The potential of behavioural economics: beyond the nudge
GUEST OBSERVATION
Behavioural economics is a revolutionary field. Its champions may not have beheaded any royalty, but they have done the academic equivalent by overthrowing a paradigm that ruled our thinking and shaped our institutions.
Behavioural economists have shown that the hyper-rational, self-interested agents of standard economics are mythical creatures. Real-world people are irrational, struggle to exert control over their emotions and impulses, and are social animals with a fondness for fairness.
This paradigm shift radically changes how we think about behaviour, and it seems poised to change our institutions. But how much change will it bring?
Will the influence of behavioural economics on policy be limited to improving the wording of parking fines sent out to misbehaving motorists? Or can it also help us institute bold, sweeping reforms of the type that have helped make Australia equitable and prosperous?
The answer is: it depends.
The policy impact of behaviour economics will be limited – and may even be harmful – if public servants focus excessively on the behavioural revolution’s most famous offspring, the policymaking approach called “nudge”.
On the bright side, the focus on nudge is artificial. It does not flow from the evidence. Some of the most important regulatory reforms in Australia’s history could have been justified using behavioural economics – and such opportunities for reform will emerge again in the future.
The birth of the nudge
The “nudge” policy approach took the United States and United Kingdom by storm in 2008, when Cass Sunstein and Richard Thaler released their groundbreaking work, Nudge.
In the psychological weaknesses uncovered by behavioural economists, Sunstein and Thaler saw opportunity. They argued we can exploit cognitive biases to nudge, or gently guide, people into making choices that are better for them and for the rest of us too — while always leaving them free to choose otherwise.
Want people to sign up for organ donations? Then harness their tendency to procrastinate, by switching from a system where they must choose to opt-in to one where they must choose to opt-out.
Want people to use less electricity? Then exploit social norms by showing them how little their energy-efficient neighbours pay for electricity.
Want people to avoid payday loans? Then make the loans’ downsides salient, by directly comparing their enormous interest rates with other borrowing options.
Confusing nudges with behavioural economics
Nudges may have grown out of behavioural economics, but evidence from that very field suggests that they often fail to meet the hype. Nudges designed to discourage people from taking out ludicrously expensive payday loans, for example, tend to be toothless in practice.
In a new policy brief at the Melbourne Institute, The Potential of Behavioural Economics: Beyond the Nudge I examine what behavioural economics tells us about the promise of both nudges and tools beyond them.
There is, right under our noses, an unnoticed world of behavioural evidence showing that traditional policy tools — like taxes, transfers and mandates — are in many cases superior to nudges.
Rather than throwing out the baby with the bathwater, behavioural economics suggests we should take a fresh look at how we might use traditional policy tools to counter cognitive biases.
A nudge bias
Why has the behavioural evidence on traditional tools been so overlooked? The reasons are largely political.
Nudge was written for a US audience, in a political climate shaped by the anti-paternalistic sentiment that would later coalesce into the Tea Party.
It has become tough, in this context, to pursue rational and equitable reforms of the traditional variety. Government is seen less as the tool the people use to democratically shape the society they want, and more as a suspicious intruder interfering with the benevolent market.
Sunstein and Thaler promoted nudges specifically because they promised a “middle way”: they influenced choices without coercing choice.
This justification turns out to be fanciful. In practice, many non-coercive nudges are too weak to provide serious benefits, and making a nudge effective often means making it coercive.
Take the nudge “success” story of the Save More Tomorrow scheme, designed to counter the strong biases that affect retirement savings decisions. Its central innovation is the automatic escalation of the savings rate over time, so people sign up to save more in future years without losing a dime today.
Though certainly a clever nudge, in practice it increased the national US savings rate by a paltry 0.33%, which will barely make a dent in the enormous problem of old age poverty. Australia’s superannuation scheme leaves it in the dust, with more than twenty times the impact.
Making Save More Tomorrow effective would require increasing its coerciveness — making it compulsory for workplaces to make it the default for all employees, with a much higher default increase in the rate of savings, and a penalty for opting-out — so that it essentially becomes a mandate.
Even then, as I will touch on below, behavioural economics suggests the opt-out would make the scheme welfare-reducing compared to superannuation.
Another example is the host of nudges applied to cigarettes. We plaster packages with gore and health warnings, and we exile smokers to the cold streets where they puff away under scowls and social stigma.
People can still choose to smoke, but this regime of tough nudges is hardly non-coercive. In fact, it is effective precisely because it imposes high costs on smokers.
The point is that nudges are much like traditional tools after all: they are subject to the same trade-offs between agency and welfare.
Behavioural economics for traditional tools
Sunstein and Thaler hoped to make peace between the United States’ warring tribes of progressives and libertarians, and this is precisely how we should understand nudges: they are a political innovation, only sometimes justified by the evidence.
What does the evidence say in full? Behavioural economics has comprehensively falsified rational choice economics. We make serious mistakes, and companies know it.
It would be paradoxical if, after rational choice has been so comprehensively discredited, policymakers responded by relying more on choice-preserving tools like nudges.
Consider the policies that make Australia prosperous and equitable, like publicly-funded universities, superannuation, and financial product regulation. Perversely, the nudge philosophy would reject these as too interventionist.
A more complete behavioural economics only strengthens the case for such policy. Take retirement savings: behavioural economics tells us that people tend to be present-biased, and they succumb to the temptation to withdraw early. One solution is to penalise early withdrawals.
The US Internal Revenue Service, for example, imposes a 10% penalty on early withdrawals — but even so, for every US$1 invested, US$0.40 is withdrawn early, at great cost.
How big should the penalty be? A behavioural savings model including the effects of present bias suggests that a 100% penalty would maximise welfare, because it would stop all early withdrawals even for the most present-biased individuals.
In other words, Australia’s superannuation mandate is optimal. Savings should be made fully illiquid, with no opt-out. The mandate imposes small costs on the few who would have withdrawn rationally, but these costs are dwarfed by the vast welfare gains overall.
Cash transfers too can be reevaluated under the behavioural lens. Research explored in Mullainathan & Shafir’s Scarcity (2013) demonstrates that poverty causes IQ to plummet, because dealing with the everyday challenge of being poor takes up so much cognitive bandwidth.
The corollary is that when impoverished Indian farmers receive a cash infusion, they experience a massive boost in IQ.
The behavioural evidence suggests that poverty may be cyclic partly because it is so tough to make good decisions while in poverty. Cash transfers free cognitive resources to focus on longer-term problems like learning, finding work, and planning for the future.
A way forward
How should policymakers proceed? They should remain agnostic when choosing between nudges and more traditional policy tools.
The right tool will raise welfare as much as possible, without stopping people from making cherished decisions about their lives. The litmus test will be democratic: do people accept the policy or not?
Superannuation, notably, passes the test with flying colours — it curtails choice, but it creates such large benefits that it enjoys enormous levels of support (only 7% of people polled were against the increase in super from 9% to 12% of income).
It does not curtail a cherished choice, like what job we can take, where we can live, what we can learn, or who we can associate with. It helps most of us do what we already want to do, but struggle to do.
Nudges will play a valuable, complementary role to traditional tools. The science, however, reasserts that there is no easy middle road.
We must continue to democratically debate big ideas about the kind of society we want to build — and we should always build it upon the firm ground of evidence.
Reuben Finighan is senior research officer at The Melbourne Institute, Fellow of the ARC Life Course Centre of Excellence at University of Melbourne and author on The Conversation. He can be contacted here.
