A sudden interest in TMTGM by the BLS
Friday, February 19, 2010
Well, someone at the Bureau of Labor Statistics has been poking around here this afternoon and they spent a fair amount of time here. Hey! That's my taxpayer money, isn't it?
I've not heard from anybody on this subject yet but I did go through some calculations with some of the other categories in the most recent inflation data as a sanity check and, as far as I can tell, they've got an error in today's report as noted here earlier today.
It still seems to me that, despite what you may have read in the mainstream media and elsewhere today, monthly core inflation did not decline for the first time since 1982.
5 comments:
Holy Moly Tim, is this pans out it will be great effort on your part. Nice catch.
This will be huge.
Can you post some of the example calculations you've done so we can verify,
It's Friday, who can do math on a Friday!
It would take a good deal of time to write it up in presentable form so, maybe next week. The check that I did was fairly simple though and it used data from here.
Just take any category with multiple subcategories and, first check that all the subcategory weightings add up to the total category weighting. Then apply the subcategory weighting to the individual monthly change from the right-most column and see if the weighted subcategory total equals the overall category total.
I did this for data in the Food and beverages category and the Other goods and services category and both came out within 0.1 percentage points, whereas, it's off by about 0.4 percentage points for shelter (-0.5 instead of -0.1) which makes a big difference in core inflation since this accounts for about 40 percent of that figure.
If it worked for other categories, then I am with you. Maybe BLS will have a press release on Sunday correcting this.
Post a Comment