This study is provided by ICPSR. ICPSR provides leadership and training in data access, curation, and methods of analysis for a diverse and expanding social science research community.

National Election Pool General Election Exit Polls, 2008 (ICPSR 28123) RSS

Principal Investigator(s):


Election data for 50 states and the District of Columbia were collected through interviews conducted with voters as they left their polling places on election day, November 4, 2008. Part 1, National Data, contains data collected from a national sample. National sample respondents were asked a series of questions about their electoral choices, the issues surrounding the elections, and the factors that influenced their decisions. Questions focused on the direction of the country, national security, terrorism, the war in Iraq, the state and future of the nation's economy, gay marriage, and the George W. Bush presidency. Demographic variables of national respondents cover age, race, gender, Hispanic descent, sexual orientation, age of children in household, marital status, political party, political orientation, employment status, education, religion, sexual orientation, and family income.

Parts 2-52 contain data collected from each state and District of Columbia surveys. Respondents were asked for their opinions of Republican presidential candidate Senator John McCain, Democratic presidential candidate Senator Barack Obama, and the United States Congress, as well as for their vote choices in the relevant gubernatorial, senatorial, and congressional elections. Those queried were also asked their opinions of the candidates' spouses, Cindy McCain and Michelle Obama. Demographic variables of individual state respondents cover age, race, gender, education, voter participation history, political party, political orientation, sexual orientation, and family income.

Telephone interviews were the only type of interview conducted in Colorado, Oregon, and Washington. Telephone interviews were also used to poll absentee voters in Alaska, Arkansas, Arizona, California, Florida, Georgia, Iowa, Michigan, Montana, New Mexico, Nevada, North Carolina, Ohio, Tennessee, and Texas.

Access Notes

Study Description


National Election Pool, Edison Media Research, and Mitofsky International. National Election Pool General Election Exit Polls, 2008. ICPSR28123-v2. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor], 2013-08-26. doi:10.3886/ICPSR28123.v2

Persistent URL:

Export Citation:

  • RIS (generic format for RefWorks, EndNote, etc.)
  • EndNote XML (EndNote X4.0.1 or higher)

Scope of Study

Subject Terms:   Bush Administration (George W., 2001-2009), candidates, congressional elections, election forecasting, exit polls, gubernatorial elections, Hispanic or Latino origins, Iraq War, McCain, John, national economy, national elections, national security, Obama, Barack, Obama, Michelle, presidential elections, terrorism, voter preferences, voter turnout, voters, voting behavior, voting precincts

Geographic Coverage:   Alabama, Alaska, Arizona, Arkansas, California, Colorado, Connecticut, Delaware, District of Columbia, Florida, Georgia, Hawaii, Idaho, Illinois, Indiana, Iowa, Kansas, Kentucky, Louisiana, Maine, Maryland, Massachusetts, Michigan, Minnesota, Mississippi, Missouri, Montana, Nebraska, Nevada, New Hampshire, New Jersey, New Mexico, New York (state), North Carolina, North Dakota, Ohio, Oklahoma, Oregon, Pennsylvania, Rhode Island, South Carolina, South Dakota, Tennessee, Texas, United States, Utah, Vermont, Virginia, Washington, West Virginia, Wisconsin, Wyoming

Time Period:  

  • 2008

Date of Collection:  

  • 2008-11-04

Unit of Observation:   individual

Universe:   Voters casting a ballot in the 2008 United States general election.

Data Types:   event/transaction data, survey data

Data Collection Notes:

In eight states, questionnaires in Spanish as well as in English were available in every precinct in which Hispanics represented 20 percent or more of the total population. Spanish language questionnaires were available in some precincts in Arizona, California, Florida, Illinois, Nevada, New Mexico, New York, and Texas. In the data sets for these eight states and in the national dataset, the variable SPANISHQ (column 78) indicates the language of the questionnaire the respondent completed.

In order to standardize the questions on race and ethnic background, the 2006 and 2008 exit polls asked these questions the same way in virtually every state. In most states, this represents a change from how these questions were handled in previous years. Please see section H. of the user guide.

Note that in 2008, all respondents in California, Illinois, Maryland, Massachusetts, and New York answered one version of the national questionnaires.


Study Design:  

National: A sample of exit poll precincts was drawn from each of the 50 states and the District of Columbia. A subsample of these precincts was selected to form the national sample. The national survey was administered in a total of 300 sample exit poll precincts. Respondents in the national precincts were given one of four versions of the national questionnaire. The four versions were interleaved on pads that were handed out to respondents. Responses to the four versions are combined into one dataset. All versions have questions in common as well as questions unique to each version.

State Data: As mentioned above, a sample of exit poll precincts was drawn in each state. A subsample of these precincts was selected to form the national sample. The remaining precincts in each state made up the state sample and were given questionnaires specific to that state. Because the national questionnaire has several items in common with the state questionnaire, national respondents are included in the state exit poll dataset for these common questions. To determine which questions are on the national questionnaire, simply crosstab each question by QTYPE (found in column 13 of the ascii dataset), indicating whether the respondent completed the state or national survey. If the corresponding item did not appear on that respondent?s version of the questionnaire, it was coded as system missing in the SPSS file and will appear as a blank in the ascii dataset. Remember, as noted above, some questions on the national survey appear on multiple versions of the national and some do not. Note that in 2008 all respondents in California, Illinois, Maryland, Massachusetts, and New York answered one version of the national questionnaires.

Sample:   The samples were selected in two stages. First, a probability sample of voting precincts within each state was selected that represents the different geographic areas across the state and the partisan make-up of the state. Precincts were selected with a probability proportionate to the number of voters in each precinct. There is one exception. In some states, precincts that had large minority populations were sampled at a higher rate than other precincts. The sample weighting (described below) adjusts the representation of these precincts to their correct share of the total vote. Within each precinct, voters were sampled systematically throughout the voting day at a rate that gave all voters in a precinct the same chance of being interviewed.

Weight:   The exit poll results are weighted to reflect the complexity of the sampling design. That is, the weighting takes into account the different probabilities of selecting a precinct and of selecting a voter within each precinct. For example, minority precincts that were selected at a higher rate receive a smaller weight than other precincts of the same size. An adjustment is made for voters who were missed or refused to be interviewed, which is based on their observed age, race, and gender. Respondents are also weighted based upon the size and distribution of the final tabulated vote within geographic regions of the state or of the nation.

Mode of Data Collection:   face-to-face interview, telephone interview


Original ICPSR Release:  

Version History:

  • 2013-08-26 Parts 14 through 16 have been updated with the appropriate state filesets per user request.

Found a problem? Use our Report Problem form to let us know.