Hi doma,
I was hoping you would read the employment report and represent your own opinion having done so. There's a lot more to it than Table B1. Unless you are determined to stick tightly to Biderman and nothing more than Biderman - in which case I guess you are invincible against discussion.
Is this a conversation about whether US employment is increasing and unemployment is decreasing?
Because the answer to that is very clear in the numbers as reported, adjusted or unadjusted. If you are concerned by seasonal adjustment, then compare the unadjusted figures from January 2011 and similar figures for January 2012. That way, you self-adjust for seasonality.
Unadjusted employment figures show an increase year over year of 2.3m. They show a drop in unemployment.
There's not much else to add. If Biderman doesn't like the way the adjustment occurs, then he can employ unadjusted figures year over year and see the same trend. I don't see him challenging the unadjusted numbers. Just the method of adjustment.
And if he is suspicious of the adjustment method, that's fine, even if his suspicion is a long shot. A month to month comparison of unadjusted monthly employment is going to result in the kinds of variations you get as a result of seasonality. You explain seasonality by reference to things that occur in those seasons. Things like Christmas. Has he noticed that shops sell a lot of stuff and hire a lot of temporary staff in November and December? That is why employment grows in those months. And then, in January, those people are out of luck.
To get adjusted employment figures, economists reform these seasonal factors. This way they smooth the monthly growth curve. They can do this because they know that similar factors occur each year.
But this doesn't mean you cannot use comparable unadjusted figures from one year to the next in order to discover if the employment numbers are up. That's essentially a seasonally adjusted measure with no human intervention in the adjustment process - or to put it another way, there's no need for a seasonal adjustment when comparing the figures in a particular month in any year with the same month in another year. Every January has the same seasonal factors, right.
I think what you have reported as Biderman's opinion should be listed under the title: "Biderman discovers the seasons". There's nothing else to it that I can see. Unadjusted month to month figures simply aren't very useful.
| 