Sunday, April 29, 2012

Forecasting an Incumbent's Chances

We are just about six months away from the 2012 election.

In the early days of my collegiate career, I was a political science major, and one of the things my professors drummed into my head (and the heads of my peers) was the conviction that voters reach their conclusions about an incumbent about six months before the election.

That is a piece of conventional wisdom that has held up rather well, whereas I have found that people who rely on surveys that measure the likability of an incumbent inevitably are misled.

When times are generally good, as they were when George W. Bush ran against Al Gore, voters can afford the luxury of picking the presidential candidate with whom they would rather share a beer. (Of course, the incumbent wasn't on the ballot that year.)

But when times are not good, likability takes a backseat to competence with most voters. In 1972, for example, when Richard Nixon was seeking a second term, the Vietnam War was still unpopular and there was some dissatisfaction with the economy, but voters were influenced more by whether they thought he was doing a good job (and most did) than whether they liked him (and most did not).

Job approval surveys didn't make their debut until Franklin Roosevelt had already won his second term — and FDR has always struck me as being a special case, having been elected president four times. The margins were different each time, and, of course, one of the issues when Roosevelt sought his third and fourth terms was whether any American president should be allowed to serve more than two terms. Some people who had supported him the first two times opposed him the second two times for that very reason.

Those campaigns for the third and fourth terms qualify as unique cases, therefore, and job approval surveys were still evolving, anyway, so the conventional wisdom of which my professors spoke isn't really applicable.

Nor, for that matter, is it applicable to 1948, when Harry Truman defeated Tom Dewey in an "upset."

I don't mean to suggest that Truman's victory wasn't a surprise (it was) or that voters were insincere when they told pollsters they didn't approve of the job he was doing.

I'm just mindful of the fact that Truman was completing FDR's fourth term. He hadn't been elected president. He was elected vice president.

And, while I can't speak for everyone, I can say that I have never based my presidential preference on the identity of the running mate. Those who choose which ticket to support on that basis are all but assuming that the guy at the top of the ticket won't complete the term.

That strikes me as being the same thing as trying to prove a negative, and, at least in my opinion, it is the wrong way to choose a president.

(In 1980, I did know some people who supported Ronald Reagan because George H.W. Bush was his running mate, and they figured that there was no way Reagan could survive the term. They became increasingly frustrated as Reagan simply refused to die — even after he had been shot — and they wound up having to wait eight years until Bush was elected on his own.

(There are sure to be some who have voted on that basis in other elections. I'm confident there were those who voted against some Republican tickets because Sarah Palin or Dan Quayle were on them. But my gut feeling is that their numbers were few.)

Besides, job approval surveys were still evolving, as I say, and the 1948 election would have a significant influence on how such surveys were conducted in the future.

So I don't include Roosevelt or Truman in such comparisons. Approval polling methods were still primitive when they occupied the White House.

But, by 1956, when Dwight Eisenhower sought his second term, a lot had been learned.

For one thing, pollsters kept polling right up until Election Day. They didn't stop polling long before the election, as they had in 1948 because it was a foregone conclusion that Truman would lose.

In early May 1956, Gallup found that 69% of respondents approved of the job Eisenhower had been doing.

A lot of that may have been due to something of a wave of sympathy for Ike. He had suffered a heart attack about eight months earlier. If the May approval rating was influenced by his health, that wave crested well before Election Day 1956, but Eisenhower still carried 41 states and received 57% of the popular vote.

Eisenhower's successor, John F. Kennedy, died before the end of his term, and his vice president, Lyndon Johnson, had been president for less than a year when he won a full term.

Six months before the election, three–fourths of the respondents to a Gallup survey approved of the job that he was doing — but that, too, may have been the result of public sympathy.

Johnson went on to win the 1964 election by a landslide — but, four years later, the increasingly unpopular Vietnam War forced Johnson to drop out of the race, and a month after he did so, the approval ratings appeared to confirm the wisdom of his decision.

In April 1968, Johnson's approval rating, as reported by Gallup, dropped below 50% and never exceeded that number again. It was tumbling by May.

Consequently, Johnson's decision to drop out of the race seems prescient in hindsight. Odds are, he wouldn't have been successful, even if he had been renominated.

Nixon replaced LBJ, and his Gallup approval ratings in May 1972 hinted at what was to come.

Nixon's job approval six months before the election was somewhere between 54% (his rating in late April) and 62% (his rating in late May). On Election Day, more than three–fifths of the voters endorsed his bid for a second term.

One can argue, of course, that Nixon's campaign was tainted by the illegal activities of his staff and his own involvement in efforts to cover up those activities. But the bottom line is that the job approval ratings pretty accurately measured how voters felt about the Nixon presidency in mid–1972.

Nixon's second term was cut short by the Watergate scandal, and his replacement, Gerald Ford, enjoyed high approval ratings in his first month in office, but his ratings declined rapidly after he pardoned Nixon.

In May 1976, he received approval ratings that hinted at what would happen in the election. For members of Barack Obama's staff, it might be instructive to study the Ford campaign because Ford's approval in mid–1976 was close to where Obama's has been lately — just short of 50%.

Ford, of course, went on to lose to Jimmy Carter in a very close election. Some people have been tempted to presume that it was almost entirely because, not having voted for him to begin with, voters felt no real allegiance to Ford — but he lost largely due to his job performance, because he pardoned Nixon, not because voters stopped liking him. He was still affable Jerry Ford in the minds of most.

But voters did not like the decision to pardon Nixon.

The voters definitely soured on Carter by the time he sought a second term in 1980, and the approval ratings six months out did more than hint at that. In May 1980, Gallup found that the share of voters who approved of the job he had been doing was in the upper 30s and lower 40s.

When the voters went to the polls that November, Carter received 41% of the vote and carried six states. Reagan was elected in a landslide.

In 1984, Reagan won a second term in spite of a jobless rate that was higher than it had been for any successful incumbent in nearly 50 years.

It was conceded at the time that Reagan was widely liked by the American people, but their electoral endorsement of his presidency in November was foreshadowed by Gallup polls in May that indicated that (1) a majority approved of the job he was doing, and (2) that majority was growing, not declining.

When the votes were counted in November, Reagan received nearly 59% of the vote and carried 49 states.

Reagan's vice president, George H.W. Bush, was elected in 1988 when Reagan was constitutionally prohibited from running again, and his lightning–like victory in the Gulf War seemed to be propelling him to a second term — but a funny thing happened to Bush on the way. A recession derailed him.

The recession was mild by historical standards — certainly when compared to the one that Americans have been slogging through since late 2007 — but it was bad enough to lower Bush's approval ratings from the 50s in late 1991 to 42% by May 1992, according to Gallup.

Once again, it was a reliable predictor of the incumbent's fate. Bush went on to lose to Bill Clinton in a three–man race.

The economy was better four years later when Clinton sought his second term.

Clinton's Democrats lost control of Congress during the 1994 midterms, but Clinton, through a combination of shrewd political moves and sheer good fortune, was on an upward trajectory in May 1996. Six months before the election, CNN/Time reported that 51% of Americans approved of the job he was doing; Gallup found that 55% approved.

The endorsements were solid, if not resounding — as were November's election returns. Clinton was re–elected with 49% of the vote and the support of 31 states.

In 2000, of course, George W. Bush defeated Gore in the Electoral College but lost the popular vote. Although extremely rare, such an outcome was not unheard of — but, to be old enough to remember the last time it happened, one would have to be at least 130 years old when Bush was elected — and there were no job approval surveys in those days.

That probably made Bush something of an exception to the conventional wisdom concerning the relationship between job approval numbers and eventual electoral verdicts on incumbents — since he hadn't received the support of at least a plurality of the voters the first time, as nearly every duly elected president has.

But the job approval rule still held true when Bush sought his second term in 2004.

The job approval ratings in May 2004 warned Bush that he would face a close race in November. NBC/Wall Street Journal, Gallup and other surveys in early May found about as many Americans who disapproved of his job performance as approved.

And that was borne out in the general election, when Bush received less than 51% of the vote and the support of 31 states. In the Electoral College, he defeated John Kerry by 35, 286 to 251.

What will all this mean in the 2012 election? I guess that remains to be seen.

Next Sunday is precisely six months from Election Day, so any job approval numbers that are announced on or after that date could be said to be potential indicators of what to expect in November.

But Gallup reports that, according to his job approval average for his 13th quarter in office, Obama's ratings are below the average for presidents who went on to win re–election.

They're even below one president — Carter — who was defeated in his campaign for re–election.

J. Robert Smith makes intriguing observations in American Thinker that speak to the relevance of history — even though he doesn't connect the dots between job approval ratings and an incumbent's odds of winning a second term.

"Presidential election history gives us indications," Smith writes, "that Mr. Obama either squeaks back into the White House or gets an undignified boot in the back of his designer trousers. ... [O]nly Jerry Ford lost his re–election bid narrowly. Odds are, if Mr. Obama loses, it will probably be on the order of [Herbert] Hoover (1932) or Carter (1980)."

No comments: