Although most of their work is driven by numbers and equations, economists are inherently curious about the peculiarities of human behavior—especially when it comes to the forces that affect buying and selling. Caltech economist Marina Agranov shares this fascination. In a new study, Agranov sought to understand how emotions such as guilt and disappointment might influence the decisions of buyers and sellers.
In this study Agranov, Caltech's Rea A. and Lela G. Axline Professor of Economics, asked participants to engage in a simple communication game, in which one side (sellers) have information that the other side (buyers) need in order to make a good decision. The study introduced payoffs and costs associated with psychological states in an attempt to test the role that emotions play in these interactions.
"We started with a classical communication game," Agranov explains. "There is one person who knows some information that is important to the other person. But when the other person takes an action, it is going to determine the payoffs for both of them. One person holds the information while the other controls the action. The only way for the information to get between these two people is by talking: 'I have to tell you something, and then you're going to interpret my message and act on it.'"
This game has been used over and over again in economics, both in theory and in experiments, because it parallels the situation of a seller and buyer: The seller shares information about their product; the buyer makes a judgment about the value of that information and then decides whether or not to purchase the product. Obviously, this is crucial to the seller because it is only by convincing the buyer to make a purchase that they are able to turn a profit. Economists tweak the rules of the game in various ways depending on what they are trying to learn.
Agranov and her coauthors recruited 179 people to participate in a study. Each person was randomly assigned to be either a buyer or seller, and they remained in this role throughout the game. Sellers were told whether the goods they were trying to sell were of high or low quality and then instructed to communicate with buyers to try to make a successful deal.
"We asked participants to play three games," Agranov explains. "The first is the standard seller–buyer game where there are material payoffs for successful trade between one seller and one buyer. The second introduces psychological factors into the equation: Sellers may be told that they dislike telling lies and misleading buyers, and that they will be monetarily penalized for doing so, while buyers may be told that when they are disappointed that the seller led them to an unprofitable purchase, this imposes an additional cost. The third includes both material and psychological payoffs and costs but also adds an additional seller to see how competition between sellers affects trade."
Players had multiple interactions in all versions of the game. Sellers were told they had either high-quality or low-quality goods and then had to decide whether to pass on the correct information about the quality of their goods to buyers. Buyers had to decide whether or not to trust sellers and agree to a trade of money for goods. If a sale did not occur, both buyers and sellers received a small payoff; if a high-quality good was sold, both received a higher payoff. But if the seller convinced the buyer to purchase low-quality goods, the seller received an even higher payoff while the buyer received nothing.
In the versions of the game that included psychological factors, players were awarded money for successful trades, but they were also penalized monetarily if they incurred psychological costs. Whether or not sellers and buyers incurred penalties for emotions was determined by the persona they were assigned at the beginning of the game, something that was not known to the other player.
Agranov and her colleagues stuck to simple intuitive emotions associated with the seller–buyer interaction. "A seller may be told that the person they are playing in this game dislikes lying per se, and will feel guilty for misleading buyers," Agranov explains. "Economists define guilt in a very specific way. I will feel guilty if I tell you something that may mislead you into taking an action that is not good for you. If I tell my kids to eat broccoli, I am not going to feel guilty, because broccoli is actually good for them. But if I tell them to jump off the roof, and they do, and they break their legs, then I'll feel guilty. In this game, if sellers lie to potential buyers about the quality of their goods, and the buyer falls for it, they will feel guilty."
On the side of the buyer, Agranov and her colleagues made disappointment a psychological cost. "If the buyer learns that the seller said something that led them to do the wrong thing, they are told that they will feel disappointed," Agranov says. "We quantify this disappointment as a monetary penalty."
"The results show that when psychological payoffs are included in the game without competition between sellers, that is, when sellers and buyers are made to feel responsible for their actions, there is more trade and higher welfare for both the sellers and buyers. Everybody benefits," Agranov says.
Unfortunately, this result does not survive the introduction of competition between sellers. "The improvement in trade where psychological payoffs were added turned out to be very fragile. As soon as competition was introduced, all the good things that happened without competition unraveled. There was decreased welfare for both sellers and buyers," Agranov explains. "When there is competition, sellers start lying much more out of fear that they will be left out of interactions with the buyer. On the other hand, buyers, who were perfectly able to decode the messages given by sellers in a one-on-one, noncompetitive game, are fooled over and over again when sellers competed with one another."
Agranov suggests that there are three reasons why buyers were less able to evaluate sellers' claims in a competitive situation. "First of all, there's just an inherent belief that almost all of us hold that competition is good for buyers, that it keeps prices lower. Because of this, they are less suspicious of sellers than they should be. The second reason is that in the type of markets we are modeling, it's very hard to learn how honest each seller is. They only get to see the quality of the goods when they make a decision to buy, so they can't discover whether or not the other seller is lying. Finally, beliefs in general are very sluggish. Contrary evidence is slow to move the buyer's belief that competition will always favor them."
"In short," Agranov says, "in the game that had no competition, emotions worked in the same direction as economic forces. Boosting the moral character of interactions between sellers and buyers by imposing costs for guilt for lying and disappointment over being fooled led to more trade and better trade. But competition neutralizes the effectiveness of emotions. In essence, competition forces are stronger than psychological ones in competitive markets."
Agranov is excited about where this research can go in the future. "Very little has been done in the economics literature on the role of emotions in market communications. We are taking first steps here and trying to introduce emotions in a rigorous way," she says.
This research was published in an article titled "Trust Me: Communication and Competition in a Psychological Game," published in Journal of the European Economic Association. Coauthors are Agranov, Utteeyo Dasgupta of Fordham University, and Andrew Schotter of New York University.
