Glossary

Terms of art, each defined once in plain language. In the chapters, a marked term shows its definition on hover or focus.

Affective polarization: The gap between how warmly people feel toward their own party and how coldly they feel toward the other party. It measures feeling, not disagreement on issues.
Area Deprivation Index: A ranking of neighborhoods built from income, education, employment and housing measures in the American Community Survey. Higher ranks mean more deprived neighborhoods.
Block group: A subdivision of a census tract, usually home to between 600 and 3,000 people. It is the smallest area for which the Census Bureau publishes sample-based estimates.
Census tract: A Census Bureau neighborhood area, usually home to between 1,200 and 8,000 people. Tract boundaries were redrawn after the 2010 census and again after 2020.
CFscore: A left-right position estimated from whom a donor gives to, or from who gives to a candidate. Negative values are left and positive values are right, and a donor keeps one score across all their giving.
Clustered standard errors: Standard errors that allow answers within the same group, such as the same county or the same week, to be related. They are wider than ordinary errors when members of a group resemble each other.
Cohort: People born in the same period, such as the same decade. Comparing cohorts shows whether a difference comes from when people were born, as distinct from their age or the year they were surveyed.
Confidence interval: A range around an estimate that would contain the true value in 95% of repeated samples if the sampling model held. The book reports 95% intervals unless it says otherwise.
Correlation: A number from minus 1 to plus 1 for how closely two measures move together in a straight-line pattern. Zero means no straight-line pattern; a correlation does not say that one measure causes the other.
County two-party margin: The Democratic vote minus the Republican vote, divided by the sum of the two, so a positive value is a Democratic lead. Votes for other candidates are left out.
DW-NOMINATE: A left-right score for each member of Congress estimated from roll-call votes. Members who vote alike receive similar scores, and the first dimension mostly separates the parties.
Ecological inference: Drawing a conclusion about individuals from data on groups or places, such as counties. A pattern across counties can differ from the pattern among the people who live in them.
Effective sample size: The number of equally weighted respondents that would give the same precision as the weighted sample. It is smaller than the raw count when the weights vary a lot, and it sets how wide the confidence interval is.
EJSCREEN: The Environmental Protection Agency's environmental justice screening tool. It combines pollution and hazard measures, such as fine particles and diesel exhaust, with the demographics of each neighborhood.
Feeling thermometer: A 0 to 100 rating of a group or person, where 0 is the coldest feeling and 100 the warmest. The American National Election Studies ask respondents to rate parties, candidates and social groups this way.
Fixed effects: Regression terms that give each group, such as each country or each year, its own baseline. The other coefficients then compare respondents within the same group.
HC3 robust standard errors: A way of computing standard errors that stays valid when the spread of the outcome differs across respondents, with a correction for small samples. The book uses it for regressions that have no clusters.
Human Development Index: A country score from 0 to 1 that combines life expectancy, years of schooling and income per person, published by the United Nations Development Programme. Higher scores mean higher human development.
Ideological constraint: The degree to which a person's positions on separate issues line up along one left-right dimension. Constraint is high when knowing someone's view on one issue predicts their view on the others.
Interrupted time series: A design that fits the trend before a dated event and compares it with the trend after, to see whether the level or slope breaks at the event. It shows a break in the series; it does not show what caused the break.
Left-right self-placement: A survey question that asks people to place themselves on a scale from left to right. The scale is the same for everyone, but the meaning attached to left and right differs across people, countries and years.
Linear probability model: A regression of a yes-or-no outcome on predictors, fitted by least squares so that each coefficient is a change in the probability of yes. The book weights it and reports robust standard errors.
Loading: The correlation, inside one country, between a survey scale such as left-right self-placement and an attitude. A loading of zero means the attitude says nothing about where people place themselves, and a positive loading means people who hold it place themselves further right.
Matched counties: The counties that report a vote in both elections being compared. Changes are computed on them so that a county created, merged or renamed between the two elections does not distort the comparison.
National swing: The change between two elections in the Democratic minus Republican margin for the country as a whole. A county's own swing is compared with it to show whether the county moved with the nation or against it.
Partisan leaner: A respondent who first says they are independent but, when asked, says they lean toward one party. Most figures group leaners with the party they lean toward, and each chapter says which grouping it uses.
Party sorting: The alignment of party with other traits, such as ideology, religion, race and place, so that people who share a party also tend to share those traits.
Percentage point: The unit for the difference between two percentages. A change from one share to another is stated in points, not in percent, so that absolute and relative change are not confused.
Postmaterialism: Ronald Inglehart's term for a priority on self-expression, participation and quality of life over economic and physical security. Survey measures ask respondents to rank goals such as order, stable prices and free speech.
Precinct: The smallest area for which election officials report votes, usually the voters assigned to one polling place. Precinct boundaries change between elections, so the same precinct cannot be followed over time.
Racial resentment: A survey scale, developed by Donald Kinder and Lynn Sanders, that measures moral judgments about the work ethic and deservingness of Black Americans. Higher scores mean more resentment.
Rural-Urban Continuum Code: A USDA scheme that sorts counties into nine classes, from large metropolitan areas to remote rural ones, by population size and nearness to a metropolitan area. The 2013 and 2023 editions classify some counties differently.
Schwartz basic values: Ten broad motivations, such as security, tradition, achievement and universalism, that Shalom Schwartz's theory arranges in a circle. Neighboring values are compatible and opposite values conflict.
Standard deviation: A measure of how far values typically fall from their average. A larger standard deviation means the values are more spread out.
Standardized coefficient: A regression coefficient measured in standard deviations of the variables involved. It lets effects on differently scaled variables be compared.
Survey mode: How respondents were interviewed: in person, by phone, on paper or online. Mode can change answers, so a change in mode can look like a change in opinion.
Survey weights: Numbers that scale each respondent so the sample matches the population on known traits such as age, race, education and region. A weighted average counts a respondent with a large weight for more than one with a small weight.
Swing: The change in a place's two-party margin between two elections. In this book a positive swing is a move toward the Democrats.
ZCTA: ZIP Code Tabulation Area: the Census Bureau's area approximation of a ZIP code, built from the census blocks that fall inside it. It is the one neighborhood geography that lines up between the 2016 and 2020 vote estimates.