Papers

Rational Attitude Change by Reference Cues when Information Elaboration Requires Effort by E. Bilancini and L. Boncinelli
this version: July 2016
ABSTRACT: In this paper we develop a parsimonious model of decision-making that aims at capturing the determinants of consumers’ attitude change in response to persuasive messages that exploit reference cues. Our model is inspired by dual-process theories of information elaboration, but does not introduce purely behavioral traits. The decision-maker receives a message containing an offer of unknown quality together with a reference cue that associates the offer with a category of offers, whose average quality is known. The decision-maker initially exerts little cognitive effort in processing the message, assessing the offer quality on the sole basis of the reference cue (acting as a “coarse thinker”). The decision-maker can then take a decision on the offer without further investigating or she can exert substantial cognitive effort to scrutinize the message carefully and obtain more precise knowledge of the offer quality. The proposed model predicts that the persuader can exploit reference cues to affect attitudes both directly (by inducing acceptance of offers) and indirectly (by inducing low cognitive effort, and hence influencing acceptance). This model matches several predictions of prominent psychological models of attitude change such as the Elaboration Likelihood Model and the Heuristic-Systematic Model.

Signaling to Analogical Reasoners Who Can Costly Acquire Information by E. Bilancini and L. Boncinelli
this version: October 2016
ABSTRACT: In this paper we prove that separation in signaling games can be obtained in the absence of the single-crossing condition, in a model where the receiver reasons analogically across a pair of signaling games and can costly acquire information on the sender’s type. Beyond the standard kind of separation – where the high type sends the high signal and the low type sends the low signal – we find that also reverse separation is sustainable in equilibrium – with the high type sending the low signal and the low type sending the high signal. Further, we prove that reverse separation in one game is obtained only if ordinary separation occurs in the other game. Pooling is possible and can go along with ordinary separation in one game.

Signaling with Costly Acquisition of Signals by E. Bilancini and L. Boncinelli
this version: July 2016
ABSTRACT:In this paper we identify a novel reason why signaling may fail to separate types, which is specific to cases where the receiver has to incur a cost to acquire the signal sent by the sender. If the receiver chooses not to incur the acquisition cost, then all sender’s types find it optimal to pool on the least costly signal; also, if all sender’s types pool on the least costly signal, then the receiver finds it optimal not to incur the acquisition cost. This kind of strategic complementarity makes the resulting pooling equilibrium extremely robust, even when costs of signal acquisition are very small. Also, pooling is shown to be robust to all refinements based on out-of-equilibrium beliefs, even when the sender can engage in further signaling that can act as an “invitation” to acquire the main signal, and when acquisition costs are smooth and depend on the receiver’s effort to acquire the signal. However, the pooling outcome is not necessarily robust when signals are not purely costly or the receiver can initially commit to acquire the signal. These results provide a new source of interest in pooling equilibria

When Market Unravelling Fails and Mandatory Disclosure Backfires: Persuasion Games with Labelling and Costly Information Acquisition by E. Bilancini and L. Boncinelli
this version: October 2016
ABSTRACT: In this paper we develop a variant of the persuasion game by Milgrom and Roberts (1986) to study the emergence and the desirability of product labelling when sophisticated buyers can acquire information on the actual quality of the product by paying a cost. Labelling is modeled as the (verifiable) public disclosure of an otherwise unobservable trait of the seller that is correlated with the actual quality of the product. Our main finding is that market unravelling can fail because of the presence of many sophisticated buyers, in which case imposing mandatory disclosure can backfire. When the joint distribution of seller’s qualities and traits is exogenous, if market unravelling fails and mandatory labelling is imposed, then naive buyers gain while profits decrease for high quality sellers. Further, if the label is sufficiently informative, then also sophisticated buyers gain and profits increase for low quality sellers. When, instead, the joint distribution of qualities and traits is endogenous, mandatory labelling can not lead to an increase in quality or in buyer’s utility. Moreover, if average quality is left unaltered, then sellers’ profits are not increased and cost-inefficiencies arise.