Go to the Globe and Mail homepage

Jump to main navigationJump to main content

U.S. President Barack Obama earlier this month. Incumbent presidents win re-election when the Dow Jones industrial average rises 40 per cent or more in the preceding three years, a research paper shows (Jonathan Ernst/Reuters/Jonathan Ernst/Reuters)
U.S. President Barack Obama earlier this month. Incumbent presidents win re-election when the Dow Jones industrial average rises 40 per cent or more in the preceding three years, a research paper shows (Jonathan Ernst/Reuters/Jonathan Ernst/Reuters)

Neil Reynolds

An Obama victory? That's what the Dow says Add to ...

Over at Intrade (which dubs itself “the leading prediction market”), U.S. President Barack Obama’s chances for a second term look good. With an implied 60.8 per cent chance of victory, they look very good. But this predictor isn’t as robust as it appears. It merely means that six out of 10 random Intrade “investors” think Mr. Obama can win 50 per cent – plus one – of the roughly 100 million votes that will be cast in November, and can arrange to have them fall in such a way that he wins a majority in the Electoral College. This could be harder than it appears.

More from Neil Reynolds

Public opinion polls show that Americans anticipate an extremely close presidential election. In February, 16 organizations conducted successive national polls in which they offered respondents the choice of 50 presidential election variables (including the four candidates who still sought the Republican nomination).

The average poll result, all 50 variations included, awarded Mr. Obama 49.9909 per cent of the popular vote. In a pristine two-party race, which won’t happen, he would lose by 0.0091 per cent of the popular vote. And everyone thought that it was close in 2000 when Democrat Al Gore beat Republican George W. Bush by 500,000 votes – and lost.

Opinion polls, though, don’t predict; they report. U.S. political economist Douglas Hibbs’s Bread and Peace Index, on the other hand, purports to reveal the results that will be reported some 30 weeks hence – on the evening of Nov. 6. Mr. Hibbs (a former economics professor at Harvard and University of Gothenburg. now retired), tracks two statistics: real-dollar changes in personal disposable income (PDI) and the number of U.S. soldiers killed in military actions abroad. These two factors, he says, systematically affect voting in presidential elections.

For an incumbent president to win re-election, he says, PDI must increase significantly during the first term in office. And the number of American soldiers killed in foreign wars must be statistically negligible.

As calculated by Mr. Hibbs in the final quarter of 2011, Mr. Obama will lose – regardless who runs against him. On one hand, the number of American soldiers killed during the first 10 quarters of his presidency (January, 2009, to June, 2011) was 1,169 – at 3.8 deaths per million population, a statistically (but not humanly) negligible factor. On the other hand, people’s spending money shrank during the same period by 0.4 per cent, a statistically significant decline.

For Mr. Obama to turn things around, Mr. Hibbs says, the U.S. economy would need to boost PDI by 4 per cent in the next seven months. The odds of this, he says, are one in seven. Without this happy-ending prosperity spurt, Mr. Obama will get only 44.1 per cent of the votes cast (the worst performance since Jimmy Carter won 44.7 per cent in his landslide loss to Ronald Reagan in 1980). This result would decisively end Washington’s partisan gridlock.

On the other hand, Mr. Hibbs’ Bread and Peace Index isn’t infallible. It didn’t quite get it right in 2000 when it anticipated that Mr. Gore, not Mr. Bush, would win (which, from a ballots-only basis, he did). Flukes happen. But the academic search for a more perfect empirical predictor continues.

In January, iconoclastic Elliott Wave theorist Robert Prechter, and others, published a research paper (Social Mood, Stock Market Performance and U.S. Presidential Elections) that identifies the Dow Jones Industrial Index as “a powerful regulator of U.S. re-election outcomes.” The paper concludes that the Dow, by itself, is a more accurate predictor of presidential elections than GDP, unemployment, inflation – or personal disposable income.

The research paper argues that the U.S. “social mood,” rather than discrete events, drives the Dow. With the Dow Jones Industrial Average creatively postdated to the administration of George Washington, the paper’s authors set out to prove that incumbent presidents win re-election, often in landslides, when the Dow rises by 40 per cent or more in the preceding three years.

If this holds for November, Mr. Obama will win – massively. The Dow hit bottom (at 6,500) two months after he assumed office. It is up 6,500 points since: a 100-per-cent advance. From Mr. Prechter’s analysis, it would take an abrupt reversal of optimism, and a subsequent collapse of the Dow, to deny him a second term.

 
Live Discussion of false on StockTwits
More Discussion on false

More from Neil Reynolds

Topics:

In the know

Most popular video »

Highlights

More from The Globe and Mail

Most Popular Stories