The Dow of Elections Theory

As the presidential campaign sprints into its final, frenzied week, the outcome remains in doubt. The polls are inconclusive. The electoral map includes a patchwork of toss-up states. And the prognostications of CNN's political pundits are about as insightful and accurate as those made by their financial counterparts on CNBC.

There is, however, one indicator that has been remarkably successful in predicting the outcome of presidential elections: the Dow Jones Industrial Average. The Dow's performance over the past three months may provide a clue as to what may happen next week. After all, the Dow has seen more political campaigns than Harold "Kid" Stassen, the 93-year-old former Minnesota governor who first tried to get the Republican nomination for president in 1948, and has run in virtually every campaign since.

For the past 100 years, the performance of the Dow in the three-month period between July 31 and Oct. 31 -- the heart of the political season -- has predicted the winner with a remarkable degree of accuracy. When the market rises, the incumbent president, or the nominee of the incumbent party, usually wins. When the market falls, the challenger is more likely to win. This Dow of Elections Theory held true for 21 of the 25 elections in the 20th century. It has proved a reliable oracle in each of the last seven campaigns, in nine of the last 10, and in 14 of the last 16.

Intuitively, the theory makes sense. In 1992, when the Dow slumped 5% during the political season, it neatly mirrored incumbent President George Bush's feeble campaign, which offered few solutions to the country's economic malaise. Four years later, Bill Clinton cheerfully floated above the recumbent Bob Dole, and the Dow zipped upward 9%, smashing through the 6000 mark.

And so on. Back to the 1900 barnburner, in which incumbent William McKinley (assisted by a 4% election-season rise in the industrials) fended off the fire-breathing William Jennings Bryan for a second time.

There have been four instances in which the theory failed: 1968, 1956, 1932, and 1912. Most of these campaigns are what statisticians call outliers.

The most recent instance, 1968, was a year of serious political meltdown: protests against the Vietnam War, the assassination of Martin Luther King Jr., the abdication of Democratic incumbent Lyndon B. Johnson and the murder of presumptive Democratic nominee Robert F. Kennedy. Add a potent third-party campaign by Alabama Gov. George Wallace, whose racial message appealed to formerly solid Democratic strongholds in the South, and the fix was in. Republican Richard Nixon won, even though the market rose about 7.8% in the campaign season.

During the 1956 campaign season, incumbent Dwight D. Eisenhower cruised to re-election, even though the Dow fell during the campaign season. But the country was in the midst of a consumer-driven boom, and the Dow had risen 65% since Ike took office. (On Nov. 23, 1954, midway through Ike's first term, the Dow finally topped the Sept. 3, 1929, high water mark of 381.17.)

And like George W. Bush this year, Eisenhower in 1956 was running against an opponent who may have been too intellectual for his own good: Adlai Stevenson. A Stevenson supporter once reassured the candidate that he had the thinking man's vote locked up. "That's nice," Stevenson said, "but I want to win."

In the 1932 campaign season, even though the Dow rose 14% during the fall campaign, incumbent Republican Herbert Hoover didn't have a prayer. The Great Depression was already well under way, and the Dow had fallen a shocking 80% during Hoover's disastrous tenure.

Finally, there's the 1912 election, in which incumbent Republican William Howard Taft lost even though the Dow inched up 1.1% during crunch time -- but this election was truly bizarre. Teddy Roosevelt, who had anointed Taft as his successor back in 1908, felt a need to get back into the arena, and mounted a typically manic campaign on the Progressive Platform.

Democratic challenger Woodrow Wilson received just 42% of the popular vote but snared 435 electoral votes. Roosevelt took home 27% of the vote and won big states like California, Michigan and Pennsylvania. Poor Taft got just 23% of the votes -- the worst-ever showing for an incumbent -- and carried only Utah and Vermont.

What does the Dow of Elections Theory tell us about the 2000 campaign? Well, on July 31 -- a volatile day -- the Dow closed at 10,521.98, after ranging between 10,374 and 10,727. For much of the latter part of August and virtually all of September, the Dow stayed above that level -- and Gore led in the polls. For much of October, as the Dow fell below 10,500, Bush generally led in the polls.

Tuesday, the Dow closed at 10,971, up about 4.3% from July 31. That would seem to indicate a Gore victory on Nov. 7. But, as pollsters -- and journalists -- are fond of warning, these numbers are well within the margin of error.

---

Daniel Gross (www.danielgross.com) is the author of Bull Run: Wall Street, the Democrats, and the New Politics of Personal Finance. The New York-based journalist has written about the intersection of business and politics for The Washington Post, New York magazine, The New York Times and The New York Observer. He welcomes your feedback at [e-mail].