The Lessons of History
This post by Matthew Yglesias reminds me of something I hate about political commentary: Using the past to predict the future. In this case, we read that unemployment trends for the first six months of an election year are a sure-fire indicator of which party will win the White House.
The problem is that every election is different. I understand that political scientists need models to work from in describing voting behavior, but especially as applied through the media, the predictive value of these models seems limited at best. Remember 1998? All the talk was about how the President's party always loses Congressional seats in an off-year election. Except it didn't happen then, and it didn't happen again in 2002. So much for that rule of thumb.
Presidential elections within my memory follow a similar pattern. In 1988, we heard how sitting Vice Presidents can't win. After all, it hadn't happened since Martin van Buren. Enter Bush '41, and in 2000 Al Gore gets a free pass on the issue. In 1992, we heard how no candidate has won the Presidency after losing the New Hampshire primary. Paul Tsongas won New Hampshire, Clinton became President. And in 2000, John McCain won New Hampshire while Bush '43 became President. The status of the economy is always supposed to invariably predict election results. Well, Clinton did win in 1996, but didn't break the 50% mark in the popular vote, while Al Gore couldn't win in 2000.
With Howard Dean, we're hearing a lot of can't-win rhetoric of the type described above. They say, for example, that a candidate can't win if they oppose a war. This appears to be based on exactly one election year - 1972. They say New Englanders can't win nationally. This seems to derive solely from 1988, the only time since Kennedy one has been nominated.
We have a tough campaign ahead. But when you hear the pundits on TV explaining why he can't win, remember: History's main lesson in politics is that history doesn't predict a thing. The only election year we need to worry about is this one.
The problem is that every election is different. I understand that political scientists need models to work from in describing voting behavior, but especially as applied through the media, the predictive value of these models seems limited at best. Remember 1998? All the talk was about how the President's party always loses Congressional seats in an off-year election. Except it didn't happen then, and it didn't happen again in 2002. So much for that rule of thumb.
Presidential elections within my memory follow a similar pattern. In 1988, we heard how sitting Vice Presidents can't win. After all, it hadn't happened since Martin van Buren. Enter Bush '41, and in 2000 Al Gore gets a free pass on the issue. In 1992, we heard how no candidate has won the Presidency after losing the New Hampshire primary. Paul Tsongas won New Hampshire, Clinton became President. And in 2000, John McCain won New Hampshire while Bush '43 became President. The status of the economy is always supposed to invariably predict election results. Well, Clinton did win in 1996, but didn't break the 50% mark in the popular vote, while Al Gore couldn't win in 2000.
With Howard Dean, we're hearing a lot of can't-win rhetoric of the type described above. They say, for example, that a candidate can't win if they oppose a war. This appears to be based on exactly one election year - 1972. They say New Englanders can't win nationally. This seems to derive solely from 1988, the only time since Kennedy one has been nominated.
We have a tough campaign ahead. But when you hear the pundits on TV explaining why he can't win, remember: History's main lesson in politics is that history doesn't predict a thing. The only election year we need to worry about is this one.
Comments