Friday, July 23, 2004

"Checking back with top predictor"

From the Dayton Daily News by Martin Gottlieb:
OK, you’re probably wondering who’s going to win the election.

To put it more precisely, the predictive system that was reported here in January to be saying “Bush,” and again in March, is still saying “Bush.”

That fact is noted here because the system has arguably never been wrong.

It will be wrong eventually, of course. It’s only human. If this is the year, that is not all bad in the eyes of the system’s creator, Professor Allan J. Lichtman of American University in Washington, D.C. He is for Kerry. But he is duty bound to report what the system says.
Okay, I'm with you so far, Marty...
If he is wrong, he will take comfort in being, by his count, 6-1 in presidential elections. And he’ll take more comfort in the prospects of the nation and the world. (He is also predicting that a second George W. Bush term would be a calamity. But he has no system behind him on that one.)
Now dropping a bomb like that is just uncalled for...if you aren't going to explain what the good doctor means, you are better off leaving the comment out.
He bases his elections predictions on systematic study of all the presidential elections in the history of the current two-party system. He has found 13 factors to be most frequently present when the incumbent party wins. None of them is always present. It’s the overall pattern he’s looking for.

This system is better than systems based on one factor — the president’s midyear approval rating in the polls, say, or the state of the economy, or the height of the candidates — because 13 is more than one.
So far so good...
(This isn’t getting too technical, is it?)
Yes, you are so much more superior than the foolish readers who buy your pathetic little rag... This sort of smugness has no place in journalism.
History shows that if eight or more of the 13 factors are present, the incumbent party wins the popular vote. Why eight? No reason; that’s just the way it works out.

Bush has the following factors going for him: (1) His party has gained seats in the House of Representatives over the past two elections. (2) He will breeze to renomination. (3) He is the incumbent president (as opposed to his party putting up a new person).

(4) There is no significant third candidacy threatening him. (5) The economy in the election year is not in recession. (6) There has been no major social unrest — meaning violence — in this country on his watch.

(7) There has been no major scandal (see below). (8) He has achieved a major international success, with the war in Afghanistan going easily (see below). And (9) his opponent is not charismatic or a major national hero.
I'm with you here...mainly because you are spitting out the information from the good doctor...
Some of these are judgment calls, of course. Many Democrats are outraged at the suggestion that there’s been no major scandal, given especially that the Bush rationale for war has collapsed.

Some people, too, question whether the Afghan war should be considered either major or a success, given that Osama bin Laden has not been captured and that al-Qaida lives.
There are Bill Clinton parallels to be made here...
But Lichtman must apply the standards as they have applied in previous elections. He can’t change the definition of “scandal” or “success” now, lest the whole system collapse.

In the past, a military victory need not have been against a powerful enemy to be counted as a success. It need only have advanced American interests in the world in some substantial way. Depriving al-Qaida of its sanctuary was almost universally seen as doing that.

As for scandal, sorry, it has to be about genuine wrongdoing, not about being wrong. And it has to be generally perceived as a scandal, not just apparent to the president’s detractors. And it should be formalized by some institution, meaning indictment, conviction, or impeachment of the president or people near him.
This doctor is smart, yes?
Another predictive system is getting a lot of attention this year. It is based upon presidential approval ratings. Specifically, presidents whose public approval ratings are — like George W. Bush’s — below 50 percent in midyear lose.

However, there has been one exception: Harry Truman, who stood at 39 percent in a May/June Gallup Poll in 1948.
Like polls, approval ratings really don't mean much this early...
Moreover, there have only been a few test cases to look at, because polling is relatively new, and because the approval rating can only be considered when an incumbent president is on the ballot.

By contrast, all of Lichtman’s questions about an election can be answered about every election since the Civil War.
It's confirmed...this guy is smart!
Still and all, some people don’t want to believe. Democrats find all kinds of rationales for seeing a Kerry victory: Iraq is just such an embarrassment to Bush. And Democrats are riled up as never before in memory. And they’ve got money. And the Internet has changed everything. And Michael Moore, Al Franken and the gang are finally fighting back against the right-wing propagandists.

In truth, though, every time there’s an election that even looks like it might be close, people find reasons to believe what they want about the future. It’s easy enough to find them — always.

Which is fine. Sometimes people complain that a system like Lichtman’s threatens to take all the fun out of politics, to make everything seem predetermined and not subject to human effort. But that would be giving the system way too much credit.
The reality is that Iraq isn't an embarrassment. The Democrats ARE riled the point of coming unhinged. The Internet really hasn't changed all that much (and that is coming from a blogger!). And Michael Moore and Al Franken are preaching to the choir while the congregation has left the church.

In truth, as Hugh Hewitt says, "If it isn't close, they can't cheat." And believe me when I say that if it is close, we'll see all kinds of shenanigans. Sometimes belief isn't have to have the votes.

Lichtman's system appears to be quite accurate. I don't know about taking the fun out politics though, the politicans do that pretty well enough on their own. Human effort is still required in order to win an election. People still have to show up at the polls and do what they need to do to make the prediction a reality.
Those of us who believe in it aren’t betting any houses on its accuracy. I harp because it’s kind of fun, because it’s the best system, and because it offers important insights into how elections works. However, being a system devised by humans, it is fallible.
...or so you hope.

Me? I'm voting for Bush/Cheney and think that you should too.


Post a Comment

<< Home