11 October 2017

Why nudges are nothing to fear

By Christopher Snowdon

Richard Thaler, a professor at the University of Chicago’s Booth School of Business, was announced as this year’s winner of the Nobel Memorial Prize in Economics on Monday. Thaler is best-known for his contributions to the field of behavioural economics, alongside the 2002 winner Daniel Kahneman and former presidential adviser Cass Sunstein.

Economists have never really believed that people are ruthlessly self-interested, perfectly informed robots who are constantly balancing costs against benefits. There are not enough hours in the day for us to be perfectly informed about every decision we make and so we use shortcuts (heuristics) to help us reach an outcome that might not be perfect, but is good enough (satisficing). We use rules of thumb and best estimates. We rely on recommendations from friends, and trust brands that have served us well in the past. ‘It makes far more sense to say that people display bounded rationality than to accuse them of “irrationality”’ says Sunstein. This is not necessarily a bad thing. It is perfectly rational to settle for less than best if it saves us time and effort, particularly when the costs are low. It would be unreasonable to spend a day researching which box of matches to buy, but it could be time well spent if we were buying a house.

But what if our mental shortcuts and human frailties stop us getting what we really want? Since the 1970s, the field of behavioural economics has shown that people fall foul of a number of cognitive biases which lead to bad decision-making. These subtle but common irrational responses have been said to undermine John Stuart Mill’s faith in reason and justify a new wave of paternalism.

In 2003, Thaler and Sunstein wrote a widely cited article for the American Economic Review, the discipline’s top journal, on “libertarian paternalism.”

The ideas in that article went mainstream in 2008 when Thaler and Sunstein published their influential best-seller Nudge. In it they argue that inertia, in particular, exerts a powerful influence over people. If one option requires conscious effort while the other doesn’t, we are more likely to do nothing and settle for the default. And yet the default option does not necessarily reflect our preference when we are asked directly (that is, in a yes/no question with no default option).

Inertia is just one of the cognitive biases that lead to suboptimal decision making. Thaler and Sunstein marshal an impressive array of evidence showing that people’s actions can be significantly affected by seemingly minor details in what they call the ‘“choice architecture”. Given that default options are inevitable, Thaler and Sunstein argue that they should be designed to “influence choices in a way that will make choosers better off, as judged by themselves”.

Behavioural experiments have shown the effectiveness of all sorts of interventions in the choice architecture. Painting a picture of a fly on a urinal gives men something to aim at and reduces “spillage”. Automatic enrolment of employees into pension plans (with an easy opt out for those who don’t want to join) increases uptake and gives people more savings in later life. Sending people a letter telling them that their money is needed for “vital public services” makes them more likely to pay their income tax bill on time. Getting a patient to write down the details of their doctor’s appointment (rather than having a member of staff do it) makes them less likely to forget about it. Placing healthy food at the front of the counter in cafeterias makes it more likely to be picked up. Even putting a light above the fruit bowl can significantly increase the number of people who opt for fruit in school canteens.

If liberty is defined as the “absence of legal coercion”, as Joel Feinberg puts it, it is difficult to argue that any of these interventions are illiberal. Thaler and Sunstein lay out clear criteria for nudging to ensure that freedom of choice is respected. They define a nudge as a change to “any aspect of the choice architecture that alters people’s behaviour in a predictable way without forbidding any options or significantly changing their economic incentives”.

In other words, the nudge must be evidence-based (“alters people’s behaviour in a predictable way”), cannot involve bans (“without forbidding any options”) and cannot make the activity less enjoyable or more expensive (“[without] significantly changing their economic incentives”).

Nudge theory has been criticised for being manipulative but, as Thaler and Sunstein repeatedly point out, it is no more manipulative that any other attempt by governments, businesses and individuals to influence our decisions. Choice architecture is everywhere and eradicating defaults is not an option.

The criticism that nudging is manipulative implies that there is some sort of natural choice architecture with which the government is meddling but, as Sunstein points out in Why Nudge?, when default options exist it is “not because God or nature has so decreed” but because somebody has chosen them. For nudge theorists, it is better if the architecture helps us follow our desires rather than dragging us away from them. School canteens have to place the food somewhere, so why not place the healthier food at the front? It would be no less “manipulative” to place it at the back. Similarly, it is not obvious why an automatic opt out of a pension plan is less manipulative than an automatic opt in.

Businesses use nudge tactics all the time. Online subscriptions are often renewed unless we explicitly cancel them and libertarians have no problem with this because consumers are free to take their business elsewhere. The market will ultimately punish any company that gets a reputation for sharp practice. It could be argued that people expect a degree of manipulation and salesmanship from business but would feel patronised if the government used the same tricks in an effort to save us from ourselves. Hausman and Welch suggest that “the cacophony of invocations of irrational responses by non-governmental agents” is made tolerable by “the limits to its effectiveness and the extent to which these invocations conflict with one another and cancel one another out”. By contrast, exploitation of psychological foibles by monopolistic government is “a form of disrespectful social control”.

Since Thaler and Sunstein insist that nudging should be done openly and with full publicity, it is certainly possible that some people will feel humiliated and browbeaten by the knowledge that government is subtly influencing their personal decisions. In the view of Isaiah Berlin: “Paternalism is despotic, not because it is more oppressive than naked, brutal, unenlightened tyranny, nor merely because it ignores the transcendental reason embodied in me, but because it is an insult to my conception of myself as a human being, determined to make my own life in accordance with my own (not necessarily rational or benevolent) purposes, and, above all, entitled to be recognised as such by others.”

These concerns, which echo Mill’s, cannot be lightly dismissed, but they do not necessarily apply to nudge theory. Although government has a monopoly, Thaler and Sunstein’s agenda of “libertarian paternalism” makes it easy for people to “take their business elsewhere” by ignoring or opting out of the nudge. Besides, many of the suggestions in Nudge are aimed at the private sector. They do not recommend, for example, that governments force cafés to display healthy food at the front counter. When government action is required for the nudge, it is when the government is already involved. Tax collection, organ donation and driving licences are all within state control. If they can be made more effective and efficient by using the same persuasive techniques that are second nature in the private sector, why not do so?

Some have argued that libertarian paternalism is not, in fact, libertarian and it is true that Thaler and Sunstein occasionally seem to think that they know what is in a person’s best interest. But it could equally be faulted for not being paternalistic. Some of their most famous nudges involve no paternalism whatsoever. The fly on the urinal, for example, is primarily intended to benefit toilet cleaners rather than toilet users. The beneficiary of organ donations is the live recipient, not the dead donor. A reminder to pay one’s income tax could benefit the individual if it helps avoid a fine, but the main beneficiary is the tax office. Reminders, warnings and education are not paternalistic because, as Hausman and Welch note, “providing information and giving advice treats individuals as fully competent decision makers”. Nudges of this sort may well help people pursue their goals but that does not necessarily make them paternalistic. They can be justified by mainstream economics.

The British government has been experimenting with behavioural economics since 2010 when the Behavioural Insights Team was set up under David Cameron. Popularly known as the Nudge Unit, it began life with the commendable pledge to close itself down if it did not produce a tenfold return on its £500,000 start up costs. In his book Inside the Nudge Unit, the team’s director David Halpern describes a string of nudging successes which only serve to demonstrate the limits of “libertarian paternalism”. The most significant of them include adding a note to income tax reminders telling the recipient that “most people pay their tax on time”; adding a photo of the driver’s car to unpaid car tax bills; sending debtors a text message to tell them that the bailiffs are due to appear on their doorstep, and offering people a loft clearance service to increase uptake of subsidised home insulation.

All these nudges had the desired effect and, Halpern says, brought in tens of millions of pounds. Since they cost little to implement, they were worthwhile innovations but most of them relied on little more than a change in presentation. Only the loft clearance scheme was paternalistic (the others were principally for the benefit of the government), but it was not really a “nudge” since it changed the costs and benefits, and it was not libertarian because it forced taxpayers to pay for other people’s home improvements.

The most damning criticism of the nudge project is not that it is illiberal, but that it is insubstantial in the context of the big issues facing government. If one strips out all the nudges that are not paternalistic, not libertarian and not trivial, there is little left of the libertarian paternalist agenda.

It is precisely because Thaler and Sunstein are reluctant to use state coercion that the implications of nudge theory for public policy are so limited. There are plenty of nudges that can be adopted by businesses and individuals and yet nudging in its pure form – with the caveat that the nudge should be easy to ignore or avoid – has fewer practical applications for government. Nudges can be effective in reminding people to do things, but they do not offer solutions to the major political challenges of the day, and the assurance of a hassle-free opt out will never satisfy single-issue campaigners who see bigger gains to be made from compulsion.

The truth is that most governments are more paternalistic and less libertarian than the nudge theorists. If the principles of Nudge were rolled out across government, many existing laws would have to be repealed and few new laws would be made. From a libertarian perspective, it is unfortunate that Thaler and Sunstein do not apply their principles to such issues as gambling and narcotics, where US law goes far beyond subtle nudges. One can only speculate as to what legislative programme would emerge if a society was started from scratch based on nudge theory, but it would surely be more libertarian than any country currently in existence.

This is an extract from Christopher Snowdon’s forthcoming book ‘Killjoys’, published by the Institute of Economic Affairs on November 10th 2017.

Christopher Snowdon is the Head of Lifestyle Economics at the IEA