Ideas for Leaders #679

Stopping Tax Evasion: Detection Probability Vs Moral Persuasion

This is one of our free-to-access content pieces. To gain access to all Ideas for Leaders content please Log In Here or if you are not already a Subscriber then Subscribe Here.

Key Concept

The authors of this research teach on the CEMS Masters in Management program at the Norwegian School of Economics - and this Idea is part of our CEMS series.

A large field experiment involving Norwegian taxpayers underreporting foreign income reveals that the probability of detection deters the greatest number of potential non-compliant taxpayers; moral arguments sways fewer non-compliant taxpayers but inspires those who listen to significantly increase their reported income. 

Idea Summary

Norwegian citizens making money abroad must report this income to two tax authorities: the authorities of Norway and the authorities of the country where they made the money. In order to avoid taxes, a number of these citizens underreport their foreign income to the Norwegian authorities — not realizing that most foreign countries send the income figures they receive from Norwegian citizens to Norway! As a result, Norwegian tax authorities know when a citizen underreports foreign income.

Using this unique situation, a recent study explored different persuasive approaches for discouraging tax evasion. Specifically, the study involved sending letters to 15,000 Norwegian citizens who likely underreported their foreign income in their previous tax return. The letters were sent just before they were to report their taxable income. 

The researchers divided these 15,000 citizens into five groups as follows:

  • No letter. The members of one group did not receive a letter from the tax authorities. 
  • Base letter. The members of the second group received a base letter from the tax authorities, which explained the process for reporting foreign income. The goal was to determine how much the underreporting was due to a lack of knowledge about the tax laws and procedures.
  • Detection Probability letter. The members of the third group received a letter that included an explanation of the intra-governmental agreements — and thus the likelihood that any underreporting would be detected. This letter was labelled the detection probability letter.
  • Moral Suasion/Fairness letter. The members of the fourth group received a moral suasion letter that presented a “fairness” argument: most Norwegians report their income correctly and completely, the letter said, so to treat all citizens fairly, foreign income needs to be reported correctly and completely.
  • Moral Suasion/Societal Benefit letter. The members of the fifth group received a moral suasion letter as well, only this one advanced a societal benefit argument, emphasizing that tax dollars are used for funding services in education, health and other sectors.

The researchers then compared the response of the taxpayers in the five groups by analysing their reported foreign income in the follow-up year (the first tax return after receiving the letter) and one year later — with the latter representing a long-term response. 

The analysis of these returns yielded the following results:

  • The base letter had an impact on self-reporting. While 11% of taxpayers who did not receive a letter self-reported foreign income in subsequent years, 20% of taxpayers who received the base letter self-reported such income.
  • The detection probability letter had an impact on the largest number of taxpayers, with 33% of taxpayers who received the detection probability letter reporting foreign income.
  • The moral suasion letters had less of an impact than the detection probability letter in terms of sheer numbers of people but a greater impact in terms of income reported. 
  • The moral suasion letters also had more of an impact on taxpayers who had previously reported at least some foreign income.
  • There were no differences in the responses to the two different moral suasion letters.
  • In sum, any communication from tax authorities will have some effect on compliance. The detection probability letter had the most extensive impact — concerning the largest number of taxpayers — while the moral suasion letters had the most intensive impact, that is, the amount of reported income.

Business Application

Companies and leaders can use the insights from the results of this research to guide how they discourage law- or rule-breaking or -bending on the part of their employees and managers. As the approach with the greater extensive impact, detection probability is probably the most effective in most situations: you want the greatest number of employees and managers to respond to the threat that they will be found out. 

The moral suasion approach seems to appeal to people who have already demonstrated some scruples (in the example above, the taxpayers who had reported some foreign income, just not all of it), and draws a more intense response. Perhaps in the case of a suspected transgression by leaders or employees with a previously clean record, appealing to their moral core would have an impact.

Intuitively, and reinforced by this research, the worst option is to do nothing. Once people successfully break the rules, they rarely stop.

Contact Us




Idea conceived

  • July 2017

Idea posted

  • October 2017

DOI number



Real Time Analytics