Wednesday, September 30, 2009

Finding a Job May Be Getting Easier Than It Looks

"Sept. 29 (Bloomberg) -- Employment in the U.S. is rebounding at a “significantly faster” pace than seasonally adjusted data on jobless claims would suggest, according to Oscar Gruss & Son Inc."
First, some definitions -

Seasonally Adjusted Data:
Observations over time modified to eliminate the effect of seasonal variations. For example, students tend to flood the job market in the summer after graduating, increasing the number of workers seeking employment.

Continuing Jobless Claims: The number of U.S. residents collecting state unemployment benefits. Usually, people who have worked full time are eligible for 26 weeks of regular state unemployment benefits.

A very optimistic report.
What this news report/analysis is saying is that seasonally adjusted data on continuing jobless claims shouldn't be trusted and that unadjusted continuing jobless claims should be looked at:
“Our assumption is that the sheer brutality of the current cycle has caused the statisticians to cease to trust the ‘raw’ data and therefore fall into the trap of abusing the process of seasonal adjustment.”
My question is this: Since when have statisticians EVER trusted 'raw' data?

In Oscar Gruss & Son Inc.'s attempt to paint a positive picture of the U.S. economy and the employment picture, they've decided to look at unadjusted continuing jobless claims (to the left) versus the more accurate seasonally adjusted continuing jobless claims (to the right).

In Intro to Econometrics, we always learned that you always have to adjust numbers such as employment data for seasonal change. For example, if we were looking at an ice cream vendor, it's obvious that they would sell more ice cream in the summer versus the winter. Comparing the sales of ice cream in the winter versus the summer would be like comparing apples to oranges. It makes a whole lot of sense to adjust for the seasons.

Likewise for any employment data, students flood the job market in the summer. Summer tends to be slower for job recruitment. People tend to take holidays in the summer (and perhaps hold off looking for jobs).
It sure makes a whole lot of sense to seasonally adjust continuing jobless claims.

Looking at the subject title: "Finding a Job May Be Getting Easier Than It Looks", I can't help but think that it might not be so true after looking at the seasonally adjusted chart. Rather, if I had to look at the charts above and make my own title, I'd headline with this:

"Finding a Job Remains Difficult as Trend Of Job Losses Remains Flat"

No comments:

Post a Comment