skip to main content
Reflections on Presidential Job Approval and Re-election Odds

Reflections on Presidential Job Approval and Re-election Odds

by Frank Newport and Joseph Carroll

Is President George W. Bush's current 64% job approval rating a good indicator that he will be able to successfully defeat the Democratic candidate for president in next year's election? The answer, right now, is a resounding, "It's too early to tell."

Presidential campaigns in which an incumbent is running for re-election are essentially a referendum on the president's performance in office. This assumption has led many observers to hypothesize that a careful analysis of the presidential job approval rating -- as close a measure as we have to a vote of confidence -- should be a good predictor of an incumbent president's chances for re-election. The job approval rating is, after all, a shorthand way of asking the American public to give a "thumbs up" or a "thumbs down" to the president's performance. This assumption certainly seems to hold true as the calendar approaches Election Day, but history indicates that there is simply too much variability in approval ratings in the pre-election year to give them any significant predictive validity.

Since Gallup began systematically collecting presidential job approval ratings, there have been 10 elections in which a sitting president has sought re-election, six of which were won and three of which were lost by the incumbent. A review of Gallup Polls in the months leading up to each of these elections shows that job approval ratings of the incumbent president do not predict chances of winning the election until it is well into the election year itself. How well a president is doing at this point (18 months before the election) does not give any type of definitive indication of how likely the president is to win or lose come Election Day.

FDR to the Present

There have been 34 elections in which an incumbent president has sought re-election. Interestingly, 3 of those 34 involved one president, Franklin D. Roosevelt, leaving 31 individual presidents out of the 42 (through Bill Clinton) who have sought re-election. The other presidents either declined to run for a second term or died in office (a higher-than-might-be-expected eight presidents died while in office, including four who were assassinated).

The track record for incumbents seeking re-election is actually not all that good. Twenty out of the 34 re-election tries were successful, while 14 were not (these 14 include presidents who sought their party's nomination for re-election, but failed to obtain it).

Unfortunately, we don't have presidential job approval ratings for the majority of these situations in which an incumbent sought re-election. It would be fascinating to know what George Washington's job approval rating was as he sought re-election in 1792, or John Adams' when he was defeated in his bid for re-election in 1800 by Thomas Jefferson, or Abraham Lincoln's in 1864, or Herbert Hoover's in 1932 when he ran and was defeated by Roosevelt. But we don't have the data.

Gallup does, however, have job approval ratings for 10 of the most recent elections in which an incumbent sought to be re-elected: Roosevelt's 1940 election, Harry Truman in 1948, Dwight Eisenhower in 1956, Lyndon Johnson in 1964, Richard Nixon in 1972, Gerald Ford in 1976, Jimmy Carter in 1980, Ronald Reagan in 1984, George H. W. Bush in 1992, and Bill Clinton in 1996.

The track record for these incumbents is better than it has been in the years going back to George Washington. Only 3 of these 10 presidents seeking re-election have lost: Ford, Carter, and Bush.

Is there a pattern in presidential job approval ratings in the 18 months leading up to an election that helps us understand and/or predict the likelihood that the president would end up being re-elected? Was the pattern of job approval ratings for Ford, Carter, and Bush substantially different from the patterns for the other seven presidents who were re-elected? As noted, these questions are very much on observers' minds now because George W. Bush is just 17 months away from November 2004, and many are wondering what can be gleaned, prediction wise, from his current standing among the voters as measured by his job approval rating.

Let's look first the pattern of job approval ratings for the 10 presidents in our sample, examining job approval from May of the year before the election year in which they sought re-election, through the election year itself.

Approval Ratings of Incumbents Who Won Re-election

Franklin D. Roosevelt

Roosevelt had relatively stable approval ratings throughout the 18-month period leading up to his third re-election try in 1940. His job performance ratings began in the mid-50% range in the summer of 1939, moved into the 60% range later in 1939, dropped back into the mid-50% range in the spring of 1940, and ended up at 64% in Gallup's last poll that included job approval, in July 1940. Roosevelt went on to defeat Republican Wendell Willkie in the November election by a 55% to 45% margin.

Harry S. Truman

Truman provided a major exception to all rules of thumb relating to pre-election polling. Truman was not popular, had low approval ratings, and Gallup and other pollsters (famously) predicted, based on final pre-election horse race polls, that he would lose the election to Republican Thomas Dewey. Of course he didn't, defeating Dewey by a 50% to 45% margin (with third-party candidates receiving 5% of the vote). Truman's job approval rating had actually been in the 50% and 60% range in 1947, but by 1948 had plummeted to as low as 36% in April and 39% in late May and early June (apparently the final job approval rating Gallup obtained on Truman before the election).

Dwight D. Eisenhower

There was little mystery about Eisenhower's re-election in 1956. His job approval ratings were consistently in the high 60% and low 70% range in the months prior to Election Day, fluctuating from a low of 68% in May 1955, to a high of 78% in November 1955. Gallup's final job approval rating for Eisenhower before the election was 68% in August 1956, and he went on to defeat Democrat Adlai Stevenson by a 58% to 42% margin.

Lyndon B. Johnson

Johnson's approval ratings at the beginning of his tenure in office in November 1963 were remarkably high -- due undoubtedly to the halo effect from the assassination of President John F. Kennedy. In December 1963, following the assassination, more than three in four Americans approved of Johnson, and his approval ratings remained at this level throughout the election season. Gallup's final job approval rating for him before the election was 74% in June 1964, and he ultimately defeated Republican candidate Barry Goldwater by a 61% to 39% margin.

Richard M. Nixon

Nixon's job approval ratings were not all that high in the year before his re-election bid in 1972, hovering around the 50% level for many months during 1970 and the beginning of 1972. In fact, at the 18-month mark (May 1971) Nixon had a job approval rating of 50%, and a rating of 49% in January 1972. But Nixon's approval ratings began to increase in March and April of 1972 to the point that 62% approved of his job performance in May 1972, and 58% in the final Gallup approval rating before the election. Nixon defeated Democrat George McGovern by a 61% to 38% margin.

Ronald Reagan

Reagan's job approval ratings were actually worse than Nixon's in the year before his successful re-election bid in November 1984. Reagan's ratings lingered well below the 50% mark for most of 1983, falling as low as 43% at one point in the summer of 1983. But by the end of 1983, Reagan's approval ratings moved above the 50% mark, and stayed comfortably above that point for the rest of the election year. Gallup's job approval rating for Reagan was a relatively robust 58% in October, and he went on to defeat Democrat Walter Mondale handily, by a 59% to 41% margin.

Bill Clinton

Eighteen months prior to the 1996 presidential election, Clinton's job approval rating was below 50%, and it stayed below that point until November 1995. Clinton's approval ratings dropped briefly in January 1996, but recovered soon thereafter and stayed above 50% for the rest of the election year. Gallup's final job approval raging for Clinton before his successful defeat of Republican Bob Dole was 56% in October. Clinton beat Dole by 49% to 41% margin, with Ross Perot getting 9% of the vote.

Approval Ratings of Incumbents Who Lost Re-election

Gerald Ford

Roughly half of all Americans approved of Ford in May 1975. Ford managed to obtain several job approval ratings at or slightly above 50% during the second half of 1975 and at one point in March 1976, but for the most part, his ratings were in the 40% range. Gallup's last rating for Ford was 45% in June. He lost to Democrat Jimmy Carter by a slim 50% to 48% margin.

Jimmy Carter

The pattern of Carter's approval ratings in the 18 months prior to his election proved to be quite interesting. Eighteen months prior to the election, only about 3 in 10 Americans approved of Carter. At the end of 1979 and in the beginning of 1980, Carter's approval ratings surged to a high of 57% as Americans responded to the taking of the American hostages in Iran and the decision by the Soviet Union to invade Afghanistan. This rally ended quickly, however. By the summer and fall of 1980, Carter's performance ratings decreased to the mid-30% range and by September, in Gallup's last pre-election approval rating, Carter's was at only 37%. Carter lost to Republican Ronald Reagan by 41% to 51% margin.

George H.W. Bush

Following the success of the Persian Gulf War, the first President George Bush enjoyed a remarkably high approval rating and many observers thought he was almost certain to win re-election. But Bush's ratings began to slide shortly after the Gulf War ended in February 1991, and essentially never recovered. By January 1992, Bush's ratings were below the 50% mark, and they continued to fall into the summer and fall of 1992. Bush reached an extremely low 29% job approval rating in late July and early August 1992, and in Gallup's last poll before the election was at only 33%. He was defeated in the November election by Bill Clinton, who received 43% of the vote to Bush's 38%, with Ross Perot getting 19%.

Bottom Line

Presidential job approval data are an important piece of the puzzle in understanding an incumbent's chances of winning re-election, but only after the time frame has moved well into the election year. Before that, particularly in the year before the election, approval ratings are too subject to change to be of much use in predicting an incumbent's chances for re-election.

Presented below is a summary chart which shows the 10 incumbents' job approval ratings at various specific time points leading up to the Election Days on which they sought to be re-elected.

Presidential Approval Leading to Presidential Elections
(Based on monthly averages of approval ratings)

 

18
months before election

12
months before election

10
months before election

8
months before election

6
months before election

2
months before election

%

%

%

%

%

%

1940 (Roosevelt)

56

58

62

57

60

--

1948 (Truman)

65

54

--

36

--

--

1956 (Eisenhower)

68

78

75

69

73

--

1964 (Johnson)

64

58

76

75

74

--

1972 (Nixon)

50

49

52

54

58

--

1976 (Ford)

46

41

48

48

45

--

1980 (Carter)

35

40

54

39

32

37

1984 (Reagan)

45

53

55

54

54

56

1992 (Bush the elder)

75

55

41

41

37

38

1996 (Clinton)

51

53

53

55

55

60

The data 18 months and 12 months before the election are suggestive, but not definitive. In particular, note that Reagan's approval rating was below 50% 18 months before his 1984 re-election bid, and Nixon had approval ratings below the 50% margin at the 12-month mark. Both men won by very significant margins. At the same time, most famously, George H.W. Bush had remarkably high job approval ratings 18 months before his 1992 re-election bid, and was still above 50% at the 12-month mark. He, of course, lost to Clinton.

By January of election year, it appears that things typically begin to straighten out. But not totally. The Carter administration provides a major exception. By January 1976, Carter's image had improved, with ratings above the 50% mark, only to fall back again within months and to eventually lose the election to Reagan.

At the eight-month mark, an incumbent's job approval rating is more significantly predictive of eventual election success. With the exception of Truman in 1948, every president who won re-election had approval ratings above 50% at the eight-month mark, and every president who lost had approval ratings below 50%.

We are of course dealing with a very limited sample of elections in looking at these data, and it is obvious from situations similar to what Carter encountered in 1980 that there are many different scenarios by which an incumbent president's job approval ratings could confound historical patterns.

At the very least, however, these data suggest that one should avoid the temptation of making any assumptions about re-electability from an incumbent's job approval rating in the year before an election. At the same time, it appears that by the eight-month and in particular six-month mark, an incumbent with job approval ratings above 50% can certainly be construed to be in relatively good shape for re-election, and an incumbent president with job approval ratings below 50% would seem to have lower odds of being re-elected.

So what does all of this mean for the current President Bush? Bush's latest job approval is 64%. Bush received extremely high approval ratings after the Sept. 11 terrorist attacks and after the war in Iraq -- but in both instances his ratings began to fall in the months that followed.

In the current situation, polling conducted since the end of the major fighting in Iraq shows a continued decrease in Bush's overall approval ratings, although they remain relatively high. But, as reviewed above, the overall lesson to be learned from history is that the current 64% rating tells us relatively little about Bush's ultimate chances of re-election. If Bush's downward trend continues, as it did with his father in 1992, and if the current president's ratings fall through the 50% range and into the 40s by next spring, then it will begin to look increasingly like an incumbent upset in the making. If Bush's ratings fall into the 50s but stabilize there by next spring, or if they stay in the 60% range or even go back up again early next year, then it will be safer to assume that Bush will be in a much better position to win re-election next fall.


Gallup https://news.gallup.com/poll/8608/reflections-presidential-job-approval-reelection-odds.aspx
Gallup World Headquarters, 901 F Street, Washington, D.C., 20001, U.S.A
+1 202.715.3030