ARLINGTON, Va., Dec. 11 -- The Casualty Actuarial Society issued the following news release:
By combining economics, psychology, sociology, biology and neuroscience, the discipline of behavioral economics focuses more on the process of decision-making, not the result. And its conclusions can be incorporated by actuaries in their own work.
Behavioral economics is emerging as a leading decision science, said David Wheeler, an associate with Innodata Synodex, a New Jersey-based analytics and data solutions firm, at the Annual Meeting of the Casualty Actuarial Society, held in Lake Buena Vista, Florida, November 11-14.
Rather than seeing the entire human species as the hyper-rational homo economicus, behavioral economists see individuals along a behavioral continuum, from homo economicus to the altruistic homo reciprocan, the type Wheeler describes as "the Kumbaya-I'll-chop-the-wood-if-you-collect-the-berries" type.
Robert Wolf, managing director of the casualty actuarial practice at Willis North America, offered some examples of how our biases move our decisions away from the rational. And he showed how knowing about these irrational behaviors can help you in your business dealings.
Example 1: What causes more deaths in the United States a year - colon cancer or motor accidents?
Most people say motor accidents. The answer is colon cancer. And the reason they get it wrong is a common mental flub called the availability heuristic. When we think of several reasons for something, we put more importance on the ones that come to mind quickly.
And few things come to mind more readily than a good story. So, Wolf said, when you want to convince your client or your boss of something, tell it in a story.
Example 2: First, would you rather have $700 for certain or have an 85% chance of winning $1000? Most people want the $700. They would rather take the sure gain, Wolf explained.
Second, would you rather pay out $700 for certain or have an 85% chance of paying out $1,000? Most people want the 85% possibility. They would rather gamble that they might avoid paying anything at all.
But really, the two problems are identical, or at least flip sides of the same coin: a chance of gain versus a chance of loss. Most people prefer to lock in a gain than take a chance for a bigger one. And they would rather take a chance to avoid a loss than to lock in a smaller loss.
"We are risk-averse, supposedly," Wolf said. Then he clarified. "We are risk-averse to gains. We are risk-seeking to loss."
Example 3: In this scenario, you ask one group of financial auditors one pair of questions, and a second group of auditors a second pair.
Ask the first group of auditors if significant executive fraud happens in more or less than 10 cases per thousand. Then ask for their estimate of how often it happens.
Next, ask the second group if significant executive fraud happens in more or less than 200 cases per thousand. Then ask for an estimate of how often the fraud happens.
Researchers ran just such an experiment. The first group on average said fraud happened in 16.52 cases per thousand. The second group gave an answer more than twice as high, 43.11.
This behavior is called anchoring. The first group anchored its estimate to the 10 case per thousand estimate it had just been asked about. The second group anchored its estimate higher, because it had just been asked about 200 cases per thousand.
"We tend to be affected by numbers," Wolf said, "even if they aren't valid."
Other behavioral biases include confirmation bias (we seek out evidence that confirms our biases while ignoring evidence that contradicts them); and overconfidence (we are often too confident in estimates of events about which we know little).
Another speaker, Rick Gorvett, director of the actuarial science program at University of Illinois - Urbana-Champaign, listed more common biases:
* The endowment effect, in which we value things more highly when we own them.
* Hindsight bias, or the "I told you so" effect, meaning we reassign a higher probability to an event after it is over than we did before.
* The recency effect, in which we place more visibility and gravitas on recent events. Gorvett's example was a list of the greatest quarterbacks ever - all of whom played in the past five years.
* Representation bias - viewing an event based on how much it resembles other events. "We think we can fill in blanks," Gorvett said, "by one data point."
How can actuaries handle these anomalies? Wolf and Gorvett suggested:
* Search for disconfirming information, and have a private collection of information ready to use in group settings. And try to obtain information one on one, which helps to overcome the hindsight bias, the confirmation bias and others.
* Conduct highly structured meetings, making key decisions early, before the biases can take root.
* Ask outsiders for information and maximize diversity in getting information.
* Every company should have, at meetings, a "chief skeptical officer" who questions everything. (It doesn't have to be the same person in every meeting.)
* Refine, reframe and flip every problem you confront.
* Carefully test the models you build and seek significant peer review of your work.
* Create a diverse work force - the diversity encourages different viewpoints.
With all that, Gorvett said, it's important to recognize these are human traits that can't be expunged from our psyche. They can only be recognized and accounted for. These thought processes are products of "millions of years of evolutionary psychology" and they developed for a reason. Change is "not going to be that easy," Gorvett said.
The Casualty Actuarial Society fulfills its mission to advance actuarial science through a focus on research and education. Among its 5,700 members are experts in property-casualty insurance, reinsurance, finance, risk management, and enterprise risk management.
TNS rd43 121214-4140666 StaffFurigay