God bless you, Jared, for estimating the October value for Shelter Index CPI using the geometric mean, rather than the arithmetic mean. In this case, the difference is only 0.0002, but foot soldiers in battle for correct math never reason why.
Thanks for the useful information. It great to see a person with the background to see the hole in the analysis that others are pushing. A person who will ask the hard questions. Hope you continue with this site and columns. I been reading your material for years. Always great reading.
I think we are going to see mild inflation data and rising unemployment numbers reported from BLS no matter what through the next 5 months because the administration is hell bent on getting the Fed to lower interest rates without spooking the bond market ala April 2025.
They might get their wish, though with a sudden whiplash once it comes out that they snookered the market to do it.
In that scenario, reality will come fast and hit hard. It’s worth reminding voters that DOGE gutted the BLS to where they are no longer staffed to faithfully determine accurate numbers, combine that with firing the leadership,
It creates the conditions to siding with the admin’s coerced, preferred outcomes.
I know this is heresy, but is there any chance that somebody’s been cooking the books? Trump’s speech was Wednesday night. Then Thursday morning his administration publishes numbers that “sorta” confirm his claim that he has tamed inflation. That’s a really, really “lucky coincidence”.
"The BLS just assumed [shelter costs] were zero for October." ??? That's a big assumption. When the economy seems to be at an inflection point, it's really unfortunate to have big holes in the data. It means the Fed (and others?) will have to do more tea leaf reading than usual to determine what/when actions to take in the face of a likely slow down, that might involve stagflation.
"The BLS just assumed [shelter costs] were zero for October. I am sure they have a good technical explanation for this, but the only way you get a two-month average for rent of 0.06% and OER at 0.135% is assuming October was zero."
my mind finds all sorts of parallels these days.
Thus, the issue of the Blase/Paragon graphic that I find questionable, which I brought up yesterday, here:
I suspect Dr. Blase / Paragon has treated data they did not have conveniently available, or was otherwise inconvenient to their intended conclusion, as 0.
(That data item being the dollar value of the direct payments for Cost-Sharing-Reductions to insurers--which stopped in 2017--going into the second lowest cost silver plan for 50 year olds on the ACA exchange plans.)
(Perhaps historians with a quantitative bent 50 years hence will be able to summarize all of the errors and deceptions of the current administration and the Republicans in Congress into say 5 or 6 principles which explain every one of them. A victory for parsimony!)
"The BLS just assumed [shelter costs] were zero for October. I am sure they have a good technical explanation for this, but the only way you get a two-month average for rent of 0.06% and OER at 0.135% is assuming October was zero."
In my past life as a statistician, I, working in SAS (Statistical Analysis System) came up with some numbers that disagreed with the numbers calculated by the person running the study, who was working in MS Excel.
After tracing this to MS Excel (at the time, possibly using options set by the person running the study) treating blank (missing data) as 0 (raising the "N" used for the average), it became clear that my calculation was the correct one.
(I have made mistakes in my day, but this was not one of them!)
God bless you, Jared, for estimating the October value for Shelter Index CPI using the geometric mean, rather than the arithmetic mean. In this case, the difference is only 0.0002, but foot soldiers in battle for correct math never reason why.
The hole in shelter data makes me feel like a frog who is swimming in slowly warming water.
Thanks for the useful information. It great to see a person with the background to see the hole in the analysis that others are pushing. A person who will ask the hard questions. Hope you continue with this site and columns. I been reading your material for years. Always great reading.
I think we are going to see mild inflation data and rising unemployment numbers reported from BLS no matter what through the next 5 months because the administration is hell bent on getting the Fed to lower interest rates without spooking the bond market ala April 2025.
They might get their wish, though with a sudden whiplash once it comes out that they snookered the market to do it.
In that scenario, reality will come fast and hit hard. It’s worth reminding voters that DOGE gutted the BLS to where they are no longer staffed to faithfully determine accurate numbers, combine that with firing the leadership,
It creates the conditions to siding with the admin’s coerced, preferred outcomes.
The BLS is part of the Trump administration, right? Wouldn’t another plausible explanation be that they are cooking the books?
Yours is the ONLY rational conclusion.
I know this is heresy, but is there any chance that somebody’s been cooking the books? Trump’s speech was Wednesday night. Then Thursday morning his administration publishes numbers that “sorta” confirm his claim that he has tamed inflation. That’s a really, really “lucky coincidence”.
"The BLS just assumed [shelter costs] were zero for October." ??? That's a big assumption. When the economy seems to be at an inflection point, it's really unfortunate to have big holes in the data. It means the Fed (and others?) will have to do more tea leaf reading than usual to determine what/when actions to take in the face of a likely slow down, that might involve stagflation.
Again, on Mr. Bernstein's:
"The BLS just assumed [shelter costs] were zero for October. I am sure they have a good technical explanation for this, but the only way you get a two-month average for rent of 0.06% and OER at 0.135% is assuming October was zero."
my mind finds all sorts of parallels these days.
Thus, the issue of the Blase/Paragon graphic that I find questionable, which I brought up yesterday, here:
https://econjared.substack.com/p/just-how-worrisome-is-the-us-labor/comment/188874906
actually comes down to the same thing.
I suspect Dr. Blase / Paragon has treated data they did not have conveniently available, or was otherwise inconvenient to their intended conclusion, as 0.
(That data item being the dollar value of the direct payments for Cost-Sharing-Reductions to insurers--which stopped in 2017--going into the second lowest cost silver plan for 50 year olds on the ACA exchange plans.)
(Perhaps historians with a quantitative bent 50 years hence will be able to summarize all of the errors and deceptions of the current administration and the Republicans in Congress into say 5 or 6 principles which explain every one of them. A victory for parsimony!)
I have an anecdote around Mr. Bernstein's:
"The BLS just assumed [shelter costs] were zero for October. I am sure they have a good technical explanation for this, but the only way you get a two-month average for rent of 0.06% and OER at 0.135% is assuming October was zero."
In my past life as a statistician, I, working in SAS (Statistical Analysis System) came up with some numbers that disagreed with the numbers calculated by the person running the study, who was working in MS Excel.
After tracing this to MS Excel (at the time, possibly using options set by the person running the study) treating blank (missing data) as 0 (raising the "N" used for the average), it became clear that my calculation was the correct one.
(I have made mistakes in my day, but this was not one of them!)
All interesting for the eco-wonks, for mere sans culottes it’s the inflation and unemployment figures that drive public sentiment.
What do I notice?
Gasoline prices down
Beef prices way up
Coffee cup from my neighborhood bodega way up, owner says all his costs are way up …
Are we teetering on the edge or blue skies and a Dow continuing to charge ahead?