Wednesday, February 15, 2012

Nudging to reduce fraud, error and tax defaults

The British Cabinet office's Behavioural Insights Team recently released a research study on the application of insights from behavioral psychology to (pdf here) reduce fraud, debt and error within public systems. The report points to seven behavioral insights that can enhance public policy and help achieve desirable outcomes.

The Behavioural Insights Team conducted RCTs on these eight interventions and found impressive results. They are summarized below.

The British government's taxation department, HMRC, sent out letters with a range of different messages to 140,000 taxpayers. Residents received either a control letter (which contained no social norm) or one of a number of different social norm messages. All of the social norm letters contained the statement that '9 out of 10 people in Britain pay their tax on time', but some also mentioned the fact that most people in the recipient’s local area (or postcode) or town had already paid.

The graph shows that there was a 15 percentage point increase from the old-style control letter which contained no social norm and the localised social norm letters. In fact, the phrasing of the letter is of great relevance. Saying that '9 out of 10 people pay their tax on time' is more effective if it is followed by 'you are one of the few people who have not paid yet'. Simply including a phrase with both these elements raised the payment rates achieved by a single letter from 36.8% to 40.7%.

The studies of the Team also found that letters which contain simple, clear messages about the issue at the beginning of the letter, clearly spell out the actions required (by underlining it or highlighting it) and underline implications of noncompliance, increases the likelihood of favorable responses.

As I blogged earlier, another study involved sending out SMS messages to defaulters on court fines. Personalized text messages were sent out to defaulters with varying messages - general message to pay the fine, message indicating the person's specific fine amount, and those giving his/her name. The results have been quite impressive.

This strategy poses two challenges for policy makers contemplating the same in developing countries like India.

1. Since compliance with rules are low and defaults very high, the marginal impact of these interventions could be considerable.

2. However, on the flip side, given the weaker social norms, the nudging impact of the messages could be attentuated in these social contexts. Since in many groups, deviation is the norm, the messages may not carry the same weight.

The ultimate effectiveness of these interventions in these societies will depend on the emergent dynamics of these two forces. If the former prevails, the intervention will be a success.

No comments: