Riding a horse on public roads is a risky business. The British Horse society reports that more than 4000 horse riders and carriage drivers were hospitalised in the UK between April 1st 2013 and March 31 2014 following road traffic accidents. To mitigate the danger, some horse riders have taken to wearing reflective jackets that mimic the uniforms worn by mounted police, except emblazened with the word "Polite" rather than "Police".
Obviously, the idea is that motorists will see the jackets, think, "Uh oh, police!", and then slow down, leading to safer roads for horse riding.
These jackets seem to work well. A study by the British Horse Society estimated that drivers gained three seconds of reaction time as a result of their reduced speed. However, not everybody is happy. There have been complaints from drivers that riders are impersonating police officers, and the Association of Chief Police Officers has warned that the jackets might break the law (specifically, Section 90 of the Police Act 1996).
Whatever the legal situation, there is no doubt that deception is the key to the effectiveness of the jackets. This results in some interesting questions that are relevant for how we think about nudging.
Richard Thaler and Cass Sunstein originally defined a nudge as:
any aspect of the choice architecture that alters people’s behavior in a predictable way without forbidding any options or significantly changing their economic incentives. To count as a mere nudge, the intervention must be easy and cheap to avoid.
In these terms, the jackets qualifiy as a nudge: drivers will predictably slow down, but, other options, such as speeding up, are not precluded, and no economic penalties are incurred in avoiding the nudge. Almost certainly, the jackets also meet the criteria of libertarian paternalism in that their goal is to influence the choices of affected parties with the aim of making them better off (i.e., to make it less likely there will be an accident), yet they involve no coercion.
However, while it is true that the jackets leave all choices in play, their psychological effects are both hard to avoid and based upon a deceit. This brings to the fore the issue of rational autonomy, which refers to the control that somebody has over their own evaluations and judgments. The worry is that if people are not in control of the deliberative process that underpins rational decision-making, they are deprived of their ability to function as fully autonomous moral agents, an essential component of human dignity. In the case of the jackets, the driver has very little control over their decision to slow down. They will react before they realise - if they do - that the jacket doesn't signifiy what they thought it signified. In this sense, the driver's decision to slow down is under the control of the rider, the choice architect, who uses a perceptual trick to manipulate the driver's responses.
Maybe there is nothing particularly worrying here. It seems very likely that most drivers would agree that they should reduce their speed while passing a horse rider even if they would not normally do so. The jackets just help them get to where they want to go. Moreover, the utilitarian case for the jackets seems strong - lives are saved and nobody is really any the worse off.
But this line of thought requires us to bite some bullets that perhaps we should not be prepared to bite. It suggests that it is right to hijack a moral agent's decision-making apparatus, regardless of their own particular goals and interests, so long as most people would accept the legitimacy of the targetted behavioral change. But there is a lot that should make us nervous about such a suggestion. If deception is acceptable, then why not subliminal advertising? If a perceptual trick is okay, then why not false information? There might be distinctions here, but it is not clear that they are morally decisive if we accept it can be legitimate to nudge a person's behaviour in a particular direction using a deception designed to subvert their rational decision-making.