BLS vs. ETA

Northern_trust_1

An astute reader points us to the level headed Paul Kasriel of Northern Trust in Chicago. Paul’s rep is that of a no nonsense economist, and comments like today’s are exactly why:

"The April Employment Situation report caused a radical reassessment of the state of the labor market and of the economy. The 274,000 reported increase in April nonfarm payrolls led many analysts to conclude that the first quarter slowdown in economic growth was just a temporary “soft patch.” The monthly Employment Situation report is compiled by the Bureau of Labor Statistics (BLS), a group within the Department of Labor (DOL).

There is another group within the DOL that compiles data on unemployment – the Employment and Training Administration (ETA). This is the outfit that brings us the initial and continuing state unemployment claims data on Thursday mornings. The ETA has been publishing data of late suggesting that the labor market is cooling off and that the economy is too.

So whom do you trust – the BLS or the ETA?

We place more trust in the ETA. Why? For starters, the ETA reports on real people standing in actual unemployment insurance lines. The initial jobless claims data are not based on samples with model guesstimates folded in of the number of new hires of new businesses. Although there are revisions to the initial claims data, they tend to be smaller than the revisions to the nonfarm payrolls data. Both the BLS and ETS publish employment/unemployment data seasonally adjusted as well as unadjusted. Seasonal adjusting is an art, not a science. So, we prefer unadjusted data. By the way, we are not the only ones who seem to trust the ETA initial claims data over the BLS nonfarm payrolls data when it comes to forecasting in which direction the economy is headed tomorrow rather than where it is today. The Conference Board has chosen to put initial jobless claims in its leading index and nonfarm payrolls in its coincident index."

>
Caveat: Once again, I find mysel "selectively" responding to those datapoints that support my views. But these have been so persuasive, that its hard not to . . .

>

Source:
Daily Economic Comment, May 9, 2005
Paul L. Kasriel
Economic Research Department, The Northern Trust Company
http://www.northerntrust.com/library/econ_research/daily/us/dd050905.pdf

Print Friendly, PDF & Email

What's been said:

Discussions found on the web:
  1. Jay S. commented on May 10

    In watching these numbers after a cycle 2-4 years, I feel like a Yo–Yo or is it that I am chasing my tail.
    Why does an economist tell you “one thing then say on the other hand”. Why are there no Laws in Economics only theories that can not be proven by the numbers. Why is it Larry Kudlow’s favorite indicator is Baltic freight rates to judge how the economy is doing (shouldn’t he rely on the GDP)? Why does Greenspan look at hundreds on indicators to try and guess where the economy has been. Let alone where it is going.

    The reason “they are all bad”, to quote Arnold in True Lies. I am speaking of the numbers. As a engineer with a minor in math, before you start building the plant you think about the quality of the numbers.

    To much time is spent slicing and dicing error ….or is it air.

Posted Under