Ethical approaches to using behavioural insights in government

There has never been a more important time to ask the question of how governments can change behaviour for the public good.

The global coronavirus pandemic has highlighted the important role that behavioural science and associated "nudge" policies can play in keeping people safe, understanding risk and promoting health behaviours such as hand-washing, alternative handshakes, and compliance with social distancing rules.

But as a social researcher who has been investigating the ethics and politics of "behavioural governance" over the past decade, I also wonder whether this is the wrong question to be asking. Is changing behaviour always the right approach and how should public servants address the ethical dilemmas of applying behavioural insights? Researchers have long been asking whether behavioural public policies are ethical, and if so, how can we be sure to nudge for good?

Whether governments should nudge or not comes down to the ways in which different forms of knowledge or expertise are prioritised in shaping government policy, and are as relevant in times of crisis as any other.

When public servants use behavioural insights while designing policies, however innovative and/or vital to life and death, there are three key considerations:

• When is it appropriate to use behavioural insights?
• How will they shape relations of power?
• Will people trust the evidence, evaluation and approach taken?

Together, these considerations help us respond to the question of whether behavioural public policies are APT for citizens.

Keeping the democratic practices of citizenship at the forefront of our ethical thinking reminds us of the people who governments serve and the pivotal nature of the social contract which governs the dynamics of ideals such as autonomy, dignity and justice within public life.

Many philosophers, social and behavioural scientists have argued that behavioural public policies should be ethically evaluated on a case-by-case basis. This seems a sensible starting position, and one that urges public servants to be keenly sensitive to the specific context in which they are acting.

Central to appropriateness is thinking through what the alternative policy options and tools might be, asking whether a behavioural intervention is justified in the particular sector or context, and analysing the drivers of the problem to be solved.

A set of culturally specific assumptions about the desirability of universal civic participation here did not translate well into how the policy played out in practice

Sometimes the short-term fix to a particular issue is not the most effective or sustainable. One example is where a behavioural intervention in Mali to increase civic participation actually led to a decline in female participation, because discussing the issue in this specific context surfaced gender inequalities, with women actively discouraged by men to participate as a result of the intervention. A set of culturally specific assumptions about the desirability of universal civic participation here did not translate well into how the policy played out in practice.

Using behavioural insights therefore requires a strong understanding of the contexts in which a policy is being implemented. Policies that target behaviours already make an assumption that the problem is caused by error-prone behaviour, rather than other, more substantive constraints. Researchers have highlighted the detrimental effects of taking a de-contextualised approach in health promotion and tackling climate change.

Though it’s rarely talked about it, uneven power relations are at the heart of every kind of public policy intervention.

Who gets to decide what the “public good” is or policy goal? What is the legitimacy of the behavioural expert, policymaker or politician, and how are they held accountable for their decisions? The systems and structures that shape people’s opportunities and motivations to change their own behaviour are significant here.

The example of how the UK’s Behavioural Insights Team worked with unemployment services to help job seekers develop the "right" kinds of positive attitudes towards employment is one case where the power inequalities between job seekers and the Department for Work and Pensions (DWP) were laid bare. The scheme was intended to support people to find work, but it ended up being a controversial intervention. Anti-workfare activists and psychologists saw it as a re-imagining of unemployment as a psychological pathology, which in turn drew criticism from DWP spokespeople who claimed the activists had misunderstood the intervention.

The intervention here seemed to imply that the attitudes, personality and cognitive capabilities of individual citizens were at fault, rather than contemporary employment conditions. This was seen by some as an abuse of power. 

Some say that for these last few years we’ve been living in a "post-truth" world, in which misinformation spreads rapidly and experts are distrusted. It is therefore unsurprising that the seeming return of expertise in the coronavirus pandemic has been widely welcomed as a return to normalcy. However, the diverse national responses to the crisis show that not all types of expertise are equal, not all experts agree, and there is no simple pathway from evidence to policy action.

Ethical conduct in policymaking therefore requires a more nuanced appreciation of what counts as evidence and insight, and whose voices are being heard. For behavioural interventions, this means that policy decisions must be open and transparent and take into account a range of viewpoints. They must also avoid the perils of what is known as "political scientism", whereby the apparently rational and neutral approach of science stands in for the provision of public reasons for a particular course of action.

Behavioural interventions are likely to be more trusted where there is a high degree of participation and where their effects are genuinely empowering

Openness, transparency, deliberation and citizen engagement can actually improve trust in the basic conventions of the scientific method and improve understanding of how competing interests have to be negotiated in policy decisions. Behavioural interventions are likely to be more trusted where there is a high degree of participation and where their effects are genuinely empowering.

The development of "nudge plus" and "boosts"as transparent, reflective and autonomy-enhancing behaviour-change techniques are useful in setting out how this can be achieved.

Unlike the precepts of "nudging", these approaches do not assume that people’s decision-making capabilities are flawed. Instead they see human behaviour as intuitive, context-specific, taking mental shortcuts as a way to make a particular course of action easier. Policies based on these approaches would encourage more reflection, group deliberation, and support "more-than-rational" rather than "non-rational" modes of thinking.

Policies designed on this basis aim to empower, educate and enable people to adapt in an uncertain world. Never before has it been so important to match up citizens’ capabilities with ethically designed policies and strong civil society institutions in this way. 

This article was originally published on Apolitical