Good choice architecture

Professor Andrew Parkes, chief scientist, Behavioural Studies, TRL, translates the nudge agenda into improved transport safety

Discussions between academics at the Business School and the Law School of Chicago University 2007 may seem an unlikely place to start an article on new directions in transport policy in the UK in 2011. However, those discussions between Richard Thaler and Cass Sunstein, and the subsequent international bestseller ‘Nudge: improving decisions about health, wealth and happiness’ have had an impact far beyond the expectations of those authors.

Most of us in academia and the research community are content with a diet of regular conference papers, the occasional journal article and, if lucky, a book or two in a career that recognises that usually we are speaking to quite a small community of like minded scientists. Very few of us make the important step of making our research really accessible to the public or to policy makers and legislators. The authors of Nudge now talk to presidents and prime ministers directly.

So what is so important about this book? And what does it mean for transport, here, now, in the UK?

An international phenomenon
It is important because it is a book on economic theory and decision making that has crossed over into the mainstream, become an international phenomenon, and the source of the phrase ‘the nudge agenda’. It has had such a direct impact on the governments of both Barack Obama and David Cameron that they have put in place special think tanks to see how the agenda can be put into policy.

Within 10 Downing Street we now have a Cabinet Office Behavioural Insight Team that includes important figures such as Sir Gus O’Donnell (cabinet secretary) and Steve Hilton (director of strategy). This team, sometimes called the Nudge Unit is already influencing policy initiatives across many government departments. They are advised by Richard Thaler who encourages ministers to become, in his parlance, choice architects who, instead of forcing people to behave better through legislation and regulation, guide them to socially acceptable choices through careful presentation of options. The immediate appeal of this approach is obvious – many policies based on the nudge agenda – require no new or extra legislation, cost nothing and so do not put any burden on the tax payer.

Tried & tested
Can anything be so simple? Can it really be so different to what we had before? Well, according to my interpretation at least, those answers are yes, and no. Yes, the nudge philosophy is something real and tested and it works. No, the term is new but the approach has been tried in a variety of contexts before – we just hadn’t thought of calling it nudge.

Interestingly, Dr David Halpern, the Nudge Unit’s director, has been here before – literally – having served in Tony Blair’s Strategy Unit. He promoted a related approach that has used MINDSPACE as a useful mnemonic: Messenger (whoever gives the information influences how it is received), Incentives, Norms, Defaults, Salience, Priming (subconscious cues) Affect (feelings from emotional associations), Commitments and Ego (and how to reward it for doing what we want).

What previous authors, experts and commentators have failed to do in comparison to Thaler and Sunstein is first, get the timing right – promote a cost neutral, non-regulatory approach at a time of economic hardship, and second, almost as important, come up with such an annoyingly memorable term as ‘the nudge agenda’.

To make my position clear: I believe in nudge (not at the exclusion of all else) and think it is essential in any thoughts about improving traffic safety. However, I think there is a danger of going too far and governments (any government) abrogating certain key responsibilities for regulation where it will still be needed.

The principles
Let’s look at what Thaler and Sunstein actually said and how it applies to us. They have brought together ideas about economics, marketing and social psychology and synthesised them into six, very well thought through, principles of good choice architecture. They arranged the six principles into their own mnemonic which, with one small cheat, comes out as:
• iNcentives
• Understanding mappings
• Defaults
• Give feedback
• Expect error
• Structure complex choices

They explain each of these principles in detail and provide a host of examples of good choice architecture in practice. Unfortunately the examples come from the areas of health psychology, insurance, pensions, gambling, organ donation and school meal choices.

The only example with even a tenuous transport link is the often quoted, but somewhat trivial, example of the design of the urinals at Amsterdam airport. Here the authorities were concerned about the perennial problem of some gentleman being rather too inaccurate in their aim and the consequent standards of cleanliness of the facilities. The solution was to encourage more concentrated and accurate aim by providing a target in the guise of a picture of a fly towards the centre of the porcelain. Most men found the temptation to aim at the fly overwhelming with the reported reduction in splashes landing on the surrounding tiles being as much as 80 per cent.

It worked, it follows the principle of nudging people to do something better, took a step away from the heavy handed approach of informing and warning, was very low cost, has caught on like wildfire round the world and made the inventor a wealthy man. But though amusing and informative to a degree, it doesn’t immediately become apparent to me as an applied scientist concerned with road injuries, how to use this example to improve UK road safety.

Car safety

In trying to explore examples of real relevance to the business sector and transport I thought back to an important report from a few years ago. In 2004 there was a House of Commons Transport Committee report that looked at ‘Cars of the Future’ and reported that "half of all fatal and disabling injuries could be avoided if all cars provided the impact protection of the best cars in the same class".

This doesn’t mean all cars should have the impact protection of, say, a top line Mercedes limousine. The conclusion is much more interesting. If all super minis afforded the same protection as the best super mini, if all 4x4s were as good as the best 4x4 and so on, we could have halved the frequency of fatalities based on 2004 rates.

Well, in fact we have made huge improvements in the fatality figures in recent years and much of the improvement is undoubtedly due to clear developments in the safety performance of the vast majority of new vehicles in the fleet. Improved primary and secondary safety features are percolating throughout the model ranges rapidly.

In light of the talk about behavioural change and choice architecture it is important to think through why this improvement in safety features has moved so rapidly recently, and why it didn’t happen around ten years ago at least. Most of the systems so widespread now were first demonstrated in some form in the mid 1980s or earlier. The government has played an important role and some legislation has played its part, but the real tipping point has come from a change in market demand – behaviour has changed – buyers choices have been nudged – even before the phrase nudge agenda was born.

Consumers have come to demand safety and manufacturers are competing to satisfy that demand. In the early 1990s this wasn’t so, and it is interesting to look at what has actually changed our behaviour.


From the 1970s onwards several European governments were showing interest in improving and standardising testing procedures for crash performance of cars. This was resisted by the majority of manufacturers. In 1994 the UK Department for Transport looked at setting up a new car assessment programme (NCAP) which could later expand across Europe. Euro NCAP was set up by the Transport Research Laboratory (TRL) for the UK DfT with input from France Germany, Sweden, the Netherlands and Spain.

First results were heavily criticised by many car manufacturers who focused on the tests and the rating scales produced; claiming the tests to be so severe that no car could achieve 4 stars for occupant protection. However, Volvo with the S40 proved them wrong, being the first to achieve 4 stars, with the Renault Laguna following soon afterwards and being the first to achieve the maximum 5 star rating.

The scheme has been further developed over the years and there is discussion about the feasibility of including more comprehensive tests of primary safety features too. No doubt the scheme will continue to have vigour and influence the market. The scheme itself cannot force international car manufacturers to change their products but it certainly provides a substantial nudge.

Consumer choice

In a complex information environment where it is very difficult for the individual buyer to find out the relevant information to compare the performance of different vehicles directly, the use of the simple 5 star system with supporting performance information that maps directly on to the understanding of the intended audience enables the structuring of complex choices. When the only intelligible information for buyers centred on acceleration times and fuel use, it was understandable that buyers and producers alike focused on performance issues far more than on safety design. Both buyers and manufacturers now get incentives and feedback from this scheme.

The buyer is assured they are to spend their money appropriately, the manufacturer aims to ensure their model at least matches the performance of the rivals. Although we might expect errors in understanding or questioning from the normal buying public, the information system defaults to the important range of indicators. The pressures in the system are now always towards better safety standards not lower ones. The government primed the system, the consumer has been nudged in the right direction and the manufacturers now are fully engaged in deploying leading edge safety systems in their cars.

Environmental issues
Similar principles must influence our thinking about clean vehicle technologies for the future. Manufacturers will solve the many technical problems facing them. But at some point they will need a market to exist for their products. Very careful attention needs to be paid to the choice architecture presented to the customers for the next generation of vehicles, especially if alternative fuels are shown to be the way forward, in order to provide the right information, incentives and feedback to allow the market to grow naturally. Manufacturers are unable to create the new market effectively without support and stimulus from central government.

The current government dislikes too close a focus on targets; indeed David Cameron in his first week in office said "if you want to set new targets, set new controls, impose new rules, don’t bother because you’re likely to get the red light!" Without getting embroiled in a discussion of what exactly is a target and how it differs from a desired outcome, the change in climate is clear.

The most recent example of the government’s thinking, and one where the nudge agenda is referred to directly, is the Strategic Framework for Road Safety published in May 2011. In the introduction Philip Hammond, secretary of state for transport, said: "Much of the harm and cost is avoidable and it is not an inevitable consequence of road transport. We believe that further measures can be taken that will provide high value for money but we are clear that improvements in road safety need to be robustly analysed, considering all costs and benefits, the pressures on spending and the opportunity costs....our approach, where possible, should be based on making it easier for road users to do the right thing – improving education and training instead of resorting to more bureaucracy, targets and regulation."

Research direction
Some commentators have assumed that a combination of the nudge and the localism agendas will lead to central government stepping away from their role in controlling the direction and pace of road safety research. This need not be the case; in the same strategy paper Hammond attests "...there is still a crucial role for national government in providing leadership in road safety, delivering better driving standards and testing, enforcement, education, managing the strategic road infrastructure and through research and the collation and provision of public information to support local delivery".

The point that the transport community must adapt to is that the climate and the tools for safety improvements have changed. Types of research to be supported in the future are focused differently; the degree of analysis of cost effectiveness is increasing and the primary importance of road users choices are brought into greater relief. Choice engineers, your time is now.

For more information

Please register to comment on this article