Doing and thinking: the ethics of 'nudge'

 

 I attended this IPA event last night. Rory Sutherland and Matthew Taylor talked about the value of understanding behavioural economics for both advertising and social change. One question that came up was ‘What about the ‘dark side’ to this?’ – the idea that in knowing more about human psychology and brain functioning we can manipulate people to undue ends.

 

The first thing to say is that the power of manipulation that behavioural economics bestows is not of the 1960s-commie-paranoia-mind-washing variety. It is much weaker than that. Moreover, I actually think this question is less interesting than it sounds – its apparent power presumes something similar to the following argument: ‘don’t develop genetic therapies for diseases, because some mad scientist might create a super-race of beings and take over the world etc.’ This is faulty reasoning: the problem is not the genetic therapy but the mad scientist with the requisite power, resources and social influence to ‘take over the world’. And that problem is one we face and deal with all the time. (Think of Rupert Murdoch.)

 

A more interesting question concerns the ethics of ‘nudging’. The latter is based on two behavioural principles: that people are very bad at making decisions concerning long-term outcomes (such as their health), and that situations or ‘choice architecture’ (both social and physical) heavily influence behaviour. (The single most effective reducer of obesity? Using smaller plates.)

 

The ethical question is this: are we paternalistically interfering in people’s lives when we nudge them, or are we enabling them to do what they already want to do, but find hard? Nudging has been labelled ‘libertarian paternalism’ and the question is whether we think the two terms complement one another or that paternalism trumps.

 

One response is  the one made by Sunstein and Thaler, that when people are nudged they are simply being aided to choose their ‘preferred preferences’. For example, if the healthy eating options are first in the canteen-line, a person might opt for them at first unthinkingly – she’s hungry, she goes for the first thing she sees. But then she might realise what she is doing. If the nudge continues to work, this is because her behaviour has been guided by a ‘choice architecture’ that enables her to more easily do what she already wanted to, but found difficult. Or so the argument goes.

 

I think this is fine. But it suggests that the only way a piece of behaviour can be encouraged without paternalism is if that behaviour chimes with an antecedent deliberated-upon choice. And this seems to me like an unwarranted bit of neo-classical prejudice. What is wrong with guiding someone to a behavioural pattern that is not the result of an antecedent consciously-endorsed decision?

 

Imagine that I have always pooh-poohed the idea of healthy eating, even though I’m overweight. ‘Life is for the living’, I have said many a time: ‘I enjoy my food, and being happy is the most important thing in life.’ One day, the canteen at work puts the healthy-food options first in line. I start eating more healthily. After a few weeks, I realise the nudge has taken place – I realise what I’m doing every lunchtime. But I think, ‘actually, it’s not so bad, this healthy-eating lark; I feel better, I think I might stick with it.’

 

What has happened here is that a capability has been built - a capability to feel the benefits of healthy eating, something I could not do before. My ‘preferred preferences’ have changed rather than simply been realised. The direction of change has been from behaviour to attitudes. (I did something without thinking; after a while realised I liked it; I consequently changed my ‘higher order’ preferences about what kind of life I want to lead.)

 

Is this unethical (overly paternalistic)? I don’t think so. I can still go back to unhealthy eating if I want, but now that choice is informed by an ability to viscerally feel the benefits of healthy eating, and is much the better for it.

 

What I am trying to get at is an implicit prejudice even Thaler and Sunstein display – that all ethical behaviour-change results from behavioural patterns conforming to attitudes we already hold and have consciously endorsed. But another way that nudges can work is through developing a behavioural capability. When this occurs someone can end up consciously endorsing a behavioural pattern they actively rejected before. The nudge, through guiding behaviour, has allowed a person to see the options clearly and make a properly informed choice for the first time.

 

What stops us from having this equally benign aspect of libertarian paternalism in view is the implicit prejudice that only behaviour-determining choices can be ethical, not choice-determining behaviours – that only the former do not override our free will. But that just falsifies how we actually make lots of informed choices – not by disembodied choosing, but by realising after a period of time the benefits of doing something.