No abstract

Declining telephone response rates have forced several transformations in survey methodology, including cell phone supplements, nonprobability sampling, and increased reliance on model-based inferences. At the same time, advances in statistical methods and vast amounts of new data sources suggest that new methods can combat some of these problems. We focus on one type of data source—voter registration databases—and show how they can improve inferences from political surveys. These databases allow survey methodologists to leverage political variables, such as party registration and past voting behavior, at a large scale and free of overreporting bias or endogeneity between survey responses. We develop a general process to take advantage of this data, which is illustrated through an example where we use multilevel regression and poststratification to produce vote choice estimates for the 2012 presidential election, projecting those estimates to 195 million registered voters in a postelection context. Our inferences are stable and reasonable down to demographic subgroups within small geographies and even down to the county or congressional district level. They can be used to supplement exit polls, which have become increasingly problematic and are not available in all geographies. We discuss problems, limitations, and open areas of research.

We use multilevel modeling to estimate support for health-care reform by age, income, and state. Opposition to reform is concentrated among higher-income voters and those over 65. Attitudes do not vary much by state. Unfortunately, our poll data only go to 2004, but we suspect that much can be learned from the relative positions of different demographic groups and different states, despite swings in national opinion. We speculate on the political implications of these findings.

Research suggests that partisans are increasingly avoiding members of the other party—in their choice of neighborhood, social network, even their spouse. Leveraging a national database of voter registration records, we analyze 18 million households in the U.S. We find that three in ten married couples have mismatched party affiliations. We observe the relationship between inter-party marriage and gender, age, and geography. We discuss how the findings bear on key questions of political behavior in the US. Then, we test whether mixed-partisan couples participate less actively in politics. We find that voter turnout is correlated with the party of one’s spouse. A partisan who is married to a co-partisan is more likely to vote. This phenomenon is especially pronounced for partisans in closed primaries, elections in which non-partisan registered spouses are ineligible to participate.

We build a model of American presidential voting in which the cumulative impression left by political events determines the preferences of voters. The impression varies by voter, depending on their age at the time the events took place. We use the Gallup presidential approval‐rating time series to reflect the major events that influence voter preferences, with the most influential occurring during a voter's teenage and early adult years. Our fitted model is predictive, explaining more than 80% of the variation in voting trends over the last half‐century. It is also interpretable, dividing voters into five meaningful generations: New Deal Democrats, Eisenhower Republicans, 1960s Liberals, Reagan Conservatives, and Millennials. We present each generation in context of the political events that shaped its preferences, beginning in 1940 and ending with the 2016 election.