Norms and Conformity I

Human beings are social animals (Aronson, 2003). We constantly interact with others on a daily basis. Social norms establish the boundaries or parameters of acceptable behavior in a group, and they are indispensable for the functioning of a group. Early research in social psychology suggests that in the absence of a social norm, people have the tendency to create one. Sherif (1937) conducted an experiment showing that after repeated public pronouncements, people establish a group norm using information provided by other people’s judgment.

Bicchieri (2005) argues that social norms and conventions are mechanisms to coordinate people’s actions in a group. She maintains that there is a subtle difference between social norms and conventions. Although both social norms and conventions are coordinating device, conventions (which are stable descriptive norms according to Bicchieri) are used to solve a coordination game, while social norms transform social dilemmas into coordination games. In other words, it is consistent with an individual’s interest to follow a convention, while following a social norm may be contradicting with a person’s immediate interest. Traffic rules are a good example of conventions. As long as you know that other people drive on the right according to the convention, you will also want to drive on the right because it fits your personal interest. But being generous and acting fairly toward others, which is prescribed by the social norms, does not square with your personal interest as it incurs material cost or the renunciation of benefit.

Bicchieri’s categorization in my view is not a good one, as it does not really correspond to the essence of social norms. Social norms have a social component in them: you conform to social norms because you want to be liked by others and have a good standing in the group. Whenever an individual thinks about whether her behavior is acceptable by others, she is probably thinking of a social norm. Foot-binding of women in ancient China, as Bicchieri argued, is a convention. No matter how this practice came into existence, it quickly spread to all but the lowest classes, and was accepted as a sign of gentility and an important condition for marriage (Mackie, 1996). It is obvious, however, a woman decides to go through the pains to bind her feet is mainly concerned about the acceptance of others, and for this reason, it is better to treat it as a social norm. Conformity to social norms has a signaling function: it sends signals to others who you are.

Studies on Culture and Social Norms VI

Economists believe that sanctioning is an inherent part of social norms. Fehr and Fischbacher (2004) use third-party punishment to study the functioning and content of social norms. A third party who can take costly action to punish selfish behavior is introduced into the dictator game and the prisoners’ dilemma game. About two thirds of the third parties punished selfish offers of the dictators in a third-party dictator game, and about 60% punished defectors in the prisoners’ dilemma game. Their experiment indicates that third-party sanctions are important for the functioning of social norms. Indeed, it may well be that this third-party punishment itself is prescribed by a social norm.

Social norms can be used to explain some interesting experimental results. Xiao and Houser (2009) found that rejection of unfair offers is significantly less frequent in an ultimatum game when receivers can express their feelings to the proposer. The authors conclude that costly punishment may be just a way to express negative emotions. Using the social norms theory, we can speculate the social norm prescribe to show antipathy when the proposer acts unfairly. When it is possible to use negative emotions to demonstrate an aversion to the unfair behavior, the receiver does not feel necessary to engage in costly punishment. The possibility for the receivers to express negative feelings also makes the proposers more likely to give fair offer, possibly because it exerts focusing influence and draws the attention of the proposers to the social norm.

Xiao and Houser (2005) studied the effect of emotional expression in a one-shot dictator game. Their results confirm that avoiding written expression of disapproval, or negative emotion, plays an important role in promoting fair decision making. Proposers act more generously when the receivers can respond with written messages, although monetary sanctions are more effective when we compare the results from the dictator game and the ultimatum game. These results imply that others' pinion may serve the same function as sanctions in the operation of social norms.

Studies on Culture and Social Norms V

Some experiments have shown that people’s preferences on conformity to social norms may depend on context. Psychologists maintain that norms only influence behavior when one’s attention is drawn to them (focusing influence) and a norm has greater impact on one’s behavior when one observes others behaving in alignment with that norm (informational influence). Krupka and Weber (2009) use a simple one-shot binary dictator game to test this theory. In the experiment, the dictator can choose either X, which leads to a fair allocation ($5, $5), or Y, which gives rise to an unfair allocation ($7, $1). In descriptive focusing treatment, before assigning roles and making decisions, each subject was asked to guess the percentages of subjects who chose X and Y in a previous session. In Bicchieri (2005)’s terms, the subjects were focused on empirical expectations. In the informational focusing treatment, the subjects were asked to guess what percentages of subjects in a previous session had stated X and Y should be chosen, and hence focused on normative expectations. In informational treatment, each subject observed the choices made by four previous participants. Compared to the baseline treatment, subjects exhibited significantly more prosocial behavior in the focusing treatments. The informational treatment shows that individuals are more likely to engage in prosocial behavior when they observe others doing so, which is consistent with the results from Bicchieri and Xiao (2008). This experiment shows that contextual cues that remind the individual the existence of social norms greatly increase pro-social behavior. Hence norm compliance exhibits a focusing effect, and norms only impact behavior when an individual’s attention is drawn to them.

Cason and Mui (1998) also demonstrate the functioning of social influence similar to Krupka and Weber (2009)’s informational treatment in the setting of a sequential dictator game. In their experimental design, 4 subjects, denoted as subject 1 to 4, form a group. In stage 1, all subjects first indicate the amount of money P1 they want to take from the $40 pie in a dictator game with market setting. The information on the decision is then exchanged between subject 1 and 3, and between subject 2 and 4 in the Relevant Information treatment, while the information about the subject’s date of birth is exchanged in the Irrelevant Information
treatment between odd-numbered subjects and between even-numbered subjects. In stage 2, all subjects indicate the amount of money P1 they want to take from the $40 pie. The role of proposer and the receiver is then assigned randomly, either odd-numbered players or even-number players would be the sellers, and each seller can be the seller of only one stage, which are also randomly determined. Their results show that receiving relevant information makes subjects less likely to move to more self-regarding choices. They also find that subjects who exhibit more self-regarding behavior on their first decisions are less likely to change choices in the second stage, implying the sensitivity to social norms is heterogeneous among subjects.

Studies on Culture and Social Norms IV

Rabin (1993) notices reciprocity is a common phenomenon in human interactions. In order to model reciprocity, he uses psychological game theory, incorporate beliefs and intentions into people’s preferences. In this model, people’s intentions are reflected by the action they take. But for situations like the dictator games where there is only one decision maker, the model is inadequate as we don’t have a measure on the other person’s intentions. A more generalized model proposed by Charness and Rabin (2002), even though incorporates social welfare concern, inequality aversion and reciprocity, is still not capable of explaining the above experimental results, because none of these components can explain the fact that people exploit moral wiggle room to behave selfishly.

The inadequacy of current theories on other-regarding preferences implies the importance to incorporate social norms into individuals’ preferences. Bicchieri (2005) is among the first that propose a norm-based utility function. The utility function she proposed has two parts: the monetary payoff and the norm-related utility. A social norm in her model is defined as one prescribed action in a situation, and disutility results when the agent violates the norm. The disutility is related to the reduction in others’ payoffs caused by this violation.

While this model provides better explanation for the above experiments compared to other models with prosocial preferences, it does not consider the case where multiple actions are equally appropriate, and its definition of norms makes it incapable in explaining results from experiments such as the multi-player dictator game studied by Dana et al. (2007).

Krupka and Weber (2013) propose a more refined definition of norms. Instead of defining a single appropriate action prescribed by the norm in a situation, they consider each action has some level of social appropriateness. They design a coordination game to elicit and quantify social norms. In their norm elicitation game, the subjects are asked to rate "social appropriateness” of an action in a specific situation on a scale of 4. The subjects get rewarded if their ratings are the same as the modal rating of the group. Using the elicited norms, Krupka and Weber analyzed the experimental results of various games, including famous games studies by other authors (List, 2007; Dana et al., 2006, 2007; Lazer and Pentland, 2009). Their experimental results show that social norms can provide a good prediction of individual decision making, and
subjects have a stable willingness to sacrifice money and take socially appropriate actions. Their research contributes to the endeavor to develop a theoretical model of social norms.