How Social Preferences Shape Incentives in (Experimental) Markets for Credence Goods[Link]

Credence goods markets suffer from inefficiencies caused by superior information of sellers about the surplus‐maximising quality. While standard theory predicts that equal mark‐up prices solve the credence goods problem if customers can verify the quality received, experimental evidence indicates the opposite. We identify a lack of robustness with respect to heterogeneity in social preferences as a possible cause of this and conduct new experiments that allow for parsimonious identification of sellers’ social preference types. Our results confirm the assumed heterogeneity in social preferences and provide strong support for our explanation of the failure of verifiability to increase efficiency.

everywhere. Note that this class subsumes altruists, as modelled by Andreoni and Miller (2002), for instance, as well as surplus maximisers, as discussed by Engelmann and Strobel (2004), among others. A spiteful (SP) expert is willing to give up own monetary pay-off to decrease the pay-off of her trading partner if the 'price of taking' is not too high. Thus, the indifference curves of a SP seller in (p s , p c ) space are upward sloped everywhere. Note that this class includes spiteful experts, as modelled by Levine (1998), for instance; competitive sellers a la Charness and Rabin (2002); and experts with concerns for relative income a la Duesenberry (1949). An inequality averse (IA) expert wants to see the pay-off of her customer increased if she is better off than the customer but she wants to see the customer's pay-off decreased if the opposite is the case. Thus, the indifference curves of an IA expert in (p s , p c ) space are downward sloped below the equal-monetary-pay-offs line and upward sloped above that line. Note that this class includes inequity or inequality averse agents, as modelled by Fehr and Schmidt (1999), or Bolton and Ockenfels (2000), for instance; agents with egalitarian motives, as discussed by Dawes et al. (2007) and by Fehr et al. (2008), among others; as well as difference averse agents a la Charness and Rabin (2002) and Fisman et al. (2007). An inequality loving (IL) expert is willing to sacrifice own material pay-off to increase the difference between the pay-offs of the two trading partners. Thus, the indifference curves of an IL expert in (p s , p c ) space are upward sloped below the equal-monetary-payoffs line and downward sloped above that line. Note that this class includes equity averse agents as discussed, for example by Charness and Rabin (2002) and by Fershtman et al. (2012).
The above distinction between five classes of social preferences omits types that are selfish in one of the two domains and pro or anti-social in the other (see Kerschbamer, 2015 for a distinction between nine archetypes of social preferences). In the experimental section (Section 5), we allow for such hybrid types. We do so by extending the definitions of IA and IL as follows: We assign a subject to IA if her behaviour is consistent with the combination: (a) @U/@p c ≥ 0 for p s ≥ p c ; (b) @U/@p c ≤ 0 for p s < p c ; and (c) at least one of the two derivatives is different from 0.
And we assign a subject to IL if her behaviour is consistent with the combination: (a) @U/@p c ≤ 0 for p s ≥ p c ; (b) @U/@p c ≥ 0 for p s < p c ; and (b) at least one of the two derivatives is different from 0.

Roles and 16 Rounds
This experiment consists of 16 rounds, each of which consists of the same sequence of decisions. This sequence of decisions is explained in detail below. There are two kinds of roles in this experiment: player A and player B. At the beginning of the experiment you will be randomly assigned to one of these two roles. On the first screen of the experiment you will see which role you are assigned to. Your role remains the same throughout the experiment.
A player A interacts with a player B. This pair of players changes for each round. Therefore, you are interacting in every round with a new player (of the other role).
All participants get the same information on the rules of the game, including the costs and payoffs for both players.

Overview of the Sequence of Decisions in a Round
Each round consists of a maximum of two decisions which are made consecutively. Decision 1 is made by player B and decision 2 is made by player A. In each round, 2 prices will be announced before players make their decisions. These prices are set for a given round. This price setting is referred to in the following as 'Decision 0'.  Detailed Illustration of the Decisions and Their Consequences Regarding Pay-offs Decision 0

Short Overview of the Sequence of Decisions in a Round
In case of an interaction player A has to choose between two actions, action I and action II, in Decision 2. Each chosen action causes costs which are as follows: Action I results in a cost of 2 points (=currency of the experiment) for player A.
Action II results in a cost of 6 points for player A.
Player A receives from player B the valid price for the action he/she chooses in Decision 2 if player B decides to interact with him/her. At Decision 0 the valid prices for action I and action II for this round are announced to both players.

Decision 1
Player B decides whether he/she wants to interact with player A.
If he/she wants to do so, then player A chooses an action in Decision 2 and he/she receives the valid price for this action from player B.
If he/she does not want to interact, then this round ends and both players get a payoff of 1.6 points for this round.

Decision 2
Before Decision 2 is made (in case player B chose 'Yes' at Decision 1) a type is randomly assigned to player B. Player B can be of one of two types: type I or type II. This type is determined a new in each round. With a probability of 50% player B is of type I, and with a probability of 50% he/she is of type II. Imagine that a coin is tossed in each round. If, for example the result is 'heads', player A is of type I, if it is 'tails' he/she is of type II.
Player A gets to know the type of player B before he/she makes Decision 2. Then player A chooses an action, either action I or action II, and receives the corresponding price (valid for the respective round).
An action is sufficient under the following conditions: (i) player B is a type I player and player A chooses either action I or action II; (ii) player B is a type II player and player A chooses action II.
An action is not sufficient if player B is of type II and player A chooses action I. Player B receives 10 points, if the action chosen by player A is sufficient. Player B receives 0 points if the action chosen by player A is not sufficient. In both cases player B has to pay the valid price for the chosen action.
At no time will player B be informed about whether he/she is of type I or a type II in any given round.

Payoffs
If player B chooses not to interact in Decision 1 (decision 'No' of player B) then both players receive 1.6 points for this round.
Otherwise (decision 'Yes' by player B) the payoffs are as follows: Player A receives the price (denoted in points, as announced in Decision 0) for the action chosen in Decision 2, less the cost of this action.
Player B's pay-off depends on whether the Decision 2 of player A was sufficient or not.
(i) If the action was sufficient, player B gets 10 points less the price for the action chosen by player A in Decision 2.
(ii) If the action was not sufficient, player B has to pay the price for the action chosen by player A in Decision 2.
At the beginning of the experiment you receive an initial endowment of 6 points. With this endowment you are able to cover losses that might occur in some rounds. Losses can also be compensated by gains in other rounds. If your total pay-off sums up to a loss at the end of the experiment you will have to pay this amount to the supervisor of the experiment. By participating in this experiment you agree to these terms. Please note that there is always a possibility to avoid losses in this experiment.
To calculate the final pay-off the initial endowment and the profits of all rounds are added up. This sum is then converted into cash using the following exchange rate: 1 point = 25 euro-cents (i.e. 4 points = 1 euro)