‘Nudging’, or using insights from behavioural science to influence behaviour, has become popular with policymakers in the UK and elsewhere. However, some take issue with the manipulative aspect of nudges. Exploring the ethics of nudging, Martin Wilkinson defines the circumstances where it may be considered manipulative and whether this is bad in itself.
Nudging uses the clever tricks of modern psychology and economics to manipulate people. We don’t like manipulation when it’s done to sell us things; we shouldn’t like manipulation when our governments do it to us. Nudging uses state of the art insights of psychology and economics to help better achieve either what we want or what benefits us all. Nudging doesn’t coerce anyone or even cost much money – what’s not to like?
At one extreme, you can have a hysterical accusation of mind control; at the other, a bland reassurance that nudging uses only the informational aspect of tools. The truth about nudging doesn’t exactly lie in the middle. It’s more accurate to say that some nudges manipulate in some circumstances and others don’t. Which nudges when, and why does it matter?
Some nudges obviously don’t manipulate. The government might want people to spend less on credit and people might indeed spend less once they know what the cards cost them. Suppose the government makes credit card companies tell customers clearly and visibly how much they pay in fees and interest. It may be bad for the companies but it doesn’t manipulate their customers to have otherwise-baffling information given in an easily understandable form.
Some nudges obviously do manipulate. Suppose the government makes job-seekers take bogus psychometric tests to boost their resilience and the job-seekers don’t know they are bogus. The job-seekers have been deceived and deceit is the paradigm of manipulation.
What’s wrong with manipulation?
People can be manipulated in their own interests. People can even be manipulated into doing what they would do anyway. The ethical problem of manipulation is not intrinsically in what it gets people to do, it’s in how it works. Manipulation has a manipulator – an agent who wants to control someone else’s beliefs and actions in some underhand way. Victims of manipulation have their autonomy violated. In a way, manipulation is worse than coercion. At least you can make up your own mind when the government taxes, regulates, and makes laws backed up by punishment. Your options may be changed but how you react to these options is up to you. With manipulation, you have lost control of your beliefs, your actions, or both. That’s why the supporters of nudging can’t just say ‘nudging is libertarian and no one is coerced’. If the nudges are manipulative, that’s bad too.
When do nudges manipulate?
But then not all nudges are manipulative. How can we decide which ones are and which aren’t? Let me show why deciding is not easy before being more constructive.
First, the same techniques may be manipulative in some cases and not others. Take framing. Framing shapes thoughts and sometimes action. You might think it manipulates when it’s used to scare people, as in ‘X doubles your risk of a stroke’ instead of ‘X increases your risk from 0.001 per cent to 0.002 per cent’. But some framing is essential to making use of countless pieces of information. Someone who wants you to think of whether nudges manipulate frames the discussion simply by raising the question, and what’s wrong with that?
Second, whether nudges manipulate depend on how they work (assuming they do work). My water bill in New Zealand tells me how much water is used by the average family of various sizes. Some people might reduce their water consumption to comply with the implicit norm, the sort of motivation stressed by nudgers. But when I found out we used more water than the norm, I looked for a leak in my pipes and reduced my consumption by stopping it. You might think the norm-compliers are manipulated, but people like me (in this case but maybe not others) acted in a standard rational choice way.
So we can’t say ‘these nudges manipulate because they use manipulative techniques’. Even if we could pick out potentially manipulative techniques, sometimes they still wouldn’t manipulate. So we don’t have a complete answer. The reasons why we don’t, I think, are because ‘nudging’ covers so many different methods and, more philosophically, because we don’t fully understand what manipulation is. But even without a complete answer, we can say something.
Whether nudging is manipulative depends on what the nudgers intend. Are they trying to get their targets to act a certain way? Some nudgers may not want people to act any particular way. Suppose they insist on cooling-off periods for hire purchase. Nudgers may want to protect people from being committed to what they bought on impulse without wanting them to change their minds and send the goods back. Or take opt out pensions. Nudgers may recognise that some people will have good reasons not to want to save or at least not that way and actually want them to opt out. Neither the cooling off nor the opting out is manipulative – not because the techniques don’t manipulate but because the intentions of the nudgers aren’t manipulative. They are not trying to control their targets.
But some nudgers probably are trying to control the nudged. The ones who say ‘we’ll try nudging but if it doesn’t work, we’ll regulate’ clearly want to change behaviour a certain way. If they also have a manipulative method then their nudging would manipulate.
Manipulative nudges are not always wrong
When nudging manipulates, it is wrong in one important way. Does that mean it is wrong, full stop? I don’t think so; some ends justify some means and a minor amount of manipulation may be justified if it produces enough benefit. Painting dashed lines on the road may produce the illusion of speed, slow drivers down, and reduce crashes. If that is manipulation, it’s a tiny amount and it seems justified to me.
I also think some manipulation may not be even a bit wrong – but only in the special case where the target agrees with it. People are often well aware of their own weaknesses and want to get around them. Personally, I would have been glad of an opt out pension scheme since it took me eight years of employment before kicking myself into joining one. If people agree to, or with, the nudges, they might be manipulated but their autonomy remains unviolated. But that would leave all the people who did not agree and all the nudges that people do not notice. When nudgers have manipulative intentions and use manipulative techniques on people who did not agree for the sake of less than major benefits – their nudging is probably unethical.
Note: This article gives the views of the author, and not the position of the British Politics and Policy blog, nor of the London School of Economics. Please read our comments policy before posting.
Martin Wilkinson teaches politics at the University of Auckland, New Zealand. His research is mainly on the ethics of organ transplants. He wrote `Ethics and the Acquisition of Organs’ (Oxford University Press, 2011).