Signaling Goodness: Social Rules and Public Choice

8. A Study of Political Positions

CHAPTER 8 A Study of Political Positions Hypotheses In this chapter we test the theory developed in chapter 6 by focusing on three implications of that theory. First, the lower the cost of signaling "goodness," the more people will adopt "progoodness" political positions . This proposition cannot simply be derived from the downwardsloping demand curve because in our case there is a contrary force. When others know of an increase in the price to an individual if he adopts a given political position out of goodness, that individual signals more goodness by espousing such a position. So both the costs and the returns of signaling goodness increase with an increase in the pnce. However, as shown in chapter 3, the amount ofgoodness signaled is the price of adopting a political position out of goodness times the political position itself (all, of course, measured in appropriate units). Assume the desired amount of goodness signaled is invariant with respect to price. That assumption corresponds to the standard assumption that utility is independent of price, holding real income constant. Then, to keep the goodness signaled constant, an increase in one of its components, price, must lead to a proportional decrease in its other component, political position. The price elasticity ofdemand for goodness signaling in political positions should be -1. The political positions associated with goodness should increase with a decrease in the price of goodness. Another proposition has been developed in the previous chapter. People use more resources to loudly proclaim "good" than "bad" positions . This proposition not only applies to demonstrations and what is generally meant by political activism. It also applies to any occupational choice that is in part determined by the goodness motivations. There are some occupations in which one can preach about political positions. One would expect people who wish to preach goodness to be more likely to choose those occupations. 133 134 Signaling Goodness A third implication of our theory follows easily from this second proposition. Relative to private discussions, some people will get a higher proportion oftheir information about the political views of others from loud activists and "do-gooder" occupations. The imitative behavior of those thus informed will, hence, make them choose political positions with a larger goodness component. As developed in this, the previous chapter, and the next chapter, the information from loud activists and "do-gooder" occupations includes the media, education, and books. Besides testing these propositions, this chapter discusses a large number of empirical regularities that have gone unexplained. Though most economists studying voting behavior have long been aware of their existence, economists seem a remarkably uncurious lot. If something cannot be easily explained by our simplest models, we just assume it is part of "taste." In the case of voting behavior that means we ignore a lot. "Why are the young, college teachers, and residents of large cities more liberal?" These questions and others have been unanswered , that is, attributed to "tastes." Our goodness theory provides more satisfactory answers, answers that allow us to explore subtler features of these regularities. Before we can test these propositions and answer these questions, however, we must select a data set that allows these tests and controls for the other relevant variables that also have an impact on political positions. Data and Issues Our procedure is to run regressions on answers to public policy questions against characteristics of respondents and their families given by data for the United States from the General Social Survey, 1972-1996 (NaRC 1996). Currently, the preferred procedure in the public choice literature for running such regressions is parsimony, but those working with the simple self-interest model usually cannot resist the inclusion of at least a few variables, such as race, region or city size, that they cannot justify on theoretical grounds. We include a large number of variables . That inclusion is justified by the theory we are testing: that concern with what others are thinking is crucial in the determination of voter behavior. There are two main manifestations of that concern: (I) political positions as imitation; (2) political positions chosen to be "good." In this chapter we concentrate primarily on the latter, since the former has been more thoroughly examined in chapter 5. A Study of Political Positions 135 Our approach is to examine seventeen different issues, opinions about which will be potentially affected by goodness. We use the commonly accepted liberal versus conservative characterization of views about these...