Yesterday's diary listed half a dozen well known regularities in voter turnout. They can all be derived from a relatively simple general theory about turnout that is based on three key insights about what is involved in voting.
The first is that voting is an "expressive" act. Rationalistic theories of voting that treat it as a straightforward way to influence what government does usually fail because the probability that any single voter's vote will change the outcome of any election is almost always miniscule -- and would-be voters know this. Thus, given would-be voters' time, money or information processing costs, rationalistic theories tend to predict tiny voter turnout, and generally are ill-equipped to deal with the obvious fact that millions of people vote in a national election. Although turnout does appear to be higher in close elections (one of our rules of thumb from yesterday) the effect is modest and not what one would expect if voters were highly rationalistic.
By contrast, if we treat voting as a way of expressing solidarity with the "right" team -- as something akin to wearing a Carolina Tarheels baseball cap if one is a UNC fan -- a great deal about voting behavior falls into place.
The second insight is disarmingly simple: because people are not born wanting to express themselves politically, the desire to vote must be acquired, either by learning about the candidates, by using party identication as a cognitive shortcut, or by contact from a trusted source. And of course what people learn is partly a function of what they are taught, as well as how, if at all, they search for political information and use it when they get it.
This points to a consideration of political socialization -- in the home, in the schools, in other institutions, such as churches, and in in society. The last-named point is particularly important, because peer groups or other social reference groups help novice political voters decide which "team" to "cheer for" -- and to vote for. People who are "bowling alone" -- not well anchored in social networks of friends, relatives and acquaintances -- lack this important source of cues about proper voting beavior and thus are less lkely to vote.
In the context of a single election, this insight points to the importance of parties devoting effort to preserving their "brand image" and adjusting it carefully. "Flip-flopping," for example, is a problem not just because it conveys that a candidate is indecisive and is mindlessly pandering to public opinion, but because it blurs in voters' minds the image of what the candidate -- and by extension the candidate's party -- stands for. "Staying on message" is a way of policing candidates' and officials' pronouncements so that the party "brand image" is not blurred by messages that are spread out over the political spectrum. This is probably an even more important insight for Democrats than for Republicans (even though Republicans seem to be more self-conscious about using it), because Democrats have to rely more heavily on the votes of those who are not highly educated and who are not sophisticated information processors. For such voters, knowing the party of the candidate needs to be a reliable guide to policy consequences. If it is, then party ID is an excellent "cognitive short-cut." If it is not, and voters are confused about the differences between the parties, or where the parties stand, then voting turnout declines. Strong party "brand image" is necessary for voters to engage in "low-information rationality" -- voting decisions that reflect their objective interests, but which are based on extremely simple information gathering and information processing routines.
Third, voters learn about the candidates and parties over their entire life-span. In contrast to older models that saw political attitudes as relatively fluid in youth but fixed in early adulthood, we now understand that political learning continues throughout the life of the voter. Older voters have more experience watching the performance of the parties when they are in office. They are like folk statisticians who use a larger sample of observations to make more accurate and more confident predictions about how candidates will behave once in office. Like statisticians, their estimates are more stable as their sample size grows. Young voters with limited political experience are relatively easily swayed to support a different party or candidate, but as they accumulate more experience with the consequences of elections, their political leanings become more stable.
Tomorrow we'll consider more regularities in voting behavior, some of which are considerably less intuitive than the ones that we have previously covered.