Does the taller candidate always win the election?

SHARE Does the taller candidate always win the election?

Dear Cecil: Has anyone ever made a study of the comparative heights of winners and losers in elections, and if so, what were the results? Wistfully, Eli “Shorty” Pindsey, Chicago


Illustration by Slug Signorino

Cecil replies:

Well, being a dwarf sure didn’t help that little Greek fella in 1988. (What was his name now?) And no question everybody in politics thinks being tall is an advantage. Five-foot-nine-and-a-half-inch Jimmy Carter’s handlers went to great lengths to prevent him from having to stand next to 6’1″ Gerald Ford during the 1976 presidential debates.

But the data on this question is notoriously flaky. U of Chicago statistics professor William Kruskal, who used to keep track of this sort of thing in his spare time, says he never found anybody whose height estimates could be relied upon, particularly when it came to the heights of the losers in earlier elections.

That caveat aside, various sources do claim that the taller candidate usually wins. For example, in Language on Vacation (1965), word and number buff Dmitri Borgmann claimed that in the 19 U.S. presidential elections between 1888 and 1960, the taller candidate won the popular vote all but once, when 6’2″ Franklin Roosevelt beat 6’2-1/2″ Wendell Willkie in 1940. In 1888 5’11” Grover Cleveland beat 5’6″ Benjamin Harrison at the polls but was cuffed in the electoral college, and in 1896 and 1900 both candidates were the same height.

In his 1982 book Too Small, Too Tall psychologist John Gillis presents similar results: in the 21 presidential elections from 1904 to 1984, the taller candidate won 80 percent of the time. What’s more, he says, in the whole history of the Republic, only two presidents — Harrison and James Madison (5’4″) — were appreciably shorter than the average height in their day.

We glean further insight on this issue from a delightful book called The Height of Your Life (1980) by Ralph Keyes (5’7.62″). Keyes notes that a survey of the U.S. Senate in 1866 found the average height of the members to be 5’10-1/2″, well above average for men at the time. Keyes’s own survey of 27 senators found the average height had risen to 6’0.33″, 3.33″ taller than the average American male. A similar survey of 31 governors found the average height to be 6’0.46″.

So, case closed, right? Not so fast. Consider an alternative theory of presidential electoral success — the Longer Name Hypothesis, which is also discussed in the Borgmann book. Of the 22 elections between 1876 and 1960, the candidate with more letters in his last name won the popular vote 20 times.

In two cases, Tilden-Hayes in 1876 and Cleveland-Harrison in 1888, the winners of the popular vote lost in the electoral college. In 1916 Wilson and Hughes had the same number of letters in their names, so the voters obviously chose on the basis of the issues. The only time the longer-named candidate lost was in 1908, when Taft whomped Bryan. However, Taft weighed more than 300 pounds, and probably attracted votes by force of gravity alone.

The situation has been somewhat muddied in the seven presidential elections since 1960, with only one victory for the long-named candidate, five defeats, and one case in which both candidates had names of equal length. This just shows you the difficulty of doing good science in the face of an uncooperative fact situation. As far as I’m concerned, the Longer Name Hypothesis remains at least as persuasive as the Longer Body Hypothesis. Until scientific measurements can be adduced for all dimensions of the presidential person (something that can surely be expected any day in light of the Thomas-Hill hearings), I think we must admit the question remains open.

Picking presidents: A new paradigm

Dear Cecil:

I read with interest your article discussing the Longer Body (Greater Height) and Longer Name hypotheses in predicting the outcome of American presidential elections. To help you further with the political education of your readers, I would like to share a rule which correctly predicts every election since the modern American politico-economic system began under FDR: a president with an unusual first name always alternates with a president with a common first name — Franklin, Harry, Dwight, John, Lyndon, Richard, Gerald, Jimmy, Ronald, George.

Ignore the occasional nitpicker who does not think Gerald is all that unusual a name, and you have a powerful analytical tool. While either John (Kennedy) or Richard (Nixon) could have followed Dwight (Eisenhower), and George (Bush) and Michael (Dukakis) both had a chance after Ronald, Adlai Stevenson could no more unseat Dwight than Hubert (Humphrey) could succeed Lyndon.

The acid test of any theory, of course, is in predicting the future. To put it on the line: neither Paul (Tsongas) nor Bill (Clinton) nor Tom (Harkin) nor Jerry (Brown) nor Dick (Gephardt) nor, for that matter, Pat (Buchanan) has a prayer of replacing George. [Editor’s note: This was originally written in mid-1992.] Mario (Cuomo) or Jesse (Jackson) could, but they may want to wait until 1996, when the Republicans will be stuck with Dan (Quayle), Jack (Kemp), or Dick (Cheney). Pierre duPont would be their only hope, unless he pursues the chimera of populism by insisting on being called Pete.

— Juozas Algimantas Kazlas (a common name, but only in Lithuania), New York

Cecil replies:

Not to give anybody ideas, Joe. But assuming Mario and Jesse aren’t available, there’s always (blush) Cecil. I am also obliged to note that Bill did defeat George. But your theory may still hold up, because right now his name is mud.

Cecil Adams

Send questions to Cecil via