(no subject)

[weaktwos]

The weather in the Sacramento area is lousy today! Lots of rain and high winds (20-40mph). It made the drive to work simply precious.

In the news:
http://www.salon.com/tech/feature/2004/02/25/jobs_economy/index.html

Why is this man smiling?
There are few new jobs -- and even fewer new good jobs. But the economy is growing, and if history holds, George W. Bush will be reelected easily.
…
In the mid-1970s, after poring over economic data, Fair developed a mathematical equation that factors in economic growth rates, inflation and an incumbent's general record on economic matters to arrive at an estimate of the popular vote on Election Day. His model rests on a simple truth -- the prevailing economic conditions of an election year provide an excellent guide to voters' behavior in November -- and it is stunningly accurate. Fair's equation predicts the popular vote to within 2 or 3 percentage points for almost every presidential election since 1920. He forecast the toss-up in 2000, the reelections of Bill Clinton and Ronald Reagan, George H.W. Bush's victory in 1988, and Jimmy Carter's loss in 1980. And based on the current health of the U.S. economy, Fair predicts an easy win for George W. Bush in November.
This article goes on to give some sage advice that the Democrats best not rest their entire platform on the unemployment rate.

Comments

[imlac.livejournal.com]

You know, I hear so many of these models that claim to be able to make such predictions, but it's all just Nostrodamus, in my opinion.

One model says if you look at all the presidential races wherein the president was elected by less than a 2% margin, in the next election the encumbant will not be reelected. This holds true for about a dozen cases dating back to the start of the office, and there is not a single counterexample to it. Based on that, Bush should definitely loose.

There are probably dozen of models with equally accurate retrodictive success, half proclaiming one one, the other half proclaiming the other. Proving their PREdictive abilities is another matter entierly. Like Hume says, the future is under no obligation to mimick the past.

[weaktwos.livejournal.com]

I agree. I don't think there is enough data to justify any reasonable predictability. I was more intrigued (though I did not take the time to express this) by some other issues this article brought up over and above election predictability.

Namely, the economy is growing by 3.1 percent this year, yet unemployment is not drastically improving, our deficit is bulking up, Bush runs away from his bold predictions of new jobs by the CEA, and finally, Greenspan recommends we cut social security to prevent more economic problems in the future.

The article does question whether numbers are being massaged, because these various factors do not seem to add up properly.