Presentation A Classroom Experiment on Risk and Uncertainty John C. Bernard Food and Resource Economics University of Delaware Presented at 16th Annual Conference Teaching Economics: Instruction and Classroom Based Research February 17-19, 2005 Robert Morris University, Moon Township, PA Daria J. Bernard Policy Analysis and Management Cornell University Course Information Economics of Biotechnology and New Technologies Primarily juniors Ranges from 15 to 30 students Prerequisite: Intro. Microeconomics Risk Experiment In today’s experiment we will be examining people’s decision making under risk and uncertainty. Below, you are presented with two different situations. For each situation, please select the choice that you would prefer. Almost no students have had intermediate Many students from outside majors 1 Risk Decision Situation #1 Choice A You will get 3 extra points on exam one. Risk Decision Choice B You have an 80% chance of getting 4 extra points on exam one, 20% chance of getting 0 extra points Situation #2 Choice A* You have a 25% chance of getting 3 extra points on exam one, 75% chance of getting 0 extra points Purpose of Experiment Understand economic concepts: Rational Expected Value and Expected Utility Risk Averse (Risk Neutral / Risk Loving) Choice B* You have a 20% chance of getting 4 extra points on exam one, 80% chance of getting 0 extra points Purpose of Experiment Understand limits to theories and alternative ideas: Allais Paradox Framing and Certainty Effects Prospect Theory Anomalies 2 Design with Context: Farmer Facing Risk Purpose of Experiment Understand course content: Situation #1 Relate to concerns over new technologies Especially GM Foods Difference in lay person and scientists’ views of risks and uncertainty Situation #2 Economic Rational Behavior Theory: a rational decision maker will make choices to maximize expected utility or expected value. More is better! Remind them that economics is about dealing with scarce resources A: Plant non-GM crops and earn $3,000 B: Plant GM crops with 80% chance of earning $4,000 and 20% of earning 0 due to consumer rejection of GM foods A*: Plant non-GM crops with 25% of earning $3,000 and 75% chance of earning 0 because crop fails B*: Plant GM crops with 20% chance of earning $4,000 and 80% of earning 0 due to consumer rejection of GM foods Expected Value (EV) EV = (probability event occurs) * (value of the event) So, which choices in each situation had the highest expected values? 3 Class Experiment Situation 1: Class Experiment Situation 2: Theory therefore states that rational decision makers should select B and B* Most pick A and B* Violates standard theory Choice A: Expected Value = 1 (3) = 3 Choice B: Expected Value = 0.8 (4) = 3.2 Choice A*: Expected Value = 0.25 (3) = .75 Choice B*: Expected Value = 0.2 (4) = .8 Past Results Course FREC 316 Spring Spring Spring FREC 806 Spring TOTAL Class Irrational? A-A* A-B* B-A* B-B* 01 02 03 2 2 3 5 6 7 0 1 2 1 0 0 01 0 4 2 3 7 22 5 4 Was the class irrational by their selections in the Risk Decision Experiment? Why or why not? 4 Possible Reasons: Avoid risk A risk averse decision maker would select A and A* This would be consistent with Expected Utility Theory under risk aversion Possible Reasons: Can’t overlook the sure thing! Called ‘Certainty Effect’ People often want to take a sure gain option and avoid a sure loss option Debate: does this hold in low-stake situations? Possible Reasons: View B as chance to lose 3 points rather than to gain 4 points. Prospect Theory Relates to ‘Framing Effect’ Typically seen in subjects making opposite choices when options phrased in terms of gains (positive frame/risk averse) compared to losses (negative frame/risk seeking) Designed to explain behavior inconsistent with EU theory Uses concept of a value function, defined in terms of gains and losses (deviations from a reference point). The function has a different shape for gains and losses. For losses it is convex and relatively steep, for gains it is concave and not quite so steep. Framing and reference point matter 5 Anomalies Still have a choice (B-A*) that isn’t explainable by either Expected Utility theory or by Prospect Theory Demonstrates limits to economic theory in explaining human behavior Uncertainty and GM Foods Consider utility a consumer gets from an Non-GM good (known, no risk considered) For GM version: Two factors affect their expected utility: p = scientific probability GM version posses a health hazard w = probability value for p is wrong (subjective, goes beyond prospect theory) Perception and Consumer Utility If either p or w > 0, consumer has lower expected utility, needs income compensation Calculating Risk People have trouble judging when: From experiment, know the amount of compensation may need to be larger than expected utility theory would suggest Dealing with probabilities Media coverage is unbalanced Especially true for very low levels How has media coverage been for GM foods? They’ve had misleading personal experiences Must rely on intuition 6 Factors Influencing People’s Risk Perceptions Compared with Experts (1987) Activity / Technology Smoking Nuclear power Motor vehicles Alcoholic beverages Surgery Pesticides Electric power plants (non-nuclear) Police work Genetically modified foods X-rays Irradiated foods Swimming Food preservatives Vaccinations Antibiotics Food coloring Class Rank 1 2 3 4 5 6 7 (tie) 7 (tie) 9 10 11 12 13 14 15 16 Experts Rank 2 11 1 3 4 6 7 10 5 8 9 14 13 12 Previous Classes (2) 1 (tie) 4 1 (tie) 7 5 6 9 3 15 8 10 12 (tie) 14 11 12 (tie) 16 Educating the Public Claim: if public is educated / informed, they will have the same risk assessments as experts. Is this realistic? Studies say no! Novelty Nature of possible consequences / Ethical concerns Scientific findings, evidence Reaction from: Government Activists Previous bad experiences Voluntary / Involuntary Attitude of proponents Risk Views: Public and Experts Lay People (the public): use all the factors we’ve discussed Experts: base strictly on technical estimates of possible fatalities Lay people will give rankings similar to experts if asked to rank by fatality possibilities instead of by risk 7 Thank You! Copies of the paper are available under working papers on my web page: www.udel.edu/FREC/bernard/index.htm 8