Behavioral Economics and Lessons for FCPA, AML, and Tax Compliance

July 10th, 2013

Under the standard economic theory of crime, compliance with laws is a mix of two important factors.  One: the penalty that results if the offender is caught and 2: the probability of the offender getting caught in the first place.  If the fine is proportional to the crime, but the probability of being caught is almost certain, few will risk it.  In the same way, if the probability of being caught is low, but the penalty is very high, again few will risk it.

Gary Becker—the libertarian economist who wrote Crime and Punishment: An Economic Approach, an authoritative economic theory on crime—has described his own encounter with these tradeoffs. In one inspirational moment for his research, while teaching at Columbia University, Becker asked himself whether he should park in a spot that was closer to campus, and illegal, or in a lot which was somewhat further away. He notes that he had to make a calculation: What was the likelihood that he’d be caught if he parked down the street, versus the time and the money that would be lost by parking further away?

Economists often use this framing device when talking about a whole host of crimes—including white collar crime—and, I have discussed these tradeoffs as well, in particular with respect to the Foreign Corrupt Practices Act.

Of course we can’t all be as rational as Gary Becker and there are other experts who do not believe criminals’ decisions are this cut and dry. There is a body of research in behavioral and experimental economics that has found—in addition to the rational, external motivation for criminal behavior—humans face internal and social motivations for honest and dishonest behavior. For example, Mazar, Amir, and Ariely (2005) test whether or not study participants are willing to cheat on a test at different levels of probability of being caught. They give the same test to four groups and promise each one $.10 per correct answer. The first group (the control) has their test graded by an examiner. The remaining groups have different methods of reporting, including self-reporting with turning in their answer sheets, and self-reporting where the participants shred their answer sheets. Each group has an opportunity to cheat and face different probabilities of being caught.

Traditional economic theory would suggest that the participants would respond differently in each scenario—cheating more when the probability of being caught decreases. Indeed, the authors do find that the self-reporters have test scores about 20% higher than the control group, but they do not find any differences between the self-reporting groups.

The findings suggest that people’s internal reward mechanisms govern behavior to a large degree, particularly after a certain threshold. These mechanisms are likely inactive at low levels of dishonesty, but kick in at a certain threshold, preventing us from engaging in additional dishonest behavior.

In a similar set of studies, Dan Ariely, the behavioral economist and author of Predictably Irrational, finds that people are more ethical when dealing with paper money than with objects of similar monetary value. For example, college students are more likely to take soda cans from a public fridge than equivalent dollar amounts. And office workers are more likely to take pens and pencils from a supply closet than a couple quarters from the petty cash drawer. Ariely also shows that removing people from paper money only symbolically—for example by using electronic money—has a similar effect on honesty and ethical behavior.

The implications for financial crimes—including tax evasion, money laundering, and foreign bribery—are striking. We can easily imagine that the Swiss banker, who happily flies to America to encourage his clients to avoid taxes with a Swiss bank account, would be unwilling to fly across the Atlantic Ocean with the American’s money in his suitcase. We can imagine that the HSBC bankers, who “willfully failed” to apply money laundering controls to $881 million in drug trafficking proceeds, would never have accepted paper money from a representative of a Mexican drug cartel. We imagine that the executives of Wal-Mart, who were aware of and concealed their companies’ illegal bribery activities in Mexico, would not have handed cash to the foreign officials themselves.

These findings suggest that increasing penalties and investigations (that is, the probability of being caught) will not necessarily solve the problem. Because much of our unethical behavior is caused by social and cultural norms, we must also invest in socialization that increases in the strength of the internal reward mechanism. Anti-corruption compliance programs among multi-national corporations that educate staff about FCPA are one example. Another is the social and media pressure exerted after a company violates the letter, or the spirit, of a law, as was the case with Apple earlier this year. Stronger non-pecuniary penalties, including jail time for offenders, would also provide a strong deterrent and would change some of the socialization dynamics that lead to this type of crime.

While these types of policy changes are not as easy as increasing penalties and investigations, for the behavioral dynamics listed above, they may be just as important to change behavior in the long-run. At a minimum, when addressing these problems, it is important to remember that humans are not always the rational, utility-maximizing robots we expect them to be. Penalties for crimes should reflect both our human—and our robot—tendencies.

Written by Ann Hollingshead

Follow @FinTrCo