Voters, Political Parties, and Elections

Bill Miller

Purpose:

Introduction to the vocabulary of voting behavior and elections

Introduction to available sources of information about voting and elections

1. Overview: elections <parties <voters

American election results are determined by our two political parties

Party performance is determined by voting behavior

We will look at voting behavior, party performance, and elections in turn

2. Voting Behavior of Different Groups

Polling/opinion research, survey research: its purpose, its methods (market research, politics)

Different polls: VEP, VAP, Exit, self-reported data, objective data

Cleavages => Groups

Factors: sex, age, race, region, education, income, and more

Predicting group behavior

3. Political Parties as Coalitions of Different Groups

Alignment of coalitions, realignment, and de-alignment

Democratic Southern Coalition, etc.

Voter Values Determine Political Affiliation, by Tom Edsall, 2001

America's Class Politics Turned Upside Down, by Roge Karma, 2024

Focus on White Women Voters by Virginia Kruta, 2024

4. Patterns in American National Elections

Presidential elections: Electoral College

Congressional elections: Incumbency Re-election Rates

5. Population (in millions) on October 6 of each following year:

2024: 337.2M (U.S. Census Population Clock)

2022: 333.8M

2020: 331.8M

2016: 327.2M (Macrotrends.net)

2012: 316.7M

2008: 305.7M

White: 222.2M or 65.9% Black: 42.2M or 12.5%

Asian: 19.6M or 5.8% Hispanic: 61.3M or 19% (2021)

Voting Behavior of Different Groups

Polling/opinion research, survey research: purpose, methods (market research, politics)

Different polls: VEP, VAP, EXIT POLLS

Cleavages=Groups

Factors: sex, age, race, region, education, income

Predicting groups behavior

Polling Methods

Purpose of polling: to find significant voting similarities among members of definable groups of people, and then to target those groups with a persuasive message. �Similar� here means �predictable voting behavior.�

Defining groups by finding divisions or cleavages within the population, then� finding the cross-cutting cleavages among those groups.

Examples:

Sex: Do female voters behave differently from male voters? If so, do all women voters behave similarly? Or, perhaps, do white female voters behave differently from black female voters?

Race: Do Hispanic or Asian voters behave differently from white or black voters? Do all Hispanic or Asian voters behave similarly?

Eg., Cuban-, Puerto Rican-, Central American-Hispanic voters.

Chinese-, Japanese-, Vietnamese-Asian voters.

In this way significant voting groups are discovered.

Pollsters then establish a universe or representative sample made up of the different groups and poll representative numbers of each group in that population. (Special secret sauce!)

Example (from above): A universe of 1,000 people to be polled (cheaper and quicker than a universe of 2,000 but a greater margin of error!): If Blacks make up 20% of the Voting Eligible Population (or �Citizen Population� in Census Bureau statistics), then 200 of those interviewed should be Blacks. If 60% of the Black voters are women, then 120 of the 200 blacks to be interviewed should be women; and so on dividing up the arbitrary universe of 1,000 interviewees into a proportionate number of representatives of each significant group.

Exit polls:

� ask voters leaving the polls a series of questions

� use the groups defined in survey research

CNN 2020 Exit Poll:

https://www.cnn.com/election/2020/exit-polls/president/national-results

Problems: participant cooperation, participant truthfulness. The social acceptability bias aka the Bradley effect.

Voter Participation

Self-reported data v. External, objective, statistics

VAP (Voting Age Population) v. VEP (Voting Eligible Population)

Census Bureau: https://www.census.gov/topics/public-sector/voting/data/tables.html

U.S. Elections Project: https://www.electproject.org/home

Voter Partisanship