PRINCETON, NJ -- Gallup's presidential job approval measure has been used as a shorthand summation of the public's opinion of a president for over 50 years. The job approval measure came under particular scrutiny and criticism during 1998, based primarily on the seemingly paradoxical fact that the measure did not drop precipitously as Clinton's Lewinsky assignations became public and as Bill Clinton became only the second president in U.S. history to be impeached.
The facts are straightforward: Bill Clinton received the highest job approval ratings of his administration during the Lewinsky/impeachment controversy. As the Lewinsky situation unfolded, Clinton's job approval went up, not down, and his ratings remained high for the duration of the impeachment proceedings:
- Bill Clinton's mean job approval rating, 1st quarter 1993 through 1st quarter, 1999 was 53.8
- Bill Clinton's mean job approval rating for the five years preceding 1998 was 51.3
- Bill Clinton's mean job approval rating in 1998 was 63.8
- Bill Clinton's average job approval rating for 1998 was thus 10 points above his overall administration to-date average
- Bill Clinton's average job approval rating for 1998 was thus 12.5 points above his administration average for the five years preceding 1998
- Bill Clinton's average job approval rating for 1998 was 5.7 points above that of the previous year, 1997, which in turn was higher than that of any of the four years which preceded it
- Bill Clinton's job approval rating in the first quarter during which the Lewinsky situation became public knowledge (1st quarter 1998) jumped 5.6 points compared to the immediately preceding quarter
Thus, the onset of the publicity surrounding the Lewinsky revelations was correlated with a significant jump in Clinton's job approval rating, and the two quarters during which the House and Senate debated impeachment and conviction -- 4th quarter 1998 and 1st quarter 1999 -- saw the public give Bill Clinton the highest job approval ratings of any of the 25 quarters of the Clinton administration to date. In what some might see as a paradoxical twist, Clinton's job approval ratings have fallen after the impeachment process ended with his acquittal, and are -- in Gallup's most recent May poll -- at 53%, the lowest since August of 1996.
The average presidential approval rating for every year from the Eisenhower administration to the present was 56.1. Bill Clinton's 1998 average, 63.8, is thus significantly above this 40-year average, and is higher than the averages of all but 11 of the 40 years served by the eight presidents who preceded him.
The 11 years with higher average approval ratings came in the administrations of four presidents, three of whom served a quarter of a century or longer ago: the first four years of the Eisenhower administration, the first two years of JFK's term, the first two years of LBJ's term, and the first three years of George Bush's administration.
What does a president's job approval rating from the public represent? The Gallup job approval measure was first developed in the mid-1930s, and grew out of Dr. George Gallup's attempts to rate how Americans felt about President Franklin Delano Roosevelt. The presidential vote is the one conclusive way in which American citizens can register their approval or disapproval of presidential candidates or incumbents. However, this vote occurs only every four years. The initial attempt to develop the job approval measure, therefore, was an effort to create a "between elections" measure that replicated or predicted the vote.
Gallup's initial formulations asked "which candidate -- Roosevelt or one of his challengers -- respondents would vote for at the next presidential election or, alternatively, if they planned to vote for or against Roosevelt." As the question evolved, the wording was changed to ask if the respondent would vote for or against FDR if a presidential election were held "today." From that point, the question was changed to asking if the respondent was "for or against Roosevelt today," to an "approval or disapproval" of FDR, to the final "approve or disapprove of the way Roosevelt is handling his job as president today." This is roughly the format still in use to this moment in time.
The job approval measure, then, was developed as a shorthand way of estimating the public's confidence in the job being done by the president, and thus stood as a surrogate indicator of the public's willingness to have an incumbent stay in office if a new election were held on the day of the interview.
The job approval measure, in fact, turns out to be an excellent predictor of an incumbent's chances of re-election. Professor Michael S. Lewis-Beck of the University of Iowa has calculated that the incumbent president's job approval measure in the summer before an election has a .87 correlation with the incumbent party's share of the two-party popular vote (based on the incumbent party's reelection percentage in each applicable election from 1948 to 1996).
The job approval measure, in short, allows the public to summarize its perceptions of the way in which an incumbent president is accomplishing the job for which he was elected. The factors that a respondent could take into account in making that judgement are theoretically limitless, including anything that could conceivably, in the mind of the respondent, affect the way in which the president was executing his duties. The public presumably arrives at a summary judgement in a process similar to that which occurs when a voter summarizes various factors in making the vote decision on Election Day.
It is thus important to remember that there are many ways in which the public can evaluate its incumbent president, ranging from his performance on highly specific single issues to his overall affability and likability. Pollsters have developed measures for many of these dimensions of a president and they are reported regularly. The overall job approval measure, as reviewed above, is intended as a summary measure that approximates the one way in which the citizens of the United States make their evaluation of a president known -- through the ballot box. (In the case of Bill Clinton, respondents' choices on the job approval variable in 1998 were, in fact, highly correlated with their views on whether or not Clinton should be retained in office.)
There is no theoretical reason why a personal scandal, personal characteristics, or any other criteria cannot be taken into account in making summary assessments of a president. It is highly likely, for example, that Watergate affected Richard Nixon's job approval measures, as did the Iran-Contra affair for Ronald Reagan.
When the Lewinsky revelations burst onto the public scene in January 1998, then, the key question became one of assessing the degree to which the public considered the implications of the scandal to be "job-related." To the degree to which the scandal was seen as related to Clinton's performance in office, the assumption is that his job approval rating would fall. Observers, including those of us at The Gallup Poll, watched Clinton's job approval measure and waited.
In spite of the scandal, there were, in 1998, many reasons to predict that Clinton's job approval rating would be high. The Lewinsky revelations and the impeachment process occurred during a time period in which the American public was unusually positive in its ratings of the economy and of its overall satisfaction with the way things were going in the country:
- Americans registered high levels of satisfaction in 1998 with the way things were going in the country (although not as high as immediately after the Gulf War, and about as high as measured during the "Morning in America" years after the '84 Reagan election).
- Ratings of Bill Clinton's handling of the economy shot up to as high as 81% during this time period.
- The basic ratings of the economic conditions in this country were high throughout this period.
- Personal financial wellbeing measures were at one point as high as they have been in Gallup's history of measuring them, with 7 out of 10 polled in March 1998 saying they anticipated being better off over the next year.
- A substantial percentage of Americans (71% in one January 1999 poll) said that this was the best economy of their lifetime.
The following attitudes about the state of the nation were highly significant predictors of presidential job approval:
- The correlation between satisfaction with the way things were going in the country and presidential job approval from 1993 through early 1999 was .94
- The correlation between the public's assessment of the economy and Clinton's overall job approval for his administration was .84
- The correlation between approval of the economy and Clinton's overall job approval was .88
Additional measures (other than job approval per se) that assessed the administration and managerial aspects of Bill Clinton were also positive in 1998:
- Over seventy percent of Americans said that Bill Clinton had been an effective manager
- From 67% in January 1998 to 82% in January 1999 said that Bill Clinton "can get things done"
- By January 1999, 75% said they were very or somewhat confident in Bill Clinton's abilities to carry out his duties as president, with only 13% saying they were not too or not at all confident
- In January 1998, 71% said that his presidency had been a success, while 81% said that his presidency had been a success by January 1999
In summary, it is clear that the fundamentals were in place in 1998 to provide an "everything else being equal" foundation for high job approval ratings for the incumbent president. Were it not for the scandal and impeachment proceedings, the trend in Clinton's job approval ratings would be expected and wholly understandable: the public, extraordinarily upbeat about the economy and the state of the nation, and not perceiving significant foreign policy problems for the country, was rewarding the leader with the highest job approval ratings of his administration.
It is not that the public was naïve and unaware of the Lewinsky situation, but rather that the revelations had little effect on Clinton's job approval ratings. The public early on concluded that Bill Clinton had had sexual relations with Monica Lewinsky, that he had lied under oath, and that he had obstructed justice. In other words, Americans agreed with most of the charges that formed the basis for impeachment:
- By February 1998, 63% of Americans felt that Clinton definitely or probably had had sexual relations with Monica Lewinsky
- By April, 60% felt that Clinton had lied under oath while president, a percentage that grew to 74% by August
- By August, 53% felt that Clinton had definitely or probably participated in an effort to obstruct justice while president
- More generally, in a January, 1999 poll, a majority of the public agreed with the two charges brought against Clinton by the House, including 79% who agreed with the perjury charge
Additionally, measures of Bill Clinton other than job approval began to reflect public perception of his failings relating to the scandal:
- Less than 50% of Americans had a favorable opinion of Bill Clinton "as a person" in most months except for October
- The percentage rating Bill Clinton as "honest and trustworthy" was at only 38% in January, and that number slowly dropped to 24% by January 1999
- Substantially less than half of the public said that Bill Clinton "shares [its] values," a figure that dropped from 42% in January 1998 to 35% in January 1999
- Less than half said that Bill Clinton shows good judgement
But Clinton's job approval measures remained high, in essence signaling a definitive verdict from the public that the scandalous behavior was not related to the way in which Clinton was doing his job as president. An intriguing and dramatic finding that underscores this fact is the correlation between Clinton's honesty rating and overall job approval: -.49. In other words, examined across the first six years of the Clinton administration, the lower the public's assessment of Clinton's honesty and trustworthiness, the higher its approval levels of the job he was doing as president.
An analysis of three key measures of Bill Clinton -- his overall job performance rating, his handling of the economy, and his rating as honest and trustworthy -- also helps put these data in perspective. The public gave Clinton increasingly positive ratings on his handling of the economy, increasingly negative ratings on being "honest and trustworthy," and modestly increasing ratings in terms of his overall job approval as 1998 unfolded. These trends suggest that the economy drove the approval rating, and that the fact of Clinton's being perceived as dishonest did little to mitigate this powerful relationship.
In short, Clinton's job approval ratings in 1998 basically remained high in the midst of impeachment because the public perceived that Clinton had been highly successful in doing the job for which he had been elected. A primary driving force in the high job approval ratings was the positive perception of the way in which Clinton was handling the economy. At the same time, the public downgraded its evaluation of Clinton on moral conduct-related measures. Americans were well aware of Clinton's Lewinsky-related behavior, including lying under oath, but apparently came to the conclusion that this behavior was not directly related to his ability to perform his duties as president.
The fact that the public came to this conclusion is an important focus for continuing discussion and analysis. There are a number of theories that purport to show why Clinton's moral failings did not have more of an effect on his job approval rating in the eyes of the public. Most likely the 1998 impeachment crisis was unique to this particular president, this particular year, and to the particular cast of characters involved in the high drama on both sides of the issue. At some future point, a president accused of the same failings may well find public opinion strongly against him.
The public's ability to divorce this particular president's morality and ethical behavior from assessments of his ability or right to continue to govern has created in some observers consternation, puzzlement and a sense of "moral outrage." But this outrage should not be directed at the pollster's measures. The basic tools used by pollsters to measure the president, including the job approval measure, performed their function well during the crisis. The classic job approval rating, despite criticism, remains a highly important measure of the public's confidence in its president and allows the public to summarize its opinions with a measure that approximates what its vote would be if elections were held more frequently than every four years. In this situation, the verdict from the public was clear. Clinton's high job approval numbers, maintained in spite of all that he did relating to Monica Lewinsky, translated into a strong sentiment that he should remain in office.