Campbell, Converse, Miller, and Stokes: The American voter
From WikiSummary, the Free Social Science Summary Database
- Congress: Elections
- Congress: Parties
- Ideological Traditions
- Voter Sophistication
- Voter Turnout
Campbell, Converse, Miller, and Stokes. 1960. The American voter. New York: John Wiley & Sons, Inc..
Party ID is the major influence on voters' perceptions of political choice as well as their final vote. Party ID is characterized by stability and resistance to contrary influence. Furthermore, it is formed early in life (i.e., it is inherited).
Place in the Literature
Campbell at al. represent the (University of) Michigan (psychological/party ID) approach to voter behavior. They can be seen as arguing against Berelson et al.'s Columbia studies (for whom societal pressures play the largest role), Fiorina's (1981) issue-based retrospective voting, and the broader rational choice approach of Downs (1957) and Ordeshook (1968).
Though this model has been influential enough that researchers still cite Campbell et al. decades later, The American Voter has faced substantial criticism, such as in Key (1966) and Popkin et al. (1976).
The Funnel Model
Campbell et al. argue that the best predictor (X) of whether an individual will vote Republican or Democratic is the funnel model. The funnel works like this: First, you learn your party ID from parents and socialization. You form a psychological attachment to this party. As such, your partisanship shapes the development of your attitudes; because you like your party, you adopt its positions. Your (underlying) attitudes are then reflected in your positions on the six attitudinal dimensions: the personal attributes of the Democratic candidate (Stevenson), the personal attributes of the Republican (Eisenhower), the groups involved in politics and the questions of group interest affecting them, the issues of domestic policy, the issues of foreign policy, and the comparative record of the two parties in managing the affairs of government. Finally, these issue positions are the proximal cause of your voting decision. In fact, these six issue positions predict voting decisions with 87% accuracy--which is even better than asking voters who they intend to vote for.
Not surprisingly, feelings across these six dimensions tend to be highly correlated. For Campbell and his coauthors, this correlation occurs because partisan feelings are strongly shaped by party identification. (Party ID leads to partisan feelings, not the reverse.) In this book, party ID is treated as a psychological force or tie through which voters interpret political issues (each of the aforementioned dimensions). The authors write that "Identification with a party raises a perceptual screen [i.e. selective perception] through which the individual tends to see what is favorable to his partisan orientation." In this sense, the party acts as a supplier of cues by which the individual may evaluate the elements of politics.
Origins of Partisanship
They claim that individuals "inherit" a party ID from their parents and the social milieu in which they are raised and that this party ID is characterized by stability and resistance to contrary influence. (They do recognize that objective events and conditions can lead a voter to modify her party ID or vote against it if her evaluation of the current elements of politics does not agree with her initial allegiances.)
Issues Don't Matter
In their interviews, Campbell et al. find that policies and issues play a small part in most voters' decisions, that only a small fraction of the electorate (12%) displays anything resembling an ideology (i.e., most people when asked about their positions on specific policy issues do not have a consistent pattern of responses in terms of a liberal-conservative dimension), and that voters frequently do not know which party stands for what. These findings cast doubt on the efficacy of voting as a mechanism of democratic control of government.
Changes in Partisanship
Changes in party ID are possible. These changes result from either personal forces (usually changes in an individual's social milieu) or social forces (usually the result of experiences related to great national crises or those experiences related to progress through the life cycle older voters tend to be more conservative).
Technique: Longitudinal data from three presidential elections (1948, 1952, and 1956). This data was collected through interviews with voters. Campbell et al. determine an individual's party id and partisan preferences/leanings based upon responses to interview questions (i.e., they rely on the self-identification of voters) rather than past voting behavior.
The surveys that Campbell et al. started have continued to be administer during every presidential election (albeit with significantly modified questions) and constitute the American National Election Studies (ANES, NES).
Outlines the general approach. Uses survey data from 1948, 1952, and 1956 (not much from 1948).
Chapter 2 details the "funnel analogy," the central argument in this book. The funnel works like this:
Political socialization (mainly your parents' party identification) determines party ID, which determines your political attitudes, which determines how you actually vote.
Party ID is seen as an "enduring psychological attachment." The political attitudes it determines are measured along six dimensions:
- How you feel about the Democratic candidate
- How you feel about the Republican candidate
- How well each party manages the affairs of government
- Group interests ("he represents business owners" or "little people": people like me vote for so-and-so).
- Domestic policy issues
- Foreign policy issues
Chapter 3: What our attitudes are
- Generalization: Our attitudes (both cognitive and affective) bleed over from one thing to another. In particular, our attitudes about a party affect our attitudes about a candidate.
- Threshold of awareness: It takes a lot for the masses to take note of something. Although Stevenson had a clear, frequently stated position on foreign policy issues, most people didn't know them. They pay far more attention to you once you're president than they do to you as a candidate.
- Social bases of stability: Some perceptions are stable, others fleeting. This can depend on several factors. For example, it took two decades for citizens to forgive Republicans for causing the Depression and give another Republican a chance. But other perceptions seem to fade much more quickly.
Chapter 4: How our attitudes affect our voting behavior
- See pg 74 (note 7): you can predict voting behavior well based on attitudes about the candidates. Predictions improve by including attitudes about domestic and partisan issues.
- See figures starting on pg 82. The authors asked five questions on which people could take a partisan position (or not). Some people took a partisan stand on all five, some on fewer. Some had consistently Republican/Democratic stands, others did not. Based on number and coherence of stands, you can predict party-line voting, passion about the election, how early the respondent will make a decision about who to vote for, etc. [See Zaller and Feldman 1992.]
Chapter 6: Partisanship
- Vote Choice is the sum of a field of forces.
- Partisanship is an antecedent to those forces.
- As shown on p 137, partisanship correlates well with voting but does not explain all of the vote choice.
- Strong party identifiers tend to be more interested in politics.
- Even Democrats have generally Republican attitudes about foreign policy, and even Republicans have generally Democratic attitudes about domestic policy (see p 129-130).
Chapter 7: Origins of partisanship
- Early socialization (family influences)
- Partisanship is stable over time
- Youth lean Democrat
- Minorities lean Democrat (Catholics, Jews, blacks); perhaps they are drawn to Democratic ideas of social equality.
Chapter 8: Public policy and political preference
Before an issue affects your vote, three things must happen:
- You must be aware of and know something about the issue. (It must be "cognized.")
- You must care about it, at least minimally. (Operationalized: You must have an opinion about a specific piece of legislation. E.g. pg 172: Are you in favor of or opposed to the Taft-Hartley Act?)
- You must know what the parties say about the issue (does your party support or oppose Taft-Hartley).
Basically, this chapter requires you to know a LOT about an issue in order for it to affect your vote. You need to know which laws deal with it and so on. Really, the only laws anybody has heard of in the past five years are the Patriot Act and No Child Left Behind. But not only do you need to know the name of the bill--you also need to know the details (not just the subject) of the bill.
Familiarity with politics varies; some people know a good bit about several issues, some know very little about any issues. However, individuals do tend to have a firm sense of how the parties differ. This, combined with an assumption that voters have stable underlying values, would explain why voters tend to stick with a single political party.
Some individuals form their views on individual issues from an underlying set of general principles, however small it may be. Many, though, simply take each new thing as it comes, with little overriding ideology. We should not assume that an individual has a coherent ideology simply because his views are congruent, though.
Attitudes about foreign and domestic policy tend not to correlate (positively or negatively). One who favors domestic interventionism does not necessarily favor foreign interventionism as well. "Internationalist" ideals also do not correlate with membership in either party, despite the parties' reputations. However, "interventionist" (domestically) ideals do correlate with membership in the Democratic party.
Many people have "non-scalar" opinions (they don't appear to have congruent views). However, non-scalar patterns occur in both parties (though they are least frequent among strong partisans, not independents).
Ideological sorting does occur: those who have "liberal" views tend identify more strongly with Democrats. However, this doesn't mean that they are ideological. Instead, people seem to concern themselves with "primitive self-interest." Only some people appear to be ideological: low-status Republicans and high-status Democrats. (Wealthy Republicans and poor Democrats, on the other hand, are just sorting according to class, not ideology).
Frequently, analysts assume that most voters are (1) sensistive to their policy position on a left-right continuum and (2) sensitive to both parties' shifting positions along that continuum (pg 217). Thus, they speak of the results of an election as indicating an ideological shift in the electorate or by one of the parties. However, only a minority of the electorate actually meet these two conditions. This ideological behavior is only one of several "frames of reference."
The "frames of references" (analogous to Converse's (1964) "levels of conceptualization"):
Level A: Ideology and Near-Ideology
An abstract level of thinking about politics: what is good and bad (principles). You don't have to use the correct intellectual terms, but you need to be forming your opinions about specific issues in relation to broader, more abstract notions.
Comment: In reality, all the examples they provide of "ideologue" interviews seem to differ from the "group benefits" interviews in only one respect: the "ideologues" used the words "liberal" or "conservative" to describe the parties, while the "group benefits" people spoke about issues. This seems to be a strange criterion to use in sorting people into levels of political sophistication. All we really learn is that only around 6 percent of people use the words "liberal" and "conservative" when speaking about the parties.
Level B: Group Benefits
A more concrete way of thinking about politics; "ideology by proxy." We view candidates or parties as friendly or hostile to people like us (our group). We worry little about "long-range plans for social betterment" (234), and instead think about whether a party/candidate is "for" my group: farmers, the working class, the poor, etc.
Level C: The "Goodness" and the "Badness" of the Times
Unlike Level D, these people do make some reference, "however nebulous or fragmentary," to some public policy issue (240). And unlike Level B, these people don't really perceive group interests. They form opinions based on whether times are good. The incumbents are doing well if my family is doing well. The incumbents are doing poorly if the war is hurting our country.
Level D: Absence of Issue Content
Although these people fail to comment on anything related to a current political debate, they made up 17 percent of the voters sampled in 1956. They tend to overlook the policy differences between the parties and worry instead about the candidates' personal characteristics ("their popularity, their sincerity, their religious practices, or home life", pg 244).
- Example: One interviewee, when asked what she likes about the Democrats, responds only, "I'm a Democrat, that's all I know" (p 247). What do you like about Stevenson? "Stevenson is a good Democrat."
The following summaries link (or linked) to this one: