the theory of decision making - semantic scholar · bar. the economic theory of decision making is...

PSYCHOLOGICAL BULLETINVol. 51, No. 4, 1954

THE THEORY OF DECISION MAKING1

WARD EDWARDSThe Johns Hopkins University

Many social scientists other thanpsychologists try to account for thebehavior of individuals. Economistsand a few psychologists have pro-duced a large body of theory and afew experiments that deal with indi-vidual decision making. The kind ofdecision making with which this bodyof theory deals is as follows: giventwo states, A and B, into either one ofwhich an individual may put himself,the individual chooses A in prefer-ence to B (or vice versa). For in-stance, a child standing in front of acandy counter may be consideringtwo states. In state A the child has$0.25 and no candy. In state B thechild has $0.15 and a ten-cent candybar. The economic theory of decisionmaking is a theory about how to pre-dict such decisions.

Economic theorists have been con-cerned with this problem since thedays of Jeremy Bentham (1748-1832). In recent years the develop-ment of the economic theory of con-sumer's decision making (or, as the

1 This work was supported by ContractN5ori-166, Task Order I, between the Officeof Naval Research and The Johns HopkinsUniversity. This is Report No. 166-1-182,Project Designation No. NR 145-089, underthat contract. I am grateful to the Depart-ment of Political Economy, The Johns Hop-kins University, for providing me with anoffice adjacent to the Economics Librarywhile I was writing this paper. M. Allais,M. M. Flood, N. Georgescu-Roegen, K. O.May, A. Papandreou, L. J. Savage, and es-pecially C. H. Coombs have kindly mademuch unpublished material available to me.A number of psychologists, economists, andmathematicians have given me excellent, butsometimes unheeded, criticism. Especiallyhelpful were C. Christ, C. H. Coombs, F.Mosteller, and L. J. Savage.

380

economists call it, the theory of con-sumer's choice) has become exceed-ingly elaborate, mathematical, andvoluminous. This literature is almostunknown to psychologists, in spite ofsporadic pleas in both psychological(40, 84, 103, 104) and economic(101, 102, 123, 128, 199, 202) litera-ture for greater communication be-tween the disciplines.

The purpose of this paper is to re-view this theoretical literature, andalso the rapidly increasing number ofpsychological experiments (performedby both psychologists and econo-mists) that are relevant to it. Thereview will be divided into five sec-tions: the theory of riskless choices,the application of the theory of risk-less choices to welfare economics, thetheory of risky choices, transitivity indecision making, and the theory ofgames and of statistical decisionfunctions. Since this literature is un-familiar and relatively inaccessible tomost psychologists, and since I couldnot find any thorough bibliographyon the theory of choice in the eco-nomic literature, this paper includesa rather extensive bibliography of theliterature since 1930.

THE THEORY OF RISKLESS CHOICES"Economic man. The method of

those theorists who have been con-2 No complete review of this literature is

available. Kauder (105, 106) has reviewed thevery early history of utility theory. Stigler(180) and Viner (194) have reviewed theliterature up to approximately 1930. Samuel-son's book (164) contains an illuminatingmathematical exposition of some of the con-tent of this theory. Allen (6) explains the con-cept of indifference curves. Schultz (172) re-

THEORY OF DECISION MAKING 381

cerned with the theory of decisionmaking is essentially an armchairmethod. They make assumptions,and from these assumptions they de-duce theorems which presumably canbe tested, though it sometimes seemsunlikely that the testing will everoccur. The most important set ofassumptions made in the theory ofriskless choices may be summarizedby saying that it is assumed that theperson who makes any decision towhich the theory is applied is aneconomic man.

What is an economic man like? Hehas three properties, (a) He is com-pletely informed. (6) He is infinitelysensitive, (c) He is rational.

Complete information. Economicman is assumed to know not onlywhat all the courses of action open tohim are, but also what the outcome ofany action will be. Later on, in thesections on the theory of risky choicesand on the theory of games, this as-sumption will be relaxed somewhat.(For the results of attempts to in-troduce the possibility of learninginto this picture, see 51, 77.)

Infinite sensitivity. In most of theolder work on choice, it is assumedthat the alternatives available to anindividual are continuous, infinitelydivisible functions, that prices areinfinitely divisible, and that economicman is infinitely sensitive. The onlypurpose of these assumptions is tomake the functions that they lead to,

views the developments up to but not includ-ing the Hicks-Allen revolution from the pointof view of demand theory. Hicks's book (87)is a complete and detailed exposition of mostof the mathematical and economic content ofthe theory up to 1939. Samuelson (167) hasreviewed the integrability problem and the re-vealed preference approach. And Wold (204,205, 206) has summed up the mathematicalcontent of the whole field for anyone who iscomfortably at home with axiom systems anddifferential equations.

continuous and differentiable. Stone(182) has recently shown that theycan be abandoned with no seriouschanges in the theory of choice.

Rationality. The crucial fact abouteconomic man is that he is rational.This means two things: He canweakly order the states into which hecan get, and he makes his choices soas to maximize something.

Two things are required in orderfor economic man to be able to put allavailable states into a weak ordering.First, given any two states into whichhe can get, A and B, he must alwaysbe able to tell either that he prefersA to B, or that he prefers B to A, orthat he is indifferent between them.If preference is operationally definedas choice, then it seems unthinkablethat this requirement can ever beempirically violated. The secondrequirement for weak ordering, amore severe one, is that all prefer-ences must be transitive. If economicman prefers A to B and B to C, thenhe prefers A to C. Similarly, if he isindifferent between A and B andbetween B and C, then he is in-different between A and C. It is notobvious that transitivity will alwayshold for human choices, and experi-ments designed to find out whetheror not it does will be described in thesection on testing transitivity.

The second requirement of ra-tionality, and in some ways the moreimportant one, is that economic manmust make his choices in such a wayas to maximize something. This isthe central principle of the theory ofchoice. In the theory of risklesschoices, economic man has usuallybeen assumed to maximize utility. Inthe theory of risky choices, he is as-sumed to maximize expected utility.In the literature on statistical de-cision making and the theory ofgames, various other fundamental

382 WARD EDWARDS

principles of decision making areconsidered, but they are all maximi-zation principles of one sort or an-other.

The fundamental content of thenotion of maximization is that eco-nomic man always chooses the bestalternative from among those opento him, as he sees it. In more techni-cal language, the fact that economicman prefers A to B implies and isimplied by the fact that A is higherthan B in the weakly ordered setmentioned above. (Some theories in-troduce probabilities into the abovestatement, so that if A is higher thanB in the weak ordering, then eco-nomic man is more likely to choose Athan B, but not certain to choose A.)

This notion of maximization ismathematically useful, since it makesit possible for a theory to specify aunique point or a unique subset ofpoints among those available to thedecider. It seems to me psychologi-cally unobjectionable. So many differ-ent kinds of functions can be maxi-mized that almost any point actuallyavailable in an experimental situationcan be regarded as a maximum ofsome sort. Assumptions about maxi-mization only become specific, andtherefore possibly wrong, when theyspecify what is being maximized.

There has, incidentally, been al-most no discussion of the possibilitythat the two parts of the concept ofrationality might conflict. It is con-ceivable, for example, that it mightbe costly in effort (and therefore innegative utility) to maintain a weaklyordered preference field. Under suchcircumstances, would it be "rational"to have such a field?

It is easy for a psychologist to pointout that an economic man who hasthe properties discussed above is veryunlike a real man. In fact, it is soeasy to point this out that psycholo-

gists have tended to reject out ofhand the theories that result fromthese assumptions. This isn't fair.Surely the assumptions contained inHullian behavior theory (91) or inthe Estes (60) or Bush-Mosteller(36, 37) learning theories are no morerealistic than these. The most usefulthing to do with a theory is not tocriticize its assumptions but ratherto test its theorems. If the theoremsfit the data, then the theory has atleast heuristic merit. Of course, onetrivial theorem deducible from theassumptions embodied in the conceptof economic man is that in anyspecific case of choice these assump-tions will be satisfied. For instance,if economic man is a model for realmen, then real men should alwaysexhibit transitivity of real choices.Transitivity is an assumption, but itis directly testable. So are the otherproperties of economic man as amodel for real men.

Economists themselves are some-what distrustful of economic man(119, 156), and we will see in subse-quent sections the results of a num-ber of attempts to relax these as-sumptions.

Early utility maximization theory.The school of philosopher-economistsstarted by Jeremy Bentham andpopularized by James Mill and othersheld that the goal of human action isto seek pleasure and avoid pain.Every object or action may be con-sidered from the point of view ofpleasure- or pain-giving properties.These properties are called the utilityof the object, and pleasure is givenby positive utility and pain by nega-tive utility. The goal of action, then,is to seek the maximum utility. Thissimple hedonism of the future iseasily translated into a theory ofchoice. People choose the alternative,from among those open to them, that

THEORY OP DECISION MAKING 383

leads to the greatest excess of positiveover negative utility. This notion ofutility maximization is the essence ofthe utility theory of choice. It willreappear in various forms through-out this paper. (Bohnert [30] dis-cusses the logical structure of theutility concept.)

This theory of choice was embodiedin the formal economic analyses of allthe early great names in economics.In the hands of Jevons, Walras, andMenger it reached increasingly so-phisticated mathematical expressionand it was embodied in the thinkingof Marshall, who published the firstedition of his great Principles ofEconomics in 1890, and revised it atintervals for more than 30 yearsthereafter (137).

The use to which utility theory wasput by these theorists was to estab-lish the nature of the demand forvarious goods. On the assumptionthat the utility of any good is amonotonically increasing negativelyaccelerated function of the amount ofthat good, it is easy to show that theamounts of most goods which a con-sumer will buy are decreasing func-tions of price, functions which areprecisely specified once the shapes ofthe utility curves are known. This isthe result the economists needed andis, of course, a testable theorem. (Formore on this, see 87, 159.)

Complexities arise in this theorywhen the relations between theutilities of different goods are con-sidered. Jevons, Walras, Menger,and even Marshall had assumed thatthe utilities of different commoditiescan be combined into a total utilityby simple addition; this amounts toassuming that the utilities of differentgoods are independent (in spite ofthe fact that Marshall elsewhere dis-cussed the notions of competinggoods, like soap and detergents, and

completing goods, like right and leftshoes, which obviously do not haveindependent utilities). Edgeworth(53), who was concerned with suchnonindependent utilities, pointed outthat total utility was not necessarilyan additive function of the utilitiesattributable to separate commodities.In the process he introduced the no-tion of indifference curves, and thusbegan the gradual destruction of theclassical utility theory. We shall re-turn to this point shortly.

Although the forces of parsimonyhave gradually resulted in the elimi-nation of the classical concept ofutility from the economic theory ofriskless choices, there have been afew attempts to use essentially theclassical theory in an empirical way.Fisher (63) and Frisch (75) have de-veloped methods of measuring margi-nal utility (the change in utility [u]with an infinitesimal change inamount possessed [Q], i.e., du/dQ)from market data, by making assump-tions about the interpersonal simi-larity of consumer tastes. RecentlyMorgan (141) has used several vari-ants of these techniques, has dis-cussed mathematical and logical flawsin them, and has concluded on thebasis of his empirical results that thetechniques require too unrealisticassumptions to be workable. Thecrux of the problem is that, for thesetechniques to be useful, the com-modities used must be independent(rather than competing or complet-ing), and the broad commodity clas-sifications necessary for adequatemarket data are not independent.Samuelson (164) has shown that theassumption of independent utilities,while it does guatantee interval scaleutility measures, puts unwarrantablysevere restrictions on the nature ofthe resulting demand function. Else-where Samuelson (158) presented,

384 WARD EDWARDS

primarily as a logical and mathe-matical exercise, a method of measur-ing marginal utility by assumingsome time-discount function. Sinceno reasonable grounds can be foundfor assuming one such function ratherthan another, this procedure holds nopromise of empirical success. Mar-shall suggested (in his notion of "con-sumer's surplus") a method of utilitymeasurement that turns out to bedependent on the assumption of con-stant marginal utility of money, andwhich is therefore quite unworkable.Marshall's prestige led to extensivediscussion and debunking of thisnotion (e.g., 28), but little positivecomes out of this literature. Thur-stone (186) is currently attemptingto determine utility functions forcommodities experimentally, but hasreported no results as yet.

Indifference curves. Edgeworth'sintroduction of the notion of in-difference curves to deal with theutilities of nonindependent goods wasmentioned above. An indifferencecurve is, in Edgeworth's formula-tion, a constant-utility curve. Sup-pose that we consider apples andbananas, and suppose that you get

25

_J 200_QL< 15

10LdCO

0 5 10 15 20 25

NUMBER OF BANANAS

FIG. 1. A HYPOTHETICAL INDIFFERENCE MAP

the same amount of utility from10-apples-and-l-banana as you dofrom 6-apples-and-4-bananas. Thenthese are two points on an indiffer-ence curve, and of course there arean infinite number of other points onthe same curve. Naturally, this is notthe only indifference curve you mayhave between apples and bananas. Itmay also be true that you are in-different between 13-apples-and-5-bananas and 5-apples-and-15-banan-as. These are two points on another,higher indifference curve. A wholefamily of such curves is called an in-difference map. Figure 1 presentssuch a map. One particularly usefulkind of indifference map has amountsof a commodity on one axis andamounts of money on the other.Money is a commodity, too.

The notion of an indifference mapcan be derived, as Edge worth derivedit, from the notion of measurableutility. But it does not have to be.Pareto (146, see also 151) was seri-ously concerned about the assump-tion that utility was measurable upto a linear transformation. He feltthat people could tell whether theypreferred to be in state A or state B,but could not tell how much theypreferred one state over the other. Inother words, he hypothesized a utilityfunction measurable only on an ordi-nal scale. Let us follow the usualeconomic language, and call utilitymeasured on an ordinal scale ordinalutility, and utility measured on aninterval scale, cardinal utility. It ismeaningless to speak of the slope, ormarginal utility, of an ordinal utilityfunction; such a function cannot bedifferentiated. However, Pareto sawthat the same conclusions which hadbeen drawn from marginal utilitiescould be drawn from indifferencecurves. An indifference map can bedrawn simply by finding all the com-


binations of the goods involvedamong which the person is indiffer-ent. Pareto's formulation assumesthat higher indifference curves havegreater utility, but does not need tospecify how much greater that utilityis.

It turns out to be possible to de-duce from indifference curves all ofthe theorems that were originally de-duced from cardinal utility measures.This banishing of cardinal utility wasfurthered considerably by splendidmathematical papers by Johnson(97) and Slutsky (177). (In moderneconomic theory, it is customary tothink of an w-dimensional commodityspace, and of indifference hyper-planes in that space, each such hyper-plane having, of course, n— 1 dimen-sions. In order to avoid unsatisfactorypreference structures, it is necessaryto assume that consumers alwayshave a complete weak ordering for allcommodity bundles, or points in com-modity space. Georgescu-Roegen[76], Wold [204, 205, 206, 208],Houthakker [90], and Samuelson[167] have discussed this problem.)

Pareto was not entirely consistentin his discussion of ordinal utility.Although he abandoned the assump-tion that its exact value could beknown, he continued to talk aboutthe sign of the marginal utility co-efficient, which assumed that someknowledge about the utility functionother than purely ordinal knowledgewas available. He also committedother inconsistencies. So Hicks andAllen (88), in 1934, were led to theirclassic paper in which they attemptedto purge the theory of choice of itslast introspective elements. Theyadopted the conventional economicview about indifference curves as de-termined from a sort of imaginaryquestionnaire, and proceeded to de-rive all of the usual conclusions about

consumer demand with no referenceto the notion of even ordinal utility(though of course the notion of anordinal scale of preferences was stillembodied in their derivation of in-difference curves). This paper wasfor economics something like the be-haviorist revolution in psychology.

Lange (116), stimulated by Hicksand Allen, pointed out another incon-sistency in Pareto. Pareto had as-sumed that if a person consideredfour states, A, B, C, and D, he couldjudge whether the difference betweenthe utilities of A and B was greaterthan, equal to, or less than the differ-ence between the utilities of C and D.Lange pointed out that if such acomparison was possible for any A,B, C, and D, then utility was car-dinally measurable. Since it seemsintrospectively obvious that suchcomparisons can be made, this paperprovoked a flood of protest and com-ment (7, 22, 117, 147, 209). Never-theless, in spite of all the comment,and even in spite of skepticism by adistinguished economist as late as1953 (153), Lange is surely right.Psychologists should know this atonce; such comparisons are the basisof the psychophysical Method ofEqual Sense Distances, from whichan interval scale is derived. (Samuel-son [162] has pointed out a very in-teresting qualification. Not onlymust such judgments of difference bepossible, but they must also be transi-tive in order to define an intervalscale.) But since such judgments ofdifferences did not seem to be neces-sary for the development of consumerdemand theory, Lange's paper didnot force the reinstatement of cardi-nal utility.

Indeed, the pendulum swungfurther in the behavioristic direction.Samuelson developed a new analyticfoundation for the theory of con-

386 WARD EDWARDS

sumer behavior, the essence of whichis that indifference curves and hencethe entire structure of the theory ofconsumer choice can be derivedsimply from observation of choicesamong alternative groups of pur-chases available to a consumer (160,161). This approach has been ex-tensively developed by Samuelson(164, 165, 167, 169) and others (50,90, 125, 126). The essence of the ideais that each choice defines a pointand a slope in commodity space.Mathematical approximation meth-ods make it possible to combine awhole family of such slopes into anindifference hyperplane. A family ofsuch hyperplanes forms an indiffer-ence "map."

In a distinguished but inaccessibleseries of articles, Wold (204, 205, 206;see also 208 for a summary presenta-tion) has presented the mathematicalcontent of the Pareto, Hicks and Al-len, and revealed preference (Samu-elson) approaches, as well as Cassel'sdemand function approach, and hasshown that if the assumption aboutcomplete weak ordering of bundles ofcommodities which was discussedabove is made, then all these ap-proaches are mathematically equiva-lent.

Nostalgia for cardinal utility. Thecrucial reason for abandoning cardi-nal utility was the argument of theordinalists that indifference curveanalysis in its various forms could doeverything that cardinal utility coulddo, with fewer assumptions. So faras the theory of riskless choice is con-cerned, this is so. But this is only anargument for parsimony, and parsi-mony is not always welcome. Therewas a series of people who, for onereason or another, wanted to rein-state cardinal utility, or at leastmarginal utility. There were severalmathematically invalid attempts to

show that marginal utility could bedefined even in an ordinal-utilityuniverse (23, 24, 163; 25, 114).Knight (110), in 1944, argued ex-tensively for cardinal utility; hebased his arguments in part on in-trospective considerations and in parton an examination of psychophysicalscaling procedures. He stimulated anumber of replies (29, 42; 111). Re-cently Robertson (154) pleaded forthe reinstatement of cardinal utilityin the interests of welfare economics(this point will be discussed againbelow). But in general the indiffer-ence curve approach, in its variousforms, has firmly established itself asthe structure of the theory of risklesschoice.

Experiments on indifference curves.Attempts to measure marginal utilityfrom market data were discussedabove. There have been three experi-mental attempts to measure indiffer-ence curves. Schultz, who pioneeredin deriving statistical demand curves,interested his colleague at the Univer-sity of Chicago, the psychologistThurstone, in the problem of in-difference curves. Thurstone (185)performed a very simple experiment.He gave one subject a series of com-binations of hats and overcoats, andrequired the subject to judge whetherhe preferred each combination to astandard. For instance, the subjectjudged whether he preferred eighthats and eight overcoats to fifteenhats and three overcoats. The sameprocedure was repeated for hats andshoes, and for shoes and overcoats.The data were fitted with indifferencecurves derived from the assumptionsthat utility curves fitted Fechner'sLaw and that the utilities of thevarious objects were independent.Thurstone says that Fechner's Lawfitted the data better than the otherpossible functions he considered, but


presents no evidence for this asser-tion. The crux of the experimentwas the attempt to predict the in-difference curves between shoes andovercoats from the other indifferencecurves. This was done by using theother two indifference curves to inferutility functions for shoes and forovercoats separately, and then usingthese two utility functions to predictthe total utility of various amountsof shoes and overcoats jointly. Theprediction worked rather well. Thejudgments of the one subject used areextraordinarily orderly; there is verylittle of the inconsistency and vari-ability that others working in thisarea have found. Thurstone says,"The subject . . . was entirely naiveas regards the psychophysical prob-lem involved and had no knowledgewhatever of the nature of the curvesthat we expected to find" (18S, p.154). He adds, "I selected as subjecta research assistant in my laboratorywho knew nothing about psycho-physics. Her work was largelyclerical in nature. She had a veryeven disposition, and I instructed herto take an even motivational attitudeon the successive occasions . . . I wassurprised at the consistency of thejudgments that I obtained, but I ampretty sure that they were the resultof careful instruction to assume a uni-form motivational attitude."3 Fromthe economist's point of view, themain criticism of this experiment isthat it involved imaginary ratherthan real transactions (200).

The second experimental measure-ment of indifference curves is reportedby the economists Rousseas and Hart(157). They required large numbersof students to rank sets of three com-binations of different amounts of ba-

8 Thurstone, L. L. Personal communication,December 7, 1953.

con and eggs. By assuming that allstudents had the same indifferencecurves, they were able to derive a com-posite indifference map for bacon andeggs. No mathematical assumptionswere necessary, and the indifferencemap is not given mathematical form.Some judgments were partly or com-pletely inconsistent with the final map,but not too many. The only conclu-sion which this experiment justifies isthat it is possible to derive such acomposite indifference map.

The final attempt to measure anindifference curve is a very recent oneby the psychologists Coombs andMilholland (49). The indifferencecurve involved is one between riskand value of an object, and so will bediscussed below in the section on thetheory of risky decisions. It is men-tioned here because the same meth-ods (which show only that the in-difference curve is convex to theorigin, and so perhaps should not becalled measurement) could equallywell be applied to the determinationof indifference curves in risklesssituations.

Mention should be made of theextensive economic work on statisti-cal demand curves. For some reasonthe most distinguished statistical de-mand curve derivers feel it necessaryto give an account of consumer'schoice theory as a preliminary to thederivation of their empirical demandcurves. The result is that the twobest books in the area (172, 182) areeach divided into two parts; the firstis a general discussion of the theoryof consumer's choice and the seconda quite unrelated report of statisticaleconomic work. Stigler (179) hasgiven good reasons why the statisticaldemand curves are so little related tothe demand curves of economictheory, and Wallis and Friedman(200) argue plausibly that this state

388 WARD EDWARDS

of affairs is inevitable. At any rate,there seems to be little prospect ofusing large-scale economic data to fillin the empirical content of the theoryof individual decision making.

Psychological comments. There areseveral commonplace observationsthat are likely to occur to psycholo-gists as soon as they try to apply thetheory of riskless choices to actualexperimental work. The first is thathuman beings are neither perfectlyconsistent nor perfectly sensitive.This means that indifference curvesare likely to be observable as in-difference regions, or as probabilitydistributions of choice around acentral locus. It would be easy toassume that each indifference curverepresents the modal value of a nor-mal sensitivity curve, and that choicesshould have statistical propertiespredictable from that hypothesis asthe amounts of the commodities(locations in product space) arechanged. This implies that the defi-nition of indifference between twocollections of commodities should bethat each collection is preferred overthe other 50 per cent of the time.Such a definition has been proposedby an economist (108), and used inexperimental work by psychologists(142). Of course, SO per cent choicehas been a standard psychologicaldefinition of indifference since thedays of Fechner.

Incidentally, failure on the part ofan economist to understand that ajust noticeable difference (j.n.d.) is astatistical concept has led him toargue that the indifference relation isintransitive, that is, that if A is in-different to B and B is indifferent toC, then A need not be indifferent to C(8, 9, 10). He argues that if A and Bare less than one j.n.d. apart, then Awill be indifferent to B; the same ofcourse is true of B and C; but A and

C may be more than one j.n.d. apart,and so one may be preferred to theother. This argument is, of course,wrong. If A has slightly more utilitythan B, then the individual willchoose A in preference to B slightlymore than SO per cent of the time,even though A and B are less thanone j.n.d. apart in utility. The 50 percent point is in theory a preciselydefined point, not a region. It may infact be difficult to determine becauseof inconsistencies in judgments andbecause of changes in taste with time.

The second psychological observa-tion is that it seems impossible evento dream of getting experimentallyan indifference map in w-dimensionalspace where n is greater than 3. Eventhe case of w = 3 presents formidableexperimental problems. This is lessimportant to the psychologist whowants to use the theory of choice torationalize experimental data thanto the economist who wants to de-rive a theory of general static equilib-rium.

Experiments like Thurstone's (185)involve so many assumptions that itis difficult to know what their empiri-cal meaning might be if these assump-tions were not made. Presumably,the best thing to do with such ex-periments is to consider them as testsof the assumption with the least facevalidity. Thurstone was willing toassume utility maximization and in-dependence of the commodities in-volved (incidentally, his choice ofcommodities seems singularly un-fortunate for justifying an assump-tion of independent utilities), and soused his data to construct a utilityfunction. Of course, if only ordinalutility is assumed, then experimentalindifference curves cannot be usedthis way. In fact, in an ordinal-utility universe neither of the prin-cipal assumptions made by Thurstone


can be tested by means of experi-mental indifference curves. So theassumption of cardinal utility, thoughnot necessary, seems to lead to con-siderably more specific uses for ex-perimental data.

At any rate, from the experimentalpoint of view the most interestingquestion is: What is the observedshape of indifference curves betweenindependent commodities? This ques-tion awaits an experimental answer.

The notion of utility is very similarto the Lewinian notion of valence(120, 121). Lewin conceives ofvalence as the attractiveness of anobject or activity to a person (121).Thus, psychologists might considerthe experimental study of utilities tobe the experimental study of valences,and therefore an attempt at quantify-ing parts of the Lewinian theoreticalschema.

APPLICATION OF THE THEORY OFRISKLESS CHOICES TO WEL-

FARE ECONOMICS4

The classical utility theorists as-sumed the existence of interpersonallycomparable cardinal utility. Theywere thus able to find a simple an-swer to the question of how to de-termine the best economic policy:That economic policy is best whichresults in the maximum total utility,summed over all members of theeconomy.

The abandonment of interpersonalcomparability makes this answer use-less. A sum is meaningless if theunits being summed are of varyingsizes and there is no way of reducingthem to some common size. This

4 The discussion of welfare economics givenin this paper is exceedingly sketchy. For apicture of what the complexities of modernwelfare economics are really like (see 11, 13,14, 86, 118, 124, 127, 139, 140, 148, 154, 155,166, 174).

point has not been universally recog-nized, and certain economists (e.g.,82, 154) still defend cardinal (but notinterpersonally comparable) utilityon grounds of its necessity for wel-fare economics.

Pareto's principle. The abandon-ment of interpersonal comparabilityand then of cardinal utility produceda search for some other principle tojustify economic policy. Pareto(146), who first abandoned cardinalutility, provided a partial solution.He suggested that a change shouldbe considered desirable if it lefteveryone at least as well off as hewas before, and made at least oneperson better off.

Compensation principle. Pareto'sprinciple is fine as far as it goes, butit obviously does not go very far.The economic decisions which can bemade on so simple a principle are fewand insignificant. So welfare eco-nomics languished until Kaldor (98)proposed the compensation prin-ciple. This principle is that if it ispossible for those who gain from aneconomic change to compensate thelosers for their losses and still havesomething left over from their gains,then the change is desirable. Ofcourse, if the compensation is actuallypaid, then this is simply a case ofPareto's principle.

But Kaldor asserted that thecompensation need not actually bemade; all that was necessary wasthat it could be made. The fact thatit could be made, according toKaldor, is evidence that the changeproduces an excess of good over harm,and so is desirable. Scitovsky (173)observed an inconsistency in Kaldor'sposition: Some cases could arise inwhich, when a change from A to Bhas been made because of Kaldor'scriterion, then a change back from Bto A would also satisfy Kaldor's

390 WARD EDWARDS

criterion. It is customary, therefore,to assume that changes which meetthe original Kaldor criterion are onlydesirable if the reverse change doesnot also meet the Kaldor criterion.

It has gradually become obviousthat the Kaldor-Scitovsky criteriondoes not solve the problem of welfareeconomics (see e.g., 18, 99). It as-sumes that the unpaid compensationdoes as much good to the person whogains it as it would if it were paid tothe people who lost by the change.For instance, suppose that an in-dustrialist can earn $10,000 a yearmore from his plant by using a newmachine, but that the introduction ofthe machine throws two people ir-retrievably out of work. If the salaryof each worker prior to the changewas $4,000 a year, then the in-dustrialist could compensate theworkers and still make a profit. Butif he does not compensate the work-ers, then the added satisfaction hegets from his extra $10,000 may bemuch less than the misery he pro-duces in his two workers. This ex-ample only illustrates the principle;it does not make much sense in thesedays of progressive income taxes, un-employment compensation, high em-ployment, and strong unions.

Social welfare functions. From hereon the subject of welfare economicsgets too complicated and too remotefrom psychology to merit extensiveexploration in this paper. The linethat it has taken is the assumptionof a social welfare function (21), afunction which combines individualutilities in a way which satisfiesPareto's principle but is otherwiseundefined. In spite of its lack ofdefinition, it is possible to drawcertain conclusions from such a func-tion (see e.g., 164). However, Arrow(14) has recently shown that a socialwelfare function that meets certain

very reasonable requirements aboutbeing sensitive in some way to thewishes of all the people affected,etc., cannot in general be found inthe absence of interpersonally com-parable utilities (see also 89).

Psychological comment. Some econ-omists are willing to accept thefact that they are inexorably com-mitted to making moral judgmentswhen they recommend economicpolicies (e.g., 152, 153). Others stilllong for the impersonal amorality of autility measure (e.g., 154). Howeverdesirable interpersonally comparablecardinal utility may be, it seemsUtopian to hope that any experi-mental procedure will ever give in-formation about individual utilitiesthat could be of any practical use inguiding large-scale economic policy.

THE THEORY OF RISKY CHOICES"Risk and uncertainty. Economists

and statisticians distinguish between

6 Strotz (183) and Alchian (1) present non-technical and sparkling expositions of the vonNeumann and Morgenstern utility measure-ment proposals. Georgescu-Roegen (78) criti-cally discusses various axiom systems so as tobring some of the assumptions underlying thiskind of cardinal utility into clear focus. Allais(3) reviews some of these ideas in the course ofcriticizing them, Arrow (12, 14) reviews partsof the field.

There is a large psychological literature onone kind of risky decision making, the kindwhich results when psychologists use partialreinforcement. This literature has been re-viewed by Jenkins and Stanley (96). Recentlya number of experimenters, including Jarrett(95), Flood (69, 70), Bilodeau (27), and my-self (56) have been performing experiments onhuman subjects who are required to chooserepetitively between two or more alternatives,each of which has a probability of rewardgreater than zero and less than one. The prob-lems raised by these experiments are too com-plicated and too far removed from conven-tional utility theory to be dealt with in thispaper. This line of experimentation may even-tually provide the link which ties togetherutility theory and reinforcement theory.


risk and uncertainty. There does notseem to be any general agreementabout which concept should be as-sociated with which word, but thefollowing definitions make the mostimportant distinctions.

Almost everyone would agree thatwhen I toss a coin the probabilitythat I will get a head is .5. A proposi-tion about the future to which a num-ber can be attached, a number thatrepresents the likelihood that theproposition is true, may be called afirst-order risk. What the rules are forattaching such numbers is a muchdebated question, which will beavoided in this paper.

Some propositions may depend onmore than one probability distribu-tion. For instance, I may decide thatif I get a tail, I will put the coin backin my pocket, whereas if I get a head,I will toss it again. Now, the prob-ability of the proposition "I will geta head on my second toss" is a func-tion of two probability distributions,the distribution corresponding to thefirst toss and that corresponding tothe second toss. This might be calleda second-order risk. Similarly, risks ofany order may be constructed. It is amathematical characteristic of allhigher-order risks that they may becompounded into first-order risks bymeans of the usual theorems for com-pounding probabilities. (Some econo-mists have argued against this pro-cedure [83], essentially on the groundsthat you may have more informationby the time the second risk comesaround. Such problems can best bedealt with by means of von Neumannand Morgenstern's [197] concept ofstrategy, which is discussed below.They become in general problems ofuncertainty, rather than risk.)

Some propositions about the futureexist to which no generally acceptedprobabilities can be attached. What

is the probability that the followingproposition is true: Immediatelyafter finishing this paper, you willdrink a glass of beer? Surely it isneither impossible nor certain, so itought to have a probability betweenzero and one, but it is impossible foryou or me to find out what that prob-ability might be, or even to set upgenerally acceptable rules about howto find out. Such propositions areconsidered cases of uncertainty, ratherthan of risk. This section deals onlywith the subject of first-order risks.The subject of uncertainty will ariseagain in connection with the theoryof games.

Expected utility maximization. Thetraditional mathematical notion fordealing with games of chance (and sowith risky decisions) is the notionthat choices should be made so as tomaximize expected value. The ex-pected value of a bet is found bymultiplying the value of each possibleoutcome by its probability of oc-currence and summing these prod-ucts across all possible outcomes. Insymbols:

where p stands for probability, $stands for the value of an outcome,and pi+p*+ • • • +£n = l.

The assumption that people ac-tually behave the way this mathe-matical notion says they should iscontradicted by observable behaviorin many risky situations. People arewilling to buy insurance, even thoughthe person who sells the insurancemakes a profit. People are willing tobuy lottery tickets, even though thelottery makes a profit. Considerationof the problem of insurance and of theSt. Petersburg paradox led DanielBernoulli, an eighteenth centurymathematician, to propose that theycould be resolved by assuming that

392 WARD EDWARDS

people act so as to maximize expectedutility, rather than expected value(26). (He also assumed that utilityfollowed a function that more than acentury later was proposed by Fech-ner for subjective magnitudes ingeneral and is now called Fechner'sLaw.) This was the first use of thenotion of expected utility.

The literature on risky decisionmaking prior to 1944 consists pri-marily of the St. Petersburg paradoxand other gambling and probabilityliterature in mathematics, some liter-ary discussion in economics (e.g., 109,187), one economic paper on lotteries(189), and the early literature of thetheory of games (31, 32, 33, 34, 195),which did not use the notion ofutility. The modern period in thestudy of risky decision making beganwith the publication in 1944 of vonNeumann and Morgenstern's monu-mental book Theory of Games andEconomic Behavior (196, see also197), which we will discuss more fullylater. Von Neumann and Morgen-stern pointed out that the usual as-sumption that economic man canalways say whether he prefers onestate to another or is indifferent be-tween them needs only to be slightlymodified in order to imply cardinalutility. The modification consists ofadding that economic man can alsocompletely order probability com-binations of states. Thus, supposethat an economic man is indifferentbetween the certainty of $7.00 and a50-50 chance of gaining $10.00 ornothing. We can assume that hisindifference between these two pros-pects means that they have the sameutility for him. We may define theutility of $0.00 as zero utiles (theusual name for the unit of utility, justas sone is the name for the unit ofauditory loudness), and the utilityof $10.00 as 10 utiles?, These two

arbitrary definitions correspond todefining the two undefined constantswhich are permissible since cardinalutility is measured only up to a lineartransformation. Then we may cal-culate the utility of $7.00 by usingthe concept of expected utility as fol-lows:

17(17.00) = .5 £7($10.00) +.5 E7($0.00)= .5(10)+.5(0) = 5.

Thus we have determined the cardi-nal utility of $7.00 and found that itis 5 utiles. By varying the probabil-ities and by using the already foundutilities it is possible to discover theutility of any other amount of money,using only the two permissible arbi-trary definitions. It is even moreconvenient if instead of +$10.00,— $10.00 or some other loss is used asone of the arbitrary utilities.

A variety of implications is em-bodied in this apparently simple no-tion. In the attempt to examine andexhibit clearly what these implica-tions are, a number of axiom systems,differing from von Neumann andMorgenstern's but leading to thesame result, have been developed(73, 74, 85, 135, 136, 171). Thispaper will not attempt to go intothe complex discussions (e.g., 130,131, 168, 207) of these various al-ternative axiom systems. One recentdiscussion of them (78) has con-cluded, on reasonable grounds, thatthe original von Neumann and Mor-genstern set of axioms is still the best.

It is profitable, however, to ex-amine what the meaning of this no-tion is from the empirical point ofview if it is right. First, it means thatrisky propositions can be orderedin desirability, just as riskless onescan. Second, it means that the con-cept of expected utility is behavior-ally meaningful. Finally, it means

choices among risky alternatives


are made in such a way that theymaximize expected utility.

If this model is to be used to pre-dict actual choices, what could gowrong with it? It might be that theprobabilities by which the utilitiesare multiplied should not be the ob-jective probabilities; in other words, adecider's estimate of the subjectiveimportance of a probability may notbe the same as the numerical value ofthat probability. It might be thatthe method of combination of proba-bilities and values should not besimple multiplication. It might bethat the method of combination ofthe probability-value products shouldnot be simple addition. It might bethat the process of gambling hassome positive or negative utility ofits own. It might be that the wholeapproach is wrong, that people justdo not behave as if they were tryingto maximize expected utility. Weshall examine some of these pos-sibilities in greater detail below.

Economic implications of maximiz-ing expected utility. The utility-measurement notions of von Neu-mann and Morgenstern were en-thusiastically welcomed by manyeconomists (e.g., 73, 193), though afew (e.g., 19) were at least tempo-rarily (20) unconvinced. The mostinteresting economic use of them wasproposed by Friedman and Savage(73), who were concerned with thequestion of why the same person whobuys insurance (with a negative ex-pected money value), and therefore iswilling to pay in order not to takerisks, will also buy lottery tickets(also with a negative expected moneyvalue) in which he pays in order totake risks. They suggested that thesefacts could be reconciled by a doublyinflected utility curve for money, likethat in Fig. 2. If / represents theperson's current income, then he is

clearly willing to accept "fair" in-surance (i.e., insurance with zero ex-pected money value) because theserious loss against which he is insur-ing would have a lower expectedutility than the certain loss of theinsurance premium. (Negatively ac-celerated total utility curves, likethat from the origin to /, are whatyou get when marginal utility de-creases; thus, decreasing marginal

UJ

DOLLARSFIG. 2. HYPOTHETICAL UTILITY CURVE FOR

MONEY, PROPOSED BY FRIEDMAN AND SAVAGE

utility is consistent with the avoid-ance of risks.) The person would alsobe willing to buy lottery tickets, sincethe expected utility of the lotteryticket is greater than the certain lossof the cost of the ticket, because ofthe rapid increase in the height of theutility function. Other considera-tions make it necessary that theutility curve turn down again. Notethat this discussion assumes thatgambling has no inherent utility.

Markowitz (132) suggested an im-portant modification in this hy-pothesis. He suggested that theorigin of a person's utility curve formoney be taken as his customary

394 WARD EDWARDS

financial status, and that on bothsides of the origin the curve be as-sumed first concave and then convex.If the person's customary state ofwealth changes, then the shape of hisutility curve will thus remain gen-erally the same with respect to wherehe now is, and so his risk-taking be-havior will remain pretty much thesame instead of changing with everychange of wealth as in the Friedman-Savage formulation.

Criticism of the expected-utilitymaximization theory. It is fairly easyto construct examples of behaviorthat violate the von Neumann-Morgenstern axioms (for a partic-ularly ingenious example, see 183). Itis especially easy to do so when theamounts of money involved are verylarge, or when the probabilities orprobability differences involved areextremely small. Allais (5) has con-structed a questionnaire full of itemsof this type. For an economist in-terested in using these axioms as abasis for a completely general theoryof risky choice, these examples maybe significant. But psychological in-terest in this model is more modest.The psychologically important ques-tion is: Can such a model be used toaccount for simple experimental ex-amples of risky decisions?

Of course a utility function derivedby von Neumann-Morgenstern meansis not necessarily the same as a classi-cal utility function (74, 203; see also82).

Experiment on the von Neumann-Morgenstern model. A number of ex-periments on risky decision makinghave been performed. Only the firstof them, by Mosteller and Nogee(142), has been in the simple frame-work of the model described above.All the rest have in some way oranother centered on the concept ofprobabilities effective for behavior

which differ in some way from theobjective probabilities, as well as onutilities different from the objectivevalues of the objects involved.

Mosteller and Nogee (142) carriedout the first experiment to apply thevon Neumann-Morgenstern model.They presented Harvard undergradu-ates and National Guardsmen withbets stated in terms of rolls at pokerdice, which each subject could acceptor refuse. Each bet gave a "hand"at poker dice. If the subject couldbeat the hand, he won an amountstated in the bet. If not, he lost anickel. Subjects played with $1.00,which they were given at the be-ginning of each experimental session.They were run together in groups offive; but each decided and rolled thepoker dice for himself. Subjects wereprovided with a table in which themathematically fair bets were shown,so that a subject could immediatelytell by referring to the table whethera given bet was fair, or better orworse than fair.

In the data analysis, the first stepwas the determination of "indiffer-ence offers." For each probabilityused and for each player, the amountof money was found for which thatplayer would accept the bet SO percent of the time. Thus equality wasdefined as SO per cent choice, as itis likely to be in all psychological ex-periments of this sort. Then theutility of $0.00 was defined as 0utiles, and the utility of losing anickel was defined as — 1 utile. Withthese definitions and the probabilitiesinvolved, it was easy to calculate theutility corresponding to the amountof money involved in the indifferenceoffer. It turned out that, in genera),the Harvard undergraduates haddiminishing marginal utilities, whilethe National Guardsmen had in-creasing marginal utilities.


The utilities thus calculated wereused in predicting the results of morecomplex bets. It is hard to evaluatethe success of these predictions. Atany rate, an auxiliary paired-comparisons experiment showed thatthe hypothesis that subjects maxi-mized expected utility predictedchoices better than the hypothesisthat subjects maximized expectedmoney value.

The utility curve that Mostellerand Nogee derive is different fromthe one Friedman and Savage (73)were talking about. Suppose that asubject's utility curve were of theFriedman-Savage type, as in Fig1. 2,and that he had enough money to puthim at point P. If he now wins orloses a bet, then he is moved to adifferent location on the indifferencecurve, say Q. (Note that the amountsof money involved are much smallerthan in the original Friedman-Savageuse of this curve.) However, the con-struction of a Mosteller-Nogee utilitycurve assumes that the individual isalways at the same point on hisutility curve, namely the origin. Thismeans that the curve is really of theMarkowitz (132) type discussedabove, instead of the Friedman-Savage type. The curve is not reallya curve of utility of money in general,but rather it is a curve of the utility-for-w-more dollars. Even so, it mustbe assumed further that as the totalamount of money possessed by thesubject changes during the experi-ment, the utility-for-«-more dollarscurve does not change. Mosteller andNogee argue, on the basis of detailedexamination of some of their data,that the amount of money possessedby the subjects did not seriouslyinfluence their choices. The utilitycurves they reported showed chang-ing marginal utility within theamounts of money usdd in their ex-

periment. Consequently, their con-clusion that the amount of moneypossessed by the subjects was notseriously important can only be trueif their utility curves are utility-for-w-more dollars curves and if theshapes of such curves are not affectedby changes in the number of dollarson hand. This discussion exhibits atype of problem which must alwaysarise in utility measurement andwhich is new in psychological scaling.The effects of previous judgments onpresent judgments are a familiarstory in psychophysics, but they areusually assumed to be contaminatinginfluences that can be minimized oreliminated by proper experimentaldesign. In utility scaling, the funda-mental idea of a utility scale is suchthat the whole structure of a subject'schoices should be altered as a resultof each previous choice (if the choicesare real ones involving money gainaor losses). The Markowitz solutionto this problem is the most practicalone available at present, and thatsolution is not entirely satisfactorysince all it does is to assume thatpeople's utilities for money operatein such a way that the problem doesnot really exist. This assumption isplausible for money, but it getarapidly less plausible when othercommodities with a less continuouscharacter are considered instead.

Probability preferences. In a seriesof recent experiments (55, 57, 58,59), the writer has shown that subjects,when they bet, prefer some probabil-ities to others (57), and that thesepreferences cannot be accounted forby utility considerations (59). Allthe experiments were basically of thesame design. Subjects were requiredto choose between pairs of bets ac-cording to the method of paired com-parisons. The bets were of threekinds: positive expected value, nega-

396 WARD EDWARDS

tive expected value, and zero ex-pected value. The two members ofeach pair of bets had the same ex-pected value, so that there was never(in the main experiment [57, 59]) anyobjective reason to expect that choos-ing one bet would be more desirablethan choosing the other.

Subjects made their choices underthree conditions: just imagining theywere betting; betting for worthlesschips; and betting for real money.They paid any losses from their ownfunds, but they were run in extrasessions after the main experiment tobring their winnings up to $1.00 perhour.

The results showed that two fac-tors were most important in deter-mining choices: general preferences ordislikes for risk-taking, and specificpreferences among probabilities. Anexample of the first kind of factor isthat subjects strongly preferred lowprobabilities of losing large amountsof money to high probabilities oflosing small amounts of money—theyjust didn't like to lose. It also turnedout that on positive expected valuebets, they were more willing to acceptlong shots when playing for realmoney than when just imagining orplaying for worthless chips. An ex-ample of the second kind of factoris that they consistently preferredbets involving a 4/8 probability ofwinning to all others, and consistentlyavoided bets involving a 6/8 prob-ability of winning. These preferenceswere reversed for negative expectedvalue bets.

These results were independent ofthe amounts of money involved inthe bets, so long as the condition ofconstant expected value was main-tained (59). When pairs of bets whichdiffered from one another in expectedvalue were used, the choices were acompromise between maximizing ex-

pected amount of money and bettingat the preferred probabilities (58).An attempt was made to constructindividual utility curves adequate toaccount for the results of several sub-jects. For this purpose, the utility of$0.30 was defined as 30 utiles, and itwas assumed that subjects cannotdiscriminate utility differences small-er than half a utile. Under these as-sumptions, no individual utility curvesconsistent with the data could bedrawn. Various minor experimentsshowed that these results were relia-ble and not due to various possibleartifacts (59). No attempt was madeto generate a mathematical model ofprobability preferences.

The existence of probability prefer-ences means that the simple vonNeumann-Morgenstern method ofutility measurement cannot succeed.Choices between bets will be deter-mined not only by the amounts ofmoney involved, but also by thepreferences the subjects have amongthe probabilities involved. Only anexperimental procedure which holdsone of these variables constant, orotherwise allows for it, can hope tomeasure the other. Thus my experi-ments cannot be regarded as a wayof measuring probability preferences;they show only that such preferencesexist.

It may nevertheless be possible toget an interval scale of the utility ofmoney from gambling experiments bydesigning an experiment which meas-ures utility and probability prefer-ences simultaneously. Such experi-ments are likely to be complicatedand difficult to run, but they can bedesigned.

Subjective probability. First, aclarification of terms is necessary.The phrase subjective probability hasbeen used in two ways: as a namefor a school of thought about the


logical basis of mathematical prob-ability (51, 52, 80) and as a name fora transformation on the scale ofmathematical probabilities which issomehow related to behavior. Onlythe latter usage is intended here. Theclearest distinction between thesetwo notions arises from considera-tion of what happens when an objec-tive probability can be denned (e.g.,in a game of craps). If the subjectiveprobability is assumed to be differentfrom the objective probability, thenthe concept is being used in its sec-ond, or psychological, sense. Otherterms with the same meaning havealso been used: personal probability,psychological probability, expecta-tion (a poor term because of thedanger of confusion with expectedvalue). (For a more elaboratetreatment of concepts in this area,see 192.)

In 1948, prior to the Mosteller andNogee experiment, Preston andBaratta (149) used essentially similarlogic and a somewhat similar experi-ment to measure subjective prob-abilities instead of subjective values.They required subjects to bid com-petitively for the privilege of takinga bet. All bids were in play money,and the data consisted of the winningbids. If each winning bid can be con-sidered to represent a value of playmoney such that the winning bidderis indifferent between it and the bethe is bidding for, and if it is furtherassumed that utilities are identicalwith the money value of the playmoney and that all players have thesame subjective probabilities, thenthese data can be used to construct asubjective probability scale. Prestonand Baratta constructed such ascale. The subjects, according to thescale, overestimate low probabilitiesand underestimate high ones, with anindifference point (where subjective

equals objective probability) at about0.2. Griffith (81) found somewhatsimilar results in an analysis ofparimutuel betting at race tracks, asdid Attneave (17) in a guessing game,and Sprowls (178) in an analysis ofvarious lotteries. The Mosteller andNogee data (142) can, of course, beanalyzed for subjective probabilitiesinstead of subjective values. Mostel-ler and Nogee performed such ananalysis and said that their resultswere in general agreement withPreston and Baratta's. However,Mosteller and Nogee found no in-difference point for their Harvardstudents, whereas the NationalGuardsmen had an indifference pointat about 0.5. They are not able toreconcile these differences in results.

The notion of subjective probabil-ity has some serious logical difficulties.The scale of objective probability isbounded by 0 and 1. Should a sub-jective probability scale be similarlybounded, or not? If not, then manydifferent subjective probabilities willcorrespond to the objective proba-bilities 0 and 1 (unless some trans-formation is used so that 0 and 1 ob-jective probabilities correspond toinfinite subjective probabilities, whichseems unlikely). Considerations ofthe addition theorem to be discussedin a moment have occasionally ledpeople to think of a subjectiveprobability scale bounded at 0 butnot at 1. This is surely arbitrary.The concept of absolute certainty isneither more nor less indeterminatethan is the concept of absolute im-possibility.

Even more drastic logical problemsarise in connection with the additiontheorem. If the objective probabilityof event A is P, and that of A notoccurring is Q, then P+Q=1. Shouldthis rule hold for subjective proba-bilities? Intuitively it seems neces-

398 WARD EDWARDS

sary that if we know the subjectiveprobability of A, we ought to be ableto figure out the subjective proba-bility of not-^4, and the only reason-able rule for figuring it out is sub-traction of the subjective probabilityof A from that of complete certainty.But the acceptance of this additiontheorem for subjective probabilitiesplus the idea of bounded subjectiveprobabilities means that the subjec-tive probability scale must be identi-cal with the objective probabilityscale. Only for a subjective proba-bility scale identical with the objec-tive probability scale will thesubjective probabilities of a collec-tion of events, one of which musthappen, add up to 1. In the specialcase where only two events, A andnot-A, are considered, a subjectiveprobability scale like SI or S2 inFig. 3 would meet the requirementsof additivity, and this fact has led tosome speculation about such scales,particularly about 51. But suchscales do not meet the additivity re-quirements when more than twoevents are considered.

One way of avoiding these diffi-

H_JCO<COoOLQ.

L.J0.5

OLd

CODCO

52.

Si

0 0.5 I

OBJECTIVE PROBABILITYFIG. 3. HYPOTHETICAL SUBJECTIVE PROB-

ABILITY CURVES

culties is to stop thinking about ascale of subjective probabilities and,instead, to think of a weightingfunction applied to the scale of objec-tive probabilities which weights theseobjective probabilities according totheir ability to control behavior. Pre-sumably, I was studying this abilityin my experiments on probabilitypreferences (55, 57, 58, 59). There isno reason why such weighted proba-bilities should add up to 1 or shouldobey any other simple combinatoryprinciple.

Views and experiments which com-bine utility and subjective probability.The philosopher Ramsey publishedin 1926 (reprinted in 150) an essayon the subjective foundations of thetheory of probability; this containedan axiom system in which both utilityand subjective probability appeared.He used 0.5 subjective probability asa reference point from which to de-termine utilities, and then used theseutilities to determine other sub-jective probabilities. Apparently,economists did not discover Ramsey'sessay until after von Neumann andMorgenstern's book aroused interestin the subject. The only other formalaxiom system in which both utilityand subjective probability play apart is one proposed by Savage(171), which is concerned with un-certainty, rather than risk, and usesthe concept of subjective probabilityin its theory-of-probability sense.

The most extensive and importantexperimental work in the whole fieldof decision making under risk anduncertainty is now being carried outby Coombs and his associates at theUniversity of Michigan. Coombs'sthinking about utility and subjectiveprobability is an outgrowth of histhinking about psychological scalingin general. (For a discussion of hisviews, see 43, 44, 45, 46, 47.) The


essence of his work is the attempt tomeasure both utility and subjectiveprobability on an ordered metricscale. An ordered metric scale has allthe properties of an ordinal scale,and, in addition, the distances be-tween some or all of the stimuli can.be rank ordered. Coombs has de-veloped various experimental pro-cedures for obtaining such informa-tion about the spacings of stimuli.

In the most important article onutility and subjective probability tocome out of the Coombs approach,Coombs and Beardslee (48) presentan analysis of gambling decisions in-volving three independent variables:utility for prize, utility for stake, andsubjective probability. All three areassumed measurable only up to anordered metric, although it is as-sumed that the psychological prob-ability of losing the stake is one minusthe psychological probability ofwinning the prize, an assumption thatlimits the permissible underlyingpsychological probability functionsto shapes like those in Fig. 3. Anelaborate graphic analysis of the in-difference surfaces in this three-dimensional space is given, contain-ing far too many interesting relation-ships to summarize here. An ex-periment based on this model was de-signed. Coombs is reluctant to usesums of money as the valuable ob-jects in his experiments because ofthe danger that subjects will respondto the numerical value of the amountof dollars rather than to the psycho-logical value. Therefore he usedvarious desirable objects (e.g., aradio) as stimuli, and measured theirutility by the techniques he has de-veloped to obtain ordered metricscales. He used simple numericalstatements of probability as theprobability stimuli, and assumed thatsubjective probability was equal to

objective probability. The subjectfrom whose judgments the orderedmetric utility measurement was con-structed was then presented withimaginary bets involving these ob-jects and probabilities, and it turnedout that she almost always chose theone with the higher expected utility.This experiment is significant onlyas an illustration of the applicationof the method; the conclusion thatsubjects attempt to maximize ex-pected utility cannot very comfort-ably be generalized to other subjectsand to real choices without betterevidence.

Coombs and Milholland (49) did amuch more elaborate experiment inwhich they established ordered metricscales, both for the utilities of a col-lection of objects and for the subjec-tive probabilities of a collection ofstatements (e.g., Robin Roberts willwin 20 games next year). Statementsand objects were combined into"bets," and the two subjects forwhom the ordered metric scales hadbeen established were asked to makejudgments about which bet theywould most, and which they wouldleast, prefer from among varioustriads of bets. These judgments wereexamined to discover whether or notthey demonstrated the existence ofat least one convex indifference curvebetween utility and subjective prob-ability (the requirements for demon-strating the convexity of an in-difference curve by means of orderedmetric judgments are fairly easy tostate). A number of cases consistentwith a convex indifference curve werefound, but a retest of the orderedmetric data revealed changes whicheliminated all of the cases consistentwith a convex indifference curve forone subject, and all but one case forthe other. It is not possible to makea statistical test of whether or not

400 WARD EDWARDS

that one case might have come aboutby chance. No evidence was foundfor the existence of concave indiffer-ence curves, which are certainly in-consistent with the theory of riskydecisions. This experiment is a fineexample of the strength and weak-ness of the Coombs approach. Itmakes almost no assumptions, takesvery little for granted, and avoidsthe concept of error of judgment; asa result, much of the potential in-formation in the data is unused andrarely can any strong conclusions bedrawn.

A most disturbing possibility israised by experiments by Marks (133)and Irwin (94) which suggest that theshape of the subjective probabilityfunction is influenced by the utilitiesinvolved in the bets. If utilities andsubjective probabilities are not inde-pendent, then there is no hope of pre-dicting risky decisions unless theirlaw of combination is known, and itseems very difficult to design an ex-periment to discover that law of com-bination. However, the main dif-ferences that Marks and Irwin foundwere between probabilities attachedto desirable and undesirable alterna-tives. It is perfectly possible thatthere is one subjective probabilityfunction for bets with positive ex-pected values and a different one forbets with negative expected values,just as the negative branch of theMarkowitz utility function is likelyto be different from the positivebranch. The results of my probabil-ity preference experiments showedvery great differences between theprobability preference patterns forpositive and for negative expected -value bets (57), but little differencebetween probability preferences atdifferent expected-value levels solong as zero expected value was notcrossed (59). This evidence supports

the idea that perhaps only two sub-jective probability functions are nec-essary.

Santa Monica Seminar. In thesummer of 1952 at Santa Monica,California, a group of scientists con-ferred on problems of decision mak-ing. They met in a two-month semi-nar sponsored by the University ofMichigan and the Office of NavalResearch. The dittoed reports ofthese meetings are a gold mine ofideas for the student of this problem.Some of the work done at this semi-nar is now being prepared for a bookon Decision Processes edited by R. M.Thrall, C. H. Coombs, and R. L.Davis, of the University of Michigan.

Several minor exploratory experi-ments were done at this seminar.Vail (190) did an experiment in whichhe gave four children the choice ofwhich side of various bets theywanted to be on. On the assumptionof linear utilities, he was able to com-pute subjective probabilities for thesechildren. The same children, how-ever, were used as subjects for anumber of other experiments; so,when Vail later tried them out onsome other bets, he found that theyconsistently chose the bet with thehighest probability of winning, re-gardless of the amounts of money in-volved. When 50-50 bets were in-volved, one subject consistently chosethe bet with the lowest expectedvalue. No generalizable conclusionscan be drawn from these experiments.

Kaplan and Radner (100) tried outa questionnaire somewhat likeCoombs's method of measuring sub-jective probability. Subjects wereasked to assign numbers to variousstatements. The numbers could beanything from 0 to 100 and were torepresent the likelihood that thestatement was true. The hypothesesto be tested were: (a) for sets of ex-


haustive and mutually exclusivestatements in which the numbers as-signed (estimates of degree of belief)were nearly equal, the sums of thesenumbers over a set would increasewith the number of alternatives (be-cause low probabilities would be over-estimated) ; (b) for sets with the samenumbers of alternatives, those withone high number assigned would havea lower set sum than those with nohigh numbers. The first predictionwas verified; the second was not.Any judgments of this sort are somuch more likely to be made on thebasis of number preferences andsimilar variables than on subjectiveprobabilities that they offer verylittle hope as a method of measuringsubjective probabilities.

Variance preferences. Allais (2, 3,4) and Georgescu-Roegen (78) haveargued that it is not enough to applya transform on objective value and onobjective probability in order to pre-dict risky decisions from expectedutility (see also 188); it is also neces-sary to take into account at least thevariance, and possibly the highermoments, of the utility distribution.There are instances in which thisargument seems convincing. Youwould probably prefer the certaintyof a million dollars to a 50-50 chanceof getting either four million or noth-ing. I do not think that this prefer-ence is due to the fact that the ex-pected utility of the 50-50 bet is lessthan the utility of one million dollarsto you, although this is possible. Amore likely explanation is simplythat the variances of the two propo-sitions are different. Evidence infavor of this is the fact that if youknew you would be offered this choice20 times in succession, you wouldprobably take the 50-50 bet eachtime. Allais (5) has constructed anumber of more sophisticated exam-

ples of this type. However, from asimple-minded psychological point ofview, these examples are irrelevant.It is enough if the theory of choicecan predict choices involving familiaramounts of money and familiarprobability differences—choices suchas those which people are accustomedto making. It may be necessary foreconomic theory that the theory ofchoice be universal and exceptionless,but experimental psychologists neednot be so ambitious. This is fortu-nate, because the introduction of thevariance and higher moments of theutility distribution makes the prob-lem of applying the theory experi-mentally seem totally insoluble. It isdifficult enough to derive reasonablemethods of measuring utility alonefrom risky choices; when it also be-comes necessary to measure subjec-tive probability and to take thehigher moments of the utility dis-tribution into account, the problemseems hopeless. Allais apparentlyhopes to defeat this problem by usingpsychophysical methods to measureutility (and presumably subjectiveprobability also). This is essentiallywhat Coombs has done, but Coombshas recognized that such proceduresare unlikely to yield satisfactoryinterval scales. The dollar scale ofthe value of money is so thoroughlytaught to us that it seems almost im-possible to devise a psychophysicalsituation in which subjects wouldjudge the utility, rather than the dol-lar value, of dollars. They mightjudge the utility of other valuableobjects, but since dollars are theusual measure of value, such judg-ments would be less useful, and eventhese judgments would be likely to becontaminated by the dollar values ofthe objects. I would get more utilityfrom a new electric shaver than Iwould from a new washing machine,

402 WARD EDWARDS

but because of my knowledge of therelative money values of these ob-jects, I would certainly choose thewashing machine if given a choicebetween them. Somewhat similararguments can be applied againstusing psychophysical methods tomeasure subjective probability. Afinal point is that, since these subjec-tive scales are to be used to predictchoices, it would be best if they couldbe derived from similar choices.

Other approaches. Shackle (175)has proposed a theory of decisionmaking under risk and uncertainty.This theory is unique in that it doesnot assume any kind of maximizingbehavior. For every possible out-come of a decision made in a risky oruncertain situation, Shackle assumesthat there is a degree of potentialsurprise that this, rather than someother, outcome would occur. Everyoutcome-potential surprise pair isranked in accordance with its abilityto stimulate the mind (stimulation in-creases with increasing outcome anddecreases with increasing potentialsurprise). The highest-ranking posi-tive outcome-potential surprise pairand the highest-ranking negative pairare found, and these two possibilitiesalone determine what the individualwill do. Semi-mathematical methodsare used to predict the outcome ofconsideration of possible lines of ac-tion. Although attempts have beenmade to relate it to Wald's minimaxprinciple for statistical decision func-tions (see below), the fact remainsthat most critics of the Shackle pointof view have judged it to be either toovague to be useful, or, if specified indetail, too conducive to patently ab-surd predictiona (e.g., 201).

Shackle's point of view was de-veloped primarily to deal with uniquechoices—choices which can be madeonly once. Allais (3) has similarly

criticized conventional utility the-ory's attack on this problem. Sincethe usual frequency theory of prob-ability conceives of the probability asthe limit of the outcomes of a largenumber of similar trials, it is ques-tionable that notions which use prob-ability in the ordinary sense (like thenotion of maximizing expected util-ity) are applicable to unique choices.However, this seems to be an experi-mental problem. If notions which useordinary probability are incapable ofpredicting actual unique choices,then it will be necessary to seek othertheoretical tools. But so long as agenerally acceptable probability canbe defined (e.g., as in the unique tossof a coin), it is not necessary to as-sume a priori that theories based onconventional probabilities will be in-adequate. When no generally ac-ceptable probability can be defined,then the problem becomes very dif-ferent.

Cartwright and Festinger (38, 41)have proposed a theory about thetime it takes to make decisions whichis in some ways similar to those dis-cussed in this section. The main dif-ference is that they add the conceptof restraining forces, and that theyconceive of all subjective magnitudesas fluctuating randomly around amean value. From this they deducevarious propositions about decisiontimes and the degree of certaintywhich subjects will feel about theirdecisions, and apparently these prop-ositions work out experimentallypretty well (38, 3Q, 61, 62). TheLewinian theoretical orientationseems to lead to this kind of model;Lewin, Dembo, Festinger, and Sears(122) present a formally similartheory about level of aspiration. Ofcourse, the notion of utility is verysimilar to the Lewinian notion ofvalence.


Landahl (115) has presented amathematical model for risk-takingbehavior based on the conceptualneurology of the mathematical bio-physics school.

Psychological comments. The areaof risky decision making is full offascinating experimental problems.Of these, the development of a satis-factory scale of utility of money andof subjective probability must comefirst, since the theory of risky de-cision making is based on these no-tions. The criterion for satisfactori-ness of these scales must be that theysuccessfully predict choices otherthan those from which they were de-rived. To be really satisfactory, it isdesirable that they should predictchoices in a wide variety of differingsituations. Unlike the subjectivescales usually found in psychophys-ics, it is likely that these scales willdiffer widely from person to person,so a new determination of each scalemust be made for each new subject.It can only be hoped that the scalesdo not change in time to any seriousdegree; if they do, then they areuseless.

Once scales of utility and subjec-tive probability are available, thenmany interesting questions arise.What about the addition theorem forsubjective probabilities? Does gam-bling itself have utility, and howmuch? To what extent can these sub-jective scales be changed by learning?To what degree do people differ, andcan these differences be correlatedwith environmental, historical, orpersonality differences? Finally, psy-chologists might be able to shed lighton the complex economic problem ofinteracting utilities of different goods.

The area of risky decision making,like the area of the theory of games,tends to encourage in those inter-ested in it the custom of carrying out

small pilot experiments on their sons,laboratory assistants, or secretaries.Such experiments are too seldomadequately controlled, and are al-most never used as a basis for larger-scale, well-designed experiments.Whether an ill-designed and hap-hazardly executed little experiment isbetter than no experiment at all isquestionable. The results of suchpilot experiments too often are pickedup and written into the literaturewithout adequate warning about theconditions under which they wereperformed and the consequent limita-tions on the significance of the results.

THE TRANSITIVITY OF CHOICESIn the section on riskless choices

this paper presented a definition ofeconomic man. The most importantpart of this definition can be summedup by saying that economic man isrational. The concept of rationalityinvolves two parts: that of a weakordering of preferences, and that ofchoosing so as to maximize some-thing. Of these concepts, the onewhich seems most dubious is the oneof a weakly ordered preference field.This is dubious because it impliesthat choices are transitive; that is, ifA is preferred to B, and B is preferredto C, then A is preferred to C.

Two economists have designed ex-periments specifically intended totest the transitivity of choices. Pap-andreou performed an elaborate andsplendidly controlled experiment(145) designed to discover whether ornot intransitivities occurred in im-agined-choice situations. He pre-pared triplets of hypothetical bun-dles of admissions to plays, athleticcontests, concerts, etc., and requiredhis subjects to choose between pairsof bundles. Each bundle consisted ofa total of four admissions to twoevents, e.g., 3 plays and 1 tennis

404 WARD EDWARDS

tournament. In the main experi-ment, each bundle is compared withtwo others involving the same kindsof events, but in the better designedauxiliary experiment, a total of sixdifferent events are used, so that eachbundle has no events in common withthe other two bundles in its triplet.Since there are three bundles in eachtriplet, there are three choices be-tween pairs for each triplet, andthese choices may, or may not, betransitive. The subjects were per-mitted to say that they were indiffer-ent between two bundles; conse-quently there were 27 possible con-figurations of choices, of which only13 satisfied the transitivity axiom.In the main experiment, 5 per centof the triplets of judgments wereintransitive; in the auxiliary experi-ment, only 4 per cent. Papandreoudevelops a stochastic model forchoices under such conditions; theresults are certaihly consistent withthe amount of intransitivity per-mitted by his model. Papandreouconcludes that at least for his specificexperimental conditions, transitivitydoes exist.

May (138), using different kinds ofstimuli in a less elaborate experiment,comes up with results less consistentwith transitivity. May required aclassroom group to make pairwisechoices between three marriage part-ners who were identified only bysaying how intelligent, good looking,and rich they were. Judgments ofindifference were not permitted. Theresults were that 27 per cent of thesubjects gave intransitive triads ofchoices. May suggests, very plausi-bly, that intransitive choices may beexpected to occur whenever morethan one dimension exists in thestimuli along which subjects mayorder their preferences. However,May would probably have gotten

fewer intransitivities if he had per-mitted the indifference judgment. Ifsubjects are really indifferent amongall three of the elements of a triad ofobjects, but are required to choosebetween them in pairs and do so bychance, then they will choose in-transitively one-fourth of the time.Papandreou's stochastic model givesone theory about what happenswhen preferences diverge just slightlyfrom indifference, but presumably amore detailed model can be workedout. Papandreou's model permitsonly three states: prefer A to B,prefer B to A, and indifferent. Itought to be possible to base a modelfor such situations on the cumulativenormal curve, and thus to permit anydegree of preference. For every com-bination of degrees of preference,such a model would predict the fre-quency of intransitive choices.

In the paired comparisons amongbets (57) described in the section onrisky choices, quite elaborate in-transitivities could and did occur.However, it is easy to show that anyintransitivity involving four or moreobjects in a paired comparisonsjudgment situation will necessarilyproduce at least one intransitivity in-volving three objects. Consequently,the intransitive triplet or circulartriad is the best unit of analysis forintransitivities in these more com-plicated judgment situations. Icounted the frequency of occurrenceof circular triads and found that theyregularly occurred about 20 per centof the total number of times theycould occur. (Of course, no indiffer-ence judgments could be permitted.)The experiment fulfills May's cri-terion for the occurrence of intransi-tivities, since both probability andamount of money were present ineach bet, and subjects could be ex-pected to take both into account


when making choices. It might besupposed that the difference betweenthe imaginary choices of the Papan-dreou and May experiments and thereal choices in my experiment wouldlead to differences in the frequency ofoccurrence of intransitivities, butthere were no substantial differencesin my experiment between the fre-quencies of occurrence in the just-imagining sessions and in the realgambling sessions, and what differ-ences there were, were in the directionof greater transitivity when reallygambling. These facts should facili-tate further experiments on this prob-lem.

In one sense, transitivity can neverbe violated. A minimum of threechoices is required to demonstrateintransitivity. Since these choiceswill necessarily be made in sequence,it can always be argued that the per-son may have changed his tastes be-tween the first choice and the third.However, unless the assumption ofconstancy of tastes over the period ofexperimentation is made, no experi-ments on choice can ever be mean-ingful, and the whole theory of choicebecomes empty (see 184 for a similarsituation). So this quibble can be re-jected at once.

Utility maximization will not workexcept with a transitive preferencefield. Consequently, if the modelsdiscussed in this paper are to predictexperimental data, it is necessarythat intransitivities in these data beinfrequent enough to be consideredas errors. However, from a slightlydifferent point of view (54) the occur-rence or nonoccurrence of transitivechoice patterns is an experimentalphenomenon, and presumably a law-ful one. May has suggested whatthat law is: Intransitivities occurwhen there are conflicting stimulusdimensions along which to judge.

This notion could certainly be testedand made more specific by appropri-ate experiments.

A final contribution in a related,but different, area is Vail's stochasticutility model (191). Vail assumesthat choices are dependent on utili-ties that oscillate in a random man-ner around a mean value. From thisassumption plus a few other reason-able ones, he deduces that if theover-all preference is 1>2>3, and if1 is preferred to 2 more than 2 ispreferred to 3, then the frequencies ofoccurrence of the six possible transi-tive orderings should be ordered asfollows: 123>132>213>312>231>321. This result is certainly easyto test experimentally, and soundsplausible.

THE THEORY OF GAMES AND OFDECISION FUNCTIONS'

This section will not go into thetheory of games or into the intimatelyrelated subject of statistical decisionfunctions at all thoroughly. Theseare mathematical subjects of a highly

6 Marschak (134), Hurwicz (92), Neisser(143), Stone (181), and Kaysen (107) pub-lished reviews of The Theory of Games andEconomic Behavior which present the funda-mental ideas in much simpler language thanthe original source. Marschak works out indetail the possible solutions of a complicatedthree-person bargaining game, and therebyillustrates the general nature of a solution. Thetwo volumes of Contributions to the Theory ofGames (112, 113), plus McKinsey's book onthe subject (129), provide an excellent bibliog-raphy of the mathematical literature. McKin-sey's book is an exposition of the fundamentalconcepts, intended as a textbook, which issimpler than von Neumann and Morgensternand pursues certain topics further. Wald'sbook (198) is, of course, the classical work onstatistical decision functions. Dross's book(35) presents the fundamental ideas aboutstatistical decision functions more simply, andwith a somewhat different emphasis. Girshickand Blackwell's book (79) is expected to be avery useful presentation of the field.

406 WARD EDWARDS

technical sort, with few statementswhich lend themselves to experimen-tal test. Rather, the purpose of thissection is to show how these subjectsrelate to what has gone before, to givea brief summary of the contents ofTheory of Games and Economic Be-havior by von Neumann and Morgen-stern (197), and to describe a few ex-periments in the area of game playing—experiments which are stimulatedby the theory of games although notdirectly relevant to it.

The theory of games. The theory ofgames probably originated in thework of Borel (31, 32, 33, 34; see also71, 72) in the 1920's. In 1928, vonNeumann (195), working independ-ently of Borel, published the firstproof of the fundamental theorem inthe theory, a theorem that Borel hadnot believed to be generally true.However, the subject did not becomeimportant until 1944, when vonNeumann and Morgenstern pub-lished their epoch-making book (196).(A second edition, with an appendixon cardinal utility measurement,came out in 1947 [197].) Their pur-pose was to analyze mathematically avery general class of problems, whichmight be called problems of strategy.Consider a game of tic-tac-toe. Youknow at any moment in the gamewhat the moves available to your op-ponent are, but you do not knowwhich one he will choose. The onlyinformation you have is that hischoice will not, in general, be com-pletely random; he will make a movewhich is designed in some way to in-crease his chance of winning and di-minish yours. Thus the situation isone of uncertainty rather than risk.Your goals are similar to your op-ponent's. Your problem is: whatstrategy should you adopt? Thetheory of games offers no practicalhelp in developing strategies, but itdoes offer rules about how to choose

among them. In the case of tic-tac-toe, these rules are trivial, sinceeither player can force a draw. Butin more complicated games of strat-egy, these rules may be useful. Inparticular, the theory of games maybe helpful in analyzing proper strat-egy in games having random ele-ments, like the shuffling of cards, orthe throwing of dice. It should benoted that the concept of a game is anexceedingly general concept. A scien-tist in his laboratory may be con-sidered to be playing a game againstNature. (Note, however, that wecannot expect Nature to try to defeatthe scientist.) Negotiators in a labordispute are playing a game againstone another. Any situation in whichmoney (or some valuable equivalent)may be gained as the result of aproper choice of strategy can be con-sidered as a game.

To talk about game theory, a fewtechnical terms are necessary. Astrategy is a set of personal rules forplaying the game. For each possiblefirst move on your part, your op-ponent will have a possible set of re-sponses. For each possible responseby your opponent, you will have a setof responses, and so on through thegame. A strategy is a list which speci-fies what your move will be for everyconceivable previous set of moves ofthe particular game you are playing.Needless to say, only for the simplestgames (e.g., matching pennies) doesthis concept of strategy have anyempirical meaning.

Associated with strategies are im-putations. An imputation is a set ofpayments made as a result of a game,one to each player. In general, differ-ent imputations will be associatedwith different sets of strategies, butfor any given set of strategies theremay be more than one imputation(in games involving coalitions).

Imputation X is said to dominate


imputation F if one or more of theplayers has separately greater gains(or smaller losses) in X than in F andcan, by acting together (in the case ofmore than*one player), enforce theoccurrence of X, or of some other im-putation at least as good. The rela-tionship of domination is not transi-tive.

A solution is a set of imputations,none of which dominates another,such that every imputation outsidethe solution is dominated by at leastone imputation within the solution.Von Neumann and Morgenstern as-sert that the task of the theory ofgames is to find solutions. For anygame, there may be one or more thanone. One bad feature of the theory ofgames is that it frequently gives alarge, or even infinite, number of solu-tions for a game.

The above definitions make clearthat the only determiner of behaviorin games, according to this theory, isthe amounts of money which may bewon or lost, or the expected amountsin games with random elements. Thefun of playing, if any, is irrelevant.

The minimax loss principle. Thenotions of domination and of solutionimply a new fundamental rule fordecision making—a rule sharply dif-ferent from the rule of maximizingutility or expected utility with whichthis paper has been concerned up tothis section. This rule is the rule ofminimizing the maximum loss, or,more briefly, minimax loss. In otherwords, the rule is to consider, for eachpossible strategy that you couldadopt, what the worst possible out-come is, and then to select that strat-egy which would have the least ill-effects if the worst possible outcomehappened. Another way of puttingthe same idea is to call it the principleof maximizing the minimum gain, ormaximin gain. This rule makes con-siderable sense in two-person games

when you consider that the otherplayer is out to get you, and so willdo his best to make the worst possibleoutcome for you occur. If this rule isexpressed geometrically, it assertsthat the point you should seek is asaddle-point, like the highest point ina mountain pass (the best rule forcrossing mountains is to minimize themaximum height, so explorers seekout such saddle-points).

Before we go any further, we needa few more definitions. Games maybe among any number of players, butthe simplest game is a two-persongame, and it is this kind of gamewhich has been most extensively andmost successfully analyzed. Funda-mentally, two kinds of payoff ar-rangements are possible. The sim-plest and most common is the one inwhich one player wins what the otherplayer loses, or, more generally, theone for which the sum of all the pay-ments made as a result of the game iszero. This is called a zero-sum game.In nonzero-sum games, analyticalcomplexities arise. These can be di-minished by assuming the existenceof a fictitious extra player, who winsor loses enough to bring the sum ofpayments back to zero. Such a ficti-tious player cannot be assumed tohave a strategy and cannot, of course,interact with any of the other players.

In zero-sum two-person games,what will happen? Each player, ac-cording to the theory, should pick hisminimax strategy. But will this re-sult in a stable solution? Not always.Sometimes the surface representingthe possible outcomes of the gamedoes not have a saddle-point. In thiscase, if player A chooses his minimaxstrategy, then player B will have anincentive not to use his own minimaxstrategy, because having found outhis opponent's strategy, he can gainmore by some other strategy. Thusthe game has no solution.

408 WARD EDWARDS

Various resolutions of this problemare possible. Von Neumann andMorgenstern chose to introduce thenotion of a mixed strategy, which is aprobability distribution of two ormore pure strategies. The fundamen-tal theorem of the theory of games isthat if both players in a zero-sumtwo-person game adopt mixed strat-egies which minimize the maximumexpected loss, then the game will al-ways have a saddle-point. Thus eachperson will get, in the long run, hisexpected loss, and will have no in-centive to change his behavior evenif he should discover what his op-ponent's mixed strategy is. Since A isalready getting the minimum possibleunder the strategy he chose, anychange in strategy by B will only in-crease A's payoff, and therefore causeB to gain less or lose more than hewould by his own minimax strategy.The same is true of B.

Games involving more than twopeople introduce a new element—thepossibility that two or more playerswill cooperate to beat the rest. Sucha cooperative agreement is called acoalition, and it frequently involvesside-payments among members of thecoalition. The method of analysis forthree-or-more-person games is to con-sider all possible coalitions and tosolve the game for each coalition onthe principles of a two-person game.This works fairly well for three-per-son games, but gets more complicatedand less satisfactory for still morepeople.

This is the end of this exposition ofthe content of von Neumann andMorgenstern's book. It is of courseimpossible to condense a tremendousand difficult book into one page. Themajor points to be emphasized arethese: the theory of games is not amodel of how people actually playgames (some game theorists will dis-

agree with this), nor is it likely to beof any practical use in telling youhow to play a complicated game; thecrux of the theory of games is theprinciple of choosing tfre strategywhich minimizes the maximum ex-pected financial loss; and the theorydefines a solution of a game as a setof imputations which satisfies thisprinciple for all players.

Assumptions, In their book vonNeumann and Morgenstern say "Wehave . . . assumed that [utility] isnumerical . . . substitutable and un-restrictedly transferable between thevarious players." (197, p. 604.) Gametheorists disagree about what thisand other similar sentences mean.One likely interpretation is that theyassume utility to be linear with thephysical value of money involved ina game and to be interpersonallycomparable. The linear utility curvesseem to be necessary for solving two-person games; the interpersonal com-parability is used for the extension ton persons. Attempts are being madeto develop solutions free of these as-sumptions (176).

Statistical decision functions. VonNeumann (195) first used the mini-max principle in his first publicationon game theory in 1928. Neymanand Pearson mentioned its applic-ability to statistical decision prob-lems in 1933 (144). Wald (198), whoprior to his recent death was thecentral figure in the statistical deci-sion-function literature, first seriouslyapplied the minimax principle to sta-tistical problems in 1939. Appar-ently, all these uses of the principlewere completely independent of oneanother.

After Theory of Games and Eco-nomic Behavior appeared in 1944,Wald (198) reformulated the problemof statistical decision making as oneof playing a game against Nature,


The statistician must decide, on thebasis of observations which costsomething to make, between policies,each of which has a possible gain orloss. In some cases, all of these gainsand losses and the cost of observingcan be exactly calculated, as in in-dustrial quality control. In othercases, as in theoretical research, it isnecessary to make some assumptionabout the cost of being wrong and thegain of being right. At any rate, whenthey are put in this form, it is obviousthat the ingredients of the problem ofstatistical decision making have agamelike sound. Wald applied theminimax principle to them in a wayessentially identical with game the-ory.

A very frequent criticism of theminimax approach to games againstNature is that Nature is not hostile,as is the opponent in a two-persongame. Nature will not, in general,use a minimax strategy. For thisreason, other principles of decisionmaking have been suggested. Thesimple principle of maximizing ex-pected utility (which is the essence ofthe Bayes's theorem [15, 198] solutionof the problem) is not always applica-ble because, even though Nature isnot hostile, she does not offer anyway of assigning a probability to eachpossible outcome. In other words,statistical decision making is a prob-lem of uncertainty, rather than ofrisk. Savage has suggested the prin-ciple of minimaxing regret, where re-gret is defined as the difference be-tween the maximum which can begained under any strategy given acertain state of the world and theamount gained under the strategyadopted. Savage believes (170, alsopersonal communication) that neithervon Neumann and Morgenstern norWald actually intended to proposethe principle of minimaxing loss; they

confined their discussions to cases inwhich the concepts of minimax lossand minimax regret amount to thesame thing. Other suggested princi-ples are: maximizing the maximumexpected gain, and maximizing someweighted average of the maximumand minimum expected gains (93).None of these principles commandsgeneral acceptance; each can bemade to show peculiar consequencesunder some conditions (see 170).

Experimental games. The conceptsof the theory of games suggest a newfield of experimentation: How dopeople behave in game situations?Such experimentation would centeron the development of strategies, par-ticularly mixed strategies, and, inthree-or-more-person games, on thedevelopment of coalitions and on thebargaining process. You should* re-member that the theory of gamesdoes not offer a mathematical modelpredicting the outcomes of suchgames (except in a few special cases);all it does is offer useful concepts andlanguage for talking about them, andpredict that certain outcomes willnot occur.

A few minor experiments of thiskind have been conducted by Flood,a mathematician, while he was atRand Corporation. He usually usedcolleagues, many of whom were ex-perts in game theory, and secretariesas subjects. The general design ofhis experiments was that a group ofsubjects were shown a group of de-sirable objects on a table, and toldthat they, as a group, could have thefirst object they removed from thetable, and that they should decideamong themselves which object tochoose and how to allocate it. In thefirst experiment (64) the allocationproblem did not arise because enoughduplicate objects were provided sothat each subject could have one of

410 WARD EDWARDS

the kind of object the group selected.The subjects were Harvard under-graduates, and the final selection wasmade by negotiation and voting. Inthe second experiment (65), in whichthe subjects were colleagues and sec-retaries, a long negotiation processeliminated some of the objects, but atime limit forced a selection by lotfrom among the rest. Further negoti-ations to solve the allocation problemwere terminated by a secretary, whosnatched the object, announced thatit was hers, and then tried to sell it.No one was willing to buy, so the ex-periment terminated. Other experi-ments (66, 67) showed that coalitionssometimes form, that a sophisticatedsubject could blackmail the group foran extra side-payment by threateningto change his vote, and that thelarcenous secretary, having succeededonce, had to be physically restrainedin subsequent sessions to preventmore larceny. The general conclusionsuggested by all these experiments isthat even experts on game theory areless rational and more conventionalthan game theory might lead experi-menters to expect.

Psychological comments. The mostnutritive research problems in thisarea seem to be the social problems ofhow bargaining takes place. Flood'sexperiments left bargainers free andused physical objects, whose utilitiesprobably vary widely from subject tosubject, as stimuli to bargain over.This is naturalistic, but producesdata too complex and too nonnumeri-cal for easy analysis. A simpler situa-tion in which the possible communi-cations from one bargainer to an-other are limited (perhaps by meansof an artificial vocabulary), in whichthe subjects do not see one another,and in which the object bargainedover is simple, preferably beingmerely a sum of money, would be

better. Physical isolation of one sub-ject from another would make it pos-sible to match each subject against astandard bargainer, the experimenteror a stooge, who bargains by a fixedset of rules that are unknown to thesubject. Flood (personal communica-tion) is conducting experiments ofthis sort. For three-or-more-persongames, Asch's (16) technique of usinga group consisting of only one realsubject and all the rest stooges mightwell be used. It would be interesting,for instance, to see how the probabil-ity of a coalition between two playerschanges as the number and power ofplayers united against them increase.

The theory of games is the areaamong those described in this paperin which the uncontrolled and casu-ally planned "pilot experiment" ismost likely to occur. Such experi-ments are at least as dangerous hereas they are in the area of risky de-cision making. Flood's results sug-gest that it is especially important touse naive subjects and to use themonly once, unless the effects of expert-ness and experience are the majorconcern of the experiment.

SUMMARYFor a long time, economists and

others have been developing mathe-matical theories about how peoplemake choices among desirable alter-natives. These theories center on thenotion of the subjective value, orutility, of the alternatives amongwhich the decider must choose. Theyassume that people behave rationally,that is, that they have transitivepreferences and that they choose insuch a way as to maximize utility orexpected utility.

The traditional theory of risklesschoices, a straightforward theory ofutility maximization, was challengedby the demonstration that the mathe-


matical tool of indifference curvesmade it possible to account for risk-less choices without assuming thatutility could be measured on an in-terval scale. The theory of risklesschoices predicted from indifferencecurves has been worked out in detail.Experimental determination of indif-ference curves is possible, and hasbeen attempted. But utility meas-ured on an interval scale is necessary(though not sufficient) for welfareeconomics.

Attention was turned to riskychoices by von Neumann and Mor-genstern's demonstration that com-plete weak ordering of risky choicesimplies the existence of utility meas-urable on an interval scale. Hostellerand Nogee experimentally deter-mined utility curves for money fromgambling decisions, and used them topredict other gambling decisions.Edwards demonstrated the existenceof preferences among probabilities ingambling situations, which compli-cates the experimental measurementof utility. Coombs developed a model

for utility and subjective probabilitymeasured on an ordered metric scale,and did some experiments to test im-plications of the model.

Economists have become worriedabout the assumption that choicesare transitive. Experiments haveshown that intransitive patterns ofchoice do occur, and so stochasticmodels have been developed whichpermit occasional intransitivities.

The theory of games presents anelaborate mathematical analysis ofthe problem of choosing from amongalternative strategies in games ofstrategy. This paper summarizes themain concepts of this analysis. Thetheory of games has stimulated in-terest in experimental games, and afew bargaining experiments whichcan be thought of in game-theoreticalterms have been performed.

All these topics represent a newand rich field for psychologists, inwhich a theoretical structure has al-ready been elaborately worked outand in which many experiments needto be performed.

REFERENCES1. ALCHIAN, A, The meaning of utility

measurement. Amer. econ. Rev., 1953,43, 26-50.

2. ALLAIS, M. Fondements d'une thebriepositive des choix comportant unrisque et critique des postulats etaxiomes de 1'ecole americaine. Col-logue Internationale du Centre Na-tional de la Recherche scientifique, 1952,No. 36.

3. ALLAIS, M. Le comportement de 1'hom-rae ratlonnel devant le risque: critiquedes postulats et axiomes de 1'ecoleamericaine. Econometrica, 1953, 21,503-546.

4. ALLAIS, M. L'Extension des theories de1'equilibre economique gen6ral et durendement social au cas du risque.Econometrica, 1953, 21, 269-290.

5. ALLAIS, M. La psychologic de 1'hommerationnel devant le risque: La theorieet I'exp6rience. J, soc. Statist,, Paris,1953, 94, 47-73.

6. ALLEN, R. G. D. The nature of indiffer-ence curves. Rev. econ. Stud., 1933, 1,110-121.

7. ALLEN, R. G. D. A note on the determi-nateness of the utility function. Rev.econ. Stud., 1934, 2, 155-158.

8. ARMSTRONG, W. E. The determinatenessof the utility function. Econ. J., 1939,49, 453-467.

9. ARMSTRONG, W. E. Uncertainty and theutility function. Econ. J., 1948, 58,1-10.

10. ARMSTRONG, W. E. A note on the theoryof consumer's behavior. Ox}, econ.Pap., 1950, 2, 119-122.

11. ARMSTRONG, W. E, Utility and thetheory of welfare. Ox}, econ. Pap.,1951, 3, 259-271.

12. ARROW, K. J. Alternative approaches tothe theory of choice in risk-takingsituations. Econometrica, 1951, 19,404-437.

13. ARROW, K. J. An extension of the basic

412 WARD EDWARDS

theorems of classical welfare econom-ics. In J. Neyman (Ed.), Proceedingsof the second Berkeley symposium onmathematical statistics and probability.Berkeley: Univ. of Calif. Press, 1951.Pp. S07-S32.

14. ARROW, K. J. Social choice and indi-vidual values. New York: Wiley, 1951.

15. ARROW, K. J., BLACKWELL, p., &GIRSHICK, M. A. Bayes and minimaxsolutions of sequential decision prob-lems. Econometrica, 1949, 17, 213-244.

16. ASCH, S. E. Social psychology. NewYork: Prentice-Hall, 1952.

17. ATTNEAVE, F. Psychological probabilityas a function of experienced frequency./. exp. Psychol, 1953, 46, 81-86.

18. BAUMOL, W. J. Community indifference.Rev. econ. Stud., 1946, 14, 44-48.

19. BAUMOL, W. J. The Neumann-Morgen-stern utility index—an ordinalistview. /. polit. Econ., 1951, 59, 61-66.

20. BAUMOL, W. J. Discussion. Amer. econ.Rev. Suppl, 1953, 43, 415-416.

21. BERGSON (BURK), A. Reformulation ofcertain aspects of welfare economics.Quart. J. Econ., 1938, 52, 310-334.

22. BERNARDELLI, H. Note on the determi-nateness of the utility function. Rev.econ. Stud., 1934, 2, 69-75.

23. BERNARDELLI, H. The end of marginalutility theory? Economica, 1938, 5,192-212.

24. BERNARDELLI, H. A reply to Mr.Samuelson's note. Economica, 1939, 6,88-89.

25. BERNARDELLI, H. A rehabilitation of theclassical theory of marginal utility.Economica, 1952, 19, 254-268.

26. BERNOULLI, D. Specimen theoriaenovae de mensura sortis. ComentariiAcademiae Scientiarum Imperiales Pe-tropolitanae, 1738, 5, 175-192. (Trans,by L. Sommer in Econometrica, 1954,22, 23-36.)

27. BILODEAU, E. A. Statistical versus In-tuitive confidence. Amer. J. Psychol.,1952, 65, 271-277.

28. BISHOP, R. L. Consumer's surplus andcardinal utility. Quart. J. Econ., 1943,57, 421-449.

29. BISHOP, R. L. Professor Knight and thetheory of demand. /. polit. Econ.,1946, 54, 141-169.

30. BOHNERT, H. G. The logical structure ofthe utility concept. In R. M. Thrall,C. H. Coombs, & R. L. Davis (Eds.),Decision Processes. New York: Wiley,in press.

31. BOREL, E. La thforie du jeu et les equa-tions integrates & noyau symfitrique.C. R. Acad. Set., Paris, 1921, 173,1304-1308. (Trans, by L. J. Savage inEconometrica, 1953, 21, 97-100.)

32. BOREL, E. Sur les jeux oil interviennent1'hasard et I'habilitfi des joueurs. InE. Borel, Theorie des probability.Paris: Librairie Scientifique, J. Her-mann, 1924. Pp. 204-224. (Trans, byL. J. Savage in Econometrica, 1953, 21,101-115.)

33. BOREL, E. Algebre et calcul des proba-bilitfe. C. R. Acad. Sci., Paris, 1927,184, 52-53. (Trans, by L. J. Savage inEconometrica, 1953,21, 116-117.)

34. BOREL, E. Traitt du calcul des probabil-ites et de ses applications, applicationsdes jeux de hasard. Vol. IV, No. 2.Paris: Gauthier-Villars, 1938.

35. BROSS, I. Design for decision. NewYork: Macmillan, 1953.

36. BUSH, R. R., & MOSTELLER, F. Amathematical model for simple learn-ing. Psychol. Rev., 1951, 58, 313-323.

37. BUSH, R. R., & MOSTELLER, F. A modelfor stimulus generalization and dis-crimination. Psychol. Rev., 1951, 58,413-423.

38. CARTWRIGHT, D. Decision-time in rela-tion to differentiation of the phenome-nal field. Psychol. Rev., 1941, 48, 425-442.

39. CARTWRIGHT, D. The relation of de-cision-time to the categories of re-sponse. Amer. J. Psychol., 1941, 54,174-196.

40. CARTWRIGHT, D. Survey research: psy-chological economics. In J. G. Miller(Ed.), Experiments in social process.New York: McGraw-Hill, 1950. Pp.47-64.

41. CARTWRIGHT, D., & FESTINGER, L. Aquantitative theory of decision. Psy-chol. Rev., 1943, 50, 595-621.

42. CLARK, J. M. Realism and relevance inthe theory of demand. J. polit. Econ.,1946, 54, 347-353.

43. COOMBS, C. H. Psychological scalingwithout a unit of measurement. Psy-chol. Rev., 1950, 57, 145-158.

44. COOMBS, C. H. Mathematical models inpsychological scaling. /. Amer.statist. Ass., 1951, 46, 480-489.

45. COOMBS, C. H. A theory of psychologi-cal scaling. Bull. Engng Res. Inst.Univer. Mich., 1952, No. 34.

46. COOMBS, C. H. Theory and methods ofsocial measurement. In L. Festinger &D. Katz (Eds.), Research methods in


the behavioral sciences. New York:Dryden, 1953. Pp. 471-S3S.

47. COOMBS, C. H. A method for the studyof interstimulus similarity. Psycho-metrika, in press.

48. COOMBS, C. H., & BEARDSLEE, D. C.Decision making under uncertainty.In R. M. Thrall, C. H. Coombs, &R. L. Davis (Eds.), Decision processes.New York: Wiley, in press.

49. COOMBS, C. H., & MILHOLLAND, J. E.Testing the "rationality" of an indi-vidual's decision making under un-certainty. Psychometrika, in press.

50. CORLETT, W. J., & NEWMAN, P. K. Anote on revealed preference and thetransitivity conditions. Rev. econ.Stud., 1952, 20, 156-158.

51. DE FINETTI, B. La prevision: ses loislogiques, ses sources subjectives. Ann.Inst. Poincare, 1937, 7, 1-68.

52. DE FINETTI, B. Recent suggestions forthe reconciliation of theories of prob-ability. In J. Neyman (Ed.), Proceed-ings of the second Berkeley symposiumon mathematical statistics and prob-ability. Berkeley: Univer. of Calif.Press, 1951.

53. EDGEWORTH, F. Y. Mathematicalpsychics. London: Kegan Paul, 1881.

54. EDWARDS, W. Discussion. Econometrica,1953, 21, 477. (Abstract)

55. EDWARDS, W. Experiments on economicdecision-making in gambling situa-tions. Econometrica, 1953, 21, 349-350. (Abstract)

56. EDWARDS, W. Information, repetition,and reinforcement as determiners oftwo-alternative decisions. Amer.Psychologist, 1953, 8, 345. (Abstract)

57. EDWARDS, W. Probability-preferencesin gambling. Amer. J. Psychol., 1953,66, 349-364.

58. EDWARDS, W. Probability preferencesamong bets with differing expectedvalues. Amer. J. Psychol., 1954, 67,56-67.

59. EDWARDS, W. The reliability of prob-ability preferences. Amer. J. Psychol.,1954, 67, 68-95.

60. ESTES, W. K. Toward a statisticaltheory of learning. Psychol. Rev.,1950, 57, 94-107.

61. FESTINGER, L. Studies in decision: I.Decision-time, relative frequency ofjudgment and subjective confidenceas related to physical stimulus differ-ences. J. exp. Psychol., 1943, 32, 291-306.

62. FES,TINGE;R, L. Studies in decision: II.

An empirical test of a quantitativetheory of decision. J. exp. Psychol.,1943,32,411^23.

63. FISHER, I. A statistical method formeasuring "marginal utility" andtesting the justice of a progressive in-come tax. In J. Hollander (Ed.),Economic essays contributed in honor ofJohn Bates Clark. New York: Macmil-lan, 1927. Pp. 157-193.

64. FLOOD, M. M. A preference experiment.Rand Corp. Memo., November 1951,No. P-256.

65. FLOOD, M. M. A preference experiment(Series 2, Trial 1). Rand Corp. Memo.,December 1951, No. P-258.

66. FLOOD, M. M. A preference experiment(Series 2, Trials 2, 3, 4). Rand Corp.Memo., January 1952, No. P-263.

67. FLOOD, M. M. A preference experiment(Series 3). Unpublished memorandum,Rand Corporation. February 25, 1952.

68. FLOOD, M. M. Some experimentalgames. Rand Corp. Memo., March1952, No. RM-789-1. (Revised June1952.)

69. FLOOD, M. M. Testing organizationtheories. Rand Corp. Memo., Novem-ber 1952, No. P-312.

70. FLOOD, M. M. An experimental multi-ple-choice situation. Rand Corp.Memo., November 1952, No, P-313.

71. FR£CHET, M. Emile Borel, initiator ofthe theory of psychological games andits application. Econometrica, 1953,21, 95-96.

72. PRICKET, M.,& VON NEUMANN, J. Com-mentary on the three notes of EmileBorel. Economelrica, 1953, 21, 118-126.

73. FRIEDMAN, M., & SAVAGE, L. J. Theutility analysis of choices involvingrisk. J. polit. Econ., 1948, 56, 279-304.(Reprinted with minor changes inG. J. Stigler & K. E. Boulding [Eds.],Readings in price theory. Chicago:Richard D. Irwin, 1952. Pp. 57-96.)

74. FRIEDMAN, M., & SAVAGE, L. J. Theexpected-utility hypothesis and themeasurability of utility. /. polit.Econ., 1952, 60, 463-475.

75. FRISCH, R. New methods of measuringmarginal utility. In R. Frisch, Beitragezur okonomischen'Theorie. Tubingen:Mohr, 1932.

76. GEORGESCU-ROEGEN, N. The puretheory of consumer's behavior. Quart.J. Econ., 1936, 50, 545-593.

77- GE.QRGESCU-RQEGEN, N- The theory of

414 WARD EDWARDS

choice and the constancy of economiclaws. Quart. J. Econ., 1950, 64, 125-138.

78. GEORGESCU-ROEGEN, N. Utility, ex-pectations, measurability, prediction.Paper read at Econometric Soc.,Kingston, September, 1953.

79. GIRSHICK, M. A., & BLACKWELL, D.Theory of games and statistical de-cisions. New York: Wiley, 1954.

80. GOOD, I. J. Probability and the weighingof evidence. London: Griffin, 1950.

81. GRIFFITH, R. M. Odds adjustments byAmerican horse-race bettors. Amer.J. Psychol., 1949, 62, 290-294.

82. HARSANYI, J. C. Cardinal utility in wel-fare economics and in the theory ofrisk-taking. /. polit. Econ., 1953, 61,434-435.

83. HART, A. G. Risk, uncertainty, and theunprofitability of compounding prob-abilities. In 0. Lange, F. Mclntyre,& T. O. Yntema (Eds.), Studies inmathematical economics and econo-metrics. Chicago: Univer. of ChicagoPress, 1942. Pp. 110-118.

84. HAYES, S, P., JR. Some psychologicalproblems of economics. Psychol. Bull.,1950, 47, 289-330.

85. HERSTEIN, I. N., & MILNOR, J. Anaxiomatic approach to measurableutility. Econometrica, 1953, 21, 291-297.

86. HICKS, J. R. The foundations of welfareeconomics. Econ. J., 1939, 49, 696-712.

87. HICKS, J. R. Value and capital. Oxford:Clarendon Press, 1939.

88. HICKS, J. R., & ALLEN, R. G. D. A re-consideration of the theory of value.Economica, 1934, 14, 52-76, 196-219.

89. HILDRETH, C. Alternative conditions forsocial orderings. Econometrica, 1953,21, 81-94.

90. HOUTHAKKER, H. S. Revealed prefer-ence and the utility function. Eco-nomica, 1950, 17, 159-174.

91. HULL, C. L. Principles of behavior, anintroduction to behavior theory. NewYork: D. Appleton-Century, 1943.

92. HURWICZ, L. The theory of economic be-havior. Amer. econ. Rev., 1945, 35,909-925. (Reprinted in G. J. Stigler &K. E. Boulding [Eds.], Readings inprice theory. Chicago: Richard D.Irwin, 1952. Pp. 505-526.)

93. HURWICZ, L. What has happened to thetheory of games? Amer. econ. Rev.Suppl., 1953, 43, 398^05.

94. IRWIN, F. W. Stated expectations as

functions of probability and desirabil-ity of outcomes. /. Pers., 1953, 21,329-335.

95. JARRETT, JACQUELINE M. Strategies inrisk-taking situations. UnpublishedPh.D. thesis, Harvard Univer., 1951.

96. JENKINS, W. O., & STANLEY, J. C., JR.Partial reinforcement: a review andcritique. Psychol. Bull., 1950, 47, 193-234.

97. JOHNSON, W. E. The pure theory ofutility curves. Econ. J., 1913, 23, 483-513.

98. KALDOR, N. Welfare propositions andinter-personal comparisons of utility.Econ. J., 1939, 49, 549-552.

99. KALDOR, N. A comment. Rev. econ.Stud., 1946, 14, 49.

100. KAPLAN, A., & RADNER, R. A question-naire approach to subjective prob-ability—some experimental results.Working Memorandum 41, SantaMonica Conference on Decision Prob-lems, August 15, 1952.

101. KATONA, G. Psychological analysis ofbusiness decisions and expectations.Amer. econ. Rev., 1946, 36, 44-62.

102. KATONA, G. Contributions of psycho-logical data to economic analysis. /.Amer. statist. Ass., 1947, 42, 449-459.

103. KATONA, G. Psychological analysis ofeconomic behavior. New York: Mc-Graw-Hill, 1951.

104. KATONA, G. Rational behavior and eco-nomic behavior. Psychol. Rev., 1953,60, 307-318.

105. KAUDER, E. Genesis of the marginalutility theory from Aristotle to theend of the eighteenth century. Econ.J., 1953, 63, 638-650.

106. KAUDER, E. The retarded acceptance ofthe marginal utility theory. Quart. J.Econ., 1953, 67, 564-575.

107. KAYSEN, C. A revolution in economictheory? Rev. econ. Stud., 1946, 14, 1-15.

108. KENNEDY, C. The common sense of in-difference curves. Oxf. econ. Pap.,1950, 2, 123-131.

109. KNIGHT, F. H. Risk, uncertainty, andprofit. Boston: Houghton Mifflin,1921.

110. KNIGHT, F. H. Realism and relevancein the theory of demand. /. polit.Econ., 1944, 52, 289-318.

111. KNIGHT, F. H. Comment on Mr.Bishop's article. /. polit. Econ., 1946,54, 170-176.

112. KUHN, H. W., & TUCKER, A. W. (Eds.)Contributions to the theory of games.


Vol. I. Ann. Math. Stud., No. 24.Princeton: Princeton Univer. Press,1950. 130.

113. KUHN, H. W., & TUCKER, A. W. (Eds.)Contributions to the theory of games.Vol. II. Ann. Math. Stud., No. 28. 131.Princeton: Princeton Univer. Press,1953.

114. LANCASTER, K. A refutation of Mr.Bernardelli. Economica, 1953, 20, 132.259-262.

115. LANDAHL, H. D. A neurobiophysical in- 133.terpretation of certain aspects of theproblem of risks. Bull. Math. Bio-physics, 1951, 13, 323-335.

116. LANGE, O. The determinateness of the 134.utility function. Rev. econ. Stud., 1933,1, 218-225.

117. LANGE, O. Note on the determinatenessof the utility function. Rev. econ. 135.Stud., 1934, 2, 75-77.

118. LANGE, O. The foundations of welfareeconomics. Econometrica, 1942, 10,215-228. 136.

119. LANGE, O. The scope and methods ofeconomics. Rev. econ. Stud., 1945, 13,19-32.

120. LEWIN, K. Principles of topological psy-chology. New York: McGraw-Hill,1936.

121. LEWIN, K. Behavior and developmentas a function of the total situation. 137.In L. Carmichael (Ed.), Manual ofchild psychology. New York: Wiley, 138.1946. Pp. 791-844.

122. LEWIN, K., DEMBO, TAMARA, FESTINGER,L., & SEARS, PAULINE S. Level of 139.aspiration. In J. McV. Hunt (Ed.),Personality and the behavior disorders. 140.Vol. I. New York: Ronald, 1944. Pp.333-378.

123. LEWISOHN, S. A. Psychology in eco- 141.nomics. Polit. Sci. Quart., 1938, 53,233-238.

124. LITTLE, I. M. D. The foundations of 142.welfare economics. Oxf. econ. Pap.,1949, 1, 227-246.

125. LITTLE, I. M. D. A reformulation of the 143.theory of consumer's behavior. Oxf.econ. Pap., 1949, 1, 90-99.

126. LITTLE, I. M. D. The theory of con- 144.sumer's behavior—a comment. Oxf.econ. Pap., 1950, 2, 132-135.

127. LITTLE, I. M. D. Social choice and indi-vidual values. /. polit. Econ., 1952, 145.60, 422H132.

128. MACFIE, A. L. Choice in psychology andas economic assumption. Econ. J.,1953, 63, 352-367. 146.

129. McKiNSEY, J. C. C. Introduction to the

theory of games. New York: McGraw-Hill, 1952.

MALINVAUD, E. Note on von Neumann-Morgenstern's strong independenceaxiom. Econometrica, 1952, 20, 679.

MANNE, A. S. The strong independenceassumption—gasolene blends andprobability mixtures. Econometrica,1952, 20, 665-669.

MARKOWITZ, H. The utility of wealth.J. polit. Econ., 1952, 60, 151-158.

MARKS, ROSE W. The effect of prob-ability, desirability, and "privilege"on the stated expectations of children.J. Pers., 1951, 19, 332-351.

MARSCHAK, J. Neumann's and Morgen-stern's new approach to static eco-nomics. J. polit. Econ., 1946, 54, 97-115.

MARSCHAK, J. Rational behavior, un-certain prospects, and measurableutility. Econometrica, 1950, 18, 111-141.

MARSCHAK, J. Why "should" statisti-cians and businessmen maximize"moral expectation"? In J. Neyman(Ed.), Proceedings of the secondBerkeley symposium on mathematicalstatistics and probability. Berkeley:Univer. of Calif. Press, 1951. Pp. 493-506.

MARSHALL, A. Principles of economics.(8th Ed.) New York: Macmillan, 1948.

MAY, K. O. Transitivity, utility, andaggregation in preference patterns.Econometrica, 1954, 22, 1-13.

MELVILLE, L. G. Economic welfare.Econ. J., 1939, 49, 552-553.

MISHAN, E. J. The principle of com-pensation reconsidered. J. polit. Econ.,1952,60,312-322.

MORGAN, J. N. Can we measure themarginal utility of money? Econo-metrica, 1945, 13, 129-152.

MOSTELLER, F., & NoGEE, P. An experi-mental measurement of utility. /.polit. Econ., 1951, 59, 371-404.

NEISSER, H. The strategy of expectingthe worst. Soc. Res., 1952, 19, 346-363.

NEYMAN, J., & PEARSON, E. S. The test-ing of statistical hypotheses in rela-tion to probability a priori. Proc.Cambr. phil. Soc., 1933, 29, 492-510.

PAPANDREOU, A. G. An experimentaltest of an axiom in the theory ofchoice. Econometrica, 1953, 21, 477.(Abstract)

PARETO, V. Manuale di economic, po-litica, con una introduzione utta scienta

416 WARD EDWARDS

sociale. Milan, Italy: Societa Editrice 165.Libraria, 1906.

147. PHELPS-BROWN, E. H. Note on the de-terminateness of the utility function. 166.Rev. econ. Stud., 1934, 2, 66-69.

148. PIGOU, A. C. Some aspects of welfareeconomics. Amer. econ. Rev., 1951, 41, 167.287-302.

149. PRESTON, M. G., & BARATTA, P. Anexperimental study of the auction- 168.value of an uncertain outcome. Amer.J. Psychol., 1948, 61, 183-193.

150. RAMSEY, F. P. Truth and probability. 169.In F. P. Ramsey, The foundations ofmathematics and other logical essays.New York: Harcourt Brace, 1931.

151. RICCI, U. Pareto and pure economics. 170.Rev. econ. Stud., 1933, 1, 3-21.

152. ROBBINS, L. Interpersonal comparisonsof utility: a comment. Econ. J., 1938, 171.48, 635-641.

153. ROBBINS, L. Robertson on utility andscope. Economica, 1953, 20, 99-111.

154. ROBERTSON, D. H. Utility and all thatand other essays. London: George 172.Allen & Unwin, 1952.

155. ROTHENBERG, J. Conditions for a socialwelfare function. /. polit. Econ., 1953, 173.61, 389-405.

156. ROTHSCHILD, K. W. The meaning ofrationality: a note on Professor Lange's 174.article. Rev. econ. Stud., 1946, 14, 50-52.

157. ROUSSEAS, S. W., & HART, A. G. Ex- 175.perimental verification of a compositeindifference map. J. polit. Econ., 1951,59,288-318. 176.

158. SAMUELSON, P. A. A note on measure-ment of utility. Rev. econ. Stud., 1937,4,155-161.

159. SAMUELSON, P. A. Empirical implica- 177.tions of utility analysis. Econometrica,1938, 6, 344-356.

160. SAMUELSON, P. A. A note on the puretheory of consumer's behavior. Eco-nomica, 1938, S, 61-71.

161. SAMUELSON, P. A. A note on the puretheory of consumer's behavior. An 178.addendum. Economica, 1938, S, 353-354.

162. SAMUELSON, P. A. The numerical repre-sentations of ordered classifications 179,and the concept of utility. Rev. econ.Stud., 1938, 6, 65-70.

163. SAMUELSON, P. A. The end of marginal 180,utility: a note on Dr. Bernardelli'sarticle. Economica, 1939, 6, 86-87.

164. SAMUELSON, P. A. Foundations of eco- 181,nomic analysis. Cambridge, Mass.:Harvard Univer. Press, 1947. 182.

SAMUELSON, P. A. Consumption theoryin terms of revealed preference. Eco-nomica, 1948, 15, 243-253.

SAMUELSON, P. A. Evaluation of realnational income. Oxf. econ. Pap.,1950, 2, 1-29.

SAMUELSON, P. A. The problem of in-tegrability in utility theory. Eco-nomica, 1950, 17, 355-385.

SAMUELSON, P. A. Probability, utility,and the independence axiom. Econo-metrica, 1952, 20, 670-678.

SAMUELSON, P. A. Consumption theo-rems in terms of overcompensationrather than indifference comparisons.Economica, 1953, 20, 1-9.

SAVAGE, L. J. The theory of statisticaldecision. J. Amer. statist. Ass., 1951,46, 55-67.

SAVAGE, L. J. An axiomatic theory ofreasonable behavior in the face of un-certainty. Unpublished manuscript,Statistical Research Center, Univer.of Chicago, No. SRC-21222S14.

SCHULTZ, H. The theory and measure-ment of demand. Chicago: Univer. ofChicago Press, 1938.

SCITOVSKV, T. A note on welfare propo-sitions in economics. Rev. econ. Stud.,1941, 9, 77-88.

SCITOVSKY, T. The state of welfare eco-nomics. Amer. econ. Rev., 1951, 41,303-315.

SHACKLE, G. L. S. Expectations in eco-nomics. Cambridge, Eng.: CambridgeUniver. Press, 1949.

SHAPLEY, L. S., & SHUBIK, M. Solutionsof n-person games with ordinal utili-ties. Econometrica, 1953, 21, 348-349.(Abstract)

SLUTSKY, E. E. Sulla teoria del bilanciodel consumatore, Giornale degli econo-misti, 1915, SI, 1-26. (Trans, by O.Ragusa and reprinted in G. J. Stigler& K. E. Boulding [Eds.], Readings inprice theory. Chicago: Richard D.Irwin, 1952. Pp. 27-56.)

SPROWLS, R. C. Psychological-mathe-matical probability in relationships oflottery gambles. Amer. J. Psychol.,1953, 66, 126-130.

STIGLER, G. J. The limitations of sta-tistical demand curves. J. Amer.statist. Ass., 1939, 34, 469-481.

STIGLER, G. J. The development ofutility theory. /. polit. Econ., 1950,58, 307-327, 373-396.

STONE, J. R. N. The theory of games.Econ. J., 1948, 58, 185-201.

STONE, R. (J. R. N.) The role of measure-


ment in economics. Cambridge, Eng.:Cambridge Univer. Press, 1951.

183. STROTZ, R. H. Cardinal utility. Amer.econ. Rev. SuppL, 1953, 43, 384-405.

184. SWEEZY, A. R. The interpretation ofsubjective value theory in the writingsof the Austrian economists. Rev. econ.Stud., 1933, 1, 176-185.

185. THURSTONE, L. L. The indifferencefunction. /. soc. Psychol., 1931, 2,139-167.

186. THURSTONE, L. L. The measurement ofvalues. Psychol. Rev., 1954, 61, 47-58.

187. TINTNER, G. The theory of choice undersubjective risk and uncertainty. Econ-ometrica, 1941, 9, 298-304.

188. TINTNER, G. A contribution to the non-static theory of choice. Quart. J.Econ., 1942, 56, 274-306.

189. T6RNQVIST, L. On the economic theoryof lottery-gambles. Skand. Aktuar-Tidskr., 1945, 28, 228-246.

190. VAIL, S. V. Expectations, degrees of be-lief, psychological probabilities. Un-published manuscript Univer. of Mich-igan, Seminar on the Application ofMathematics to the Social Sciences,October 23, 1952.

191. VAIL, S. V. A stochastic model of utili-ties. Unpublished manuscript, No. 24,Univer. of Michigan, Seminar on theApplications of Mathematics to theSocial Sciences, April 23, 1953.

192. VAIL, S. V. Alternative calculi of sub-jective probabilities. In R. M. Thrall,C. H. Coombs, & R. L. Davis (Eds.),Decision processes. New York: Wiley,in press.

193. VICKREY, W. S. Measuring marginalutility by reactions to risk. Econo-metrica, 1945, 13, 319-333.

194. VINER, J. The utility concept in valuetheory and its critics. /. polit. Econ.,1925, 33, 369-387, 638-659.

195. VON NEUMANN, J. Zur Theorie derGesellschaftsspiele. Math. Ann., 1928,100, 295-320.

196. VON NEUMANN, J., & MORGENSTERN, O.

Theory of games and economic behavior.(1st Ed.) Princeton: Princeton Univer.Press, 1944.

197. VON NEUMANN, J., & MORGENSTERN, O.Theory of games and economic be-havior. (2nd Ed.) Princeton: Prince-ton Univer. Press, 1947.

198. WALD, A. Statistical decision functions.New York: Wiley, 1950.

199. WALKER, K. F. The psychological as-sumptions of economics. Econ. Rec.,1946, 22, 66-82.

200. WALLIS, W. A., & FRIEDMAN, M. Theempirical derivation of indifferencefunctions. In O. Lange, F. Mclntyre,& T. O. Yntema, (Eds.), Studies inmathematical economics and economet-rics. Chicago: Univer. of ChicagoPress, 1942.

201. WECKSTEIN, R. S. On the use of thetheory of probability in economics.Rev. econ. Stud., 1953, 20, 191-198.

202. WEISSKOPF, W. A. Psychological as-pects of economic thought. /. polit.Econ., 1949, 57, 304-314.

203. WELDON, J. C. A note on measures ofutility. Canad. J. Econ. polit. Sci.,1950, 16, 227-233.

204. WOLD, H. A synthesis of pure demandanalysis. Part I. Skand. Aktuar~Tidskr., 1943, 26, 85-118.

205. WOLD, H. A synthesis of pure demandanalysis. Part II. Skand. Aktuar-Tidskr., 1943, 26, 220-263.

206. WOLD, H. A synthesis of pure demandanalysis. Part III. Skand. Aktuar-Tidskr., 1944, 27, 69-120.

207. WOLD, H. Ordinal preferences or cardi-nal utility? Econometrica, 1952, 20,661-664.

208. WOLD, H., & JUR^EN, L. Demand analy-sis. New York: Wiley, 1953.

209. ZEUTHEN, F. On the determinateness ofthe utility function. Rev. econ. Stud.,1937, 4, 236-239.

Received for early publication A pril 8, 1954.

the theory of decision making - semantic scholar · bar. the economic theory of decision making is...

Documents