Skip to main content
Proceedings of the National Academy of Sciences of the United States of America logoLink to Proceedings of the National Academy of Sciences of the United States of America
. 2013 Apr 4;110(17):6754–6759. doi: 10.1073/pnas.1207805110

Classical subjective expected utility

Simone Cerreia-Vioglio a, Fabio Maccheroni a, Massimo Marinacci a,1, Luigi Montrucchio b,c
PMCID: PMC3637764  PMID: 23559375

Abstract

We consider decision makers who know that payoff-relevant observations are generated by a process that belongs to a given class M, as postulated in Wald [Wald A (1950) Statistical Decision Functions (Wiley, New York)]. We incorporate this Waldean piece of objective information within an otherwise subjective setting à la Savage [Savage LJ (1954) The Foundations of Statistics (Wiley, New York)] and show that this leads to a two-stage subjective expected utility model that accounts for both state and model uncertainty.


Consider a decision maker who is evaluating acts whose outcomes depend on some verifiable states, that is, on observations (workers’ outputs, urns’ drawings, rates of inflation, and the like). If the decision maker (DM) believes that observations are generated by some probability model, two sources of uncertainty affect his evaluation: model uncertainty and state uncertainty. The former is about the probability model that generates observations, and the latter is about the state that obtains (and that determines acts’ outcomes).

State uncertainty is payoff relevant and, as such, it is directly relevant for the DM’s decisions. Model uncertainty, in contrast, is not payoff relevant and its role is instrumental relative to state uncertainty. Moreover, models cannot always be observed: Whereas in some cases they have a simple physical description (e.g., urns’ compositions), often they do not have it (e.g., fair coins). For these reasons, the purely subjective choice frameworks à la Savage (1) focus on the verifiable and payoff-relevant state uncertainty. They posit an observation space S over which subjective probabilities are derived via betting behavior.

In contrast, classical statistical decision theory à la Wald (2) assumes that the DM knows that observations are generated by a probability model that belongs to a given subset M, whose elements are regarded as alternative random devices that nature may select to generate observations. [As Wald (ref. 2, p. 1) writes, “A characteristic feature of any statistical decision problem is the assumption that the unknown distribution Inline graphic is merely known to be an element of a given class Ω of distributions functions. The class Ω is to be regarded as a datum of the decision problem.”] In other words, Wald’s approach posits a model space M in addition to the observation space S. In so doing, Wald adopted a key tenet of classical statistics, that is, to posit a set of possible data-generating processes (e.g., normal distributions with some possible means and variances), whose relative performance is assessed via available evidence [often collected with independent identically distributed (i.i.d.) trials] through maximum-likelihood methods, hypothesis testing, and the like. Although models cannot be observed, in Wald’s approach their study is key to better understanding state uncertainty.

Is it possible to incorporate this Waldean key piece of objective information within Savage’s framework? Our work addresses this question and tries to embed this classical datum within an otherwise subjective setting. In addition to its theoretical interest, this question is relevant because in some important economic applications it is natural to assume, at least as a working hypothesis, that DMs have this kind of information [e.g., Sargent (3)].

Our approach takes the objective information M as a primitive and enriches the standard Savage framework with this datum: DMs know that the true model m that generates data belongs to M. Behaviorally, this translates into the requirement that their betting behavior (and so their beliefs) be consistent with M,

graphic file with name pnas.1207805110uneq1.jpg

where xFy and xEy are bets on events F and E, with xy. We do not, instead, consider bets on models and, as a result, we do not elicit prior probabilities on them through hypothetical (because models are not in general observable) betting behavior. Nevertheless, our basic representation result, Proposition 1, shows that, under Savage’s axioms P.1–P.6 and the above consistency condition, acts are ranked according to the criterion

graphic file with name pnas.1207805110eq1.jpg

where μ is a subjective prior probability on models, whose support is included in M. We call this representation classical subjective expected utility because of the classical Waldean tenet on which it relies.

The prior μ is a subjective probability that may also reflect some personal information on models that the DM may have, in addition to the objective information M. Uniqueness of μ corresponds to the linear independence of the set M. For example, M is linearly independent when its members are pairwise orthogonal. Remarkably, some important time series models widely used in economic and financial applications satisfy this condition, as discussed later in the paper. For this reason, our Wald–Savage setup provides a proper statistical decision theory framework for empirical works that rely on such time series.

Each prior μ induces a predictive probability Inline graphic on the sample space S through model averaging:

graphic file with name pnas.1207805110eq2.jpg

In particular, setting Inline graphic,

graphic file with name pnas.1207805110eq3.jpg

is the reduced form of V, its subjective expected utility (SEU) representation à la Savage. On the other hand, when M is a singleton {m}, we have Inline graphic for all priors μ and we thus get the von Neumann–Morgenstern expected utility representation

graphic file with name pnas.1207805110eq4.jpg

where subjective probabilities do not play any role. [Lucas (ref. 4, p. 15) writes that “Muth (5) … [identifies] … agents’ subjective probabilities … with ‘true’ probabilities, calling the assumed coincidence of subjective and ‘true’ probabilities rational expectations” [Italics in the original]. In our setting, this coincidence is modeled by singleton M and results in the expected utility criterion (4).] Classical SEU thus encompasses both the Savage and the von Neumann–Morgenstern representations.

In particular, the Savage criterion [3] is what an outside observer, unaware of datum M, would be able to elicit from the DM’s behavior. It is a much weaker representation than the “structural” one ([1]), which is the criterion that, instead, an outside observer aware of M would be able to elicit. For, this informed observer would be able to focus on the map Inline graphic from priors with support included in datum M to predictive probabilities. Under the linear independence of datum M, by inverting this map the observer would be able to recover prior μ from the predictive probability Inline graphic, which can be elicited through standard methods. The richer Waldean representation [1] is thus summarized by a triple (u, M, μ), with suppμM, whereas for the usual Savagean representation [3] is enough a pair (u, P).

Summing up, although the work of Savage (1) was inspired by the seminal decision theoretic approach of Wald (2), his purely subjective setup and the ensuing large literature did not consider the classical datum central in Wald’s approach. [See Fishburn (6), Kreps (7), and Gilboa (8). See Jaffray (9) for a different “objective” approach.] In this paper we show how to embed this datum in a Savage setting and how to derive the richer Waldean representation [1] by considering only choice behavior based on observables. Battigalli et al. (10) use the Wald–Savage setup of the present paper to study self-confirming equilibria, whereas we are currently using it to provide a behavioral foundation of the robustness approach in macroeconomics pioneered by Hansen and Sargent (11).

Preliminaries

Subjective Expected Utility.

We consider a standard Savage setting, where (S, Σ) is a measurable state space and X is an outcome space. An act is a map f: SX that delivers outcome f (s) in state S. Let ℱ be the set of all simple and measurable acts. [Maps Inline graphic such that Inline graphic is finite and Inline graphic for all Inline graphic.]

The DM’s preferences are represented by a binary relation ≿ over ℱ. We assume that ≿ satisfies the classic Savage axioms P.1–P.6. By his famous representation theorem, these axioms are equivalent to the existence of a utility function u: X → ℝ and a (strongly) nonatomic finitely additive probability P on S such that the SEU evaluation Inline graphic represents ≿. [Strong nonatomicity of P means that for each Inline graphic and Inline graphic there exists Inline graphic such that Inline graphic and Inline graphic. See ref. 12, p. 141–143 for the various definitions and properties of nonatomicity of finitely additive probabilities.] In this case, u is cardinally unique and P is unique.

Given any f, g ∈ ℱ and E ∈ Σ, f Eg is the act equal to f on E and to g otherwise. The conditional preference ≿E is the binary relation on ℱ defined by fE g if and only if f EhgEh for all h ∈ ℱ. By P.2, the sure thing principle, ≿E is complete. An event E ∈ Σ is said to be null if ≿E is trivial (ref. 1, p. 24); in the representation, this amounts to P(E) = 0 (E is null if and only if it is P-null).

For each nonnull event E, the conditional preference ≿E satisfies P.1–P.6 because the primitive preference does [e.g., Kreps (ref. 7, Chap. 10)]. Hence, Savage’s theorem can be stated in conditional form by saying that ≿ satisfies P.1–P.6 if and only if there is a utility function u: X → ℝ and a nonatomic finitely additive probability P on S such that, for each nonnull event E,

graphic file with name pnas.1207805110eq5.jpg

represents ≿E, where P(⋅|E) is the conditional of P given E.

Models, Priors, and Posteriors.

As usual, we denote by Δ = Δ(S, Σ) the collection of all (countably additive) probability measures on S. Unless otherwise stated, in the rest of this paper all probability measures are countably additive.

In the sequel, we consider subsets M of Δ. Each subset M of Δ is endowed with the smallest σ-algebra ℳ that makes the real valued and bounded functions on M of the form mm(E) measurable for all E ∈ Σ and that contains all singletons. In the important special case M = Δ, we write Inline graphic instead of ℳ.

Probability measures μ on Δ are interpreted as prior probabilities. The observation of a (non-Inline graphic-null) event E allows us to update prior μ through the Bayes rule

graphic file with name pnas.1207805110uneq2.jpg

for all DInline graphic, thus obtaining the posterior of μ given E.

A finite subset M = {m1, … , mn} of Δ is linearly independent if, given any collection of scalars {α1, … , αn} ⊆ ℝ,

graphic file with name pnas.1207805110eq6.jpg

Two probability measures m and m′ in Δ are orthogonal (or singular), written mm′, if there exists E ∈ Σ such that m(E) = 0 = m′(Ec). A collection of models M ⊆ Δ is orthogonal if its elements are pairwise orthogonal.

If E ∈ Σ and m(E) = 0 imply m′(E) = 0, m′ is absolutely continuous with respect to m and we write m′ ≪ m.

Finally, we denote by Δna the collection of all nonatomic probability measures. By the classical Lyapunov theorem, the range {(m1(E), … , mn(E)): E ∈ Σ} of a finite collection Inline graphic of nonatomic probability measures is a convex subset of ℝn.

Representation

Basic Result.

The first issue to consider in our normative approach is how DMs’ behavior should reflect the fact that they regard M as a datum of the decision problem. To this end, given a subset M of Δ, say that an event E is unanimous if m(E) = m′(E) for all m, m′∈ M. In other words, all models in M assign the same probability to event E.

Definition 1.

A preferenceis consistent with a subset M of Δ if, given E, F ∈ Σ, with E unanimous,

graphic file with name pnas.1207805110eq7.jpg

for all outcomes xy.

Consistency requires that the DM is indifferent among bets on events that all probability models in M classify as equally likely. The next stronger consistency property requires that DMs prefer to bet on events that are more likely according to all models.

Definition 2.

A preferenceis order consistent with a subset M of Δ if, given E, F ∈ Σ, with E unanimous,

graphic file with name pnas.1207805110eq8.jpg

for all outcomes xy.

Both these notions are minimal consistency requirements among information and preference that behaviorally reveal (to an outside observer) that the DM considers M as a datum of the decision problem. Note that order consistency implies consistency because the premise of [7] implies that also F must be unanimous (this observation also emphasizes how weak an assumption is consistency).

We can now state our basic representation result, which considers finite sets M of nonatomic models.

Proposition 1.

Let M be a finite subset of Δna. The following statements are equivalent:

  • i) ≿ is a binary relation onthat satisfies P.1–P.6 and is order consistent with M;

  • ii) there exist a nonconstant utility function u: X → ℝ and a prior μ on Δ with suppμM, such that

graphic file with name pnas.1207805110eq9.jpg

represents ≿.

Moreover, u is cardinally unique for eachsatisfying statement i, whereas μ is unique for each suchif and only if M is linearly independent.

Although uniqueness of the utility function u is well known and well discussed in the literature, uniqueness of the prior μ is an important feature of this result. In fact, it pins down μ even though its domain is made of unobservable probability models. Because of the structure of Δ, it is the linear independence of M—not just its affine independence—that turns out to be equivalent to this uniqueness property. This simple, but useful, fact is well known [e.g., Teicher (13)].

Each prior μ: Inline graphic → [0, 1] induces a predictive probability Inline graphic on the sample space through the reduction [2]. The reduction map Inline graphic relates subjective probabilities on the space M of models to subjective probabilities on the sample space S, that is, prior and predictive probabilities. [Note that probability measures on S can play two conceptually altogether different roles: (subjective) predictive probabilities and (objective) probability models.] Clearly, [9] implies that

graphic file with name pnas.1207805110eq10.jpg

which is the reduced form of V, its Savage’s SEU form. As observed in the introductory section, this is the criterion that an outside observer, unaware of datum M, would be able to elicit from the DM’s behavior. It is a much weaker representation than the structural one ([9]), which can be equivalently written as

graphic file with name pnas.1207805110uneq3.jpg

because suppμM (recall that finite subsets of Inline graphic are measurable). This is the criterion that, instead, an outside observer aware of M would be able to elicit. In fact, denote by Δ(M) the collection of all priors μ: Inline graphic → [0, 1] such that suppμM. The informed observer would be able to focus on the restriction of the reduction map Inline graphic to Δ(M). If M is linearly independent, such correspondence is one-to-one and thus allows prior identification from the behaviorally elicited Savagean probability Inline graphic through inversion.

The structural representation [9] is a version of Savage’s representation that may be called classical SEU because it takes into account Waldean information, with its classical flavor. [Diaconis and Freedman (14) call “classical Bayesianism” the Bayesian approach that considers as a datum of the statistical problem the collection of all possible data-generating mechanisms.] In place of the usual SEU pair (u, P) the representation is now characterized by a triple (u, M, μ), with suppμM. According to the Bayesian paradigm, the prior μ quantifies probabilistically the DM’s uncertainty about which model in M is the true one. This kind of uncertainty is sometimes called (probabilistic) model uncertainty or parametric uncertainty.

In the introductory section, we observed that when datum M is a singleton, the classical SEU criterion [9] reduces to the von Neumann–Morgenstern expected utility criterion [4], which is thus the special case of classical SEU that corresponds to singleton data. In contrast, when M is nonsingleton but the support of a prior μ is a singleton, say suppμ = {m′} ⊆ M, then it is the DM’s personal information that prior μ reflects, which leads him to a predictive that coincides with Inline graphic. In this case,

graphic file with name pnas.1207805110uneq4.jpg

is a Savage’s SEU criterion.

Support.

In Proposition 1 the support of the prior is included in M; i.e., suppμM. In fact, because of consistency, models are assigned positive probability only if they belong to datum M. However, the DM may well decide to disregard some models in M because of some personal information. This additional information is reflected by his subjective belief μ, with strict inclusion and μ(m) = 0 for some mM. (In fact, the interpretation of μ is purely subjective, not at all logical/objective à la Carnap and Keynes.)

Next we behaviorally characterize suppμ as the smallest subset of M relative to which ≿ is consistent. These are the models that the DM believes to carry significant probabilistic information for his decision problem. In this perspective it is important to remember that M is a datum of the problem whereas suppμ is a subjective feature of the preferences.

We consider a linearly independent M in view of the identification result of Proposition 1.

Proposition 2.

Let M be a finite and linearly independent subset of Δna andbe a preference represented as in point ii of Proposition 1. A model mM belongs to suppμ if and only ifis not consistent with M\m.

Therefore, consistency arguments not only reveal the acceptance of a datum M, but also allow us to discover what elements of M are subjectively maintained or discarded.

Variations.

We close by establishing the conditional and orthogonal versions of Proposition 1. We begin with the conditional version, i.e., with the counterpart of representation [5] under Waldean information.

Proposition 3.

Let M be a finite subset of Δna. The following statements are equivalent:

  • i) ≿ is a binary relation onthat satisfies P.1–P.6 and is order consistent with M;

  • ii) there exist a nonconstant utility function u: X → ℝ and a prior μ on Δ with suppμM, such that

graphic file with name pnas.1207805110eq11.jpg

represents Inline graphic for all non-Inline graphic-null events Inline graphic.

Moreover, u is cardinally unique for each Inline graphic satisfying statement i, whereas μ is unique for each such Inline graphic if and only if M is linearly independent.

The representation of the conditional preferences Inline graphic thus depends on the conditional models Inline graphic and on the posterior probability Inline graphic that, respectively, update models and prior in the light of E. Criterion [11] shows how the DM currently plans to use the information he may gather through observations to update his inference on the actual data-generating process. [As Marschak (ref. 15, p. 109) remarked “to be an ‘economic man’ implies being a ‘statistical man’.” Some works of Jacob Marschak (notably refs. 15 and 16 and his classic book, ref. 17, with Roy Radner) have been a source of inspiration of our exercise, as we discuss in ref. 18. Our work addresses, inter alia, the issue that he raised in ref. 16, in which he asked how to pin down subjective beliefs on models from observables. In so doing, our analysis also shows that to study general data M, possibly linearly dependent, it is necessary to go beyond betting behavior on observables.]

The conditional predictive probability is

graphic file with name pnas.1207805110eq12.jpg

and therefore the reduced form of [11] is

graphic file with name pnas.1207805110eq13.jpg

The conditional representations [11] and [13] are, respectively, induced by the primitive representations [9] and [10] via conditioning.

Orthogonality is a simple, but important, sufficient condition for linear independence that, as the next section shows, some fundamental classes of time series models satisfy. Because of its importance, the following result shows what form the classical SEU representation of Proposition 1 takes in this case.

Proposition 4.

Let M be a finite and orthogonal subset of Inline graphic. The following statements are equivalent:

  • i) Inline graphic is a binary relation onthat satisfies P.1–P.6 and is consistent with M;

  • ii) there exist a nonconstant utility function Inline graphic and a prior μ on Δ with Inline graphic, such that

graphic file with name pnas.1207805110uneq5.jpg

represents Inline graphic.

Moreover, for each Inline graphic satisfying i, u is cardinally unique and μ is unique.

Note that here consistency suffices and that the prior μ is automatically unique because of the orthogonality of M. In ref. 18 we also show that a representation with an infinite M can be derived in the orthogonal case.

The reduction map Inline graphic between prior and predictive probabilities is easily seen to be affine. More interestingly, in the orthogonal case it also preserves orthogonality and absolute continuity.

Proposition 5.

Under the assumptions of Proposition 4, two priors μ and ν on Δ with support in M are orthogonal (resp., absolutely continuous) if and only if their predictive probabilities Inline graphic and Inline graphic on S are orthogonal (resp., absolutely continuous).

Intertemporal Analysis

Setup.

Consider a standard intertemporal decision problem where information builds up through observations generated by a sequence Inline graphic of random variables taking values on observation spaces Inline graphic. For ease of exposition, we assume that the observation spaces are finite and identical, each denoted by Inline graphic and endowed with the σ-algebra Inline graphic.

The relevant state space S for the decision problem is the sample space Inline graphic. Its points are the possible observation paths generated by the process Inline graphic. Without loss of generality, we identify Inline graphic with the coordinate process such that Inline graphic for each Inline graphic.

Endow Inline graphic with the product σ-algebra Inline graphic generated by the elementary cylinder sets Inline graphic. These sets are the observables in this intertemporal setting. In particular, the filtration Inline graphic, where Inline graphic is the algebra generated by the cylinders Inline graphic, records the building up of observations. Clearly, Inline graphic is the σ-algebra generated by the filtration Inline graphic.

In this intertemporal setting the pair Inline graphic is thus given by Inline graphic. The space of data-generating models Δ consists of all probability measures m on Inline graphic. The outcome space X has also a product structure Inline graphic, where Inline graphic is a common instant outcome space. Acts Inline graphic can thus be identified with the processes Inline graphic of their components. When such processes are adapted, the corresponding acts are called plans [here Inline graphic is the outcome at time t if state s obtains]. By Proposition 3, the conditional version of the classical SEU representation at Inline graphic is

graphic file with name pnas.1207805110eq14.jpg

where Inline graphic and Inline graphic are, respectively, the conditional model and the posterior probability given the observation history Inline graphic. Under standard conditions, the intertemporal utility function Inline graphic in [14] has a classic discounted form Inline graphic, with subjective discount factor Inline graphic and bounded instantaneous utility function Inline graphic.

Stationary Case.

The next known result (e.g., ref. 19, p. 39) shows that models are orthogonal in the fundamental stationary and ergodic case, which includes the standard i.i.d. setup as a special case.

Proposition 6.

A finite collection M of models that make the process Inline graphic stationary and ergodic is orthogonal.

By Proposition 4, if Inline graphic satisfies P.1–P.6 and is consistent with a finite collection M of nonatomic, stationary and ergodic models, then there are a cardinally unique utility function u and a unique prior μ, with Inline graphic, such that [14] holds. Its reduced form Inline graphic features a predictive probability Inline graphic that is stationary (exchangeable in the special i.i.d. case).

Because a version of Proposition 6 holds also for collections of homogenous Markov chains, we can conclude that time series models widely used in applications satisfy the orthogonality conditions that ensure the uniqueness of prior μ. The Wald–Savage setup of this paper provides a statistical decision theory framework for empirical works that rely on such time series (as is often the case in the finance and macroeconomics literatures).

Under these orthogonality conditions, there is full learning. Formally, denoting by

graphic file with name pnas.1207805110uneq6.jpg

the continuation value at Inline graphic of any act f and by Inline graphic the true model, it can be shown that

graphic file with name pnas.1207805110uneq7.jpg

for Inline graphic almost every z in Inline graphic. As observations build up, DMs learn and eventually behave as SEU DMs who know the true model that generates observations. The above convergence result shows how the Classical SEU framework of this paper allows to formalize, in terms of learning, the common justification of rational expectations according to which “with a long enough historical data record, statistical learning will equate objective and subjective probability distributions” [Sargent and Williams (ref. 20, p. 361)]. Further intertemporal results are studied in ref. 18 (the working paper version of this paper), which we refer the interested reader to.

Appendix: Proofs and Related Analysis

Letting M be a subset of Inline graphic, a probability measure Inline graphic is said to be a predictive of a prior on M (or to be M-representable) if there exists Inline graphic such that Inline graphic. If in addition such μ is unique, then P is said to be M-identifiable (13).

We state the next result for any M because the proof for the finite case is only slightly simpler. We say that a subset M of Inline graphic is measure independent if, given any signed measure Inline graphic,

graphic file with name pnas.1207805110uneq8.jpg

If M is finite, measure independence reduces to the usual notion ([6]) of linear independence.

Lemma 1.

Let Inline graphic. The following statements are equivalent:

  • i) every predictive of a prior on M is M-identifiable;

  • ii) the map Inline graphic from Inline graphic to Inline graphic is injective;

  • iii) M is measure independent.

Proof:

The equivalence of statements i and ii is trivial.

Statement iii implies ii. If Inline graphic are such that Inline graphic, then Inline graphic is a signed measure on M and

graphic file with name pnas.1207805110uneq9.jpg

Because M is measure independent, it follows that Inline graphic; i.e., Inline graphic.

Statement ii implies iii. Assume, per contra, that M is not measure independent. Then, there is a signed measure γ on M such that

graphic file with name pnas.1207805110eq15.jpg

By the Hahn–Jordan decomposition theorem, Inline graphic, where Inline graphic and Inline graphic are, respectively, the positive and negative parts of γ. By [15],

graphic file with name pnas.1207805110uneq10.jpg

Because Inline graphic, this implies that Inline graphic. Then Inline graphic, Inline graphic (else Inline graphic), and, by [15], for each Inline graphic

graphic file with name pnas.1207805110uneq11.jpg

Therefore, Inline graphic, negating injectivity. ▪

Lemma 2.

If Inline graphic is finite, then

graphic file with name pnas.1207805110uneq12.jpg

Moreover, the map Inline graphic from Inline graphic to Inline graphic is injective if and only if M is linearly independent.

Proof:

The equality Inline graphic follows from the fact that ℳ contains all singletons. Next we show that Inline graphic contains all singletons. Note that if Inline graphic in M, there exists Inline graphic such that Inline graphic. Then for each Inline graphic,

graphic file with name pnas.1207805110uneq13.jpg

is a finite intersection of Inline graphic-measurable sets and so it is measurable too.

Recall that Inline graphic whereas Inline graphic is the set of all probability measures Inline graphic.

Let Inline graphic. Setting Inline graphic for all Inline graphic, it follows that Inline graphic and Inline graphic. Denote by Inline graphic the restriction of Inline graphic to Inline graphic and note that Inline graphic, where Inline graphic is the restriction of Inline graphic (defined on Inline graphic) to ℳ, and that Inline graphic. If M is linearly independent, then Inline graphic implies Inline graphic. By Lemma 1, Inline graphic. Thus, Inline graphic for all Inline graphic and Inline graphic. This proves injectivity.

Conversely, if M is not linearly independent, by Lemma 1 there exist Inline graphic and Inline graphic in Inline graphic such that Inline graphic but Inline graphic. Now, setting Inline graphic for Inline graphic, it follows that Inline graphic but Inline graphic. This negates injectivity. ▪

Proof of Proposition 1:

Statement i implies ii. By the Savage representation theorem, there are a nonconstant function Inline graphic and a unique (strongly) nonatomic and finitely additive probability P on S such that setting Inline graphic,

graphic file with name pnas.1207805110uneq14.jpg

By assumption, each m is nonatomic. By the Lyapunov theorem, there is a unanimous event Inline graphic, say with Inline graphic for all Inline graphic. By order consistency, for each Inline graphic

graphic file with name pnas.1207805110eq16.jpg

and

graphic file with name pnas.1207805110eq17.jpg

By ref. 21, Theorem 20, P belongs to the convex cone generated by M, because Inline graphic for all Inline graphic, and then Inline graphic and representation [9] holds.

Statement ii implies i. Define Inline graphic. Because each Inline graphic is a nonatomic probability measure, so is P. By the Savage representation theorem, it follows that Inline graphic satisfies P.1–P.6. Finally, we show that Inline graphic is order consistent with M. Let Inline graphic and assume Inline graphic for each Inline graphic. Then for all outcomes Inline graphic, normalizing u so that Inline graphic, Inline graphic, and so Inline graphic. A fortiori order consistency is satisfied (with respect to both Inline graphic and M).

Moreover, for each Inline graphic satisfying statement i, the cardinal uniqueness of u and the uniqueness of Inline graphic follow from the Savage representation theorem. If M is linearly independent, for each Inline graphic satisfying i, Inline graphic is unique and Lemma 2 delivers the uniqueness of μ. Conversely, if M is not linearly independent, by Lemma 2 there exist two different Inline graphic such that Inline graphic; arbitrarily choose a nonconstant Inline graphic to obtain a binary relation Inline graphic satisfying i that is represented both by Inline graphic and by Inline graphic (together with u) in the sense of ii. ▪

Proof of Proposition 2:

Let Inline graphic. Replicating the last part of the previous proof, if m does not belong to Inline graphic, then Inline graphic is consistent with Inline graphic. Now assume that Inline graphic is consistent with Inline graphic. Take Inline graphic such that Inline graphic for all Inline graphic; by consistency, if Inline graphic, then

graphic file with name pnas.1207805110uneq15.jpg

If m belongs to Inline graphic, then

graphic file with name pnas.1207805110uneq16.jpg

Because each element of M is nonatomic, by ref. 21, Theorem 20, Inline graphic, which contradicts the linear independence of M. ▪

Proof of Proposition 3:

Clearly statement ii of this proposition implies point ii of Proposition 1, which in turn implies i.

Conversely, statement i of this proposition implies point ii of Proposition 1, which together with [5] implies that Inline graphic represents Inline graphic for all nonnull Inline graphic. However, Inline graphic and hence

graphic file with name pnas.1207805110uneq17.jpg

so that ii holds.

The rest follows immediately from Proposition 3. ▪

Proof of Proposition 4:

The proof of statement i implies ii of Proposition 3 has to be modified because consistency yields only [16]. Then ref. 21, Theorem 20, yields only that P belongs to the vector subspace generated by M. In any case, there exists a collection Inline graphic of scalars such that Inline graphic for all Inline graphic. From Inline graphic for all Inline graphic, it follows that Inline graphic. Moreover, by orthogonality, there exists a partition Inline graphic of S in Σ such that Inline graphic and Inline graphic for all distinct Inline graphic (see the beginning of the next proof). Hence, for each m it holds that Inline graphic, and so Inline graphic. We conclude that Inline graphic again. The rest of the proof is very similar to that of Proposition 3. ▪

Proof of Proposition 5:

We consider orthogonality and leave absolute continuity to the reader. Suppose Inline graphic, i.e., there is Inline graphic such that Inline graphic. Next we show that there exists a partition Inline graphic of S in Σ such that Inline graphic and Inline graphic for all distinct Inline graphic. [Note that Inline graphic for all Inline graphic such that Inline graphic actually follows from the fact that Inline graphic is a partition and Inline graphic for all Inline graphic.] Let Inline graphic. For Inline graphic, the result is true by definition of orthogonality. Assume Inline graphic and the result holds for Inline graphic. Then there exists a partition Inline graphic of S in Σ such that Inline graphic for all Inline graphic. However, Inline graphic for each Inline graphic, and hence there is Inline graphic such that Inline graphic. By setting Inline graphic and Inline graphic we then have Inline graphic and Inline graphic for each Inline graphic. The desired partition is obtained by setting Inline graphic.

Set Inline graphic. Clearly, Inline graphic. Moreover, Inline graphic for all Inline graphic and Inline graphic for all Inline graphic. Then,

graphic file with name pnas.1207805110uneq18.jpg

and

graphic file with name pnas.1207805110eq18.jpg

which implies Inline graphic. As to the converse, suppose Inline graphic. There is Inline graphic such that Inline graphic. Set Inline graphic. We have Inline graphic because A is finite. It holds that

graphic file with name pnas.1207805110uneq19.jpg

and so Inline graphic. Moreover,

graphic file with name pnas.1207805110eq19.jpg

whence Inline graphic for all Inline graphic because Inline graphic. We conclude that Inline graphic and Inline graphic. ▪

Acknowledgments

The financial support of ERC (advanced Grant BRSCDP-TEA) is gratefully acknowledged.

Footnotes

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

References

  • 1.Savage LJ. The Foundations of Statistics. New York: Wiley; 1954. [Google Scholar]
  • 2.Wald A. Statistical Decision Functions. New York: Wiley; 1950. [Google Scholar]
  • 3.Sargent TJ. Evolution and intelligent design. Am Econ Rev. 2008;98:3–37. [Google Scholar]
  • 4.Lucas RE., Jr . Understanding business cycles. In: Brunner K, Meltzer AH, editors. Stabilization of the Domestic and International Economy. Amsterdam: North-Holland; 1977. pp. 7–29. [Google Scholar]
  • 5.Muth JF. Rational expectations and the theory of price movements. Econometrica. 1961;29:315–335. [Google Scholar]
  • 6.Fishburn PC. Utility Theory for Decision Making. New York: Wiley; 1970. [Google Scholar]
  • 7.Kreps DM. Notes on the Theory of Choice. Boulder, CO: Westview Press; 1988. [Google Scholar]
  • 8.Gilboa I. Theory of Decision under Uncertainty. Cambridge, UK: Cambridge Univ Press; 2009. [Google Scholar]
  • 9.Jaffray J-Y. Linear utility theory for belief functions. Oper Res Lett. 1989;8:107–112. [Google Scholar]
  • 10. Battigalli P, Cerreia-Vioglio S, Maccheroni F, Marinacci M (2011) Selfconfirming equilibrium and model uncertainty. Working paper 428 (Innocenzo Gasparini Institute for Economic Research, Milan)
  • 11.Hansen LP, Sargent TJ. Robustness. Princeton: Princeton Univ Press; 2007. [Google Scholar]
  • 12.Bhaskara Rao KPS, Bhaskara Rao M. Theory of Charges. New York: Academic; 1983. [Google Scholar]
  • 13.Teicher H. Identifiability of finite mixtures. Ann Math Stat. 1963;34:244–248. [Google Scholar]
  • 14.Diaconis P, Freedman D. On the consistency of Bayes estimates. Ann Stat. 1986;14:1–26. [Google Scholar]
  • 15.Marschak J. Neumann’s and Morgenstern’s new approach to static economics. J Polit Econ. 1946;54:97–115. [Google Scholar]
  • 16.Marschak J. Personal probabilities of probabilities. Theory Decis. 1975;6:121–153. [Google Scholar]
  • 17.Marschak J, Radner R. Economic Theory of Teams. New Haven: Yale Univ Press; 1972. [Google Scholar]
  • 18. Cerreia-Vioglio S, Maccheroni F, Marinacci M, Montrucchio L (2011) Classical subjective expected utility. Working paper 400 (Innocenzo Gasparini Institute for Economic Research, Milan)
  • 19.Billingsley P. Ergodic Theory and Information. New York: Wiley; 1965. [Google Scholar]
  • 20.Sargent TJ, Williams N. Impacts of priors on convergence and escapes from Nash inflation. Rev Econ Dyn. 2005;8:360–391. [Google Scholar]
  • 21.Marinacci M, Montrucchio L. Subcalculus for set functions and cores of TU games. J Math Econ. 2003;39:1–25. [Google Scholar]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences

RESOURCES