diffusion of those ‘choice-engines’, some companies would be rewarded for their overall transparency, easy comparability of their offer, ability of proposing you the ‘right’ products (e.g. not those triggering you an allergy or another adverse reaction). Interestingly enough, the same services will ensure that those who do not offer good quality for value will be the losers in the market.

Successful examples of Smart Disclosure are the development of the GPS technology (started as US Defense project, elaborated by the industry and turned into a 100 billion business) and the US National Weather Service (NWS).

Smart disclosure makes information not only merely available, but also accessible and usable. Thus, for instance, those signing a mortgage should not be given thousands of pages written in small print, but a USB stick containing the terms of the contract. This would enable anyone to play with the text, search for keywords, and eventually be in control of the transaction by mitigating market asymmetries.

How to persuade companies to engage into Smart Disclosure? This seems a tough call as companies do not seem to have incentives to tell you what they know about your consumption patterns because if everybody will do that they might end up competing just on prices (i.e. less profit margin). Indeed, should companies sign up to the Smart Disclosure agenda, the world would look like that typically described in conventional neoclassical economic textbooks. Yet there seems to be a case for progressively requiring companies to disclose the information they already have. This might enable the attainment of valuable policy objectives while mitigating compliance costs.

David Halpern, the Director of BIT, followed up with a short talk illustrating the long history of nudging, since the introduction of road strips and slow down signs. He discussed the many challenges (media pressure, ideological divide on the issue, methodological, etc) of setting up the first dedicated behavioral policy-making unit in the world and illustrated how its team successfully began to engage into in-house generated experiments. After finding itself under the media pressure, the BIT quickly became aware that that the only manner to gain (further) support within the government and the overall public was to make some ‘quick wins’. This seemed to have been the priority of the BIT over the last year.

In order to illustrate how the BIT moved from the lab to policy making, Owain Service lined up an impressive number of (mostly young) researchers working on different areas of policy intervention. Virtually all of them have been running pioneering experiments showing that behavioral insights are not only lab-based evidence but that they can deliver dazzling results in policies as diverse as criminal enforcement (e.g. let’s communicate with a convicted via SMS not standard letters), tax collection, energy saving and public health. It was the UK counterpart of Malcom Gladwell, the undercovered economist, Tim Harford, who introduced the individual presentations. As you may know, Tim has been quite skeptical about the ability of BIT to effectively integrate behavioral research into policy making.

Although the BIT’s first experiments all took place in uncontroversial territories (more likely to lead to ‘quick wins’), they suggest a serious and honest commitment to ‘test, learn, adapt’ while adapting public policies. The newly released report (which carries such a self-explanatory title) identifies Randomised Control Trials as the privilege tool to test policies whose ambitious is to inter alia integrate behavioral research findings. Lacking a unifying method for integrating behavioral insights into policy making (only in public health dozens of methods compete), the rather neutral RCT might offer a reasonably accessible methodology to continue the experiments.

The narrative chosen by the BIT to present its challenges to the nudged-enthusiastic audience gathered in London clearly had the merit to raise important questions. Yet there seemed to be a broader issue animating the debate and that is today at the center of the BIT experimentation: how governments and their administrations can make evidence closer to the policy cycle? In particular, what is needed to make a new promising body of knowledge, such as behavioral research, to become a source of policy advice? Isn’t there a duty for public administrations to nudge the political process to allocate public resources in the most effective manner while designing policies? Yet, in the affirmative, how to persuade career civil servants and politicians to experiment new forms of regulatory intervention that ontologically question their role, expertise and understanding of public authorities’ role?