Why 2012 election predictions are rubbish: Fear the Black Swan!
You want to know who’s going to be the next president of the United States? Happy to oblige.
Just tell me who’s going to win Ohio. No Republican has ever won the White House without winning Ohio. And only one Democrat has done it—JFK by a whisker—in the past 50 years.
Or tell me what will happen to real personal income growth in the third quarter of 2012.
Or tell me what the jobless rate will be in the fall, since (all together now), no incumbent since FDR has been re-elected when the unemployment rate has been higher than 7.2 percent.
What’s that? You can’t do that because it’s only April?
That doesn’t stop an army of soothsayers — including ones at Yahoo! — from offering up formulas to calculate, with scientific precision, the shape of the November vote. As common-sense guides, they make sense: incumbents and incumbent parties suffer when the economy is bad; a deeply divided party has a hard time winning a general election. As “laws” with the predictive capacity of knowing when ice melts … not so much. (Back in 2000, the most trusted academic models of the election forecast a comfortable-to-overwhelming Democratic popular vote victory based on the glowing economy; what we got was an effective tie).
I received an early lesson in caution after boldly predicting that John Lindsay would win the White House in 1972. Even stronger lessons were provided over the years by the appearance of a hugely influential factor in Presidential elections: the Black Swan.
The term comes, not from that Natalie Portman ballet movie, but from a best-selling book in 2007 by Nassim Nicholas Taleb that examines our persistent “ability” to ignore the potentially huge effects of unlikely, random events. Given what happened a year later–when we woke up on a mid-September day to find the financial universe on the brink of collapse–the book seemed prescient. In political terms, “Black Swans” have shown up often enough to make even the boldest soothsayer hold his tongue.
Think back to 1960, when Republicans could still compete for the black vote, and when an influential figure like Martin Luther King Sr. endorsed Richard Nixon out of concern about a Catholic in the White House. Then, on October 25, King’s son was arrested on a bogus parole-violation charge and transferred to a rural state prison where, his family feared, his life might be endangered. After John Kennedy called King’s wife, and Robert Kennedy called the governor of Georgia (and after Richard Nixon’s efforts to have the Justice Department intercede were ignored), King was released, and his father announced he was transferring his “suitcase full of votes” to Kennedy. On Election Day, black voters were crucial to Kennedy’s razor-thin margins not just in Illinois (8,000 controversially counted votes), but also in Michigan, New Jersey and Missouri.
Or consider 1968, when Hubert Humphrey had closed the once-cavernous gap between himself and Richard Nixon. With days to go before Election Day, the United States and North Vietnam were very close to an agreement on peace negotiations. Thanks to the intervention by Anna Chennault, an unofficial but well-connected Nixon campaign emissary, the South Vietnamese government balked. Had that deal been concluded by the Lyndon B. Johnson administration, there’s good reason to think that Vice President Humphrey would have won the election.
Go back to the last days of the 2000 campaign, and the disclosure of a drunk-driving arrest of a young George W. Bush. Karl Rove maintained that the story cost Bush the popular vote by keeping a few million evangelicals away from the polls. And for Democrats, that butterfly ballot in Palm Beach County will always be a Black Swan of pterodactyl-sized proportions.
Or look again at the financial collapse of mid-September 2008. I’m skeptical of claims that John McCain could have won that contest under any circumstances, given the financial resources of Barack Obama’s campaign and the country’s unhappiness with President Bush. Without question, though, the fear of economic meltdown meant a shift in the tenor of the campaign, one that that redounded in Obama’s favor.
Not every late-breaking event changes the outcome of an election. John Kerry believed that the release of an Osama Bin Laden video just before the 2004 election cost him the White House; I lean more toward a superior get-out-the-vote operation in Ohio by the Bush campaign.
And it’s not that fundamental things don’t apply. If you think in terms of probabilities rather than predictive certainty, the fall economic data is a sound guide for placing bets.
But until someone can take a quick trip into the future and tell me how Ohio’s going to vote, I’ll say no sooth.