Thursday, March 04, 2004

Taking down bithead ... an economic argument

NOTE: This posting originally went up at Dan Drezner's weblog where I defend the notion that during the 90's the composition of the unemployment numbers said a very different story from what they show today.

Mr. Bithead,

You wrote:
You provide only one point on which to judge, a point which as presented, gives a slanted view with no possiblility of accurate comparison with previous administaton actions and I'm projecting narrow thinking?

I was making a quite similar point to what Stout makes.. in his words "...why does the media only pull numbers comparable to yours during Republican administrations and not during Democratic administrations?"

I happen to know they're not. You see, I've actually looked, instead of taking my numbers form the Democrats. (Even the raw numbers at their peak values, are worse, much less the calculated ones you're attempting to tar Mr. Bush with) Whence comes the questions that I asked.

I'd be curious to see the numbers that you have. You see if you look at the historical Household Survey data from the BLS some things become clear.

For instance, let me set you up a chart


Now if you 1992 to the present you can see that the difference between the non-institutional population and the Percent Employment in 2003 is closest in appearance to 1994.

Now if you look here at labor force particapation numbers you can see that the non-institutional pop numbers are almost identical to the labor force participation numbers for ages 16&up.

So what we're comparing is labor force participation against percent employment. If you do that, you can see that there has been a strong drop - the DIFF - column from peak.

Moreover, this drop is stronger if you compare the maximum labor force participation with the current employment percentage:
67.1% - 62.3% = 4.8%.

So from peak we can see that normalized to population there is a 4.8% gap in the people who could work and the people who have work.

That's significantly comparable only with 1992, which had a 4.9% DIFF.

Finally, we look at average and median time unemployed ... if you look here at this BLS time series request and ask for the average and median time unemployed ... you can see that the time from 1992 declined while the average and median time unemployed was still growing through the beginning of 2004.

So what does all this mean? Well it means that according to these statistics that normed to the population, that from 1992+ the unemployment rate dropped and that the gap between the percent of people who could work and had worked declined. Likewise, from 2000+ the gap between the people who had work and could work grew, and that the median and average time unemployed has been steadily growing even as the nominal unemployment rate has contracted.

If you ask on the BLS request page for the graph and for the time series to be set from 1992-2002 the results are striking. You can also play with the percent unemployed at 15+ weeks. It confirms that through the Clinton years, unemployment and time unemployed steadily declined while it's been steadily increasing 2000+ during the Bush years.

These are verifiable and presented facts. I don't know what the hell you were looking at, but your arrogant attitude is completely unsupported by actual numbers.

You might want to consider Bithead that the reason that the oldman argues so authoratatively is that he is right so often. Confidence comes from success time and time again. You might want to try instead of sounding off like a wanker. Or maybe you should stop getting your numbers from the RNC.


Post a Comment

<< Home