Brief 39: Attitudes Towards Risk and Illegal Behavior

This study investigates the relationship between risk attitudes and illegal behavior in South African fishing communities. The authors found that people who were more inclined to gamble – in economics terms, who were less risk-averse – were also more likely to have broken the law. They also examined the behavior and risk attitudes of women, as well as those with and without fishing licenses.

Read Full Study

Category: Natural Resources, Public Service Provision

Tags: fishing, risk, behavior, licenses, lotteries, overfishing

Date of Publication: Wednesday, February 24, 2016

EGAP Researcher: Justine Burns

Other Authors: Kerri Brick, Martine Visser

Partners: Photo credit South African Tourism office, available via Creative Commons BY 2.0

Geographical Region: Africa

Research Question: What is the relationship between attitudes towards risk and illegal fishing among South African fishers?

Preparer: Seth Ariel Green

English

Background:

This study investigates what sorts of people, among South African fishing communities, are more or less likely to overfish. Overfishing is both a global problem and a particular problem in South Africa. A majority of people whom the authors surveyed said that the process by which fishing licenses and quotas were determined in South Africa was unfair and/or corrupt.

The average income of the sample, taken from 9 coastal fishing communities, was $47 a month. Most of the sample identified as “Coloured,” “Black,” or “Other.” Average educational attainment was 8 years. Of the 555 people in the sample, 258 (46%) had been allocated a quota and/or permit or, alternatively, work for a crew that has been allocated a quota and/or a permit.


Research Design:

First, subjects were surveyed on a variety of topics, including their main sources of income, attitudes towards the law, and whether they had ever broken any laws concerning limits on fishing. The survey also included questions on whether they had ever fished more than official quotas allowed. Approximately one third of respondents either admitted to overfishing, or to being charged with overfishing.

Then, a month later, subjects played a game in which they chose between 8 separate pairs of lotteries, one of which was a guaranteed payoff of a certain amount, and the other of which was a risky bet. As the lotteries proceeded, the sure payoff became smaller and smaller while the risky bet stayed the same, making the risky bet progressively more attractive. In the first lottery, subjects chose between a guaranteed payoff of R20 (about $3), and a gamble of either winning R20 or nothing at 50/50 odds. The last lottery had subjects choose between a sure payoff of just R2 and either winning R20 or nothing at 50/50 odds.

The complete list of lotteries is as follows:

Lottery Number

Lottery A

Lottery B

Expected Payoff of A Expected Payoff of B
1

R20, guaranteed

R20, 50% odds; R0, 50% odds 20 10
2 R15, guaranteed R20, 50% odds; R0, 50% odds 15 10
3 R12, guaranteed R20, 50% odds; R0, 50% odds 12 10
4 R10, guaranteed R20, 50% odds; R0, 50% odds 10 10
5 R8, guaranteed R20, 50% odds; R0, 50% odds 8 10
6 R6, guaranteed R20, 50% odds; R0, 50% odds 6 10
7 R4, guaranteed R20, 50% odds; R0, 50% odds 4 10
8 R2, guaranteed R20, 50% odds; R0, 50% odds 2 10
 

At the end of the experiment, a lottery was chosen at random for subjects, and they actually either won the amount they had chosen or took the bet they had chosen for that lottery.

If a subject took the bet on lottery 1, they could reasonably be called risk-seeking, or someone who got intrinsic value from gambling (the bet in lottery 1 had zero chance of being better, in terms of money earned, for participants than the sure payoff). If a subject took the guaranteed payoff in lottery 8, then they were very risk-averse.

In total, subjects’ attitudes towards risks were determined by identifying the point at which they switched over to the risky bet. A perfectly risk-neutral subject would choose the sure payoff on lotteries 1-3, be indifferent between the two options in lottery 4, and choose the bet in lotteries 5-8.

The outcome of interest was the relationship between law-breaking behavior and attitudes towards risk.
 
Results:

The authors found three substantive results with policy relevance. First, men were less risk-averse, and more likely to break the law, than women. While only 17% of female fishers acknowledged breaking the law, 50% of male fishers did, and men were more likely to take risky bets in the experimental task.

Second, this result holds for licensed fishers. Women who had legal permission to fish were both less-risk averse and less likely to break the law than men who had the legal right to fish.

Third, 79% of overfishers had the legal right to fish (i.e., were overfishers, not poachers). Legally employed fishers were both less-risk averse and more likely to break the law than those who did not have permission to fish.
 
Policy Implications:

First, policymakers might want to take into account systematic differences in risk-taking attitudes when making labor and allocation decisions. As every car insurer knows, young men are especially prone to risky decisions; one way to avoid this problem is to hire more women and more people who are older.

Second, there is a strong relationship between perceived legitimacy and law-breaking. Most of the subjects in the experiment—both those who report breaking the law and those who do not—report that they do not trust the government to make fair decisions. In an environment where more people trusted policy, compliance could very well be higher.

Finally, policymakers should pay attention to the amount of overfishing that occurred among participants with the legal right to fish. This finding suggests that regulation—especially regulation from a body perceived as illegitimate—can crowd out intrinsic motivation towards pro-social behavior, namely not to overfish.