Two recent papers have again brought into question the value and effectiveness of ‘nudge’ based policy interventions. Tony Hockley argues that as much as these studies reveal about nudge policies, they say more about the complex way social and behavioural science concepts have moved into the mainstream of policymaking.
Pursuing impact can be a disturbing balancing act between spin and substance. Underdo the spin whilst maintaining substance and the impact will likely be zero, but credibility is upheld. Overdo the spin and risk the substance being diluted by marketing and misappropriation. The story of “Nudge” offers insights into what can happen when research has an unpredictably large impact in the world of politics and policy.
Has “Nudge” overdone the spin, and how much is a one-word book title to blame if it has? It is certainly true that the usual academic balancing act of spin versus substance was tipped by a publisher’s suggestion of snappy title instead of the usual academic tongue-twister intelligible only to the initiated. Under the title “Nudge” the book found a receptive audience of policymakers looking to fix problems easily and on the cheap after the 2008 economic crash, and a public policy community eager to adopt exciting new terminology into their own areas of interest. ‘Behavioural Insights Teams’ quickly sprang up around the world, dubbed (very inaccurately) as “Nudge Units”. There was little discernible push back against this high-level misappropriation of the term, the general excitement, and the loss of strict definition attached to the authors’ underlying concept for nudge policies of “libertarian paternalism”. In short, the authors had lost control of their own work. The book became a global bestseller. In 2021 it was updated and republished, in what was described as “the final edition”. Perhaps in recognition that the concept had stretched to the end of its logical road?
Under the title “Nudge” the book found a receptive audience of policymakers looking to fix problems easily and on the cheap
Gerd Gigerenzer, never a fan of libertarian paternalism for its assumptions of human irrationality, has argued that: “almost everything that affects behavior has been renamed a nudge”, rendering it meaningless. Even before Nudge was named and published Gary Becker said of “libertarian paternalism”, that it was “virtually impossible to distinguish such paternalism from plain unadulterated paternalism”. It may be that spinning the concept to a wide audience as “Nudge” not only made it easy for anyone to pick up the concept and run anywhere with it, but also exacerbated an underlying weakness of definition. Following the initial rush of public enthusiasm the critique began to build. For paternalists “Nudge” detracted from the need for firm state action against serious problems, for libertarians its adoption put the state on a slippery slope towards previously unethical levels of state paternalism, coercion and subterfuge. Thaler and Sunstein have expended great effort into reclaiming the substance of the original concept.
The pandemic provided the ideal circumstances for critics to draw attention to their concerns. In Sweden and the UK the accusation was that the alleged laissez-faire strategy to in face of a novel and deadly virus was effected through the deployment of “nudge policies” as an alternative to “hard science”. It was portrayed as the mark of an uncaring government committed to doing very little, regardless of the realities of context.
The past year has also heralded a major, but largely unedifying, showdown in behavioural public policy in the academic literature. First into the conflict was the publication of a meta-analysis of 440 “choice architecture interventions” that showed these “nudges” to be “an effective and widely applicable behavior change tool”. The backlash was prompt, focusing on the authors’ claim on just “moderate publication bias” within the sample, claiming that correcting this bias would produce “no evidence for the effectiveness of nudge”. Next along was a study showing precisely the opposite – that nudges are mostly ineffective. All authors have recognised the bias in the published literature, favouring positive results for publication over others. There is clearly an issue here that needs to be addressed by journal editors. Bakdash and Marusich, for example, conclude that this topic has a real impact on policymaking, because publication bias “impedes understanding for variations in nudge effectiveness”.
In May 2022 the authors of the original article published a substantive correction relating to two of the studies included in the meta-analysis, but noted that: “none of the corrections have had an impact on the pattern of results or any of the conclusions drawn from them”. In a later blog they make the case that, regardless of overall effect size, the core lesson is that policymakers need to field trial and assess the effectiveness of the different choice architecture interventions available to them, but that it is still very much worth doing.
What the competing narratives do show, however, is that the behavioural public policy community has perhaps neglected the discipline’s core message for policy: “it is complicated”. This is embodied in the Behavioural Insights Team core mantra of “Test – Learn – Adapt”. There is at least as much to learn from failure as from success, and each context matters considerably (Take a look at this study on nudges for Covid vaccination) This complexity and humility inherent in the application of behavioural science to public policy is probably the aspect of the approach that was most lost by the simple appeal of “Nudge”.
What the competing narratives do show, however, is that the behavioural public policy community has perhaps neglected the discipline’s core message for policy: “it is complicated”.
Nudges may be cheap and easy, because they often entail very slight alterations to existing “choice architecture, but they are just a small tool within the toolbox of behavioural science. The design of financial incentives for the energy transition away from fossil fuels is set to dominate public policy in the coming years. Behavioural science will hugely influence the design and delivery of these incentives. No doubt there will be nudges in the mix too, but there is no chance at all that they will deliver the shift alone.
Despite the heated academic debate around nudge effectiveness, it is clear that behavioural science is now deeply embedded in public policy. Behavioural teams are now working on policy within governments and in intergovernmental agencies around the world, making a difference. The popularity of “nudge” was probably a one-off, in which the circumstances of the time created an exceptional window of opportunity for huge impact, perhaps. conforming well to Kingdon’s definition of a “critical juncture”. Behavioural science and behavioural economics for policy making were not invented in 2008. Elinor Ostrom was arguing for cultural norms to be incorporated into policy back in the 1970s around the same time that Kahneman and Tversky published Prospect Theory. Some have argued that behavioural economics can be traced right back to Adam Smith. But it was popularisation by “Nudge” from 2008 that catapulted behavioural science into the world of policymaking.
In short, nudge policies now sit alongside the other behavioural and traditional interventions available to policymakers, and a new culture of trialling policies is developing. There is no going back on this positive development. The original proponents of “Nudge” could probably have done more to retain some conceptual purity as the concept was popularised, but that moment has long passed. Nonetheless, “Nudge” has transformed policymaking for good. Whilst the term “nudge”, that such was a critical factor in this, may now fade from academic use due to its dilution and divisiveness, the extraordinary experience of the past decade will offer important insights into turning research into impact, how timing and presentation matter and the serious (but fortunate) challenge of riding the wave of success.
The content generated on this blog is for information purposes only. This Article gives the views and opinions of the authors and does not reflect the views and opinions of the Impact of Social Science blog (the blog), nor of the London School of Economics and Political Science. Please review our comments policy if you have any concerns on posting a comment below.
Image Credit: LSE Impact Blog via Canva.
Surely the most useful way of looking at the “nudge” approach is to maintain that any social science initiative in the policy area that fails to take into account the choice architecture that individuals face is missing a very important part of the picture? This does not mean that in any systematic review “nudge” will come out strongly favoured or not, because the impact is surely going to be contextual rather than universal. I can think of two policy areas where failing to at least discuss and then try an amendment to the choice architecture that individuals face would be a serious analytical failure – organ donation and contributory pension schemes. In both case a switch from just informing individuals of the options to actually making it much easier to adopt the most beneficial approach from a public interest point of view makes a huge difference. Thus moving from “opt on” to “opt off” with organ donation enormously increase the number of potential donors, and a scheme of so-called “auto-enrolment” can achieve much higher levels of retirement savings uptake, not far below a mandatory scheme. A couple of such examples – and counter-examples – in this article would have helped. Actually, I think overall the “nudge” story is not one of failure but of success, but contextual rather than universal.