Asymmetric gain–loss reference dependence and attitudes toward uncertainty
 754 Downloads
Abstract
This paper characterizes a model of reference dependence, where a statecontingent contract (act) is evaluated by its expected value and its expected gain–loss utility. The expected utility of an act serves as the reference point; hence, gains (resp., losses) occur when the act provides an outcome that is better (worse) than expected. The utility representation is characterized by a belief regarding the state space and a degree of reference dependence; both are uniquely identified from behavior. We establish a link between this type of reference dependence and attitudes toward uncertainty. We show that loss aversion and reference dependence are equivalent to max–min and concave expected utility.
Keywords
Referencedependent preferences Endogenous reference points Gain–loss attitudes Subjective expected utility Belief distortionJEL Classification
D800 D8101 Introduction
In many circumstances, a decision maker (DM) may evaluate an uncertain prospect not only in absolute terms but also in relative relation to some reference point. Kahneman and Tversky (1979) first introduced the notion of reference dependence, in the seminal Prospect theory, to explain experimental violations of expected utility. Within Prospect theory, deviations from the reference point are weighted by a gain–loss value function, which has the feature, referred to as loss aversion, that losses have more negative value than equal sized gains have positive value.
A different resolution for empirical deviations from expected utility proposes models of multiple priors, in particular MaxMin Expected Utility (MMEU). MMEU, axiomatized by Gilboa and Schmeidler (1989) as an explanation of the Ellsberg Paradox,^{1} considers a DM who holds a family of beliefs regarding the likelihood of events. She evaluates uncertain prospects by the minimum expected utility consistent with any of her beliefs. As such, a MMEU DM displays uncertainty aversion (or, ambiguity aversion), the feature that she prefers to minimize her exposure to uncertainty.
At a purely intuitive level, there seems to be a connection between loss aversion and uncertainty aversion; both behaviors characterize some form of pessimism in comparison with a subjective expected utility (SEU) maximizer. A loss averse DM places more weight on the utility of “bad” events but leaves the probabilities undistorted, whereas an uncertainty averse DM places more weight on the probability of “bad” events but leaves the utilities undistorted. We show in this paper that this connection is more than superficial; there exists a formal connection between reference dependence and attitude toward uncertainty. In particular, we axiomatize a simple class of referencedependent preferences, called asymmetric gain–loss (AGL) preferences, which can be equivalently represented by a MMEU functional. Within our framework, loss aversion and uncertainty aversion produce identical choice data.
1.1 AGL preferences
1.2 Reference point formation
There are two alternative views on the formation of endogenous reference points. In the first, the DM forms a reference point before making a choice, based on the set of options that she faces.^{3} As such, her beliefs about her own actions will affect the reference point, leading these papers to generally require an equilibrium condition to account for the mutual relationship between reference points and choices.
In the second, the DM’s chosen action completely determines the reference point. Thus, each element of a choice set is associated with its own reference point.^{4} AGL preferences fall into this second category. In particular, the AGL representation is closely related to the notion of choice acclimating personal equilibrium defined by Kőszegi and Rabin (2007), excepting that there the domain of uncertainty is objective risk. A main contribution of our paper is that, by considering the case where probabilities are subjective, we show how it is possible to simultaneously identify both the reference attitude and the beliefs of the DM.
It is worth noting that when the reference point is defined using equilibrium conditions, as in Kőzsegi and Rabin (2006), the joint identification of beliefs and reference effects is generally not possible. The feedback loop between choices and the reference point can lead to intransitivity of the revealed preference, as shown in Gul and Pesendorfer (2006).
Because of this identification problem, and the intrinsic complexity surrounding equilibrium conditions, the latter notion or reference point determination has proved more suitable for applications. Indeed, many applications use AGL preferences: Lange and Ratan (2010) explore how reference dependence can increase the optimal bid in sealed bid auctions (to be more in line with empirical evidence); Herweg et al. (2010) show that loss aversion can explain prevalence of binary incentive schemes (i.e., bonuses) in moral hazard environments; Abeler et al. (2011) show that, in an effort provision experiment, expectationsbased reference dependence best explains their data; Karle and Peitz (2014) consider the competition of differentiated firms when buyers exhibit loss aversion. Each of the abovementioned papers assumed a kinked, piecewise linear gain/loss function and assumed the reference point was the expected consumption utility—exactly the characterization given here. Our work provides the foundational restrictions for such consumer behavior, and shows that beliefs regarding uncertainty and reference effects can be jointly identified.
1.3 A simple example of AGL preferences
We employ the following numerical example to explain the intuition behind the representation, and show how asymmetric gain–loss preferences can explain different types of behavior regarding uncertainty.
Now suppose that the buyer has gain–loss preferences: in addition to the expected value she wants to avoid losses, so she subtracts any expected losses from the expected consumption utility to determine the valuation (this corresponds to the parametrization \(\lambda = 1\)). A loss for her is any outcome where her expost utility is worse than the expected value; therefore, outcomes in which she does not obtain the item are considered losses.
When the buyer in this simple takeitorleaveit example takes into consideration expected gains and losses in addition to the standard expected utility, she is better off increasing her bid. Intuitively, she sacrifices her payoff in good outcomes (where she obtains the item) in order to decrease the chance of bad outcomes (not obtaining the item). While her payoff is smaller contingent on obtaining the good, the outcome is favorable more often and she increases her exante utility.
1.4 Structure of the paper
The rest of the paper is structured as follows. Section 2 provides an axiomatic characterization of the preferences, discusses the concept of the alignment of acts, which is instrumental for the endogenous determination of a reference point, and formally defines the utility representation. Section 3 explores the link between gain–loss and ambiguity attitudes. Section 4 puts forth comparative statics results. Section 5 contains a literature review. All proofs are contained in the “Appendices.”
2 Axiomatization
In this section, we formally present the choice environment and a set of axioms which prove to be necessary and sufficient for the representation presented later in this section.
Let \(S = \{s_1,s_2,\ldots ,s_n\}\) be a finite set of states of the world that represent all possible payoffrelevant contingencies for the DM; any \(E\subseteq S\) is called an event. Define \(\mathcal {E} = \mathcal {P}(S){\setminus }\{\emptyset , S\}\) as the set of all nontrivial events. Denote by \(\mathcal {F} = {{\mathbb {R}}}_{+}^S\) the set of all acts, that is, functions \(f:S\rightarrow {{\mathbb {R}}}_{+}\) (endowed with the standard Euclidean topology). We interpret the act f as providing the payoff f(s) in state \(s \in S\) and assume it is the utility received by the DM when f is chosen and s is realized.^{6}
Take the mixture operation on \(\mathcal {F}\) as the standard pointwise mixture, where for any \({\alpha }\in [0,1]\), \({\alpha }f+(1{\alpha })g\in \mathcal {F}\) gives \({\alpha }f(s)+(1{\alpha })g(s)\in {{\mathbb {R}}}_{+}\) for any \(s\in S\). Abusing notation, any \(c \in {{\mathbb {R}}}_{+}\) can be identified with the constant act \(c(s) = c\) for all \(s\in S\). Let \(\mathcal {F}_c \cong {{\mathbb {R}}}_{+}\) be the set of constant acts. Preferences on \(\mathcal {F}\) are denoted by the binary relation \(\succsim \); \(\succ \) and \(\sim \) represent, respectively, the asymmetric and symmetric components of \(\succsim \). For each \(f\in \mathcal {F}\), if there is some \(c_f \in \mathcal {F}_c\), such that \(f\sim c_f\), then call \(c_f\) the certainty equivalent of f. Before we can specify the behavioral restrictions on preference that correspond to the AGL utility representation, we need to consider some particular structures in the choice domain.
2.1 Balanced pairs of acts
A particularly important type of act to study AGL preference is given by those that provide perfect hedges against uncertainty. Hedging gets rid of uncertainty, and therefore, it also removes all possible gain–loss considerations from the act. Call a pair of acts \((f,\bar{f})\)balanced if they provide a perfect hedge and are indifferent to each other.^{7} The importance of balanced acts is that eliminating subjective gain–loss considerations allows an analyst to identify beliefs from preferences.
Definition 1
When the notation \(\bar{f}\) is used, it is always in reference to the balancing act of \(f\in \mathcal {F}\). The conditions imposed on preferences below guarantee that \(c_f\) and \(e_f\) are unique and well defined for each f.
2.2 Act alignment: separating positive and negative states
Balanced acts will provide a behavioral way of separating gains and losses. We require that when the outcome in state s is considered a gain for f, the outcome on state s is considered a loss for \(\bar{f}\). This is a natural requirement given that f and \(\bar{f}\) provide a perfect hedge to the DM. Hence, \(\bar{f}\) has the exact opposite gain–loss composition of f. For an act, define positive states as those states that deliver gains, and negative states as those states that deliver losses.
Definition 2
Let \((f,\bar{f})\) be a balanced pair. Say \(s\in S\) is a positive state for f if \(f(s) \ge \bar{f}(s)\), and a negative state for f if \(\bar{f}(s) \ge f(s)\). If a state is both positive and negative (i.e., \(f(s) = \bar{f}(s)\)) say s is a neutral state for f.
Any balanced pair of acts induces a set of partitions, each of which splits the state space into two events: one event that contains only positive states for f (\(\{s\in Sf(s) \ge \bar{f}(s)\}\)) and one event that contains only negative states for f (\(\{s\in S\bar{f}(s)\ge f(s)\}\)). We use the convention that neutral states can be labeled as either positive or negative (but not both).^{8} When there are no neutral states, each act has a unique way of partitioning the states into positive and negative. These partitions associated with each act are called the alignment of the act. We use the convention that the alignment of the act is represented by the event that includes the positive states E (the complement is the negative states) rather than saying that the alignment is represented by the partition \(\{E,E^c\}\).
Definition 3
Let \((f,\bar{f})\) be a balanced pair. For any \(f\in \mathcal {F}\), say f is aligned with the event \(E\in \mathcal {E}\) if for all \(s\in E\), \(f(s)\ge \bar{f}(s)\) and for all \(s\in E^c\), \(\bar{f}(s) \ge f(s)\).
For every \(E\in \mathcal {E}\), there is a set of acts that is aligned with E.
Definition 4
Note that any constant act is its own balancing act, and therefore, constant acts are aligned with all partitions of the state space. It is useful to consider acts that have only one alignment, which are called singlealignment acts. These acts are important because they are acts where small perturbations on outcomes do not change the alignment.
Definition 5
Let \((f,\bar{f})\) be a balanced pair. Then, f is singlealignment act if for no \(s\in S\), \(f(s) = \bar{f}(s)\).
If the event E represents an alignment of f, every subset of E or \(E^c\) is called a nonoverlapping event. These are the events where all the states are either all positive or all negative for f, so there is no overlap between positive and negative states for f. Nonoverlapping events provide a way of specifying situations where there is no tradeoff between positive and negative states, only across one type of state.
Definition 6
Given \(f\in {\mathcal {F}}\), event \(F\subset S\) is a nonoverlapping event for f if every state in F is aligned in the same way.
With these definitions in mind, we can now specify the behavioral restrictions that are necessary and sufficient to be represented by the AGL functional, as given by Eq. (1.2).
2.3 Standard axioms

A1. (Weak Order). \(\succsim \) is complete and transitive.

A2. (Continuity). For all \(f\in \mathcal {F}\), the sets \(\{g\in \mathcal {F}: g\succsim f\}\) and \(\{g\in \mathcal {F}: f\succsim g\}\) are closed.

A3. (Strict Monotonicity). If for all s, \(f(s) \ge g(s)\) then \(f \succsim g\). If \(f(s) > g(s)\) for some s and \(f(s) \ge g(s)\) for all s, then \(f\succ g\).
2.4 New axioms: mixture conditions
The standard subjective expected utility model from Anscombe and Aumann (1963) is characterized by some version of A1–A3, plus the independence axiom. Independence requires that \(f\succsim g\) if and only if \({\alpha }f + (1{\alpha }) h \succsim {\alpha }g + (1{\alpha }) h\) for any \(h\in \mathcal {F}\), and any \({\alpha }\in (0,1)\).
The independence axiom does not hold for AGL preferences because it does not allow for gains and losses to be evaluated differently; convex combinations of acts can change the gain–loss composition of acts, therefore changing the assessments as well. AGL preferences relax independence, but impose three consistency requirements for mixtures of acts.

A4. (Alignment Independence). If \(f,h\in {\mathcal {F}}^E\) and \(g,h \in {\mathcal {F}}^F\) for some \(E,F \in \mathcal {E}\), then \(f\succsim g\) if and only if \({\alpha }f+ (1{\alpha })h \succsim {\alpha }g + (1{\alpha })h\) for all \({\alpha }\in [0,1]\).
Under the full independence axiom, the preference between an \(\alpha \) mixture of f and h or f and \(h'\) would depend only on the preference between h and \(h'\). Here, on the other hand, mixing acts may change valuations in nonlinear ways: This could happen if either (i) the mixture changes the alignment of states, or (ii) the mixture provides an opportunity to hedge by improving loss states and worsening gain states (or the opposite, if the DM is gain seeking).
 A5. (Local Mixture Consistency). For any singlealignment acts \(f \in {\mathcal {F}}^E\) and \(g\in {\mathcal {F}}^{E'}\), any event F, which is nonoverlapping for bothE and \(E'\), and any \(h,h'\in \mathcal {F}\) such that \(h(s) = \bar{h}(s) = h'(s) = \bar{h}'(s)\) for all \(s\not \in F\), there exists \(\alpha ^* < 1\) such thatfor all \({\alpha }\in ({\alpha }^*,1)\).$$\begin{aligned} {\alpha }f + (1{\alpha })h\succsim {\alpha }f + (1{\alpha })h' \ \Longleftrightarrow {\alpha }g + (1{\alpha }) h \succsim {\alpha }g + (1{\alpha }) h', \end{aligned}$$
 A6. (Antisymmetry). For any acts f and g such that \(f\sim g\), for every \(h\in \mathcal {F}\), and for any \(\alpha \in (0,1)\), \( \alpha h + (1\alpha ) f \succsim \alpha h + (1\alpha )g\) implieswhere \(\bar{f}\), \(\bar{g}\), and \(\bar{h}\) are balancing acts of f, g, and h, respectively.
 (i)
\( \alpha h + (1\alpha ) \bar{f} \precsim \alpha h + (1{\alpha }) \bar{g}\), and,
 (ii)
\({\alpha }\bar{h} + (1{\alpha }) f \precsim {\alpha }\bar{h} + (1{\alpha }) g\),
 (i)
2.5 Representation results
This section provides the main representation results of the paper. Theorem 2.1 introduces the AGL representation as characterized by the above axioms. This section also outlines important preliminary results that highlight the role of particular axioms and elucidate the relation between the AGL representation and other decision theoretic models.
Theorem 2.1
The bound on \(\lambda \) is a consequence of strict monotonicity: Increasing the payoff in any state must increase the valuation of the act. The bound on \(\lambda \) ensure that the marginal increase in consumption utility outweighs any negative marginal decrease in gain/loss utility. The parameter \(\lambda \) captures the difference between the weight placed on gains and the weight placed on losses, which for the representation is unique. An important application of the representation result from Theorem 2.1 is that it provides an index for reference dependence (\(\lambda \)) that is decoupled from risk attitudes, and can be easily estimated. The fact that \(\mu \) lies in the interior \(\Delta (S)\) is also a consequence of strict monotonicity.
2.5.1 Sketch of the proof and preliminary results
The result is proven in two steps. First, Lemma 2.2 provides a SEU representation on \(\mathcal {F}^E\) established by Axioms A1–A4 (that is, excluding Local Mixture Consistency and Antisymmetry), which can be extended to aggregate preferences across families of mutually aligned acts. Then, we utilize the properties of Axioms A5 and A6 to generate the final result.
Lemma 2.2
Every prior in \(\{\mu _E\}_{E\in \mathcal {E}}\) is different, and Alignment Independence does not imply any structure on the priors. To derive the main result from the representation of Lemma 2.2, Local Mixture Consistency and Antisymmetry are used to guarantee that every prior in the set \(\{\mu _E\}_{E\in \mathcal {E}}\) can be written as functions of one unique prior \(\mu \).
If we mix f with a small amount of noise (either the act h or \(h'\)), where the noisy acts exhibit variation only on F nonoverlapping with E, then Local Mixture Consistency guarantees that the preference over the two mixtures depends only on the expected consumption utility of the noise. Hence, for all \(E\in \mathcal {E}\), the conditional distributions—conditioned on F nonoverlapping with E—are the same for all \(\mu _E\). That is, for E and \(E'\in \mathcal {E}\), \(\mu _E(\cdot F) = \mu _{E'}(\cdot F)\) whenever F is nonoverlapping for E and \(E'\). In fact, there is a single \(\mu \) such that \(\mu (s  F) = \mu _{E}(s  F)\), keeping the condition of F.
Since E is nonoverlapping with itself, this last point implies that for any \(s,s'\in E\), \(\frac{\mu (s)}{\mu (s')} = \frac{\mu _E(s)}{\mu _E(s')}\), which holds only if for all \(s\in E\), \(\mu (s)= \gamma \mu _E(s)\), \(\gamma \in {{\mathbb {R}}}_{++}\). So on E, \(\mu _E\) is just a constant perturbation of the original prior \(\mu \). Since \(E^{c}\) is nonoverlapping with E the same holds for \(E^{c}\).
3 Relation to ambiguity attitude
3.1 Maxmin expected utility
According to Theorem 2.1, the DM who abides by the AGL axioms is probabilistically sophisticated but displays some reference effect. That is, she holds some unique belief, \(\mu \), regarding the state space, and evaluates each act according to this belief and her preferences for outcomes. Nonetheless, Lemma 2.2 states that the same preferences can be represented by a family of distributions, \(\{\mu _E\}_{E\in \mathcal {E}}\), each of which is a distortion of the original belief, \(\mu \). This alludes to a possible relationship between reference effects and attitudes toward uncertainty, which has classically been modeled by a DM who considers a (nonsingleton) set of priors.
Definition 7
MMEU, axiomatized by Gilboa and Schmeidler (1989), is characterized by two key conditions: certainty independence and uncertainty aversion. Certainty independence requires that \(f\succsim g\) if and only if for all \({\alpha }\in [0,1]\), \({\alpha }f + (1{\alpha }) c \succsim {\alpha }g + (1{\alpha }) c\) where \(c\in \mathcal {F}_c\). Mixing two acts with a common constant act does not reverse the preference between them. Since constant acts are aligned with every \(E\in \mathcal {E}\), Alignment Independence implies certainty independence.
Uncertainty aversion requires that for all f, g such that \(f\sim g\) for any \({\alpha }\in (0,1)\), \({\alpha }f + (1{\alpha }) g \succsim f\). If uncertainty aversion is changed for uncertainty seeking preferences,^{10} then the representation is a Maxmax representation, where the DM evaluates an act according to the prior that maximizes her expectations. It is clear from the representation that if \(\lambda \le 0\) then the DM is uncertainty averse, and if \(\lambda \ge 0\) then she is uncertainty seeking.
Uncertainty aversion can be characterized as a preference for hedging, as hedging reduces the exposure to uncertainty. Moreover, hedging reduces the exposure to negative states. Pushing the utility value in each state closer to the average has more effect on the negative states (because of the loss bias) and hence weakly improves the act.
The formal connection is captured by the following result, which states that asymmetric gain–loss preferences always admit a Maxmin or Maxmax representation and that the set of priors C has a specific structure that is related to the distortion of the (unique) beliefs of the DM.
Theorem 3.1
Suppose \(\succsim \) admits an AGL representation \((\mu , \lambda )\), with \(\lambda < 0\), then \(\succsim \) admits a MMEU representation. Moreover, \(C ={\text {conv}}(\{\mu _E\}_{E\in \mathcal {E}})\), as defined in Lemma 2.2.
Theorem 3.1 has several implications. First, it shows that this form of reference dependence is always tied to a particular attitude toward uncertainty.^{11} So, preferences studied in this paper will always be either uncertainty averse or uncertainty seeking. Second, it gives a precise form to the belief distortion that takes place when gain–loss consideration affect a probabilistically sophisticated DM.
While every AGL representation can be faithfully captured within the MMEU framework, the converse is not true. In the AGL framework, the distorted beliefs keep the relative likelihood of states among gains and among losses unchanged, but, depending on the sign of \(\lambda \), increase or decrease the total weight given to gains (and losses) proportional to the baseline belief. This distortion is a function only of the degree of reference dependence, \(\lambda \), and the baseline prior \(\mu \). In addition, Antisymmetry implies the set of priors is symmetric with respect to all hyperplanes (in the \(S  1\) dimensional simplex) which divide the state space into positive and negative states and which pass through the baseline prior. See Fig. 1; the dashed lines show such symmetries.
3.2 Concave expected utility
Alignment Independence imposes more structure than certainty independence, and therefore, AGL also shares a connection to a class of ambiguity models outside of MMEU. In particular, any loss adverse AGL preference is also a concave expected utility (cavEU) preference. CavEU is a capacitybased model, which considers all possible decompositions of an act into bets over events (where a bet of magnitude \(a_{E} \in {{\mathbb {R}}}_{++}\) on E, is an act that is constant on E and 0 off E, i.e., \(a_{E}\mathbb {1}_{E}\) where \(\mathbb {1}_{E}\) is the characteristic function on E). The preference \(\succsim \) is cavEU if it can be represented by a concave integral introduced by Lehrer (2009).
Definition 8
The concave integral returns the maximum value of all possible decompositions, when aggregated according to the capacity v. Lehrer and Teper (2015) show that \(\succsim \) is cavEU if and only if it satisfies A1–A3 plus uncertainty aversion, independence with respect to the constant act 0, and codecomposable independence. This last requirement states, for every nonbet act f, there exist a bet \(a_{E}\) and an act \(f'\) such that (i) \(f = \alpha a_{E} + (1\alpha )f'\) for \(\alpha \in (0,1)\), and (ii) \(\succsim \) satisfies independence over \(\{\alpha a_{E} + \beta f'  \alpha ,\beta \in {{\mathbb {R}}}_{+}\}\).
Theorem 3.2
Suppose \(\succsim \) admits an AGL representation \((\mu , \lambda )\), with \(\lambda < 0\), then \(\succsim \) admits a cavEU representation with \(v: \mathcal {P}(S) \rightarrow [0,1]\) defined by \(v: E \mapsto \min _{F \in \mathcal {E}}\mu _F(E)\).
The property that AGL preference admits cavEU representations stems from the fact that each act, \(f \in \mathcal {F}^{E}\) can always be decomposed into a bet on E and another act in \(\mathcal {F}^{E}\). Since all these acts share the same alignment, independence holds within the convexcone generated thereby. As with the set of priors in the MMEU representation, the capacity v is characterized by the lower envelope of the distorted beliefs arising themselves from Lemma 2.2. Of course, this must be, since these functionals represent the same preferences! Note, cavEU and MMEU are not nested models; AGL preferences reside in the nontrivial intersection.^{12}
4 Comparative gain/loss attitudes
This section advances comparative statics results relating behavior to elements of the AGL representation. For an act f, recall the hedge, \(e_{f}\), is the constant act which provides the expected consumption utility in every state; the constant equivalent, \(c_{f}\), is the constant act which provides the expected total utility in every state, in other words taking into account gain/loss considerations.
A natural measure for the degree and direction of reference effects is the gap between \(e_f\) and \(c_f\), the hedge and the certainty equivalent. For a loss averse DM, the difference between the hedge and the constant equivalent is how much, in utility terms, she is will sacrifice to avoid having to feel a loss. In the standard SEU model, \(e_f = c_f\), so the SEU model is the baseline case for reference effects.
Definition 9
Let \(\succsim \) be a preference over \(\mathcal {F}\). Say \(\succsim \) is gainbiased if for all \(f\in \mathcal {F}\), \(c_f \succsim e_f\). Say \(\succsim \) is lossbiased if for all \(f\in \mathcal {F}\), \(e_f \succsim c_f\).
Remark 1
An AGL decision maker is gainbiased (respectively, lossbiased) if and only if she is uncertainty seeking (resp., uncertainty averse) if and only if \(\lambda \ge 0\). (resp., \(\lambda \le 0\)).
Remark 1 follows immediately from the observation that \({{\mathbb {E}}}_{\mu }[f] = e_{f}\) and examination of the representing functionals.
Problematically, however, the hedge and constant equivalent of an act depend on the DM’s beliefs, so if we want to be able to compare two DM’s degree of reference dependence we want to disentangle reference dependence from beliefs. To do this, we define \(f\vee \bar{f}\), the join of a balanced pair \((f,\bar{f})\), as the act that gives the DM the best outcome between f and \(\bar{f}\) for each \(s\in S\).
Definition 10
From the AGL representation, gain–loss utility depends on how much the act deviates state by state from \(e_f\). \(f\vee \bar{f}\) provides the absolute value of the state by state deviations of f from \(e_f\). Thus, the hedge of the join, \(e_{f\vee \bar{f}}\), is the average deviation of f from \(e_{f}\). Then, to capture reference dependence behaviorally across DMs, we focus on acts that have the same hedge: If acts have different hedges, the reference effects can be confounded by the beliefs.
The intuition behind our comparative notion of “more loss biased” is that, holding the hedge constant, the DM prefers an act f with smaller expected losses. Conversely, a DM with gain–bias prefers acts with larger gains. Since we want to consider acts that have the same hedge, the comparative notions of “more gainbiased” and “more lossbiased” depend on a possibly different act for each DM: f for DM 1 and g for DM 2. We use the notation where \(e_f^i\) denotes the hedge of f for DM i. If, in addition, \(e^{1}_{f\vee \bar{f}} = e^{2}_{g\vee \bar{g}}\), then g (evaluated according to \(\mu _{1}\)) has the same variance as g (according to \(\mu _{2}\)). So we say DM 1 is more lossbiased than DM 2, if the exposure to the same variance, keeping the expected consumption utility the same, produces a harsher utility penalty.
Definition 11
Given two preference orders \(\succsim _1\) and \(\succsim _2\), say that \(\succsim _1\) is more lossbiased than \(\succsim _2\) (and \(\succsim _2\) is more gainbiased than \(\succsim _1\)) if for any f, g with \(e_f^1 = e_g^2\) and \(e^1_{f\vee \bar{f}} = e^2_{g\vee \bar{g}}\), then for any \(c\in \mathcal {F}_c\), \(f \succsim _1 c\) implies \(g \succsim _2 c\) and \(f \succ _1 c\) implies \(g \succ _2 c\).
Theorem 4.1
Let \(\succsim _i\) admits an AGL representation given by \((\mu _i,\lambda _i)\) for \(i=1,2\). Then, \(\succsim _1\) is more lossbiased than \(\succsim _2\) if and only if \(\lambda _1 \le \lambda _2\).
When \(\mu _1 = \mu _2\), we can say more than Theorem 4.1.
Remark 2
Let \(\succsim _i\) admits an AGL representation given by \((\mu ,\lambda _i)\) for \(i=1,2\). Further assume \(\lambda _{1} \le \lambda _{2} \le 0\). Then, (i) \(f \succsim _1 c\) implies \(f\succsim _2 c\), for all acts f and constant acts c, and (ii) \(C_2 \subseteq C_1\) for the equivalent Maxmin/Maxmax representation from Theorem 3.1.
These addition equivalences stem from the fact that when DMs have the same belief, then for any \(f\in \mathcal {F}\), \(e_f^1 = e_f^2\). In such circumstances, the degree of loss bias is equivalent to the comparative notion of ambiguity aversion from Ghirardato and Marinacci (2002). Remark 2 furthers this link: Whenever \(\succsim _i\) is gainbiased or lossbiased for both DMs, the notion of loss bias is consistent with the representation of comparative ambiguity aversion derived from Gilboa and Schmeidler (1989) (that the more ambiguity averse DM should have a larger set of priors). This observation establishes a clear connection between the idea of “loss aversion” that has been prevalent since Prospect theory, and uncertainty aversion.
These comparative statics results establish an unexplored link between the absolute and comparative notions of gain or loss bias, and existing notions of uncertainty aversion which is worth further exploring. The initial motivation for studying uncertainty was due to the Ellsberg (1961) idea that DMs are not able to formulate unique probabilities over uncertain events. Many models with multiple priors have been developed to capture what is considered “Ellsbergian behavior.” Nonetheless, even if the DM is able to form a unique prior, having gain–loss considerations can appear to contaminate her prior in a way that gives rise to behavior embodied by some multiple priors model. Hence, for AGL preference a probabilistically sophisticated DM can appear to have multiple priors due to gain–loss asymmetry.
5 Related literature
This paper links reference dependence and attitudes toward ambiguity. We show the notion of choice acclimating personal equilibrium (CPE) from Kőszegi and Rabin (2007)—where the reference point is the expectation of consumption utility—provides a clean way to link these two concepts in the domain of choice under uncertainty. Bell (1985), Loomes and Sugden (1986) and Kőzsegi and Rabin (2006) also provide various models where the DM is loss averse with respect to a reference point given by her expected consumption utility of an uncertain prospect.
In many decision theory models, the status quo has been interpreted as a reference point. Giraud (2004a), Masatlioglu and Ok (2005), Sugden (2003), Sagi (2006), Rubinstein and Salant (2007), Apesteguia and Ballester (2009), Ortoleva (2010), Riella and Teper (2014) and Masatlioglu and Ok (2013) provide models of reference dependence, where the reference point is exogenously given. Along with Kőzsegi and Rabin (2006) and Kőszegi and Rabin (2007), other papers that tackle the problem of endogenous reference point determination are Giraud (2004b), Sarver (2011), Ok et al. (2014), and Werner and Zank (2017). The approach in Ok et al. (2014) investigates reference point determination problem under a very general framework, where they do not need an equilibrium condition to characterize reference dependence. Nonetheless in their framework it is impossible to identify reference points and reference effects uniquely.
In Gul (1991), outcomes of a (n objective) lottery are considered either a disappointment or an elation depending on whether they are less than or more than the certainty equivalent. The DM suffers a utility penalty when an outcome is considered disappointing. In contrast, we assume an outcome is disappointing if it is dispreferred to the hedge, rather than the certainty equivalent. Blavatskyy (2010) extends this to a domain where certainty equivalents need not exist.
Dillenberger (2010) shows that Gul’s disappointment averse preferences satisfy negative certainty independence and so admit a cautious expected utility representation a la CerreiaVioglio et al. (2015). The later paper also shows that cautious expected utility is, in the objective risk domain, the analogy to MMEU in the subjective uncertainty domain. The connection between AGL and MMEU is therefore the subjective counterpart to the connection between disappointment aversion and cautious EU. AGL preferences (with \(\lambda < 0\)) satisfy negative certainty independence.^{13}
In a similar spirit to our paper, Masatlioglu and Raymond (2016) provide an complete characterization of CPE within the domain of objective risk. They show that CPE is exactly the intersection of quadratic preferences and rankdependent expected utility preferences.
For AGL preferences, the evaluation of acts depends on the state by state variation of the act. Although some papers have studied attitudes toward variation in the context of risk and uncertainty, none relates such attitudes to reference dependence. In the risk domain, Quiggin and Chambers (1998, 2004) measure attitudes toward risk, which depend on the expectation of the lottery and a risk index of the lottery that depends on the variation of the distribution.
From the vantage of attitudes toward ambiguity, AGL preferences are a clear special case of meandispersion preferences: Grant and Polak (2013) axiomatize a very general model of meandispersion preferences, where an act is evaluated by the representation \(V(f) = \mu \rho (d)\), where \(\mu \) is the expected consumption utility of f with respect to a given probability, d is the vector of statebystate utility deviations from the mean, and \(\rho (\cdot )\) is a measure of (aversion to) dispersion.
Many wellknown families of preferences such as Choquet EU (Schmeidler 1989), Maxmin EU (Gilboa and Schmeidler 1989), invariant biseparable preferences (Ghirardato et al. 2004), variational preferences (Maccheroni et al. 2006), and Vector EU (Siniscalchi 2009) belong to this family of preferences. Our paper (under loss aversion) corresponds to the specification where \(\rho = \lambda {{\mathbb {E}}}(\min \{d(s),0\})\). The interest in studying this special case is twofold. First, meandispersion preferences are so general that it is predominantly not possible to identify the DM’s baseline prior. (Although some authors do provide various additional restrictions that facilitate identification.) The additional structure imposed in this paper precipitates not only the identification of beliefs, but also the comparative statics results presented in Sect. 4. By taking a stand on way dispersion affects utility (i.e., via linear loss aversion), we can more thoroughly relate the parameters of the representation to behavioral patterns. The second motivation is the ubiquity of AGL (or very similar) preferences in applications. As outlined in Sect. 1.2, linear loss aversion with respect to expected consumption utility has proven to be a popular way of representing reference dependence in applied work. This paper, therefore, precisely outlines the tacit assumptions made in such applications.
Footnotes
 1.
Because it is well established in the literature, we refer the reader to Gilboa and Schmeidler (1989) for a formal discussion of the Ellsberg Paradox and its resolution by MMEU.
 2.When \(\lambda >0\), the DM is gain seeking and receives a utility bonus. At first glance, it may seem counterintuitive that a utility bonus occurs in states where the outcome is below the reference utility. However, because gains and losses are relative to the expectation, expected gains must equal expected losses. Exploiting this symmetry of gains and losses, the above expression is equivalent to having the utility bonus occur in states that surpass the expectations. The equivalent expression given by Eq. (1.3) more clearly illustrates the role of the gain–loss term when \(\lambda \) is positive.
 3.
 4.
 5.
 6.
Note, we are tacitly assuming the decision makers cardinal utility has already been identified via standard means, i.e., the examination of preferences over objective lotteries. We could just as easily add a second stage of objective randomization into acts, à la Anscombe and Aumann (1963), but this would require additional notation, and the elicitation of utility values is not central to our model.
 7.
Siniscalchi (2009) calls a pair of acts that provide perfect hedging as complementary acts. We strengthen the definition of complementary acts to further require the acts to be indifferent.
 8.
We do not allow the neutral states to be labeled both positive and negative, instead when there are neutral states a balanced pair induces more than one partition.
 9.
Strict monotonicity implies nontriviality, stateindependence, and that no state is null, i.e., the DM puts positive probability on every state occurring.
 10.
For all \(f,g\in \mathcal {F}\), \(f\sim g\) implies for all \({\alpha }\in (0,1)\)\(f\succsim {\alpha }f + (1{\alpha }) g\).
 11.
Of course, if \(\lambda > 0\), then Theorem 3.1 holds when Maxmin is replaced with Maxmax.
 12.
AGL is a strict subset of the intersection of MMEU and cavEU. A6 implies a specific form to the set of distributions (or, the capacity) in the equivalent representations.
 13.Negative Certainty Independence (adapted to our domain):for \(f,g \in {\mathcal {F}}\), \(c \in {\mathcal {F}}_{c}\) and \(\alpha \in (0,1)\). Assume \(f \succsim c\). Notice that when \(\lambda \le 0\) the AGL functional is concave. So we have$$\begin{aligned} f \succsim c \implies \alpha f + (1\alpha )g \succsim \alpha c + (1\alpha )g \end{aligned}$$(NCI)where the last equality is a consequence of A4 (that independence is preserved over similarly aligned acts, and in particular, constant acts).$$\begin{aligned} V(\alpha f + (1\alpha )g)&\ge \alpha V(f) + (1\alpha )V(g) \\&\ge \alpha V(c) + (1\alpha )V(g) \\&=V(\alpha c + (1\alpha )g) \end{aligned}$$
 14.
This cannot be violated in case (iii), and in case (ii), E must have more than two elements. Further, note that if \(F =1\), \(\gamma _{\emptyset }\) would not be defined, but in that case if \(S\ge 3\), \(F^c = n1\) and the result can follow from reversing the roles of F and E, with \(E^c\) and \(F^c\) and Proposition A.3.
References
 Abeler, J., Falk, A., Goette, L., Huffman, D.: Reference points and effort provision. Am. Econ. Rev. 101, 470–492 (2011)Google Scholar
 Anscombe, F.J., Aumann, R.J.: A definition of subjective probability. Ann. Math. Stat. 34, 199–205 (1963)Google Scholar
 Apesteguia, J., Ballester, M.: A theory of referencedependent behavior. Econ. Theory 40(3), 427–455 (2009)Google Scholar
 Bell, D.E.: Disappointment in decision making under uncertainty. Oper. Res. 33(1), 1–27 (1985)Google Scholar
 Blavatskyy, P.R.: Expected utilitymean absolute semideviation model of individual decision making under risk. Mimio (2010)Google Scholar
 CerreiaVioglio, S., Dillenberger, D., Ortoleva, P.: Cautious expected utility and the certainty effect. Econometrica 83(2), 693–728 (2015)Google Scholar
 Dillenberger, D.: Preferences for oneshot resolution of uncertainty and allaistype behavior. Econometrica 78(6), 1973–2004 (2010)Google Scholar
 Ellsberg, D.: Risk, ambiguity, and the savage axioms. Q. J. Econ. 75(4), 643–669 (1961)Google Scholar
 Ghirardato, P., Marinacci, M.: Ambiguity made precise: a comparative foundation. J. Econ. Theory 102(2), 251–289 (2002)Google Scholar
 Ghirardato, P., Maccheroni, F., Marinacci, M.: Differentiating ambiguity and ambiguity attitude. J. Econ. Theory 118(2), 133–173 (2004)Google Scholar
 Gilboa, I., Schmeidler, D.: Maxmin expected utility with nonunique prior. J. Math. Econ. 18(2), 141–153 (1989)Google Scholar
 Giraud, R.: Referencedependent preferences: rationality, mechanism and welfare implications. Mimeo (2004a)Google Scholar
 Giraud, R.: Framing under risk : endogenizing the reference point and separating cognition and decision. Cahiers de la maison des sciences economiques (2004b)Google Scholar
 Grant, S., Polak, B.: Meandispersion preferences and constant absolute uncertainty aversion. J. Econ. Theory 148(4), 1361–1398 (2013)Google Scholar
 Gul, F.: A theory of disappointment aversion. Econometrica 59(3), 667–686 (1991)Google Scholar
 Gul, F., Pesendorfer, W.: The revealed preference implications of reference dependent preferences. Mimeo (2006)Google Scholar
 Herstein, I.N., Milnor, J.: An axiomatic approach to measurable utility. Econometrica 21(2), 291–297 (1953)Google Scholar
 Herweg, F., Müller, D., Weinschenk, P.: Binary payment schemes: moral hazard and loss aversion. Am. Econ. Rev. 100(5), 2451–2477 (2010)Google Scholar
 Kahneman, D., Tversky, A.: Prospect theory: an analysis of decision under risk. Econometrica 47(2), 263–291 (1979)Google Scholar
 Karle, H., Peitz, M.: Competition under consumer loss aversion. Rand J. Econ. 45(1), 1–31 (2014)Google Scholar
 Kőszegi, B., Rabin, M.: Referencedependent risk attitudes. Am. Econ. Rev. 97(4), 1047–1073 (2007)Google Scholar
 Kőzsegi, B., Rabin, M.: A model of referencedependent preferences. Q. J. Econ. 121(4), 1133–1165 (2006)Google Scholar
 Lange, A., Ratan, A.: Multidimensional referencedependent preferences in sealedbid auctionshow (most) laboratory experiments differ from the field. Games Econ. Behav. 68(2), 634–645 (2010)Google Scholar
 Lehrer, E.: A new integral for capacities. Econ. Theory 39(1), 157–176 (2009)Google Scholar
 Lehrer, E., Teper, R.: Subjective independence and concave expected utility. J. Econ. Theory 158, 33–53 (2015)Google Scholar
 Loomes, G., Sugden, R.: Disappointment and dynamic consistency in choice under uncertainty. Rev. Econ. Stud. 53(2), 271–282 (1986)Google Scholar
 Maccheroni, F., Marinacci, M., Rustichini, A.: Ambiguity aversion, robustness, and the variational representation of preferences. Econometrica 74(6), 1447–1498 (2006)Google Scholar
 Masatlioglu, Y., Ok, E.A.: Rational choice with status quo bias. J. Econ. Theory 121(1), 1–29 (2005)Google Scholar
 Masatlioglu, Y., Ok, E.A.: A canonical model of choice with initial endowments. Rev. Econ. Stud. 81(2), 851–883 (2013)Google Scholar
 Masatlioglu, Y., Raymond, C.: A behavioral analysis of stochastic reference dependence. Am. Econ. Rev. 106(9), 2760–2782 (2016)Google Scholar
 Ok, E.A., Ortoleva, P., Riella, G.: Revealed (p) reference theory. Am. Econ. Rev. 105(1), 299–321 (2014)Google Scholar
 Ortoleva, P.: Status quo bias, multiple priors and uncertainty aversion. Games Econ. Behav. 69, 411–424 (2010)Google Scholar
 Quiggin, J., Chambers, R.G.: Risk premiums and benefit measures for generalized–expectedutility theories. J. Risk Uncertain. 17(2), 121–137 (1998)Google Scholar
 Quiggin, J., Chambers, R.G.: Invariant risk attitudes. J. Econ. Theory 117(1), 96–118 (2004)Google Scholar
 Riella, G., Teper, R.: Probabilistic dominance and status quo bias. Games Econ. Behav. 87, 288–304 (2014)Google Scholar
 Rubinstein, A., Salant, Y.: \((a,f)\) choice with frames. Technical report, Mimeo (2007)Google Scholar
 Sagi, J.S.: Anchored preference relations. J. Econ. Theory 130(1), 283–295 (2006)Google Scholar
 Sarver, T.: Optimal anticipation. Mimeo (2011)Google Scholar
 Schmeidler, D.: Subjective probability and expected utility without additivity. Econometrica 57(3), 571–587 (1989)Google Scholar
 Shalev, J.: Loss aversion equilibrium. Int. J. Game Theory 29(2), 269–287 (2000)Google Scholar
 Siniscalchi, M.: Vector expected utility and attitudes toward variation. Econometrica 77(3), 801–855 (2009)Google Scholar
 Stiemke, E.: Über positive Lösungen homogener linearer Gleichungen. Math. Ann. 76, 340–342 (1915)Google Scholar
 Sugden, R.: Referencedependent subjective expected utility. J. Econ. Theory 111(2), 172–191 (2003)Google Scholar
 Werner, K.M., Zank, H.: A revealed reference point for prospect theory. Econ. Theory (2017). https://doi.org/10.1007/s0019901710962 Google Scholar
Copyright information
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.