I believe that President Bush's attitude toward polls must have been about the same as it was toward speeches ... that it was not legitimate, that it was not real leadership, that it was somehow phony and artificial.
--Bush press secretary Marlin Fitzwater (1)
In 1988, George H. W. Bush became the first sitting vice president since Martin Van Buren (1837) to be elected president of the United States. This historic victory ushered in a man determined to create his legacy in a vastly different manner than the president under whom he served. As Bush's former deputy chief of staff Andy Card said, "You have to understand, President Bush didn't want to do things the way President Reagan did. He didn't want to be dictated by a communications strategy." (2) Bush's effort to distance himself was so complete that on entering the White House, all photographs of Ronald Reagan were removed from the West Wing (Boyd and Molotsky 1989). Interior decorating choices notwithstanding, Bush's decision to differentiate his style of leadership was perhaps most evident in the way he chose to integrate--and not integrate--his private polling apparatus into the process of rhetorical invention. Card says that President Bush had a name for polls: he called them "reflections of yesterday" because he believed "that a poll was only as good as the day it was taken." (3) However, Bush's attitude toward polls and his use of them represent but one link in an otherwise long chain of relationships between presidents and public opinion research.
That presidents and polls enjoy an inextricable relationship should come as no surprise (Geer 1996; Herbst 1993; Hogan and Smith 1991; Jacobs and Shapiro 1994, 1995, 1996, 1999, 2000, 2001; Lavrakas, Taugott, and Miller 1995; Ratzan 1989). In 1939, Franklin Delano Roosevelt requested that Eugene Meyer, publisher of the Washington Post, ask George Gallup to conduct a poll to reveal Americans' views toward U.S. involvement in the war in Europe and report his findings back to the White House (Steele 1974; Sussman 1988). Since then, presidents of both major parties have consulted polls in varying degrees and for differing purposes. Harry Truman found little use for polls (Heith 1998). Eisenhower was curious about survey findings but did not allocate full-time resources to conduct opinion research in the White House (Jacobs and Shapiro 1995). Lawrence R. Jacobs and Robert Y. Shapiro (1995) note that "starting with John Kennedy, however, the White House's sensitivity to public opinion became an enduring institutional characteristic of the modern presidency" (p. 165). Lyndon Johnson attacked and leaked polls and also used personal relationships with pollsters to influence their findings (Altschuler 1986, 1990). "In the Nixon administration," according to Diane Heith (1998), "public opinion data was both tightly controlled and publicly displayed" (p. 170). Ford chose to use it privately rather than publicly. Jimmy Carter personally parceled out polling information to his advisers who frequently misread public opinion findings and thus often misused data in decision making (Heith 1998). I have shown how presidential pollster Richard Wirthlin, working on behalf of Ronald Reagan, created a complex computer system called the Political Information System as well as used "PulseLines" (also known as dial groups) extensively to pinpoint the most favorable lexical constructions in the president's rhetoric and then "recycled" them in subsequent speeches throughout his presidency (Hall 2002). George H. W. Bush was thought not to use polls nearly as much as his predecessor, while Jacobs and Shapiro have shown how Bill Clinton created "crafted talk"--a strategy whereby decision makers "use research on public opinion to pinpoint the most alluring words, symbols, and arguments in an attempt to move public opinion to support their desired policies" (Jacobs and Shapiro 2000, xv)--using his pollsters' private research (Hall and King 2002). …