Archive for October, 2010

Lies, Damn Lies and Statistics

October 7, 2010

I’m always suspicious when people make claims about Presidential elections based on patterns and history.  This is an event so rare and infrequent that only 43 people in our country’s history have ever won. (Grover Cleveland was President #22 and 24).

So when I read Peter Beinart’s article titled “Obama’s a Lock in 2012” this morning, I was — to say the least — skeptical.  The strongest predictor Beinart claims is this:

For starters, American presidents usually get reelected. In the last 75 years, incumbents have lost a grand total of three times: in 1976, 1980, and 1992. And what did Gerald Ford, Jimmy Carter, and George H.W. Bush all have in common? They had serious primary challenges within their own party (from Ronald Reagan, Ted Kennedy, and Pat Buchanan, respectively). The last president who lost reelection without a major primary challenge was Herbert Hoover in 1932.

This is silly on a number of levels.  First, a strong primary challenge is not a CAUSE of vulnerability, it’s a RESULT.  Does anyone think Reagan won re-election in 1984 because he didn’t have a strong primary challenger?  Of course not.  He didn’t have a strong primary challenger because he (unlike Obama) actually was a lock for re-election.  The only person foolish enough to run against him was Walter Mondale who, in spite all of his campaigning, only got 13 more electoral votes than I did.  Serious primary challengers emerge when a President is already vulnerable, not the other way around.  Granted, the expense and infighting of a primary challenge is not good for the incumbent, but a primary challenger is unlikely to see the light of day if the incumbent is strong to begin with.

But the real knee slapper in Beinart’s writing is his selective horizon of incumbent re-election prospects.  Note he says, “In the last 75 years, incumbents have lost a grand total of three times: in 1976, 1980, and 1992.”  Hmmmm . . . if all three happened in the last 24 years, why go back 75 years for the statistic?  Here’s another equally accurate way of stating the same information:

In the last 24 years, half the incumbents to seek the Presidency lost.

Ford ’76 – defeated
Carter ’80 – defeated
Reagan ’84 – won
Bush ’92 – defeated
Clinton ’96 – won
Bush ’04 – won

To the limited extent past Presidential elections have predictive value for future elections, one would think that more recent elections are better predictors than those of the 30’s, 40’s, 50’s and 60’s.  But Beinart lumps in elections my grandparents were too young to vote in so as to make Obama’s re-election  seem inevitable.

2012 is a long way off in political times.  But to state that Obama’s re-election is “a lock” and “that we’re still probably in the early stages of an era of Democratic dominance” isn’t just whistling past the graveyard. It’s willfully ignoring the will of the people.  Which, come to think of it, seems to be Obama’s legislative strategy in the first place.