Published on January 15, 2008
Valuing Environmental Resources:A Constructive Approach(Alternative Title:Valuing Nature and Its Protection: Psychological Perspectives): Valuing Environmental Resources: A Constructive Approach (Alternative Title: Valuing Nature and Its Protection: Psychological Perspectives) Paul Slovic [email protected] September,2006 Slide2: A Wandering Tale The psychological complexity of values and preferences and their resulting lability (instability) Empirical studies: From preference reversals to preference construction A method for constructing quantified values (if you must) Slide3: The Measurement Problem Measurement of physical qualities (e.g., weight, distance) requires the satisfaction of description invariance and procedure invariance. Measures of preference systematically violate both of these fundamental requirements. Slide4: 2. Procedure Invariance Different (but equivalent) measures for eliciting preferences should lead to the same preference ordering, just as different but equivalent measures of heaviness—for example, placing two objects on opposite sides of a pan balance to see which is heavier versus measuring each object separately on a scale—should yield the same ordering of weight. Slide5: (like choice) Slide6: (like rating, matching, or pricing) Slide7: Preference measures often violate procedure invariance. Object A may be clearly preferred over object B under one method of measurement (e.g., choice), while B is clearly preferred under a different but presumably equivalent measurement procedure (e.g., matching or pricing). This fact has significant implications for the valuation of environmental resources and many other goods as well. Slide8: Violation of Procedure Invariance: Early Studies Preference Reversals 1971-1973 P bet 35/36 win $4.00 1/36 lose $1.00 3.86 EV $ bet 11/36 win $16.00 25/36 lose $1.50 3.85 EV Many Ss choose the P bet but attach higher $ values (buying and selling prices; cash equivalents) to the $ bet. (Slovic & Lichtenstein, 1971 and 1973) Slide11: A body of data and theory has been developing within psychology which should be of interest to economists. Taken at face value the data are simply inconsistent with preference theory and have broad implications about research priorities within economics. The inconsistency is deeper than the mere lack of transitivity or even stochastic transitivity. It suggest that no optimization principles of any sort lie behind even the simplest of human choices and that the uniformities in human choice behavior which lie behind market behavior may result from principles which are of a completely different sort from those generally accepted. This paper reports the results of a series of experiments designed to discredit the psychologists’ works as applied to economics. -Grether and Plott, 1979, p. 623 Slide12: Criticism of Lichtenstein and Slovic Studies: 1) Misspecified incentives 2) Income effects 3) Indifference not allowed 4) Confusion and misunderstanding 5) Unsophisticated subjects 6) Experimenters were psychologists Slide13: Needless to say the results we obtained were not those expected when we began this study. Our design controlled for all the economic-theoretic explanations of the phenomenon which we could find. The preference reversal phenomenon which is inconsistent with the traditional statement of preference theory remains. It is rather curious that this inconsistency between the theory and certain human choices should be discovered at a time when the theory is being successfully extended to explain choices of nonhumans. -Grether and Plott, 1979, p. 634 Slide14: The one theory that we cannot reject…is…the least satisfactory since it allows individual choice to depend upon the context [i.e., response mode] in which the choices are made. Grether and Plott Slide15: Reactions to Grether and Plott: More attempts to make reversals disappear. Some modifications of choice theory axioms. About 400 citations of Grether and Plott in the literature. Preference reversals continually demonstrated —except in some aggressive arbitrage situations. Slide17: This review has attempted to show how preference reversals fit into a larger picture of information-processing effects that, as a whole, pose a collective challenge to preference theories far exceeding that from reversals alone. These effects seem unlikely to disappear, even under rigorous scrutiny. -Slovic and Lichtenstein, 1983, p. 603 Slide18: We urge economists not to resist these developments but, instead, to examine them for insights into the ways that decisions are made and the ways that the practice of decision making can be improved. -Slovic and Lichtenstein, 1983, p. 603 The Blossoming of Preference Reversals: The Blossoming of Preference Reversals Choose X; also reject X (Shafir, 1993) Ratings of value prices choices Wide range of stimuli - new products - ice cream - job candidates - job offers - air quality - salaries - punitive damage awards by juries - countries in which to live (income distribution) Etc. Slide20: Construction of Preference: The Broader Perspective Emerges The reality of preference reversals has gradually become accepted, even among economists (see, e.g., Camerer, 1995; Starmer, 2000; and Seidl, 2002) and most efforts today by economists and others are rightly attempting to discover the conditions that allow preference instability to exist in important settings outside the laboratory. Something significant has been occurring in the world of preference reversals. Slowly, the study of reversals has metamorphized from a curious phenomenon observed with simple gambles in laboratory settings to a method for documenting a broad, encompassing view of human behavior that has come to be known as “the construction of preference.” The Construction of Preference(published last week): The Construction of Preference (published last week) Slide22: The thesis of preference construction is that we often do not know our own values and must construct them “on the spot,” using not only our knowledge, feelings, and memory, but also many aspects of the current environment, including how the preference question is posed and what type of response is required. Some elements of preference construction:: Some elements of preference construction: Editing Elimination by aspects Prominence effects Framing Dominance structuring Choice bracketing and mental accounting Reason-based choice Miswanting Decision Utility, Predicted Utility, Experienced Utility Focalism Immune Neglect Duration Neglect Slide24: Elements of Preference Construction (cont.): Contingent weighting Distinction bias Lay rationalism Reliance on defaults Affect Evaluability Value elicitation has barely begun to address the challenges posed by preference construction Slide25: Preference Reversal Between Joint and Separate Evaluations Work by Chris Hsee and Colleagues Whereas traditional preference reversals involve different types of responses (e.g., choices and prices), differences in response modes (e.g., joint evaluation–JE–vs. separate evaluation–SE) also produce reversals. JE SE due to evaluability factors (dictionary study) or other causes. Slide26: Irwin et al. (1993) found that, in JE, people were willing to pay more for improving the air quality in Denver than for improving a consumer product such as a VCR, but in SE, WTP values were higher for improving the consumer product. Kahneman and Ritov (1994) found that, in JE, people would contribute more to programs that save human lives (e.g., farmers with skin cancers), but in SE, they would contribute more to programs that save endangered animals (e.g., dolphins) Slide27: Hsee and Zhang (2004) demonstrate that, in JE, people overpredict the differences that different values of an attribute (e.g., different salaries) will make to their happiness in SE. This is called distinction bias. Predicted utility experienced utility. Slide28: The Quarterly Journal of Economics, February 2003 Tom Sawyer and the Construction of Value: Tom Sawyer and the Construction of Value Dan Ariely George Loewenstein Drazen Prelec (in press) Slide30: The apparent orderliness in . . . choices, their stability for a given individual, and the generally correct directional response to changing incentives, encourages the belief that the choices are firmly rooted in personal likes and dislikes—in fundamental values. We suggest, in contrast, that correct directional responses to changing incentives do not provide strong support for fundamental valuation, but can follow from the fact that people try to behave in a sensible manner when it is obvious how to do so. Slide31: Modern economics assumes that exogenous consumer preferences interact with ‘technologies’ and initial endowments to produce equilibrium states of the economy—prices and production levels. This analysis falls apart if preferences are themselves influenced by the very equilibrium states that they are presumed to create. By posting a price for a new product, for instance, a firm invites consumers to consider whether they would purchase at that price and so replicates the anchoring manipulation as conducted in our experiments. If prices and other economic parameters function like public anchors, then consumer tastes no longer exist independently of prices but are endogenous to the economy. When Web Pages Influence Choice:Effects of Visual Primes on Experts and NovicesNaomi Mandel and Eric J. JohnsonJournal of Consumer Research, 2002: When Web Pages Influence Choice: Effects of Visual Primes on Experts and Novices Naomi Mandel and Eric J. Johnson Journal of Consumer Research, 2002 This article extends the idea that priming can influence preferences by making selected attributes focal. Our on-line experiments manipulate the background pictures and colors of a Web page, affecting consumer product choice. We demonstrate that these effects occur for both experts and novices. Slide36: Response modes 1. Pricing: buying price, selling price, cash equivalent 2. Choice: e.g., gamble vs. sure gain (e.g., $2) 3. Attractiveness rating Evaluating the Attractiveness of a Bet We would like you to indicate how attractive the prospect of playing the following bet is to you. The bet is 7/36 to win $9.00. This means that there are 7 chances out of 36 that you will win the bet and receive $9.00 and 29 chances out of 36 that you will win nothing. Visualize a roulette wheel on the left with 36 numbers along the circumference. If a ball lands on any of the 7 numbers between 1 and 7 inclusive, you win $9.00. If it lands on numbers 8-36, you win nothing. Strategic Equivalence:Bet Preference: Strategic Equivalence: Bet Preference Play the bet you chose Two bets will be selected randomly: Play the bet that had the higher cash equivalent (price) had the higher attractiveness rating or or Slide38: Attractiveness Rating Slide39: Preference Reversals: People choose bets with higher probabilities of winning but smaller payoffs and also rate these bets as more attractive But they assign higher buying and selling prices to bets with lower probabilities of winning larger payoffs. Example: Slide40: A conjoint study showed that rated attractiveness was influenced far more by probability then by payoff even though respondents thought they were weighting both attributes about equally. Pr obab i l i t y Payoff Slide41: A Compatibility Explanation Probability maps easily onto the 0-20 attractiveness scale Payoff is less easily mapped: e.g. how attractive is $9? Perhaps adding a small loss would make payoff more compatible (e.g. more precisely mapped into the response) and then be given more weight (see e.g. Mellers, Richards, & Birnbaum 1992: Distributional theories of impression formation) A “mutant” bet was created to enhance attractiveness: 7/36 win $9 Became: 7/36 win $9 29/36 lose 5¢ Attractiveness of Simple Gambles: Gamble 1. 7/36 chance to win $9 Otherwise win nothing Gamble 2. 7/36 chance to win $9 29/36 chance to lose 5¢ Attractiveness of Simple Gambles Mean attractiveness (0-20 scale) 9.4 14.3 Slide43: Choice Responses Enter affect:: Enter affect: A valenced quality (e.g., goodness or badness) associated with a stimulus Affect is one of many elements in preference construction (but a powerful one) Slide46: Affect Information Affect conveys meaning upon information Without affect, information lacks meaning and will not be used in judgment and decision making Affect is a key ingredient of rational behavior Affect sometimes leads to poor decision making Meaning Slide47: Evaluability Chris Hsee (University of Chicago) “To say that an attribute is hard to evaluate . . . means that people do not know whether a given value on the attribute is good or bad . . . Attributes of Two Dictionaries in Hsee’s Study: Attributes of Two Dictionaries in Hsee’s Study Source: Adapted from Hsee (1998) Slide51: An Affect Account of the 5¢ Loss Effect Probability is evaluable. 7/36 is a poor chance. Payoff is less evaluable. How good or bad is $9? Adding the 5¢ loss makes $9 “come alive with feeling” and it then becomes weighted in the judgment. Slide52: Results: April 2003 $9 is most positive when paired with the 5¢ loss (Condition B) the 5¢ loss is rated more positively than +5¢ and 0 Affect for $9 predicts attractiveness ratings – but highest predictability is in Condition B Affect for “the other outcome” predicts attractiveness only in B (-5¢): liking for –5¢ predicts in positive direction What Have We Learned from the Gambles?: What Have We Learned from the Gambles? the importance of contextual factors in determining affect and preference the fact that meaning, utility, and weighting of even a very familiar and “well-understood” monetary outcome such as $9 is not fixed, but depends greatly on these contextual factions Slide55: D 2 D 1 Y Z X Adding Z to the set of options increases the probability that X will be selected over Y. Asymmetric Dominance Affect Account Z makes X “look good” Slide56: (Psychology & Marketing, 1999) Slide57: Images Affect Behavior Psycholinguistics (Osgood – 1952) Connotative meaning of words dominated by evaluative (good/bad) dimension. Learning Theory (Mowrer – 1960) Behavior is guided and controlled by conditioned emotional reactions to images. Neurology (Damasio – 1994) Thought is made largely from images, broadly construed to include perceptual and symbolic representations. Through learning, images become “marked” by positive and negative feeling states (Somatic Markers). A negative SM linked to image is “cause for alarm” A positive SM is a “beacon of incentive” Social Psychology (Epstein – 1994) Dual Process Theories of Cognition The Bigger Picture: Converging Views Slide58: There is no dearth of evidence in everyday life that people apprehend reality in two fundamentally different ways, one variously labeled intuitive, automatic, natural, non-verbal, narrative, and experiential, and the other analytical, deliberative, verbal, and rational. Seymour Epstein; 1994, p. 710 Slide60: Some Elements of the “Affect Story” Evaluating gambles Destination preferences (vacations, jobs, retirement) Risk perception/communication (dangerousness) Marketing/advertising/promotion names, labels, images, packaging Protective measures/insurance/life saving Stigmatization of places, products, technologies Investment judgments and decisions Punitive damage awards Youth smoking Intuitive Toxicology — Main Result: Intuitive Toxicology — Main Result Many people lack dose-response sensitivity for exposure to chemicals that can produce effects that are dreaded, such as cancer (high affect). If large exposures are bad, small exposures are also bad. Terrorism and Probability NeglectCass R. SunsteinThe Journal of Risk and Uncertainty, 26(2/3); 121-136, 2003: Terrorism and Probability Neglect Cass R. Sunstein The Journal of Risk and Uncertainty, 26(2/3); 121-136, 2003 People are prone to . . . probability neglect, especially when their emotions are intensely engaged. Probability neglect is highly likely in the aftermath of terrorism. People fall victim to probability neglect if and to the extent that the intensity of their reaction does not greatly vary even with large differences in the likelihood of harm. When probability neglect is at work, people’s attention is focused on the bad outcome itself, and they are inattentive to the fact that it is unlikely to occur. Slide65: Gain Value Value functions based on calculations (dotted line) and based on feeling (solid line). The x-axis of the function is the scope of a stimulus, and the y-axis is subjective value. Source: Hsee & Rottenstreich. Strong Affect Distorts Valuation Slide67: Much behavioral research has been devoted to illustrations of choices that violate the logic of the economic model. The implied claim is that people do not have preferences, in the sense in which that term is used in economic theory. It is therefore fair to ask: if people do not have economic preferences, what do they have instead? …Statements of WTP are better viewed as expressions of attitudes than as indications of economic preferences. The core of an attitude is a valuation, which assigns to the entity an affective value that can range from extremely positive to extremely negative. (Kahneman, Ritov, and Schkade, 1993) Slide68: Valuation: Insensitivity to scope Desvousges et al. (1992) asked separate groups of participants how much they would donate to save 2,000, 20,000, or 200,000 migrating birds from drowning in uncovered oil ponds. The mean responses were $80, $78, and $88 respectively. Kahneman et al. (1999) argue that these results could be explained by the notion that these questions evoke a mental representation of a prototypical incident, perhaps an image of an exhausted oil-soaked bird and that respondents decide how much to donate based on their affective reactions to this image. Slide72: Defining Value If values are constructed during the elicitation process in a way that is strongly determined by context and has profound effects on the resultant evaluation, we should take a deliberate approach to value construction in a manner designed to rationalize the process. Slide74: Key Concept: The Construction of Preference Extensive research in human judgment and decision making indicates that people often do not hold values that are well defined in monetary units. Unless the items valued are simple and familiar—and most environmental goods are neither—individuals will construct values during the elicitation process, based on the context that is provided. Slide75: The underlying assumption of the approach to be discussed in this article is that people have strong feelings, beliefs, and values for many things that are not sold through markets. However, people’s cognitive beliefs about these values typically are not numerically quantified and, most importantly for CV, are not represented monetarily. The fact that people are not used to thinking about environmental goods in monetary units suggests that a CV approach must function as a kind of tutorial, building the monetary value as it elicits it. We therefore view a CV survey as an active process of value construction (Tversky, Sattath, & Slovic, 1988), rather than as a neutral process of value discovery. Slide76: Thus, we believe, the designers of a CV study should function not as archaeologists, carefully uncovering what is there, but as architects, working to build a defensible expression of value. In this article, we first argue that CV methods need to be changed to accommodate the constructive nature of environmental preferences. We then propose criteria to guide the selection of a defensible environmental-values-elicitation method. Next, we examine the possibility of developing a new CV method based on techniques derived from multi-attribute utility theory and decision analysis. A Constructive Method:Multi-Attribute Utility Theory: A Constructive Method: Multi-Attribute Utility Theory Structure the problem Explicit, comprehensive mixture of all components of value Consult technical experts and affected citizenry Attributes (effects on genetic diversity, tourist values, etc.) Subattributes—scenic beauty Measures of subattributes—number of dead fish/acres of beach A Constructive Method: Multi-Attribute Utility Theory (cont.): A Constructive Method: Multi-Attribute Utility Theory (cont.) 2. Assess utilities 0-100 scales (minimum/maximum) Allow non-linearity Assess relative weights for scales 3. Calculate total value (on 1 or perhaps several dimensions) 4. Perform sensitivity analysis Advantages of MAUT/CV: Advantages of MAUT/CV Accomodates the multidimensionality of value Integrates market and non-market values Lessons embedding Asks the right question—how valuable is this (not WTP)? May reduce “irrelevant” affect (experiential/System 1) Aids careful thought (analytic/System 2) Open to scrutiny and debate Case studies using constructed preference methods:: Case studies using constructed preference methods: - Gregory, R., & Failing, L. (2002). Using decision analysis to encourage sound deliberation: Water use planning in British Columbia, Canada. Journal of Policy Analysis and Management, 21, 492-499. - Gregory, R.S., & Keeney, R.L. (2002). Making smarter environmental management decisions. Journal of the American Water Resources Association, 38, 1601-1612. - Keeney, R. L., & Gregory, R. (2005). Selecting attributes to measure the achievement of objectives. Operations Research, 53(1), 1-11. - Trousdale, W., & Gregory, R. (2004). Property evaluation and biodiversity conservation: Decision support for making hard choices. Ecological Economics, 48, 279-291.