I considered many variables for measuring the extent of public familiarity with presidents’ constitutional legacies. Some were unwieldy, such as street names. Others were measurable, such as the names of schools or colleges, but they were largely geographically bound. People know the names of the schools or colleges named after presidents near where they live but rarely know the names of ones located outside their daily routines. Other variables, such as postage stamps, were measurable but had little arguable pertinence. While there are polling data on public familiarity with American leaders, they are overinclusive, since they indicate that most Americans are unfamiliar with the names of most presidents, much less their constitutional legacies.
I settled on seven variables, which were measurable and pertinent to public familiarity with presidents’ constitutional legacies. The first was how long ago a president served in office. The variable reflects the number of years since a president entered office. This is pertinent because changes in technology have affected the quality and availability of the means for memorializing a president’s public statements. Today, everything a president says or does is preserved for posterity, usually in many forms, which are accessible to almost everyone. Moreover, formal archiving is a relatively recent phenomenon, so that the more recent a presidency the more voluminous the material is likely to be. Thus, more recent presidents will be more likely to be remembered by the general public, all other things being equal (which of course may not be).
Second, the frequency of a president’s appearance in the most popular American history textbooks used in middle and high schools is important because middle and high school history classes are among the first and arguably most critical places where young people learn about presidents’ achievements and legacies. It follows that familiarity with certain presidents is developed during childhood and adolescence. Therefore, the presidents discussed the most in middle and high school textbooks are likely to achieve greater familiarity with the average American during childhood.
(p.246) I coded the second variable as follows: I gave a president a mark for each sentence in each of a representative sampling of textbooks that discusses him by name or obvious pronoun. I used a cross section of textbooks from different authors and publishers and designed for different groups that I selected from the American Textbook Council’s list of the most commonly used textbooks within American middle and high school curricula. The textbooks surveyed were Joyce Appleby, Alan Brinkley, and James Macpherson, American Journey (McGraw Hill, 2008); Robert Dallek, Jesus Garcia, Donna Ogle, and C. Frederick Risinger, American History (McDougal Little, 2007); James West Davidson, American History of Our Nation: Beginnings through 1877 (Prentice Hall, 2007); and Daniel Boorstin, A History of the United States (Prentice Hall, 1997).
Third, I calculated the average number of books about particular presidents in twelve university libraries. I substantiated this variable in two ways. First, the number of volumes represents the market for each president; less familiar presidents are unlikely to have a significant market share. The market is more likely to demand the stories of renowned presidents or even those popular through vilification. Second, it is more difficult to be familiar with something when there is less information available about it. Presidential familiarity can be captured by the sheer extent of biographies. Thus, I gave one count to each volume in the nation’s ten largest university libraries—Harvard, Yale, University of Illinois–Champaign, University of California–Berkeley, University of Texas–Austin, Stanford, University of Michigan, Columbia, UCLA, and University of Wisconsin–Madison. I added two other libraries, Duke and UNC–Chapel Hill, to ensure greater geographic diversity.
The fourth variable, presidential libraries, was a dummy variable. Presidential libraries foster and promote presidential legacies. Thus, a president who has a presidential library is more likely to retain familiarity because an organization is devoted to educating the public about his stature. Because presidential libraries are a relatively modern creation, they are imperfect. While every president has an official historic site, there are only nineteen official presidential libraries, including John Quincy Adams, Hayes, Lincoln, and McKinley. The absence of a library likely inhibits learning about a presidency and reinforces ignorance about it.
Fifth, I measured presidential greatness. The opposite of greatness is insignificance. The further a president is from greatness the more insignificant his presidency. I derive my ranking of presidential greatness from William J. Ridings and Stuart B. McIver’s Rating the Presidents (Carol Publishing Group, 1997), which polled more than 700 presidential experts to rank presidents.
The sixth variable was personality. The principal work on which I modeled my testing of personality was Dean Keith Simonton, “Presidential Style: Personality, Biography, and Performance,” 55 Journal of Personality and Social Psychology 928–36 (1986). I used two of Simonton’s variables, creativity and charisma, because they are likely to influence the probability that a president will be remembered. Charisma makes a president appealing to the public and the press, who are instrumental to maintaining his legacy and image. Creativity reflects a president’s ability to initiate new legislation or (p.247) craft creative solutions to difficult problems. I calculated one personality variable by adding both measures together.
My final variable was rhetoric. I assumed that public speaking is the medium of choice and that people will more likely remember presidents who established bonds of familiarity through their rhetorical skills. I supposed presidents who delivered bland, unimaginative, or dull speeches were more likely to be forgotten.
I considered measuring rhetoric in different ways. For example, I considered but rejected measuring the quantity of a president’s public statements, because this likely has little if any bearing on the words or images that moved people during or after a presidency. Rhetoric is a skill that requires special talent, and thus my objective was to find some indicia of the extent to which a president had it. I also considered but rejected surveying historians because their scores could be based on many other qualities, such as greatness. Nonetheless, one interesting, recent study of the presidency is Robert W. Merry’s Where They Stand: The American Presidents in the Eyes of Voters and Historians (2012), which surveys how historians and others have rated presidents. I found especially interesting how presidents ranked in terms of the numbers of times Merry mentions them in the book; the thirteen presidents who rank the lowest in terms of how often he mentions them are Garfield and William Henry Harrison (twice); Tyler and Arthur (four); Hayes, Fillmore, and Pierce (five); Taylor, Ford, and Benjamin Harrison (six); Van Buren (seven); and Taft and John Quincy Adams (nine). As readers will see, there is a significant degree of overlap between the presidents who are mentioned least in the latter study and the presidents whom I deem to be forgotten.
Consequently, I created a composite variable for rhetoric, using three quantifiable qualities—speech length, imagery, and identification words—that are theoretically indicative of effective rhetoric. For my understanding of these terms, I relied on Ryan Lee Teten’s The Evolutionary Rhetorical Presidency: Tracing the Changes in Presidential Address and Power (2011). Teten demonstrates that speech length and identification words vary by president. The rhetoric literature suggests that less verbose speeches, such as the Gettysburg Address, tend to be more effective. I borrowed data about speech word count from Teten, who tabulated the number of words communicated for each president’s State of the Union addresses. If a president served two terms, I averaged the two.
Teten also provides an analysis of “identification rhetoric.” Through the use of certain terms, such as pronouns, presidents are “able to speak to the people and [to] convince them that he is on their side, on the same page with them, and has the interests of the greater whole in mind” (39). This technique is meant to “build consensus and agreement by creating identification between the citizens of the United States and himself; if they identify with what he says, he will receive greater support and have the ability to proceed further with policy objectives.” I measured this variable as the number of times a word of identification appeared in a State of the Union address, relative to all other words.
Another study, Cynthia G. Emrich, Holly H. Brower, Jack M. Feldman, and Howard Garland, “Imagines in Words: Presidential Rhetoric, Charisma, and Greatness,”46 Administrative Science Quarterly 527–57 (2001), hypothesizes that speech imagery (p.248) indicates effectiveness because it provides a basis for connecting with the audience. The study separately tabulated the frequencies of image-based words and concept-based words in presidential inaugural addresses. Thus, considering the importance of imagery and identification, I constructed a variable that was an average between the imagery score of the landmark and inaugural speeches.
A significant issue with rhetoric is that each measure within the composite excluded some presidents. The latter study omitted Gerald Ford, John Tyler, Andrew Johnson, Millard Fillmore, and Chester A. Arthur because they were never elected to deliver inaugural addresses, while the former study omitted Garfield and William Henry Harrison because of their brief tenures. A missing value would serve to remove these presidents from the composite index. Therefore, I imputed the missing value within the rhetoric variable, using the president’s other qualities. Basically, statistical imputation assumes that there is a relationship among a group of variables. Using how others with a full set of values relate to the missing values, the method assigns a value to the missing ones based upon the studied relationship. I chose this method because it offers a solution that produces conservative estimates that are unlikely to disturb the model’s predictive ability. Imputation provides a theoretical validation for using these new values, as opposed to an arbitrary assignment.
Readers will find two different charts showing the results of my calculations of these variables. The first, figure A.1, measures presidential presence in middle and high school textbooks and university libraries. For each president in this figure, there are two columns; the left column represents volumes in university libraries and the right represents presence in textbooks.
(p.249) The four presidents with the fewest volumes about them in the major university libraries that I surveyed—Arthur, Benjamin Harrison, Fillmore, and Pierce—are in this study. Indeed, almost all the presidents in this book are in the lower half of presidents based on this metric. While Taft and Carter are near the middle of presidents on this scale, the volumes on each of them devote significant attention to their unusually active postpresidential careers. Carter also benefits from the enhanced archival requirements in the 1970s.
The second figure, A.2, shows the composite score for each president on public familiarity. The left or y-axis posits levels of familiarity on a range from 0 to 4, and the right or x-axis lists presidents in decreasing order of their familiarity. Hence, the president with the highest composite score (or most public familiarity) is on the far left of the x-axis, and the president with the lowest on the far right of the x-axis.
Several findings are noteworthy. First, the data suggest that the public is unfamiliar with most presidents, though Carter is the only president in this study ranking in the top half of presidential familiarity.
Second, the presidents in this book do not comprise the bottom thirteen based on this data. More than a handful of presidents appear to have less public familiarity than the presidents in this study, though I did not include these other presidents for several reasons. Some of the latter have higher scores than the presidents I included on certain factors I consider to be important (such as the frequency of their being mentioned in history textbooks). Several of the other presidents I did not include are marked by dramatic events that have made them more distinctive than several presidents I chose for this study. For instance, Monroe is well known for the Monroe Doctrine, while Buchanan often gets attention for being rated the worst president.
Third, I appreciate that Carter is not as forgotten as many other presidents whom I did not include. I explain elsewhere in the book why I have included Carter, though it bears repeating that public familiarity with Carter has a lot to do with his being a more (p.250) recent president, with his still being alive, and with his being an unusually active ex-president. Because I believe that Carter lost control of his narrative more than did any other post–World War II president, he, more than any of these others, has had his narrative—the story of his presidency—left for others to tell.
Last, the final rankings of presidents, in terms of presidential familiarity, seem reasonable. The least forgotten president is Lincoln, while the most forgotten is Fillmore. Though William Henry Harrison served barely a month as president, he is the second most forgotten because he has some notoriety as the first president to die in office. Fillmore has no such distinction. He did not die in office and followed another president who, by dying in office, garnered greater distinction. Fillmore served for roughly two years during which he was overshadowed by the political forces and more charismatic, energetic, and eloquent leaders with which he had to contend. It is telling that, after the many speeches Fillmore made to defend his administration’s enforcement of the Fugitive Slave Act, his rhetoric has not been memorable. This is because he was not a good orator, he was speaking to largely hostile crowds, he was conflicted about the course he had set, and subsequent defenses of federal supremacy were made for nobler causes. His efforts were not lost to history, but the tide of history cast them aside as failures. They have since been eclipsed by more dramatic events and bolder, more successful, more colorful personalities. Such is the fate of a forgotten president.