Estimating how many people are breaking the rules in a protected area

Successful conservation work is largely about getting people to follow the rules - whether it be laws against poaching the last remaining gorillas in Central Africa or regulations on building campfires in Yosemite during the wildfire season. 

A new study in the journal Biological Conservation tests an innovative survey methodology to help resource managers figure out how many people are breaking the rules. This is actually a much tougher question to figure out than one might think.

People who break rules are reluctant to admit it, which leads to two type of biases in surveys: a 'social desirability bias' in which respondents give the socially acceptable answer (i.e. they lie) and a 'response bias' in which rule breakers skip the questions all together.

To overcome these biases, Freya St. John and fellow researchers from Bangor University in the UK present a survey approach called the randomized response technique and tested it on fly fishers in the country. They set-up the survey questions to obtain data on fly fisher compliance with regulations on the activity - e.g. fishing with a license, not exceeding the bag limit, etc.

In the survey approach a die is rolled and respondents are required to either respond truthfully or give a predetermined response depending on the number that comes up, which is only visible to the respondent. This approach has been used before in other fields but rarely in the context of conservation.

In the case of the study, if the die landed on 1, 2, 3, or 4, the respondent was required to give the truth to a 'yes' or 'no' question regarding their fly fishing behavior. If the die landed on 5, the respondent was required to say 'yes,' regardless of whether it was true or not. If the die landed on 6, the respondent was required to answer 'no.' Based on the number of 'yes/no' responses and the probability that the die would land on 5 or 6, the study authors were able to extrapolate the true 'yes/no' responses about rule breaking.

The researchers hypothesized that respondents who had broken the rules would be more likely to admit it in this survey format because the randomized, secret element from the dice adds a layer of privacy. They also conducted two similar surveys with different formats: one in which they had respondents self-complete an anonymous survey and another in which they asked respondents to report about the behavior of their close friends who fly fish.

The study found that the reporting of rule breaking was significantly higher for the randomized response technique. This suggests that this surveying technique encourages more honest reporting on rule breaking behavior. For resource managers, this means having a better tool for identifying regulations that are not working and potentially figuring out why. The authors write,

"This has direct management implications with respect to designing interventions to improve compliance. In the absence of direct knowledge of rule breakers, managers have a limited ability to target interventions to change their behavior, for example through awareness schemes or through targeted enforcement activities."

--Reviewed by Rob Goldstein

St. John, F., Edwards-Jones, G., Gibbons, J., & Jones, J. (2010). Testing novel methods for assessing rule breaking in conservation Biological Conservation, 143 (4), 1025-1030 DOI: 10.1016/j.biocon.2010.01.018