Views | Ideas for India from the UK’s ‘Nudge unit’

Britain’s Conservative-led coalition government is going through something of a rough patch. The growth that front-loaded austerity was supposed to bring has proved elusive; the country is now officially back in recession. David Cameron’s personal standing has taken a beating as the phone-tapping scandal has dragged on, widened in scope, and revealed that Britain’s youngest Prime Minister in nearly two centuries thought for a long time that “LOL” was SMS-speak for “lots of love”.

But the release of a recent working paper about the working of the unit suggests that it has quietly been helping to refine the nuts and bolts of governance using behavioural economics, and doing so successfully and systematically. The working paper, which was released by the UK Cabinet Office, presents the results of eight completed and ongoing experiments designed to use behavioral principles to reduce fraud and error in various aspects of the bread and butter of the government, whether collecting overdue debt, increasing tax collection, or prevent people from defrauding the government.

What is striking about many of the experiments is just how easy and virtually costless the changes that worked were. One experiment involved tweaking the wording of a letter that was already being sent to those who owed the government back taxes. The standard letter resulted in 67.5% of late payers responding after three months. Behavioural economics, however, suggests that people are particularly keen to conforming to social norms and dislike being seen as renegades. So the government tried various versions that reminded people that the social norm was to pay up. The most impressive result involved telling people not just that 9 out of 10 of people in the country paid their taxes, but that the norm in their particular area involved paying up. This version of the “social norm letter” raised the response rate from 67.5% to 83% - an increase of 15.5 percentage points over the baseline. This is a dramatic result for a simple change in the way a letter that was already being sent was worded.

Another of the experiments was aimed at getting more doctors with outstanding tax liabilities an opportunity to set their affairs in order by using a voluntary disclosure scheme. The standard “sarkari” letter that was already being sent to a control group got a response rate of a mere 3.77%. An alternative built on the behavioural insight that simplifying letters and highlighting key actions and how to carry them should be more effective. The simpler letter increased the response rate to an altogether more respectable 35.3%. The new letters have already swelled the coffers of the tax authorities by over a million pounds.

A third experiment involved squeezing dues out of those who owed fines imposed by courts. Collecting these can be expensive, because people have to chase them down and so on. So the government decided to send debtors a text message instead. This already had impressive results over a control group who received no text – 23% of text receivers paid up, compared with 5% of the control group. But the government also used this opportunity to systematically test several versions of the text itself that built on some psychological insights. The most effective was simply addressing the debtor by name in the text – it raised the payment rate by half again, sending it to 33%. Other ideas, such as mentioning the amount owed, were not as effective.

Other experiments, and expanded versions of these ones, are going on. They show quite clearly that using psychologically-informed ways of doing what governments often do anyway (or at least, should be doing) such as sending reminders to debtors, or notices to taxpayers, can vastly increase how effective these are. By not trying these variations, money is essentially being left on the table. Uncollected taxes and fines help nobody, and governments everywhere would do well to look at the experiments carried out by the British Nudge unit for inspiration.

The scope for applying this idea to government policy in India is endless. Just imagine if tax authorities, municipalities, etc. around the country actually woke up to the idea that cleverly designed messaging could help them actually collect what they are owed, and could do so with little extra effort and very little extra cost. Of course they have to start using communication technology to begin with. So far, it seems that the only people to have cottoned on to the vast potential of using texts to communicate with people are the guys sending you endless spam about “3BHK in NRI complex @70lacs only”.

But the importance of these little tests ultimately goes beyond the specifics of what they were about. It lies instead in the fact that they took an experimental, iterative, evidence-based approach to policy design – something that ought to be standard but is in fact almost non-existent. It ought not to be.