Ours is a government by opinion, and the press is a most important part of that process.
Analyses of presidential polling during the Kennedy, Johnson, and Nixon administrations (Sudman 1982; Jacobs 1992, 1993; Jacobs and Shapiro 1994, 1995; Geer 1996; Heith 1998a, 1998b) and of political polling more generally (Sabato 1981; Herbst 1993; Eisinger and Brown 1998) reveal that the use of polls among politicians has become institutionalized. Recent scrutiny by journalists and former White House advisers (e.g., Drew 1997, 136; Morris 1997; Milbank 1999) of President Bill Clinton's polling practices has provided additional information about the frequency of presidential polling, but little is known about how presidents gauged public opinion before modern polls were developed, specifically the methods by which President Herbert Hoover's administration sought to assess citizens' attitudes.
Revealing how the Hoover administration measured public opinion provides an important context for better understanding the institutionalization of presidential polling for two reasons. First, Hoover's successor, Franklin Roosevelt, initiated private presidential polling as a means to gauge public opinion. Appreciating how presidents measured attitudes prior to polls will help explain how the institutionalization of polls germinated. Hoover, while not equipped with the surveying techniques (specifically random and quota sampling) of his successors, sought to gauge public opinion independent of his party and Congress because he did not trust these institutions to provide him with reliable assessments of citizens' attitudes. The means by which Hoover's advisers gauged public opinion therefore depended on the political interplay between the executive branch, the Republican Party, Congress, and the media. Second, the relevant scholarship demands greater analysis of the interplay between gauging public opinion and the institutions that assess citizens' attitudes; it also necessitates a historical appreciation of political polls, especially as some theories allege that polls have deleterious and pernicious effects (e.g., Ginsberg 1986; Hitchens 1992). These works beg historical comparisons, for they suggest that political polling is a recent phenomenon that mysteriously matured without any infancy or adolescence. The origin of presidential polling remains an unsolved empirical puzzle, in large part because most of the above cited academic research apropos presidents and public opinion has focused on polling by recent presidents (post-John Kennedy), not the political precursors and dynamics that have made private polls a viable method of gauging public opinion.
While Trefousse (1975) and Hilderbrand (1981) are exceptions in this regard, the discussion of how presidents use polls has concentrated on the post-Kennedy presidency. Heith (1998b) and Jacobs and Shapiro (1995) provide the richest theories explaining recent presidential polls. Collectively, they argue that recent presidents (Kennedy, Lyndon Johnson, Richard Nixon, Gerald Ford, Jimmy Carter, and Ronald Reagan) institutionalized the "polling apparatus" of the executive branch by noting how many and which staffers were involved in exchanging memoranda associated with presidential polling. According to Heith (1998b), "Public opinion represents a valuable strategic tool," and "institutional patterns" indicate that "polling aids efforts to sell both the president and his policies" (p. 187). Yet, Heith's research commences with the Nixon presidency. As a result, Heith argued that poll usage from 1969 to 1988 represented the "awkward adolescence of polling" (p. 187). This article attempts to modify this claim. Prepolling presidents were also interested in monitoring public opinion; their attempts to locate, gauge, or manipulate public opinion might explain subsequent presidents' use of polls to measure the public's mood.
This article proceeds as follows. …