Bibliography
§2.2: utility subjective/objective, as relation between man and object %}
Download 7.23 Mb.
|
- Navigate this page:
- SG higher than CE ; adaptive utility elicitation ; CE bias towards EV
- SG higher than CE ; adaptive utility elicitation , CE bias towards EV
| §2.2: utility subjective/objective, as relation between man and object %}
Mongin, Philippe & Claude d’Aspremont (1998) “Utility Theory and Ethics.” In Salvador Barberà, Peter J. Hammond, & Christian Seidl (eds.) Handbook of Utility Theory, Vol. I Principles, 371–481, Kluwer Academic Publishers, Dordrecht. {% An impressive paper giving many valuable preference foundations. They assume a two-dimensional product space i=1,…,nj=1,…,mxij. Say there is both time and uncertainty, with n states of nature and m time points. One of the components can also refer to persons or commodities or other things. The first basic result, which in itself has been known before as the authors cite, is: Assume that we only have separability of all rows and all columns. This, by Gorman’s (1968) theorem, is already enough to give full separability and an overall additive representation. This particular form of Gorman’s theorem has a long history as the problem of aggregation in economics. (Can we just take aggregate demand of every commodity in the market and then aggregate over individuals, or should we first aggregate over individuals.) My Rotterdam predecessors Van Daal & Merkies worked on this. The result is so nice because the separability of columns and rows just feels like weak monotonicity. Then, as the authors show in their Theorem 1 (p. 156), because for every row, for instance, we already have a cardinal representation, requiring ordinal identity of conditional preferences give that these rows have the same representation up to one positive factor. Doing this for columns too, we get a weighted-average representation as with EU and discounted utility while avoiding extra conditions such as bisymmetry, tradeoff consistency, or Savage’s (1954) P4. This result is not very new or very deep, but very nice and useful, and gives a host of applications and improvements on existing results. It gives a generalized version of Harsanyi (1955) and Anscombe-Aumann (1963), allowing subjective probabilities in the second stage. The authors also handle quite general subsets of product sets, as in Segal (1992) and Chateauneuf & Wakker (1993). %} Mongin, Philippe & Marcus Pivato (2015) “Ranking Multidimensional Alternatives and Uncertain Prospects,” Journal of Economic Theory 157, 146–171. {% Dutch book: extend it to many-valued events and infinitesimal probabilities. %} Montagna, Franco, Martina Fedel, & Giuseppe Scianna (2013) “Non-Standard Probability, Cherence and Conditional Probability on Many-Valued Events,” International Journal of Approximate Reasoning 54, 573–589. {% Proves that a quasi-concave separable function on an atomless space is concave. For usual additive separable representations with finite dimensions, V = V1 + ... + Vn, we have a state space S = {s1, …, sn} and a function, act, x = (x1,…,xn) and V represents preferences over acts. One can say that S is endowed with the discrete counting measure (sj) = 1 for all j and that Vj is state-dependent utility, and V state-dependent expected utility. When Wakker & Zank (1999) extended this to infinite state spaces S, one unanticipated difficulty was writing the very definition of V, in the absence of a measure on S such that V would be absolutely continuous with respect to that measure, so that V could not be written as a kind of integral. This paper studies state-dependent EU functionals on infinite, even atomless, state spaces that are endowed with a measure so that they can be written as an integral. The set of acts is taken as Lp+. The state-dependent functional is called separable. The state-dependent utility is called kernel. It cites mathematical literature on this, e.g. on continuity results. It shows that for a separable function quasi-concavity implies concavity. %} Monteiro, Paulo Klinger (1999) “Quasiconcavity and the Kernel of a Separable Utility,” Economic Theory 13, 221–227. {% Title: because responders rather accept lower share than risking being left out. %} Montero, Maria (2007) “Inequity Aversion May Increase Inequity,” Economic Journal 117, C192–C204. {% Proposes a measure of risk aversion, in addition to Pratt-Arrow, that vanishes locally under expected utility but need not vanish under nonEU. This shows that there can be first-order risk aversion under nonEU, and anticipates somewhat the first-order risk aversion in Segal & Spivak (1990). %} Montesano, Aldo (1985) “The Ordinal Utility under Uncertainty and the Measure of Risk Aversion in Terms of Preferences,” Theory and Decision 18, 73–85. {% P. 282: proposes local risk measures of 1st and 2nd order, based on normalized risk premiums, where the 2nd order agrees with Pratt-Arrow if EU and the 1st order is 0 under EU (differentiable utility). 1st order can be nonzero under nonEU. Proposes global measures by integrating over p over [0,1]. Paper is not easy to read because the mathematical derivations are not separated from their results. Incorporates multivariate measures (also studied by Bob Nau (2003). %} Montesano, Aldo (1988) “The Risk Aversion Measure without the Independence Axiom,” Theory and Decision 24, 269–288. {% %} Montesano, Aldo (1991) “Measures of Risk Aversion with Expected and Nonexpected Utility,” Journal of Risk and Uncertainty 4, 271–283. {% Explains how de Finetti (1952) had the Pratt-Arrow risk aversion index u´´/u´ as index of risk aversion. de Finetti established some local results, but not the nicest result, the one relating to lower certainty equivalents. %} Montesano, Aldo (2009) “De Finetti and the Arrow-Pratt Measure of Risk Aversion.” In Maria Carla Galavotti (ed.) Bruno de Finetti, Radical Probabilist, 115–127, College Publications, London. {% Defines uncertainty aversion as follows: if there EXISTS a subjective probability measure with EU under which all CEs (certainty equivalents) are larger (Def. 1 p. 136). So, this is the same as Ghirardato & Marinacci (2002, JET), taking probabilistic sophistication + EU as ambiguity neutrality. Under CEU (Choquet expected utility) it is equivalent to nonempty CORE. Schmeidler’s condition of preference for probabilistic mixture is called increasing uncertainty aversion (Def. 2 pp. 136-137). They show that the latter implies uncertainty aversion, but not vice versa. Section 4, nicely, proposes to relate uncertainty aversion to the nucleolus of the weighting function. It next proposes some definitions of ambiguity premiums, following up on Hilton. %} Montesano, Aldo & Francesco Giovannoni (1996) “Uncertainty Aversion and Aversion to Increasing Uncertainty,” Theory and Decision 41, 133–148. {% On his dominance search theory: in choice subjects try to (mis)perceive things such that they can claim their choice to be based on dominance. %} Montgomery, Henry (1983) “Decision Rules and the Search for a Dominance Structure: Towards a Process Model of Decision Making.” In Patrick C. Humphreys, Ola Svenson, & Anna Vari (eds.) Analyzing and Aiding Decision Processes, 343–369, North-Holland, Amsterdam. {% On his dominance search theory: in choice subjects try to (mis)perceive things such that they can claim their choice to be based on dominance. Unfortunately, he did not publish this in a journal, but only in 1983 & 1989 book chapters. %} Montgomery, Henry (1989) “From Cognition to Action: The Search for Dominance in Decision Making.” In Henry Montgomery & Ola Svenson (eds.) Process and Structure in Human Decision Making, 23–49, Wiley, Oxford. {% In augustus 1992 me aangeraden door Pat Suppes. %} Moody, Ernest A. & Marshall Clagett (1960, eds.) “The Medieval Science of Weights.” University of Wisconsin Press, Madison. {% %} Moon, John W. (1968) “Topics on Tournaments.” Holt, Rinehart and Winston, New York. {% Nice reconciliation of 3 kinds of overconfidence: (a) overestimation of one's actual performance, (b) overplacement of one's performance relative to others, and (c) excessive precision in one's beliefs. %} Moore, Don A. & Paul J. Healy (2008) “The Trouble with Overconfidence,” Psychological Review 115, 502–517. {% %} Moore, Don A, Terri Kurtzberg, Craig R. Fox, & Max H. Bazerman (1999) “Positive Illusions and Forecasting Errors in Mutual Fund Investment Decisions,” Organizational Behavior and Human Decision Processes 79, 95–114. {% Stigler (1950, end of §VII, gives a nice citation where Moore nicely formulates how economics aims to become an exact science through utility, albeit in negative terms because Moore does not like it. %} Moore, Henry L. (1914) “Economic Cycles: Their Law and Cause.” MacMillan, New York. {% Extends the “unit of measurement” method of Wold (1943) to measure cardinal utility, to nonhomothetic preferences. %} Moore, James C. (1983) “Measurable Triples and Cardinal Measurement,” Journal of Economic Theory 29, 120–160. {% %} Moore, Mike J. & W. Kip Viscusi (1990) “Models for Estimating Discount Rates for Long-term Health Risks Using Labor Market Data,” Journal of Risk and Uncertainty 3, 381–401. {% real incentives/hypothetical choice: a thorough discussion of the hypothetical bias and its literature, although focusing only on WTP. The authors propose a model where the weighting of attributes is differently for hypothetical than for real. In their data (subjects expressing WTP for apples, real or hypothetical), surprisingly, the hypothetical subjects pay more time to their decision making and ignore fewer attributes. %} Mørkbak, Morten Raun, Søren Bøye Olsen, & Danny Campbell (2014) “Behavioral Implications of Providing Real Incentives in Stated Choice Experiments,” Journal of Economic Psychology 45, 102–116. {% In what is an experienced decision task as in Barron & Erev (2003) and many follow-up pappers (although the authors do not cite this), monkeys and children prefer risky option to its expected value. This is easily explained because the risky choices provide more info (because the monkeys and children do not know the probabilities and have to find out about them) than the safe choices, and the monkeys and children do not only choose for preference value but also for obtaining more info. %} Moreira, Bruno, Raul Matsushita & Sergio Da Silva (2010) “Risk Seeking Behavior of Preschool Children in a Gambling Task,” Journal of Economic Psychology 31, 794–801. {% Subjects sample, with replacement, from risky, compound, and ambiguous urns. They weigh the new observations more (so, the prior info less) for ambiguous than for compound risk. %} Moreno, Othon M. & Yaroslav Rosokha (2016) “Learning under Compound Risk vs. Learning under Ambiguity – An Experiment,” Journal of Risk and Uncertainty 53, 137–162. {% Banks are, because of the nature of their business without physical assets, opaque in their risk; i.e., there are more unknown probabilities and there is more ambiguity as decision theorists would call it. A proxy to measure this degree of ambiguity is the disagreement between raters. Next to insurance, banks indeed have that the highest. %} Morgan, Donald P. (2002) “Rating Banks: Risk and Uncertainty in an Opaque Industry,” American Economic Review 92, 874–888. {% probability elicitation for continuous distributions. %} Morgan, M. Granger & Max Henrion (1990) “Uncertainty: A Guide to Dealing with Uncertainty in Quantitative Risk and Policy Analysis.” Cambridge University Press, New York. {% %} Morgan, John & Martin Sefton (2000) “Funding Public Goods with Lotteries: Experimental Evidence,” Review of Economic Studies 67, 783–810. {% %} Morgan, M. Granger (1993) “Risk Analysis and Management,” Scientific American 32 (July), 32–41. {% %} Morgan, Robert M. & Shelby Hunt (1994) “The Commitment-Trust Theory of Relationship Marketing,” Journal of Marketing 58, 20–38. {% Seems to have said that he and von Neumann never intended EU for very small probabilities. “For example, the probabilities used must be within certain plausible ranges and not go to .01 or even less to .001, then be compared to other equally tiny numbers such as .02, etc.” %} Morgenstern, Oskar (1979) “Some Reflections on Utility.” In Maurice Allais & Ole Hagen (eds.) Expected Utility Hypotheses and the Allais Paradox, 175–183, Reidel, Dordrecht. {% real incentives/hypothetical choice: consider simple choices between a sure outcome and a prospect. Do it both hypothetically and with real incentives. Find the usual bigger risk aversion for real incentives. But they also do EEG measurements to study neuronal effects. The abstract ends with “A higher N2 component for hypothetical payoffs revealed increased cognitive control for hypothetical decisions. These neuronal underpinnings indicate additional evaluation processes in hypothetical choice paradigms, which can explain the shift in risk attitude toward the expected value of a lottery.” They suggest that hypothetical may be cognitively better! (cognitive ability related to risk/ambiguity aversion) On hypothetical choice the authors, appropriately, write: “However, we also have to consider that there are special cases in which a realization of decision outcomes is not possible. For instance, outcomes related to questions of environmental damages, moral conflicts, losses, or very high stakes are often not realizable. In those cases, hypothetical decisions may still provide valuable information as good forecast indicators.” (p. 558; real incentives/hypothetical choice) %} Morgenstern, Ralf, Marcus Heldmann, & Bodo Vogt (2014) “Differences in Cognitive Control between Real and Hypothetical Payoffs,” Theory and Decision 77, 557–582. {% Collect data like Hey & Orme (1994), and fit four functionals: EU, disappointment aversion, RDU with power probability weighting, and RDU with the Tversky & Kahneman one-parameter family (the authors erroneously credit Quiggin 1982 for it). In their first analysis, they do within subject testing, assuming that within-subject choices are statistically independent which I find problematic. Their second test considers for each individual which theory fits best, second-best, and so on. Problem here is that close theories kill each others’ chances, in the same way as Nadar made Gore lose to Bush. According to the criteria used, EU is best, disappointment aversion second, RDU with power utility is third, and RDU with T&K weighting is fourth and last. %} Morone, Andrea & Piergiuseppe Morone (2014) “Estimating Individual and Group Preference Functionals Using Experimental Data,” Theory and Decision 77, 323–339. {% %} Morrell, Darryl R. (1993) “Epistemic Utility Estimation,” IEEE Transactions on Systems, Man, and Cybernetics 23, 129–140. {% %} Morris, Stephen (1994) “Trade with Heterogeneous Prior Beliefs and Asymmetric Information,” Econometrica 62, 1327–1347. {% Discusses much literature on the common prior assumption, such as Carnap. %} Morris, Stephen (1995) “The Common Prior Assumption in Economic Theory,” Economics and Philosophy 11, 227–253. {% Generalizes Morris & Shin (1997, ET) to nonEU. %} Morris, Stephen (1996) “The Logic of Belief and Belief Change: A Decision Theoretic Approach,” Journal of Economic Theory 69, 1–23. {% Paper shows that individuals’ willingness to bet will exhibit a bid ask spread property in the presence of heterogeneous prior beliefs and asymmetric information. Pp. 236-237: ”It is true that it is possible to imagine environments where strategic considerations are ruled out, and our individual nonetheless displays uncertainty aversion. However, it is argued that such situations are unlikely to be economically relevant.” Footnote 24: “It would be interesting to test how sensitive Ellsberg-paradox-type phenomena are to varying emphasis in the experimental designs on the experimenter’s incentives.” %} Morris, Stephen (1997) “Risk, Uncertainty and Hidden Information,” Theory and Decision 42, 235–269. {% %} Morris, Stephen (1997) “Alternative Notions of Knowledge.” In Michael Bacharach, Louis-André Gérard-Varet, Philippe Mongin, & Hyun Song Shin (1997, eds.) Epistemic Logic and the Theory of Games and Decisions, 217–234, Kluwer Academic Press, Dordrecht. {% %} Morris, Stephen, Andrew Postlewaite, & Hyun Song Shin (1995) “Depth of Knowledge and the Effect of Higher Order Uncertainty,” Economic Theory 6, 453–467. {% %} Morris, Stephen, Rafael Rob, & Hyun Song Shin (1995) “p-Dominance and Belief Potential,” Econometrica 63, 145–157. {% value of information: Savagean EU maximizer can do decision with or without further info. Info can be favorable, leading to higher EU state, or unfavorable, leading to lower EU state. (This is different thing than Blackwell-like, as authors explain p. 310 bottom.) The authors give conditions for info to be valuable. Generalizations to nonEU by Morris (1996, JET) where essentially the same results hold. Here belief is through a logical operator. %} Morris, Stephen & Hyun Song Shin (1997) “Rationality and Efficacy of Decisions under Uncertainty,” Economic Theory 9, 309–324. {% %} Morrison, Gwendolyn C. (1997) “HYE and TTO: What Is the Difference?,” Journal of Health Economics 16, 563–578. {% %} Morrison, Gwendolyn C. (1997) “Resolving Differences in Willingness to Pay and Willingness to Accept: Comment,” American Economic Review 87, 236–240. {% SG higher than CE; adaptive utility elicitation; CE bias towards EV: not exactly that, but, endowment effect induced bias of CE (certainty equivalent) towards risk seeking. Seems to find, as do Hershey & Schoemaker (1982), that in standard gamble choices people focus on the sure outcome as their reference point. %} Morrison, Gwendolyn C. (2000) “The Endowment Effect and Expected Utility,” Scottish Journal of Political Economy 47, 183–197. {% SG higher than CE; adaptive utility elicitation, CE bias towards EV: as in Morrison (2000, Scottish JPE %} Morrison, Gwendolyn C. (2000) “Expected Utility and the Endowment Effect: Some Experimental Results,” Dept of Economics, University of Nottingham, UK. {% adaptive utility elicitation %} Morrison, Gwendolyn C., Aileen R. Neilson, & Mo Malek (1998) “Improving the Sensitivity of the Time Trade-Off Method: Results of an Experiment Using Chained TTO Questions,” {% questionnaire versus choice utility: a metastudy on conversions of introspection-based measurements into revealed-preference based utilities. Impressive! Using strict selection criteria, they were left with 46 empirical studies and 16 further studies shedding light on the topic. One thing they conclude (p. 87 2nd column) is that discrepancies depend more on the domain (which disease) than on the method used. They use the term descriptive measure for introspective-based measures and the term QALY for decision-utility based. P. 67 top explains that often for practical reasons we cannot get revealed-preference based measurements and have to do with introspective measurements. %} Mortimer, Duncan & Leonie Segal (2008) “Comparing the Incomparable? A Systematic Review of Competing Techniques for Converting Descriptive Measures of Health Status into QALY-Weights,” Medical Decision Making 28, 66–89. {% risky utility u = strength of preference v (or other riskless cardinal utility, often called value): it has been widely understood that cardinality of utility has two different meanings. First, just the mathematical property of uniqueness up to unit and location. Second, that it can be given psychological interpretations. This paper discusses the issue anew, adding new literature. In the beginning of §2, the author writes “In his Manual, Pareto ([1909] 1971: 112 and 396) maintained that utility cannot be measured; i.e., that it is impossible to identify a unit of utility and express the utility of commodities as a multiple of that unit.” I regret that the author, as did so many, leaves out the crucial premise that Pareto added. Pareto made his claim only for the case where we only want to explain market demand and equilibrium. %} Moscati, Ivan (2013) “How Cardinal Utility Entered Economic Analysis, 1909-1944,” European Journal of the History of Economic Thought 20, 906–939. {% The most central idea in decision under uncertainty, and the divding line between Bayesian EU and nonEU, is the sure-thing principle. It was Savage's (1954) main invention. How did the idea come about? I have wondered since my youth. It was mainly in exchanges between Samuelson and Savage, two of the greatest minds ever. This paper carefully documents the history and origin of the idea. It is very valuable to me, answering questions I had since my youth. The crucial point why the sure-thing principle is normative, is that it concerns separability about mutually exclusive events, between which no physical interaction is possible. (The interaction is only in the, confused, minds of nonEU maximizers.) P. 225 cites a May 11 1950 letter by Marschak who points it out to Samuelson, but Samuelson’s reaction is confused. He brings in utility and is confused that utility of tea and pretzles will interact, which is besides the point. P. 227 middle cites Samuelson (1950a) on properly criticizing the Friedman-Savage EU explanation of gambling and insurance with EU. P. 229 cites Sept. 13, 1950 letter by Friedman where Friedman that under EU all preferences are completey determined by binary gambles: “Dear Paul: … It has never seemed to me obviously true or necessary that individual’s reactions to complicated gambles should be completely predictable from their reactions to two-side ones—which has always seemed to me the fundamental empirical content of the B[ernoulli]–M[arshall] hypothesis” P. 230 brings up Savage’s letter of August 12, 1950, where he first formulates the sure-thing principle as a form of event-wise monotonicity (in the same way that every separability can be written as monotonicity). On p. 231 this paper suggests that Savage (1954) used the term sure-thing principle only for his P2. But this is not so. It also included P3 (monotonicity w.r.t. outcomes) and P7. Only later it became a tradition in the field to use the term sure-thing principle only for P2, a tradition that I follow. P. 231 shows that Samuelson had changed his mind on EU, and now considered it normative, in his letter to Friedman of August 25, 1950. It is nowhere stated that the mutual exclusiveness of events played a role in Samuelson’s considerations, whereas my memory (I read the relevant letters in the early 1990s) tells it did; but I must have been confused. %} Moscati, Ivan (2016) “How Economists Came to Accept Expected Utility Theory: The Case of Samuelson and Savage,” Journal of Economic Perspectives 30, 219–236. {% Seems to find violations of RCLA; my bibliographic info below seems to be incorrect. %} Moser, Donald V., Jacob G. Birnberg, & Sangho Do (1994) “A Similarity Strategy for Decisions Involving Sequential Events,” Accounting, Organization and Society 19, 439–458. {% Gotten from Ido Erev on 5 sept. 1990; used real incentives. %} Moskowitz, Herbert (1974) “Effects of Problem Representation and Feedback on Rational Behavior in Allais and Morlat-Type Problems,” Decision Sciences 5, 225–242. {% EU analysis if probabilities and utilities are not precisely known but are only inferred up to certain limits from observed choices. %} Moskowitz, Herbert, Paul V. Preckel, & Aynang Yang (1993) “Decision Analysis with Incomplete Utility and Probability Information,” Operations Research 41, 864–879. {% %} Moskowitz, Tobias J. & Annette Vissing-Jorgensen (2002) “The Returns to Entrepreneurial Investment: A Private Equity Premium Puzzle?,” American Economic Review 92, 745–778. {% Characterize maxmin choice. %} Mosquera, Manuel, Peter Borm, M. Gloria Fiestras-Janeiro, Ignacio Garcia-Jurado, & Mark Voorneveld (2008) “Characterizing Cautious Choice,” Mathematical Social Sciences 55, 55, 143–155. {% Seems to show that under actuarially unfair coinsurance (loading factor in insurance premim) and EU with concave utility, no complete insurance is taken. %} Mossin, Jan (1968) “Aspects of Rational Insurance Purchasing,” Journal of Political Economy 76, 553–568. {% %} Mossin, Jan (1969) “A Note on Uncertainty and Preferences in a Temporal Context,” American Economic Review 59, 172–174. {% Arrow (1982): first empirical test of EU. P. 377: they deceived subjects by giving them more money than said. (deception when implementing real incentives) P. 383 mentions the Utility of gambling. Pp. 402 discusses it more. “Indeed, the writers would prefer to defer discussion of this point until a way of testing arguments about it is provided.” P. 385, end of 2nd column: a subject who violates probabilistic reduction by gambling rather on one hand than the other Real incentives: repeated gambles for money, all with real incentives. Losses were also implemented. A losses from prior endowment mechanism was used although it might in extreme cases not cover all losses. P. 399 mentions that this gives an income effect. P. 400 mentions house money effect, that subjects befome more risk seeking after prior gains. Download 7.23 Mb. Share with your friends:
|