Exit Polls: Surveying the American Electorate, 1972-2010
Publication Year: 2012
Subject: Elections & Political Campaigns
Every two years, exit polls become the most widely analyzed, written about, and discussed data-set in the United States. Although exit polls are known for their use in predicting elections, they are in fact the best tool for explaining election results. Exit polls are taken from actual voters, whereas pre-election polls that tally people’s intended votes tend to overstate the number of people who will actually go to the polls. Exit Polls: Surveying the American Electorate is a groundbreaking reference work that explores for the first time the trends in longitudinal variables asked in the national Election Day exit polls from their beginning in 1972 to the present. The book documents comparable survey items that have appeared in multiple exit polls over time. Authors ...
- Front Matter
Copyright by Sage Publications, Inc.
The genesis of this project was election night 2008, when Barack Obama defeated John McCain to become our nation's first African American president. We were poring over the exit poll results, trying to put the numbers in some sort of historical context. We knew Obama had done exceedingly well with a wide range of groups, but just how well was open to question. Did African Americans make up a disproportionate number of voters, or did Hispanics exert a far greater impact than they had in the past? Did Obama attract as much support from Catholics as Ronald Reagan did, or as many female voters as Bill Clinton did? Did he reverse Democratic fortunes among white men, rural residents, or devout Christians?
As we tried to find answers to these questions and many more, we realized quickly that there simply was no easily accessible resource that allowed us to put national exit poll results in historical perspective. Previous polls had been catalogued in archives around the country, but each was stored in a unique format that was not conducive to easy compilation or manipulation. Consequently, the results of the exit polls could not be retrieved and compared to those of others undertaken at different points in time.
Considering the sheer number of national exit polls that had been conducted through the years, we realized that this was not an unfortunate oversight, but a glaring omission in the information age. Exit polls had been administered to voters departing the polling booth in every national election conducted since 1972. Data on these polls could be used to see how voters compared not only to those from an election or two earlier but also to those from generations before. Compiling and analyzing these trends seemed likely to yield unique and important insights on electoral politics.
During the past two years, we have undertaken the painstaking process of developing a single data set to identify trends and relationships in the exit polls. We developed an original coding scheme, applied it to each data file, and merged the results together into a user-friendly spreadsheet. We then used this unique data set to explore the composition and electoral choices of American voters.
This is the product of those efforts. In this volume, we present and discuss the electoral behavior of a wide range of groups in the active electorate. We reveal a number of interesting patterns, some of which will be familiar to readers and some of which have remained hidden until now. Together, they provide a compelling portrait of American voters over the past four decades.Acknowledgments
We take full responsibility for the numbers and interpretations that follow. Nonetheless, one does not complete a project of this size and scope without the help of a great many others.
Professor Monika McDermott sparked our intellectual interest in voter feedback years earlier. She introduced us to the ins and outs of exit polling, making us aware of endless research possibilities.
The polling group at CBS News provided a venue through which we could dive into the exit poll data. Director of Surveys Kathy Frankovic offered us the opportunity to analyze exit polls for the network, ultimately planting the seed for this volume. During its preparation, she also read and critiqued portions of the manuscript, correcting misunderstandings and identifying oversights. Her colleagues at CBS News—Sarah Dutton and Kevin Hechtkopf—provided valuable feedback on pieces we wrote for the network, prompting new questions that were pursued along the way. Fellow exit poll analysts Professor David Jones and Doug Schwartz, the director of the Quinnipiac University Poll, played the foil to many analyses and interpretations during the past several elections.
Several prominent survey methodologists imparted their knowledge about the intricacies of exit polling to us. Professor Stanley Feldman passed along valuable insights about the polling process through the years. Susan Pinkus, director of the Los Angeles Times poll, and Murray Edelman, former editorial director of the Voter News Surveys, provided background about the origin and evolution of the exit polls.
Our respective universities provided us with exceptional academic homes to foment and refine our ideas. At the University of Connecticut, Professors Lyle Scruggs and Jeffrey Ladewig were great sounding boards, offering critiques and insights that only improved the manuscript. Professors Oksan Bayulgen, Kristin Kelly, Peter Kingstone, Matthew Singer, and Heather Turcotte provided alternative perspectives, bringing to bear fields outside of electoral politics and encouraging us to consider less obvious interpretations. At the University of Rhode Island, Marc Hutchison provided valuable insights from the perspective of a comparative behavioralist. Shanna Pearson-Merkowitz pushed us to consider the geographic and state-level implications of our decisions. Kristin Johnson fortunately provided us elegant solutions to seemingly unsolvable, large data set problems. And Gerry Tyler was a strong supporter of the original idea, offering sage comments on the prospectus.
We are indebted to our editorial team at CQ Press. Doug Goldenberg-Hart extolled the virtues of the project before we were even fully aware of them. January Layman-Wood shepherded the project to completion, making sure we met our deadlines and remained true to our original vision. John Martino painstakingly edited draft after draft, dramatically improving its readability in the process. Amy Marks was an outstanding and patient copy editor.
We owe considerable gratitude to our family and friends who provided unconditional support during the two years it took us to put this book together. Robert Chisholm, Everett Finklestein, Scott Herlihy, Gus Monteiro, Neath Pal, and Geoffrey Schnirman provided a layman's perspective each week at the poker table. Sean Krueger served as unknowing timekeeper, sowing [Page xvii]and reaping gargantuan gourds in his pumpkin patch to a cycle by which we monitored our progress. Our children, Alexander Best, Madeleine Best, and Melody Krueger, kept our priorities in perspective and recharged our batteries time and time again. Finally, our lovely spouses, Niki Best and Jennifer Krueger, to whom this book is dedicated, shouldered disproportionate burdens, eased our frustrations, and provided encouragement without which this book would never have been completed.
About the Authors