Essays on social influence

This dissertation explores various aspects of social influence processes in political behavior research. Specifically, three separate essays explore the ways in which communication, social conformity, cultural transmission, and other interpersonal influence processes operate to generate similar public behaviors among groups of interconnected individuals. Broadly, the dissertation project develops theoretically a set of general principles about the cognitive processes, social dynamics, and institutional structures that shape the development, transmission, and transition of social norms. The theoretical arguments are used to inform a set of experimental research designs that provide empirical evidence for some of the short-term processes at work, as well as a set of simulations that are informative about the more long-term processes. After a brief introduction in Chapter 1, Chapter 2 presents a simulation model of the cognitive tradeoff between stating a truly held internal attitude and avoiding social deviancy, as well as the dynamics of interpersonal influence in forced interaction settings. The key motivation behind the focus on these specific factors is the idea that for most people, many, if not most, day-to-day social interactions compel them to interact with people who they likely would not otherwise select to interact with absent common membership in some institution or organization. It is in these settings with unfamiliar others that the tradeoff between the expressive satisfaction of stating what one truly believes and the social satisfaction of exhibiting the group majority behavior is likely to be most obvious. One takeaway point from this essay is that when individuals place relatively equal weight on the values of internal consistency and social conformity, the behavior and overall satisfaction levels of the society at large become much less predictable in the short term, and the average overall satisfaction of the population takes its lowest value. That is, populations that place significant weight on either conformity or consistency experience more satisfaction on average. Chapter 3 presents a series of laboratory experiments on social influence. In addition to dealing with non-self-selected group interactions and conformity pressures, as in Chapter 2, several other ideas drove the design employed. One goal was to study very short-term micro-level processes of social interaction, and the benefit of the experimental setting is the ability to measure, rather than merely assume, the relative weight that individuals place on internal consistency, as opposed to social conformity. But the largest goal was to take a set of general ideas from the empirical research on social conformity, and apply them specifically to the question of individual expression of equality between groups. In a prediction that I call the Social Dissonance Hypothesis, I suggest that individuals come to the laboratory both with their own internal attitudes and some general notion about the range of attitudes that are socially acceptable in the broader population. When asked to give their views in isolation, some individuals are willing to express an attitude that runs counter to what is socially acceptable in the broader population. And among these people, there is a very strong tendency to change responses when they face a unanimously opposed majority, and they are compelled to express their attitudes publicly. Chapter 4 develops a second simulation model in which the intention is to move away from forced interaction settings. In this model, the key moving parts are the abilities of individuals to self-select into social relationships, the assumption that individuals can influence each others' types through interaction over time, and that individuals are characterized by traits in more than one dimension. Self selection into social relationships is obviously ubiquitous in reality, and the idea that individuals can influence types is meant to capture the intuitive sense that friends tend to become more alike through more interactions. The assumption that individuals are multidimensional captures the idea that social relationships might be formed on the basis of one dimension, but later other information may come to light that can make those social relationships more or less costly. The model shows that types that were relatively successful tended to be imitated more, and therefore ended up dominating the population over time. No type tended to have an advantage in general, but for any given simulation, an agent's type was consequential for the ultimate utility experienced. A final chapter offers a very brief conclusion