Application of nudge theory in conjunction with data mining to improve tax compliance

Abstract

The indirect taxation statutes have evolved over the years, providing a higher degree of freedom to the taxpayers to self-declare the information and file returns. However, the principal problem faced is that the corrective action (ensuing out of audit or other enforcement activities) is post the economic event and the taxpayer, more often than not, has to meet the tax demands from his pocket, denting his profit. This over the years has led to friction between the taxpayers and tax administration, both taking opinionated stands in favour of respective revenue, leading to a stalemate and increasing burden on the appellate and judicial forum. Such unproductive impasses may be much avoided if the approach is shifted from a ‘reactive’ sphere to a ‘proactive’ sphere, as it addresses the problem prior to its inception. One of the approaches in this direction should be to bring about an attitudinal change in the way the taxpayers perceive tax, who are on the verge of non-compliance. While the Allingham and Sandmo (AS) model (1972) assumes that taxpayers are utility maximisers and states that increasing penalty rates coupled with a higher probability of detection is a deterrent and leads to buoyancy in tax revenue, yet the same is still unproven theoretical model . On the contrary, the fiscal psychology model assumes that individuals contain an array of attitudes and beliefs which interact and respond to a social norm, and the same can be used to improve compliance. There is a need to re-couple payment and benefits arising out of tax paid which can be achieved by giving “agency” to the taxpayers and amounts to a “two-way nudge” . A number of behavioralist across the globe have conducted experiment to prove that “moral cost” (a factor that influences payment behavior, and is one of the cost a taxpayer discounts before deciding, when and how much tax it wants to commit), can be increased significantly by adding ‘descriptive norms’ and information regarding payment of taxes to the ‘nudge’ message. Further, an experiment has empirically stated that higher message specificity as part of ‘descriptive norm,’ gives an impression of greater message credibility or created a “large number effect,” which results in improved tax compliance.