>You and Doug have great faith that the statisticians are just fixing the
>seasonal and sampling problems inherent in the jobs numbers. I don't,
>which is why I posted the Crudele analysis in the first place. As to why
>the problem exists, I specifically said I don't think it's because of
>political bias. I just think the government underinvests in collecting
>good economic information and that the data is not what is needed for
>managing a modern economy. It's pathetic what this country exists in
>statistical information. Businesses like Wal-Mart invest larger chunks of
>their available capital on information, because they believe that those
>investments will help save them far more money than they spend. Yet the US
>government continues to starve information-collecting agencies far below
>their requests for additional funding.
>
>So they end up making pretty speculative adjustments to the numbers that,
>given a shifting economy, may retroactively be "correct" for the past, but
>are unlikely to be correct as the next turn in economic change undermines
>the statistical assumptions of their model.
Most of what you say here is wrong. Crudele's work isn't "analysis" - it's assertion and inuendo. He's gotten things very wrong that I've emailed him about without response. The BLS's monthly employment estimates have improved markedly with their new techniques - last year's benchmark revision was quite small by historical standards. It would be nice if they had more resources to work with, but they're doing a very fine job now. I doubt you'd be making these criticisms if there were a Dem president.
There's nothing "speculative" about the adjustments. They're based on statistical models that have been tested and refined for decades. You sound like a creationist denying the scientific evidence for evolution.
Doug