[raloxk]

I think the FAIR Model is flawed for one reason.

In the past there was a strong correlation between GDP growth and Job Growth. Thus 4% GDP often correlated to over 3m new jobs (See 1994 and 1983-84 BLS.GOV).Last 6 months of 2003 had 6.1% GDP growth but only about 250,000 new jobs. Thus if we have 4.5% GDP growth for 2004 as is forecasted, the FAIR model will predict BUSH winning in a landslide. But if job growth doesnt reach the 3m-3.5m plus that should correlate with 4.5% GDP growth, the model will overstate BUSH's popular vote total.

Given past levels of job growth, in 2003 there should have been about 2m new jobs correlating to 3.1% GDP (similar to 1986 and 1995 see BLS.GOV) plus over 3m in 2004 for a total of at least 5m new jobs the past two years. BUSH would be quite popular and would be up about 3m new jobs for his term

So the model is flawed because there is now a disconnect between strong GDP growht and Job growth that didnt exist in the past.

[AntiGuv]

You are probably correct. I think it clear by this point that in the past few years there's been a major discontinuity in the significance of any number of economic reports. If the job market in particular continues to stagnate as it has, there will likely be a lot of shocked Yale economic modellers come November.

[ijcr]

Easy peasy...We just make the Pell Grant a government paycheck with one week of compulsory service and viola 5 million new jobs.

However, you are a smart cookie to focus on this issue.

[GeronL]

It was even accurate during depressions and recessions in the past.

[RightWhale]

Is jobs a real issue? Is off-shoring of jobs a real issue? If so, and it doesn't look so, that is about the only thing that might derail the Republicans

[Nick Danger]

In the past there was a strong correlation between GDP growth and Job Growth.

What makes you think there still isn't? Six months is not a long time for these things.

Here are quarterly GDP growth rates from the Reagan era:

 1982 02  5173.6   1.901%  -1.534%
 1982 03  5159.9  -1.055%  -2.395%
 1982 04  5171.2   0.879%  -1.024%
 1983 01  5204.7   2.617%   1.076%
 1983 02  5311.9   8.497%   2.673%
 1983 03  5393.7   6.304%   4.531%
 1983 04  5484.0   6.867%   6.049%
 1984 01  5590.0   7.959%   7.403%
 1984 02  5656.0   4.807%   6.478%
 1984 03  5698.7   3.054%   5.655%

As you can see, the growth spurt started in the 2nd quarter of 1983, with four months of absolutely booming (and unsustainable) numbers in the 6-8% range.

Here are the unemployment rates for 1983. Where the GDP data was quarterly, the unemployment stats are by month:

 1983 01 10.4
 1983 02 10.4
 1983 03 10.3
 1983 04 10.2
 1983 05 10.1
 1983 06 10.1
 1983 07 9.4
 1983 08 9.5
 1983 09 9.2
 1983 10 8.8
 1983 11 8.5
 1983 12 8.3

Six months after the spurt in growth became apparent in the GDP numbers, the umployment rate had not gone down even one point. It then dropped another whole point in the following 3 months.

I think it's hilarious to watch these so-called economists from the left calling this a 'jobless recovery.' If they are even minimally competent, they know about this lag effect. It's in the numbers from every recession & recovery since dirt, and it's intuitively obvious why it occurs. So here they go on TV to throw a spear at Bush, knowing for certain that what they have to say will be overrun by events within just a few months. But they do it anyway. Why? Because they don't care about their own credibility. They only care about throwing near-term spears at Bush.

[varon]

BUMP

[nwrep]

Last 6 months of 2003 had 6.1% GDP growth but only about 250,000 new jobs.

Some recent articles have pointed to drawbacks in the current BLS method of estimating employment. For one thing, it underestimates the number of self-employed individuals, which has gone up over the last generation.

[rebel_yell2]

First, you assume that this is the most important explanatory variable in the model. Second, you believe that the establishment survey is accurate during a recovery. Third, you believe that the establishment survey will not show significant job growth during the period between now and November. Perhaps we should check your assumptions . . . ?

[ninenot]

Agreed. The lag in job-creation is still a problem, and it's already mid-February (that is, we are 1/2way through the first quarter.)

Note that Edwards is making NAFTA/WTO/PNTR for China into a centerpiece in his Wisconsin campaign, and he's beginning to pull numbers his way (anecdotal, not primary evidence..)

I still think GWB wins--but not necessarily because this 'model' says so.

[the invisib1e hand]

So the model is flawed because there is now a disconnect between strong GDP growht and Job growth that didnt exist in the past.

I assume, if this is correct, that the lower-than-predicted job growth (if that is the case, perhaps the data are lagging) would be owing in large part to increased productivity and extra-national employee outsourcing by domestic corporations. Would you agree?

IMHO, extranational outsourcing can be wise long term investment.