The distinction between manipulation and neglect is one worth making loudly, because the two get conflated constantly in public discourse right now. Firing the BLS commissioner after inconvenient revisions looks like interference, so the leap to 'the data is cooked' feels intuitive... even when the career staff are the ones actually holding the line.
Recently, Brian Moynihan, the CEO of Bank America went on a financial news channel and criticized the lack of quality of government agency data and how much better is corporate data. Fine. Maybe if BOA and other corporate leaders sought to pay a decent tax rate, federal agencies could hire more staff and upgrade software systems.
Corporate America asks that a crew build a quality house, but it won’t buy you a tape measure.
Great piece. Leaves me looking in the mirror on how to better address misunderstandings and jargon-y explanations of regular data reports. If the message doesn’t land or improve understanding, it wasn’t the right message.
"The current problem is not political interference; it is chronic and intensifying underinvestment". This could easily be construed as political interference of sorts, especially under the current regime.
As of now, we can only hope for a future government to take its responsibility seriously enough to make the investments needed.
The underinvestment into gathering sufficient and sufficiently reliable data is really important because that is going to destroy trust in economic statistics (and all government statistics) very quickly as the statistical reliability decreases. This is one of many important topics that voters should understand when making decisions about how they should vote. Thank you for this post.
I'm very grateful to all of the Substack authors spending their time to provide substantive and important information. I would like to be able to pay for subscriptions to all of them but it's just not economically feasible for most of us to subscribe to more than a few (5-10) at any one time. I subscribe to about 60 Substacks right now but can only afford to pay for 7 even though almost all seem to provide content that is likely of value commensurate with the subscription cost.
Unfortunately our media has become so fragmented that it's very difficult to find media that is well curated and provides substantive and diverse content. It would be great if there was a way that there could be one Substack that could collect multiple feeds into one for a single subscription fee much like what is available with Apple News.
There may be underinvestment into accurate data which is probably intended as the Ruling Elite want to treat us like mushrooms, and it is clear besides underinvestment, the gov't just changes definitions to suit their interests.
Speaking of, the Consumer Expenditure Survey latest release 2024:Q2 - 2025:Q1, fails to calculate after-tax income, using the TAXSIM program as it had done since 2014 - So no taxes are enumerated. I'm having to estimate, meaning most people will have no idea of US after-tax income for reference. It's a darn shame!
The Feds manipulate definitions to hide the data truth ALL the time. Government bodies just manipulate definitions to make things seem better than they are. Here's just ONE example of data manipulation by the government, their data manipulation of Unemployment rates....The 8.7% U-6 rate for January 2026 can be compared to the official U-3 unemployment rate of 4.6% for the same month , highlighting the significant number of people who are unemployed. For instance it doesn't talk about 7.5 million U3 folks who are unemployed which means 14.1 million U6 folks are unemployed...4.6% sounds better than 14.1 million unemployed folks as U6 takes into account factors like underemployed, part-time wanting full-time, and discouraged folks not looking for employment, AND if you haven't been looking for a job within the last 4 weeks, you're not included as unemployed. Where there is smoke, there is fire.
According to the official data released by the U.S. Bureau of Labor Statistics (BLS) on February 11, 2026, the U-6 unemployment rate for January 2026 was 8.7% (not seasonally adjusted). Seasonal adjustment is another manipulation, just like excluding food and energy from the CPI like government folks like to do.
Seasonal adjustment is merely an average the purpose of which is to smooth out the noise that inevitably results from large scale seasonal hiring and firing. You can call that a manipulation if you wish, but it's not done with intent to deceive.
I believe it is done with intent to deceive as the Elite treat us like children. Give us the true data and the trends will be clear and can be explained at that time.
The lower level career professionals aren't the elites. Indeed, Dr. Sahm made that point in this very article. As a general rule, statistical data is never absolutely perfect. There's always a margin of error. That's why they use averages and standard deviations.
Nevertheless, statistical data provides a window into what happened, but it's a big picture painted with broad strokes.
Of course not. The elite use the government agencies to do so through the politicians they have bought and paid for via legal machinations. I think we just disagree and any further comments are redundant and circular. Take care.
What I do like is the truth. I don't like the ruling elite oligarchs and their ring kissers manipulating data in any way ex. with definitions they create and employ to do so. I do NOT believe there's a genuine democracy on the face of the Earth. They're all Kleptocracies, including the US.
The U-6 data is broader than the U-3 and includes different things. Rather than compare unlike things against one another, compare the history of U-6 (or U-3, or any stat) against itself. In either case you’ll find those unemployment measures are quite a bit lower than the long-term average.
Hi Claudia, jobs numbers are really only growing in in home adult care. These are low paid jobs. If this trend continues, how do you see it impacting growth?
If January statistics are seasonally adjusted, and a great number of workers are laid off, what happens to them? Do they drop out of the labor force? Appear in unemployment statistics? Something else?
Thank you for saying this. It’s been so frustrating to see the lack of trust in data.
The distinction between manipulation and neglect is one worth making loudly, because the two get conflated constantly in public discourse right now. Firing the BLS commissioner after inconvenient revisions looks like interference, so the leap to 'the data is cooked' feels intuitive... even when the career staff are the ones actually holding the line.
Good insight. Rather reassuring. I only wish that the FDA and the CDC had the bureaucratic stability of the BLS.
That's hard to accomplish with a heroin addict in charge.
We all live within the economy, not the data we derive about it. Great column, CS, on so many different, but related, issues.
Recently, Brian Moynihan, the CEO of Bank America went on a financial news channel and criticized the lack of quality of government agency data and how much better is corporate data. Fine. Maybe if BOA and other corporate leaders sought to pay a decent tax rate, federal agencies could hire more staff and upgrade software systems.
Corporate America asks that a crew build a quality house, but it won’t buy you a tape measure.
Great piece. Leaves me looking in the mirror on how to better address misunderstandings and jargon-y explanations of regular data reports. If the message doesn’t land or improve understanding, it wasn’t the right message.
"The current problem is not political interference; it is chronic and intensifying underinvestment". This could easily be construed as political interference of sorts, especially under the current regime.
As of now, we can only hope for a future government to take its responsibility seriously enough to make the investments needed.
You're sympathetic with people accusing you of being part of a conspiracy, you're very open minded 😉
Very interesting post, as usuel.
The underinvestment into gathering sufficient and sufficiently reliable data is really important because that is going to destroy trust in economic statistics (and all government statistics) very quickly as the statistical reliability decreases. This is one of many important topics that voters should understand when making decisions about how they should vote. Thank you for this post.
I'm very grateful to all of the Substack authors spending their time to provide substantive and important information. I would like to be able to pay for subscriptions to all of them but it's just not economically feasible for most of us to subscribe to more than a few (5-10) at any one time. I subscribe to about 60 Substacks right now but can only afford to pay for 7 even though almost all seem to provide content that is likely of value commensurate with the subscription cost.
Unfortunately our media has become so fragmented that it's very difficult to find media that is well curated and provides substantive and diverse content. It would be great if there was a way that there could be one Substack that could collect multiple feeds into one for a single subscription fee much like what is available with Apple News.
There may be underinvestment into accurate data which is probably intended as the Ruling Elite want to treat us like mushrooms, and it is clear besides underinvestment, the gov't just changes definitions to suit their interests.
Speaking of, the Consumer Expenditure Survey latest release 2024:Q2 - 2025:Q1, fails to calculate after-tax income, using the TAXSIM program as it had done since 2014 - So no taxes are enumerated. I'm having to estimate, meaning most people will have no idea of US after-tax income for reference. It's a darn shame!
Data users should be much more suspect of data that is “imputed” because item or respondent nonresponse. Presently, both types are at all time high.
The Feds manipulate definitions to hide the data truth ALL the time. Government bodies just manipulate definitions to make things seem better than they are. Here's just ONE example of data manipulation by the government, their data manipulation of Unemployment rates....The 8.7% U-6 rate for January 2026 can be compared to the official U-3 unemployment rate of 4.6% for the same month , highlighting the significant number of people who are unemployed. For instance it doesn't talk about 7.5 million U3 folks who are unemployed which means 14.1 million U6 folks are unemployed...4.6% sounds better than 14.1 million unemployed folks as U6 takes into account factors like underemployed, part-time wanting full-time, and discouraged folks not looking for employment, AND if you haven't been looking for a job within the last 4 weeks, you're not included as unemployed. Where there is smoke, there is fire.
So you are using published government data (U-3 vs U-6) to complain about government manipulation of data.
You have not proven your point, although it is a good example of irony.
According to the official data released by the U.S. Bureau of Labor Statistics (BLS) on February 11, 2026, the U-6 unemployment rate for January 2026 was 8.7% (not seasonally adjusted). Seasonal adjustment is another manipulation, just like excluding food and energy from the CPI like government folks like to do.
Seasonal adjustment is merely an average the purpose of which is to smooth out the noise that inevitably results from large scale seasonal hiring and firing. You can call that a manipulation if you wish, but it's not done with intent to deceive.
I believe it is done with intent to deceive as the Elite treat us like children. Give us the true data and the trends will be clear and can be explained at that time.
The lower level career professionals aren't the elites. Indeed, Dr. Sahm made that point in this very article. As a general rule, statistical data is never absolutely perfect. There's always a margin of error. That's why they use averages and standard deviations.
Nevertheless, statistical data provides a window into what happened, but it's a big picture painted with broad strokes.
Of course not. The elite use the government agencies to do so through the politicians they have bought and paid for via legal machinations. I think we just disagree and any further comments are redundant and circular. Take care.
The data is there.
You like U-6? Go for it.
You dislike seasonal adjustment? Great, but then you can't compare month-to-month without a lot of statistical noise.
Fortunately there is a lot of history, so you can use year over year numbers to normalize out the seasonality.
Here you go : https://fred.stlouisfed.org/graph/?g=1SgWk
U-6 change, year over year in units of difference of percent,
So the current U-6 is half a percent worse than it was a year ago, and yes, that is about 800k people.
Pretending the data doesn't exist weakens any potential point you are trying to make. But you still haven't made a clear point.
Where there's smoke there's fire is hardly illuminating.
A nod's as good as a wink to a blind bat.
What I do like is the truth. I don't like the ruling elite oligarchs and their ring kissers manipulating data in any way ex. with definitions they create and employ to do so. I do NOT believe there's a genuine democracy on the face of the Earth. They're all Kleptocracies, including the US.
The U-6 data is broader than the U-3 and includes different things. Rather than compare unlike things against one another, compare the history of U-6 (or U-3, or any stat) against itself. In either case you’ll find those unemployment measures are quite a bit lower than the long-term average.
Hi Claudia, jobs numbers are really only growing in in home adult care. These are low paid jobs. If this trend continues, how do you see it impacting growth?
Amen. Maybe it's time these statistics are reported with "margin of error" estimates.
It is in the data releases.
If January statistics are seasonally adjusted, and a great number of workers are laid off, what happens to them? Do they drop out of the labor force? Appear in unemployment statistics? Something else?
Unemployment rate has a similar pattern. See this chart on FRED: https://fred.stlouisfed.org/graph/?g=1SllQ
Ah! Thank you.