PsyFi Search

Tuesday 21 February 2012

Have You Been Nudged Today?

Quills and Nudges

The British Civil Service is not a body known for radical change; it’s never quite got over having to give up the quill pen and tends to view with suspicion newfangled ideas like computers and women in the workplace. Despite this it’s currently at the heart of a large scale experiment on the use of techniques from behavioral science, aka, “nudge” theory.

Sitting at the center of UK government is the Behavioural Insights Team (BIT), which is tasked with using behavioral approaches “as an alternative or complement to regulation or bans”, while achieving a ten-fold return on the cost of the team. They’ve just released the results of their first set of trials on fraud, error and debt, and these suggest they’re likely to continue, as the potential savings look like they’ll be quite significant. Wherever you are, you can expect to be nudged, prodded and cajoled into being a better person; or at least one that pays their taxes and turns up to doctors' appointments on time.

Heath, Choice and Energy

The remit of the BIT is wide. It’s involved in healthcare, where it’s introduced required choice into organ donation in the driving licence program: forcing people to make a choice, rather simply making opting-out the default action is estimated to increase organ donors by a million people over the next few years. Meanwhile, trials involving patients making an active commitment to attend appointments, coupled with “normative” information, aka social norms, showing how many other people usually turned up reduced no-shows by over 30%.

The BIS is also central to a range of “Consumer Empowerment” programs, as detailed in this paper on Better Choices, Better Deals, where policies are being designed around economic insights. These aim to overcome problems with information asymmetries – where business has more information than the consumer, search space issues – where too much competition makes finding the right deal too difficult, and consumer inertia, often connected by pricing models designed to take advantage of status quo bias.

Meanwhile various behavioral methods are being trialled to improve energy efficiency. Using small scale payments up front, for instance, aims to take advantage of people’s short-term myopia about rewards, and mechanisms using social proof introduce discounts to encourage groups of people to make energy improvements. In fact, the BIT seems to have fingers in many pies as last year’s Annual Update reveals.

Fraud, Error and Debt

Its most recent publication is directly of interest to investors and savers as it tackles Applying Behavioral Insights to reduce fraud, error and debt. The value of this to the UK is plainly stated up-front:
“£21 billion is lost to fraud in the public sector, a further £9.6 billion is lost to errors, while £7–8 billion is lost in uncollected debt. These significant sums of money are ultimately borne by UK taxpayers, so any measures that reduce fraud, error and debt in a cost-effective way are to be welcomed.”
The paper identifies seven lessons from behavioral science to reduce fraud, error and debt. We might simply take these as a manifesto for nudge approaches applied to government: they’re also revealing about ways in which behavioral science may be practically applied.

The simplest lesson of all is “Make it easy”: research on tax compliance regimes across multiple countries reveals that the best performing countries pre-populate their tax returns, rather than leaving people to figure this out for themselves. As we’ve seen before (see: Retirees, Procrastinate At Your Peril), people will tend to procrastinate over stuff that’s too difficult. Other research has shown US college enrolment rates rising significantly following a similar intervention. Moreover the UK is following other regimes by building smart phone apps to streamline the tax management process; a trend we’re going to see more of in the coming years, at least until someone figures out that publishing personal details through an insecure handset may be a teeny bit risky.

Personalised Emphasis

Secondly, the BIT points out that emphasising key messages by drawing their attention to the relevant points is likely to improve compliance. Obvious though this might be it’s not something that most government form designers have ever given any thought to, as anyone who’s tried to fill one out can testify. This focuses on the concept of “cognitive ease”, Daniel Kahneman’s insight that it’s possible to simply the information processing task if data is presented clearly and simply, which in turn leads to improved response levels.

A further insight being used is that personalising messages significantly improves response rates. When Randy Garner experimented by adding personalised, handwritten Post-It notes to survey form requests the response rate soared from 48% to 76%. When the researcher added their initials and a “Thank You” the response rate went even higher – folk knowledge which any tip seeking waiter commonly takes advantage of, but which we generally consider under the concept of reciprocity (see: When A Dollar's Not Just A Dollar): people respond better to other people than to faceless bureaucracy. 

Honesty, Social Norms and Incentives 

Encouraging honesty by intervening at key moments is another area of study. As Dan Airely has shown (See Gaming The System, Game On, Basel III) we’re nowhere near as honest as we like to think we are, but we can be encouraged to be more truthful by appropriate, timely interventions – merely moving a signature declaration to the start of a form prompted people applying for car insurance to report driving 10% more miles than previously.

Broader use of social norms and conformity is also being investigated – we're strongly influenced by people around us so reminding or informing us of what other people are doing is one way of encouraging more compliance. A study in Minnesota showing that tax evasion was actually rather low increased rates of voluntary tax payments and this has been replicated in the UK in one of the BIT’s trials. Of course, as the report indicated, where there are significant levels of non-compliance you probably don’t want to emphasise the social norm and you should look to other forms of persuasion.

Incentives come under scrutiny with research suggesting that positive rewards, rather than the more usual negative punishments, may be effective in motivating better behavior. In a neat inversion of rational economic behavior (see: The Lottery of Stockpicking) the report recommends using lotteries to encourage prompt responses – so that you’re only entered for the big prize if you get your replies back by the deadline. This draws on a range of research including a neat example from China where the authorities printed lottery numbers on the back of receipts in order to reduce black market payments.

Test, Learn, Adapt

A final area being looked at is whether it’s possible to frame tax avoidance in terms of losses to specific public bodies, such by linking failure to pay local taxes to streets not being kept clean. This draws on the research of Uri Gneezy who has shown that deception is harder to maintain if you understand the personal consequences of your decisions:
"People care how much they gain from a lie, but also about how much the other side loses (this unselfish motive diminishes with the size of the gains)."
The research is also looking at how using publicity to remove the anonymity of dishonest acts may improve behavior – while bearing in mind that if you create a social norm that bad behavior is acceptable you may increase the behavior, rather than improve it.  

Show Them The Money

The BIT is now conducting a series of trials based on these insights, looking at improving tax compliance, reducing the use of unregistered cars, improving fine collection and discouraging fraudulent benefit claims. Early results are encouraging:
“Trial 1 advanced £160 million of tax debts to the Exchequer over the six-week period of the trial, the two behavioural letters in Trial 2 brought in over £1 million from doctors in additional yield to HM Revenue and Customs, while Trial 6 saved Manchester City Council up to £240,000 in council tax discounts.”
Perhaps most encouraging is the entirely sensible approach being taken to implementation of these ideas, which are largely untrialled in these kinds of broad settings, of “test, learn, adapt”. Of course, it remains to be seen whether a large scale implementation of behavioral research will yield the hoped for benefits, and whether these will be maintained in the long term. What’s for certain, though, is that cash-strapped governments the world over will be looking closely at these results. Your life is not your own: expect to be nudged by your government any time now.

Related articles: 

No comments:

Post a Comment