Imagine a school canteen. There’s a full array of food on sale, from healthy salads through to chocolate fudge brownies. But the canteen deliberately puts the salads in the children’s eye-line at the front of the counter. Economists describe such a choice as a “nudge”. Behavioural evidence suggests food placed here is more likely to be purchased. The canteen is encouraging healthy eating with this information, but without coercion. The children are, after all, still free to buy chocolate fudge brownies, should they wish.
Richard Thaler, this week’s Nobel Economics Prize winner, has made a career observing how humans deviate from perfect rationality and how applying “nudges” can alter economic decision-making. He has presented compelling evidence that humans tend to be biased towards the status quo, value things more when we already own them and are influenced by the framing of decisions.
Nudgers aim to alter our “choice architecture” to influence decisions but without restricting our freedom to choose. The UK Government has a whole unit working on this. The new policy of auto-enrolment in company pensions, requiring an active opt-out, is a “nudge” attempting to help people meet their stated desire for more saving towards retirement. Participation in workplace pensions has increased by 37 percentage points since it was introduced. Provided they are based on good evidence, do not use heavy-handed bans or change the payoffs to choices, Thaler advocates such nudges as a form of “libertarian paternalism” – guidance in decision-making, which does not restrain individual free will.
But the concept is controversial among economists. Just because some individuals are not rational does not mean regulators and politicians have better information on their circumstances or preferences. Some now auto-enrolled in pension schemes, for example, would need and prefer more cash today, but the same bias towards inertia prevents them from opting out.
In many markets regular feedback, repeat decisions and competition allow people to fulfil their preferences whilst overcoming individual-level biases. Regulators and politicians have their own motivations, too, and can be prone to groupthink and capture by vested interests. The lines between nudging and shoving are quite often blurry. Auto-enrolment might be a nudge for the employee, but it seems one hell of a shove to obliged employers threatened with fines for non-compliance.
The main issue with behavioural economics, though, is that we appear to be applying its insights to the wrong target group. The book Nudge has a whole chapter explaining the conditions under which they are likely to be effective: when the consequences of choices are delayed, choices are difficult, the choice is made infrequently, and when it is difficult to predict how the choice might affect our lives. These seem to apply most aptly to decisions politicians and regulators make all the time on our behalf.
Yes, individuals have their biases. But politicians do too. They put the status quo on a pedestal, suffer groupthink, seek to bribe the electorate, have a bias for budget deficits, continuously complicate the tax system and grow the size and scope of government. Absent a constitution that constrains them, why not change the “choice architecture” they face?
We could, for example, make mandatory five-year sunset clauses the default on new regulations to try to curb the growth of the regulatory state. Politicians would have to rubber-stamp the continuation of each regulation, enabling them to assess and reflect on their effectiveness. The same concept could be applied to all repatriated EU law.
To stop taxation by stealth, we could pass a law so that all tax thresholds rise automatically each year in line with the growth rate of nominal GDP. If politicians want to raise taxes by playing with income thresholds or through fiscal drag, they would have to do so explicitly and transparently, facing the political heat.
The opportunities to apply this thinking are endless. To deter cronyism, large political donations could be anonymised through a central clearing house, leaving complete freedom to donate to a party but dampening the incentive for politicians to appease specific donor interests. An option for politicians’ salaries to be tied to economic growth as default could be added to their contracts too, with the freedom for them to “opt out” and maintain current arrangements should they wish to signal their lack of faith in their own policies.
On the spending side, zero-based budgeting should be the norm in any comprehensive spending review. Any new policy that raises net spending above a threshold amount should trigger an OBR analysis on how much it will add to national debt over the coming 30 years, which must be read out by the relevant minister in Parliament.
“Tax trigger laws” could be passed too, meaning when revenue is much stronger than expected, the default would be to cut tax rates so revenues are simply maintained to meet government spending. This would deter the perceived “windfall” effects the Treasury can obtain from growth before budgets, often used to bribe the electorate, and would highlight the trade-off between spending and taxes.
Despite Thaler’s interesting work, history suggests bad government policy can be far more damaging to our welfare than individuals’ biases. So why not apply the insights of Nudge where it is most needed, and frame politicians’ choices to encourage the salad diet for government?