State Polls Research Articles

Using state polls to forecast presidential election outcomes in the American states

This paper presents a simple forecasting model for state-level presidential outcomes, based on statewide preference polls and a lagged vote variable. The analysis illustrates two important points. First, the candidate who is leading in a state in September usually goes on to win that state in the November election. Second, the combination of pre-election preference polls and a lagged dependent variable generates highly accurate estimates of presidential election outcomes in the states. The limits of using statewide preference polls are also discussed.

National Network of State Polls (NNSP)

National Network of State Polls (NNSP)

Gubernatorial popularity in three states: A preliminary model

This article develops and tests a model of gubernatorial popularity. Using data from three state polls in California, Iowa and Minnesota, we examine the relative importance of state-level and national-level political and economic developments for approval ratings of governors in these states. We also investigate the degree to which governors are held accountable for the performance of their party when it controls the White House. We demonstrate that the popularity of governors fluctuates directly with the president's popularity and the performance of the economy under the president's stewardship. The evidence of the relative impact of state and national level economic indicators on gubernatorial approval is mixed and in California only there is evidence of a “Honeymoon” effect.

Measuring Public Attitudes on Abortion: Methodological and Substantive Considerations

Data from a 1989 CBS News/New York Times survey are used to examine the effect that the framing of questions on abortion has on estimates of what proportions of the population support various legal positions. The nationwide data and results from six state polls show that general questions with only two or three options overestimate the proportions of respondents who either favor a ban on all abortion or who would allow abortion under all circumstances. Questions that pose specific circumstances result in movement of respondents out of extreme categories and into more moderate ones. Even respondents who indicate they would favor abortion in all specific circumstances and those who favor abortion in none are likely to moderate their views when asked if they support restrictions that have been proposed in a number of states.

Creating an Index of Public Opinion Questions

Until now researchers interested in knowing what public opinion data existed on a particular topic had a problem. There was no single source of information about survey work being done in this country. Now for the first time the American Public Opinion Index collects all the questions asked of the general public in national, state and local polls and arranges them in a topical index so that researchers will have a comprehensive and thoroughly cross listed reference for finding polling information that might be useful for their work.

State Polls: Promising Data Sources for Political Research

This research note suggests that the image of statewide polls as erratic bases for election predictions belies their reliability and usefulness as sources of information for basic scientific research. Data on partisanship from three different polling organizations, over a seven-year period, for fourteen different states, are presented to indicate that there is minimal interorganization variation among the three sources of data, a remarkable stability of sample estimates of partisanship within states through time, and a strong correlation between the state poll data and external validating indicators.