Towards Better Public Policy: a comment on Ewert, Loer and Thomann

Towards Better Public Policy: a comment on Ewert, Loer and Thomann

Alan Warde

We have yet to discover the full implications of the Covid-19 episode for social organisation and the legitimacy of government. Attempts to restrict the spread of infection has occasioned levels of intervention in the conduct of everyday life and restrictions of personal freedom on a massive scale, inconceivable before the outbreak. The results have surprised some people who are rather reassured by the high level of voluntary compliance achieved in Europe. Others deplore infringements of the regulations and fearfully anticipate ever more extensive disobedience. In the (dis-)United Kingdom speculation is mounting about Christmas celebrations, with their implications for extended social contact, amid wider assessment of the continuing capacity of the government, given its reputation for incompetence and vacillation, to maintain control over movement and contact.

Ewert, Loer and Thomann introduce their prospectus for enhancing Behavioural Public Policy (BPP) in light of the experience of Covid-19, which may do their cause a disservice because the extreme and exceptional circumstances of pandemic make it a very abnormal instance of policy-making. Their longer essay, ‘Advancing behavioural public policies’, introducing a special edition of a journal, apparently written before the outbreak of the pandemic, is in some ways instructively different in its content and tone. Their ‘Policy and Politics’ piece in this issue offers the more virulent critique of the limitations and narrowness of strategies based upon Nudge, the ‘iconic’ expression of a growing movement for applying insights from the Behavioural Sciences to policy questions.

A movement for evidence-based policy emerged in the 1990s, soon to be defined as Evidence Based Policymaking (EBPM) and pursued under the auspices of behavioural sciences which decreed that acceptable evidence emanated only from research based on specific, primarily psychological, techniques foremost among them the Randomised Controlled Trial (RCT). A very prominent representative of the movement has been the Behavioural Insight Team, employed by successive British governments, which frequently deploys the philosophy and techniques of Nudge. Finding these developments unduly narrow, Ewert, Loer and Thomann conclude that improvements in policy-making will require drawing on a wider range of evidence from outside the Behavioural Sciences. They find existing versions defective or suboptimal because they inadequately encompass social and political contexts which condition and steer individual behaviour. I agree. Behavioural approaches eliminate variations in context from their experimental designs and so are ill prepared for predicting the effects of situational differences.

In ‘Advancing behavioural public policies’, Ewert, Loer and Thomann, therefore, advocate a multi-disciplinary approach exploiting already accumulated and established knowledge of, for instance, the social sciences (they mention anthropology, human geography and sociology). One mantra is the need for an enhanced understanding of the social embeddedness of behaviour. To that end, they usefully point to methods alien to the Behavioural Sciences which generate different types of evidence. Ethnography, social network analysis, naturalistic observation, comparative institutional analysis, interviews and the like all constitute systematic and rigorous investigative and analytic procedures which produce reliable evidence.

However, having completed their excursion into distant fields, they propose to insert their discoveries back into a behavioural approach. Therein lie fundamental problems of coherence and representation. There is a gulf between the kinds of knowledge which permit ‘focusing on the macro level and adopting more of a normative perspective’ (p.3) and the principles of Behavioural Science. Theoretical differences threaten the unravelling of cooperation between disciplines whose incompatible presuppositions cannot readily be badged as an integrated behavioural programme.

The scientific basis of the prospectus for BPP seems little different from that underpinning Nudge. Their deficiencies are basically similar. A scion of Behavioural Economics, Nudge is built upon experimental psychology and developments in cognitive (neuro)science. Privileging that body of knowledge tends to detract from the value of evidence from other quarters. Yet Behavioural Insights are equipped neither to analyse the institutional features of societies nor to engage with normative perspectives. Those forms of analysis are found elsewhere, for example in comparative sociology and moral philosophy whose assumptions, procedures and protocols are far removed from the principles of behavioural analysis and resist reduction within a behavioural framing.

There is no hard and fast distinction between the practices of behavioural and social sciences but some strong differences, including of tone, technique, topic, and views of what constitutes a satisfactory analysis or explanation, militate against general endorsement for a programme of Behavioural Public Policy. Three important instances of policy-relevant differences suggest that the misrepresentation and undervaluation of the social sciences form barriers to both lucid diagnosis and effective intervention.

First, BPP, Behavioural Economics and psychological experiments have no basis in their scientific canon or bodies of specialist expertise for determining what actually constitutes the common good upon which to inform policy content. Arguably, their techniques are tools of persuasion for ensuring compliance with policies irrespective of whether they be good or bad. The irreducibly political moment in policy formulation is outside their remit and they lack the means to discriminate. Probing different and competing definitions of a social problem is vital intelligence when seeking remedies and planning to change institutional processes and social norms.

Second, to prioritise changing the behaviour of individuals, especially by altering their attitudes or by framing decision-making to facilitate self-regarding choices, is not the most efficacious way to achieve major changes within large populations. Because the USP of behavioural insight is surreptitiously to obtain voluntary compliance the tendency is to eschew regulation and compulsion. That neglect is partly justified by reference to a common contention that governments are no longer able to instruct citizens in lifestyle matters. Better therefore to try to shift responsibility for collective outcomes off the shoulders of the government and into the laps of private citizens managing their everyday activities. Such a predominant focus on the individual precludes other more effective ways to steer substantial social change.

Third, behavioural approaches are not designed to understand contexts and are unreceptive to alternative ways to conceptualise and account for situated and contextual action. Social sciences have ceased asking directly ‘what is the relationship between the individual and society?’ because the stark polarisation of an isolated individual and a more or less homogeneous external culture fails to give satisfactory description, understanding or explanation of observed variations in personal action, interaction, or collective action.

That different kinds of policy-relevant knowledge operate with fundamentally incompatible foundational assumptions probably matters little to policy makers who, I imagine, oscillate between derision and despair over disciplinary purists whose priorities are to get the science right. For policy formulation and implementation can indeed benefit from generous and open multi-disciplinary cooperation, drawing on evidence of many kinds to provide a platform for those obliged to make judgments about what to do.

Synthesis of many forms of evidence about the strength and configuration of relevant factors and the most effective points of intervention to leverage a positive outcome is required. That presumably happens in practice through the mundane processes of state-sponsored reforms across many different fields. However, the Covid-19 crisis has vividly exposed the inevitable and necessary gap between the diagnoses and prognoses of ‘the science’ presented by its expert representatives examining the spread of the virus, and making decisions which also must take into account a multitude of other pressures and considerations including economic recovery, mental health and social cohesion. When making those decisions, a task which principally falls to politicians, facts never speak for themselves and leave plenty of room for misrepresentation and miscalculation.

In this regard it is interesting to speculate on the meanings of the term a ‘new normality’ (I might wish!) which Ewert, Loer and Thomann anticipate emerging from ‘norms and rules for living together which differ significantly from routines that have shaped everyday life so far’. (p.3) It is unclear that behavioural approaches have any specialised insight (beyond common sense) into how new norms and rules will arise, or what their content should be or will be. Nor do they seem to have much purchase on currently prevailing routines which might prime or prevent individuals altering their norms, preferences or behaviours.

In sum, to capitalise fully on the valuable evidence of research communities beyond Behavioural Sciences requires greater openness towards different kinds of knowledge and evidence, as well as attention to the capabilities of collective actors rather than individuals. Why a genuinely multi-disciplinary programme cultivating many kinds of policy-relevant evidence should fly under the flag of Behavioural Public Policy is unclear. Why not recast the acronym for the programme as Better Public Policy?

The post mortem on the handling of the pandemic may provide a different answer as it should involve radical reassessment of the value of expert advice in the making of policy. A review will show that priority was given to expertise drawn from medical and allied health sciences and very little derived from either the Behavioural or the Social Sciences, despite the obvious value of evidence about how people organise their social life in the face of risk. Whether that anomaly is rectified and what role is accorded to BPP will be instructive. Hopefully it will recognise and incorporate the distinctive contributions of the social sciences.


Alan Warde is Professor of Sociology in the School of Social Sciences and Professorial Fellow of the Sustainable Consumption Institute at the University of Manchester.