Firing the BLS commissioner won’t prevent the effects of tariffs. But it will reduce American economic leadership and increase uncertainty for businesses, workers and investors.
It like you guys don’t even read the article and certainly don’t have a full understanding of how complex and vast the US economy is and the limits of forecasting and estimating employment for hundreds of millions of households given the timeframe and resources. Most likely it’s the typical MAGA mindset of if Trump is whining about it, it must be some horrible travesty without taking the time to actually understand the intricacies and details.
The report comes out the first Friday of the month, the economy is hundreds of millions of households, there is a tradeoff between precision and time as Silver discussed, how accurate can an estimate reasonably be given the resource and time constraints? The data is gathered from surveys with difficulties around response rate and sampling. The revisions months later have much more data in them from more sources. The final numbers have even more data and don’t solely rely of surveys.
Someone asked why release the early number if it’s has the most error in it? Because investors, businesses, and other decision makers would rather get a glimpse of what is happening in the short term, because despite the lack of precision the initial estimate does provide some value. It all comes down to understanding what the numbers mean, how they are derived, the underlying uncertainty in them. You don’t have to be an expert in statistical inference or Bayesian statistics to understand. You only have to break out of the strictly polarized view, not only see the black and white of the numbers are bullshit, it’s a conspiracy vs the numbers are god, worship them. The truth lives in the gray.
elitist globalist Bourgeoise luxury this actually reading and trying to understand the article.
More seriously - as an actual private equity investor, yes we want these early numbers - the actual real market consumers of this data
A. Have the staffs who are actually numerate and know how to use them - and we do know the BLS is transparent [if one feels like it, can pick it apart on stats, methodology etc] and make own-correctives.
B. Like to have the early numbers as indeed have the background and a high interest in having early indicative numbers helps in some kinds of planning (etc0
The fact that market actors with significant resources and economic interest in BLS numbers for Business and not Political reasons are not on some jihad against BLS already tells us all something - that this whole affaire is a political one and not actually about good government or anything like that.
Now the POLITICAL people and innumerate journalists who don't ... mmmm -well that's politics and journalistic story telling by probably 3/4 of journalists who maybe had at most econ 101.....
You mention zombie employment growth in the health sector. I wouldn't count on that. The trillion-dollar cutback in Medicaid and in ACA subsidies and changes in 340B ...will all close a lot of hospitals...putting a lot of folks out of work....and a lot of hospitals that don't close will endure service and staff constrictions, e.g., labor and delivery deserts (meaning no staff), ERs no longer operating 24 hours a day...no staff...you get the picture...Further, almost all small/rural hospitals will take a financial haircut and be unable to hire new/additional staff....and add in aging in the health sector.... Don't bet on employment growth
ah yes - the cut backs may indeed choke health sector growth - i don't know one should count on heavy heavy impacts, but that withdrawal of money from health is indeed likely to see private health operators at least slow-walk hiring.
During the Biden administration mainstream media spun that two quarters of negative growth was not necessarily a recession. The mainstream media spoke upon a soft landing throughout most of Biden’s term. The cheerleading for one POTUS vs the other in mainstream media, reversed on Fox news has caused most Americans to forget mainstream news entirely.
Yes but was that two quarters of negative economic growth experienced as a recession in any real sense (beyond the technical definition)? If not, the “spin” was correct. The singular economic problem during the Biden administration was inflation, not recession; economic growth was good.
Isn't three consecutive quarters of negative growth the benchmark for recession? There was no recession declaration at any point during Biden's term. Maybe the landing wasn't soft, but there was recovery without a recession (let alone a depression, as many feared in the worst phase of the pandemic) after the greatest worldwide shock to society in my lifetime (73 years). Pretty dam good imo.
There is no official definition of recession, but there is general recognition that the term refers to a period of decline in economic activity. Very short periods of decline are not considered recessions. Most commentators and analysts use, as a practical definition of recession, two consecutive quarters of decline in a country’s real (inflation-adjusted) gross domestic product (GDP)—the value of all goods and services a country produces. Although this definition is a useful rule of thumb, it has drawbacks. A focus on GDP alone is narrow, and it is often better to consider a wider set of measures of economic activity to determine whether a country is indeed suffering a recession. Using other indicators can also provide a timelier gauge of the state of the economy.
In the United States, the private National Bureau of Economic Research (NBER), which maintains a chronology of the beginning and ending dates of U.S. recessions, uses a broader definition and considers a number of measures of activity to determine the dates of recessions. The NBER’s Business Cycle Dating Committee defines a recession as “a significant decline in economic activity spread across the economy, lasting more than a few months, normally visible in production, employment, real income, and other indicators. A recession begins when the economy reaches a peak of activity and ends when the economy reaches its trough.” Consistent with this definition, the Committee focuses on a comprehensive set of measures—including not only GDP, but also employment, income, sales, and industrial production—to analyze the trends in economic activity.'
Note NBER doesn't use two quarters.
And here we have 2022 NPR - ('Main Stream media' I should think for our pretending totally not to be a Trumpie, totally pinky finger swear)
Economy
U.S. economy just had a 2nd quarter of negative growth. Is it in a recession?
So essential - NBER which is not 'Government Appointtees' and not even government and not political doesn't use 2Q, and uses multi-factor standard. Which is economicaly sensible if not politically the talking point.
I'm just an educated layperson, not an economist, who lacks your erudition. Nevertheless, I'm sticking to my general point that there was no consensus judgment that a recession occurred on Biden"s watch.
If people were informed adults, they would just look at the numbers and the underlying issues and trends.
But politicians want to spin the data, and newspapers want headlines, and average people couldn't describe GDP correctly if you spotted them 10 words in the wiki entry. (Gross domestic product is a monetary measure of the total market value of ")
So instead we get politicians counting coup with data fluctuations that have nothing to do with their policies.
And for the record, you would be hard pressed to find a softer landing than what actually happened.
Well since in mainstream not-Trumpie-agitation -and-propaganda, 2Qs is not the NBER definition of a recession, not used by NBER and not an economic standard as such, there is no spin at all.
Of course if one's concern is in fact politics and spin, then one is upset that one's opponent isn't beaten up but that's not economics, nor statistics, it[s pure tribal politics.
Slowing from *the first* round of Tariffs and a little bit from the current round, due to the *lag* on when this report is discussing, let's make that clear.
The irony is that the administration is removing the warning siren that could save them, not silencing a propaganda tool against them.
Biden, for all his stupidity, was trying to correct the economy even as he lied through his teeth about what was happening because he *did* listen to the warning. (Wiser move would be to admit it was happening and that we were in 'conditions unseen since the time of horse and buggy' and just ride it out, but I digress...)
Trump's doubling down and is going to make his problem worse.
But what do I know? I mean, I was the person crazy enough to buy AMD at $99 and invest in European defense stocks in February...
I think better to say Elderly Old Man Stubbornness - although it was stupidity for the Biden Admin to try to Spin & Inflation-Splaining their way out of their public position. I do agree it would have been much wiser to have grappled with their largely self-generated inflation with a more 'we're seeing inflation and working to address, we know the supply chains and businesses are struggling (etc)'
Denialism was a big error.
Absolutely agree that Trumpies are engaging in self-blinding that is more likely to harm themselves than not.
But see the completely foolish and self-harming idiocy of attacking the Fed / Powell to try to get lower rates in face of inflation re-acceleration.
Sorry big revisions, whether within historical norms or not, tell me your data is UNRELIABLE. Firing someone who does a bad job is reasonable. The Democrats act like government jobs are sacred. If the new appointee skews data then we can consider this a problem, but firing a bad actor is a problem. Had this data come out earlier, the excuse Powell has been making about keeping interest rates stagnant would have evaporated. Next meeting rates will go down. Businesses rely upon accurate data, not major revisions which are “still within historical norms.” She is gone and won’t be missed.
This excuse was made for polling data in 2016, 2020 and to a lesser degree 2024. Yes, data was within margins of error, but people no longer consider polls a strong indicator of who wins. Sorry, you have to get the big questions right.
Publishing timely and mostly accurate data then revising it has a lot of value… that’s why companies issue earnings projections and guidance. And why weather forecasters offer (less accurate) 5 day forecasts in addition to just telling you weather the sun is shining.
Bls projections give good, timely labor market data to everyone. If the number was only released 3 months after the fact private banks etc would have their own projections and use them to get a major advantage over smaller players. There would also be a cottage industry of forecasters doing the month by month projections (but worse ) to try to trade on the bond market.
Mostly accurate leaves a lot to be desired based on recent update. Private banks already have huge advantages. I get why you want data quickly, but not bad data. More importantly, why should anyone cry for a privileged political appointee who lost her job after a major revision? I would concede if Trump fires an appointee who begins with a bad number that will look very politically motivated.
Typical. If you do not blindly support a Biden appointee you are a Trump supporter? I would say you are a blind DNC supporter. How did that work out in 2024? If DNC wants to get back in power, stop defending bad government. Just for your information, I did not and would not ever vote for Trump. I do blame the excuse for the Democratic party for where we are today and yes, I do loathe the incapacitated Biden administration. But sorry that does not equate to being a Trumpster. 70% of America will not shed one tear for McEntarfer, the other 30% just blindly hate Trump. Trump comes across in this instance as someone holding government appointees accountable.
Utterly pathetic spin (and frankly don't believe you on your trump claim but that's neither here nor there).
no one is commenting on "Biden Appointee" (in fact a bipartisan supported technocrat who might easily have been a Bush or T-1 appointee) - people are commenting on BLS and what Nate actually wrote (but what you evidently did not read or worse, did not in any way shape or form understand).
Sorry big revisions, whether within historical norms or not, tell me your data is UNRELIABLE. Firing someone who does a bad job is reasonable. The Democrats act like government jobs are sacred. If the new appointee skews data then we can consider this a problem, but firing a bad actor is a problem. Had this data come out earlier, the excuse Powell has been making about keeping interest rates stagnant would have evaporated. Next meeting rates will go down. Businesses rely upon accurate data, not major revisions which are “still within historical norms.” She is gone and won’t be missed.
This excuse was made for polling data in 2016, 2020 and to a lesser degree 2024. Yes, data was within margins of error, but people no longer consider polls a strong indicator of who wins. Sorry, you have to get the big questions right.
My developing markets experience makes this rather... a nasty jolt. Very negative sign. The limp Republican responses are even more disappointing.
A reason 'third world shit-holes' to quote Trump are the way they are is elevated uncertainty from bad and above all manipulated or 'bullied' data (and the pretense this is about accuracy is transparently laughable as the Trump Tantrum came only a negative-for-him revision, not on the positive).
Not that the bullying is going to work outside the Pre-Convinced - any more than the Biden & Lefty Proggy Democrats self-deceptions on inflation worked for them on wider audiences. But it is long-term damage that is expensive to undo (if Americans even realize they need to undo it from their excessive comfort, boiling frogs...)
Question becomes is Trump going to succeed in his gambit to become the Orange Erdogan and destroy the Central Bank credibility by inserting sycophants (worked brilliantly for Turkey - what with inflation escalating towards triple digits annualized). This is of course a route to becoming Argentina - it's really a modernized version of Peronism in broad terms.
I do expect by Fall the numbers will start to look rather unhappy as the lead times of the feed-through tariff direct effects and secondary effects including the ricochets on TACO uncertainty (with this weeks return to Tariff Jihadism we can expect more realization that basically US has an economic policy that varies day-to-day with the whims of an aging tantrum thrower).
Of course it's not new, it's this behaviour pattern that saw his operating business go to bankruptcy. (Eastern Shuttle etc)
We are constantly told by people we must accept bad government. In fact bad government is within historical norms. The next appointee will want the data to be more accurate up front. Media should monitor said numbers more carefully and perhaps produce their own results to compare with government. You know: maybe the media should not just parrot government officials, but actually check them for accuracy. Instead mainstream media mostly thinks its job should be to parrot whatever government states. Kind of like state media.
How is issuing revisions ‘bad government’? Some companies do not answer the initial survey question because they are unsure of the answer to provide. Maybe they are planning changes to their company structure. Eventually they give BLS the jobs number, which gets folded into the revision. BLS has followed this process for decades.
Probably better to stay quiet and have people not know if you are ignorant than writing a rant and proving you are.
In fact issuing revised data is pretty Good Government (also it is Good Business and Good Management - if one's in analytics business and one needs to revise one's estimates as initial estimates were off, that's GOOD - as otherwise managers, your data consumers are using bad numbers. Initial estimates are just that - estimates.
Of course it is both Bad Government and Bad Business Management to suppress numbers and news you dislike as not flattering.
It's very much how you get failure, bankrupcty (or division of product failure).
Trump of course has long had this habit, his bankruptcies in all his operating businesses (versis his Reality TV and his brand rental / marketing) resolve back to general operating biz incompetence in contrast with his pure genuine in marketing.
If the data is bad next month, then it will be too late to shoot the messenger, but I am not interested in defending a government appointee who put out bad data, regardless of how hard the job may be. Also, it is only until Trump draws attention to this that we even begin demanding more accuracy up front. Here is an idea: if it takes 3 months to get up to date data: then wait to release it. Don’t release bad information. Quick wrong information is not helpful long term for anyone.
Impressive the level of either total and complete misunderstanding of what Nate ably described about the economic data generation OR impressive on the sheer mindless spin.
"Demanding more accuracy' is right up there in understanding with Socialist Mamdani 'demanding' affordable housing by restricting supply.
Actual real consumers of the data know the surveys, the BLS publishes clear statistical information on the surveying and modeling - and as Nate notes, private collection on this tends to be 'slightly worse' - or better - messier.
This is the nature of a real economy where we are not Stalinist planning and without a central party enforcing parties to report information, one has to in real actual private actors (of which large numbers of small private employers) that one has to estimate on as they themselves don't have statistical departments.
That is Real World differen than Reality TV or Magical TV / movie world where data magically is generated in omniscient all knowingness. (and why Socialist planning always fails)
Edited to Add (1hr later)
However this does falsify Nate's headline as very evidently Trump's denialism will fool some people who will go Hook Line & Sinker for Trump marketing spin (of course Nate wasn't being literal -and the old adage, 'there's a sucker born every minute' comes to mind.
What did I misunderstand? My only point is I could care less that a government appointee with a cozy job was fired for reporting bad data. I understand excuses and don’t care. IF, the new head of BLS is immediately fired for posting a poor and fairly accurate report, then we can condemn Trump. Until then, let’s not shed tears.
It’s clear you’re prejudiced against government employees for some reason, and assume her job is “cozy”, when more likely than not the job is incredibly difficult. Saying things like I don’t care …., just proves that your viewpoints isn’t grounded in anything but feelings/emotions,
She is an appointed official and at 1/3 of nation riled up by her firing. That’s a cozy job. But you are right, I am sure appointed government officials are some of hardest working people in nation. Of course, I have worked for a few and never seen it, but maybe! LOL
what "1/3' of the nation is doing is utterly besides the point and only apparently your concern since you're posting about your political grievances and not about the subject of Nate's analysis - which you either did not read or understood literally nothing of.
Well you evidently understood nothing of what Nate actually wrote and instead are engaging in Party Political ideological posturing about "government appointees" that has zero factual grounding on the problem.
Basically you're engaging in political shouting and whinging on about of the same quality as the Socialists side.
For anyone reading the actual content of Nate's comment the phrase "fairly accurate report" relative to the statistical work of the BLS and these revisions bascially shows you're an econometric illiterate just shouting political slogans.
Whatever. Keep your head in the sand and get back to me in 2028. I will play it fair the next few years and will acknowledge when I think Trump is acting fraudulently. Trump is acting politically and legally in firing BLS head and FEW, except angry Democrats, WILL SHED ANY TEARS FOR THIS POLITICAL APPOINTEE. Also, when numbers are consistently wrong and subject to major revisions, value is lost.
Ah the whole "Head in the Sand" rhetorical gambit.
That and repeating in Caps "Political Appointee" - so very convincing (not).
Shouting political slogans shows the response is in fact political - not business, not econometric.
Whether "few" are shedding any tears is utterly irrelevant (although in fact the Republican Senators responses are broadly negative - this in the fact of known habit of Trump violently attacking anyone criticising him - but there's your fine point of Management Incompetence and Failure as suppressing bad news and data is how you get into bankruptcy. But Trump has extensive experience in that, taking a profitable cash cow and wrecking it via operational management incompetence).
Ironically, the downward revision in jobs that Trump cried about Friday increases the likelihood that interest rate reductions will actually happen. I doubt he would have been happier if BLS had picked up the negative jobs signal faster, though.
Yeah - of course deeper irony is that public attacks on the Fed about interest rate level
A. Made Fed action on this more difficult not easier on their own institutional interest level - so impeding the very goal
B. Made the image of any rate change worse not better
C. Display a naive and child like understanding of the market rates and the degree to which a Fed cut that is not in alignment with overall market fundamentals will have any kind of 'positive' result.
(See Ergodan, Turkey)
but this isn't real policy making, it's Realy TV Showmanship.
Trump is such an immature cry baby. How he’s gone about this will not lead to accuracy. The way to do this is to investigate why the data is being revised and ask them to create systems to avoid this (eg delay returns until they’re accurate).. Give them 2-3 months to create systems and turn it around. Complaining that it’s rigged is pathetic and the sign of a bad leader. Nobody will trust the figures under Trump regardless of what he does. May as well not bother with these numbers now.
Well the way to do this is in fact learn the statistical econometrics - or get briefed on them and then come to an understanding if the reporting is "inaccurate' - and further to have engagement with the private sector market actors (investment banks, economic research consultancies, Conference Board) who are large consumers of BLS to see if they consider that there is a data problem - since they have heavy direct non-political interest in these numbers.
Inside investment circles one knows very well that there's statistical noise - the BLS is perfectly and solidly transparent on the econometrics
(now if one wants to throw more money at data collection that could be a legit thing to do - maybe)
That’s probably then a problem with how these numbers are advertised if at all. Needing to learn statistics to understand the accuracy of a model doesn’t create accessible information which the government should provided; not just for the economically literate in Wall Street.
Personally I don’t see a problem with holding onto the numbers for 3 months for accuracy.
You seem to be the kind of person who watches the house burn down and advise the inhabitants to wait and see if it spreads to the next house on the block before sending in the firemen. Are you totally clueless about the number of key government officials and experienced federal workers Trump has fired? Not to mention entire departments he has defended?
This is just another Trump attempt to hide the reality of his chaotic tariff policies, another in a long history of trying to hid his business mistakes, going back to the 1980's, something that earned him 34 felony convictions last year.
Trump has been relatively successful in covering his tracks in the past, but doing that on his individual construction projects is a much different scale than doing it with basic metrics for the United States economy.
Trump didn't just fire a shot across the bow of the US economy, he fired below the water line.
When businesses perceive that the numbers are being cooked to make Trump look like the sage his isn't, businesses will become much more cautious about increasing expenses to increase business, like hiring, infrastructure, capital investment, or any of a myriad of decisions for which they rely on data. They will likely do what they did in 2008, sit on their money. That turned out well, didn't it?
Tomorrow morning, market response will be "interesting" in the sense of the Chinese curse. The vast complexity of the US economy will require some time before the effect of this latest stunt reverberates through it, but a skittish market may react quickly.
There is of course a big difference between the Biden White House dismissing consumer fears and not acting on it...and firing an apolitical appointee for doing her job.
Trump is a great believer in the maxim, if you don't like the message, shoot the messenger. Trump is a pathetically incompetent administrator with a mile wide track record of business failures to prove it. As to personal failures, his shoddy tabloid record is there for all the world to see. His so-called trade "deals" are disastrous for the US. On one level, they are nothing more than a national sales tax. On another, they will drive trading partners to seek other markets and more reliable partners. Much of the "deals" call for certain levels if spending and investment in the US which a previous "deal" with China prove are totally unenforceable. Net net the American public will be the poorer. While the damage to our economy and standing in the world will be severely damaged, and while international consumers, angered by Trump's actions, will also result in far less demand for US made goods, the most pressing question now is will the US be able to recover post DJT's destructive reign of terror and, if so, how long will it take?
Does anyone have any reason to believe that some of what Trump is doing is a "prescribed burn" of the economy, to trigger a recession after a long expansion at a convenient time in his administration to stimulate a good rebound afterwards?
Shooting the messenger, figuratively, is a dick move. Totalitarians often do it literally.
We're starting to see a pattern coming from Trumponomics. Basically, it turns the uncertainty up to 11 so companies are scared to hire or otherwise invest, at least in the short term.
Conservative and rather Trump-skeptic commentator Erick Erickson often points out that the negative effects of tariffs are often delayed, sometimes with a brief short-term gain, as the effects work their way through the supply chain.
TACO is doing two things--one good and one bad. The bad is that it makes the uncertainty even more uncertain. The good is that the additional uncertainty is in a "good" direction.
Lastly, at some point companies and investors can't remain on the sideline forever, especially if their competitors are taking the risk. I think the market has absorbed this risk well, but there's still the "what if" if none of the tariff mess had come into play and the market might have done better.
While I agree firing the BLS commissioner was foolish, I think the post overlooks the severe defects in BLS employment numbers since CoVID.
The pandemic accelerated a restructuring of US labor markets. Gig jobs were a minor employment factor prior to 2020, in 2025 they may represent half of all jobs. Not only are businesses responding to BLS surveys less frequently, the BLS is having more trouble learning what businesses are in operation in the first place.
If Trump had said he was dissatisfied with the speed of BLS adaptations and replaced the commissioner with someone with fewer ties to legacy methods and more fresh ideas for improvement, it might have been a good thing. But firing the commissioner without a replacement in mind for publishing numbers he didn't like is a bad thing.
But no economic number is a perfect reflection of what it is supposed to measure. We learn about the economy by studying lots of numbers, plus qualitative information. The problem with BLS numbers is they don't seem to align with historical patterns, not that they favor Democrats. And all numbers published by governments should be interpreted with the knowledge they come from governments.
But the trouble is the business survey and the household survey are no longer agreeing, nor are they showing historical relations to other measures.
You're minimizing the revision. Going from 144,000 to 19,000 jobs, or from 147,000 to 14,000, does not represent the deviation from truth, but the same people estimating the same thing in the same way with a bit of extra data. If the little extra you learn in a few weeks can change your estimates by 125,000 or 133,000, it seems plausible that both your old and new estimates could be off by hundreds of thousands of jobs.
I'm not sure how you get 0.3%. 133,000 is 0.08% of the 163 million jobs in the US, but that's not an appropriate denominator. It's more meaningful to compare it to the 3 million or so people who leave jobs or take jobs in a typical month. Getting the net of of those two numbers wrong by 4% or 5% seems significant to me--especially since we're not talking about the deviation from truth, but the deviation from what you thought a few weeks ago.
0.3% is the standard MoE calculation for 95% confidence.
You can slice it to whatever subsample you want, but that isn't the sample population in the survey. As usual, MoE calculations for subpopulations differ. I'll let you do the calculation.
And while you think the surveys differ, the employed labor force of 159.5M is not really all that far from 163.1M, given the different methodology.
To your original point, the survey absolutely collects data on the gig economy, which is what I was addressing.
I read your original comment too quickly. I thought you were saying a revision of 100,000 jobs amounted to 0.3% error.
Now I see you meant that 100,000 fair coin flips will result in 50,000 heads plus or minus 300 (0.3% of 100,000) about 95% of the time.
But this vastly underestimates the error in real surveys. Samples are never fully random representations of the population, and people don't always tell the truth. Answers are slotted into categories that don't represent the full complexity of reality. Lots of adjustments are made to the raw numbers.
I'm aware that the BLS tries to accommodate gig workers but in my opinion it will require fundamental changes. Ten years ago a part-time worker who wanted more hours might have worked for Walmart and wanted more shifts. Today it could be a gal who drives for Uber, delivers for Amazon and does some web design; and could get more Uber hours anytime she wants them but wants more web design work and is planning to quit Amazon. You can't put survey answers from these two people in the same categories.
It like you guys don’t even read the article and certainly don’t have a full understanding of how complex and vast the US economy is and the limits of forecasting and estimating employment for hundreds of millions of households given the timeframe and resources. Most likely it’s the typical MAGA mindset of if Trump is whining about it, it must be some horrible travesty without taking the time to actually understand the intricacies and details.
The report comes out the first Friday of the month, the economy is hundreds of millions of households, there is a tradeoff between precision and time as Silver discussed, how accurate can an estimate reasonably be given the resource and time constraints? The data is gathered from surveys with difficulties around response rate and sampling. The revisions months later have much more data in them from more sources. The final numbers have even more data and don’t solely rely of surveys.
Someone asked why release the early number if it’s has the most error in it? Because investors, businesses, and other decision makers would rather get a glimpse of what is happening in the short term, because despite the lack of precision the initial estimate does provide some value. It all comes down to understanding what the numbers mean, how they are derived, the underlying uncertainty in them. You don’t have to be an expert in statistical inference or Bayesian statistics to understand. You only have to break out of the strictly polarized view, not only see the black and white of the numbers are bullshit, it’s a conspiracy vs the numbers are god, worship them. The truth lives in the gray.
elitist globalist Bourgeoise luxury this actually reading and trying to understand the article.
More seriously - as an actual private equity investor, yes we want these early numbers - the actual real market consumers of this data
A. Have the staffs who are actually numerate and know how to use them - and we do know the BLS is transparent [if one feels like it, can pick it apart on stats, methodology etc] and make own-correctives.
B. Like to have the early numbers as indeed have the background and a high interest in having early indicative numbers helps in some kinds of planning (etc0
The fact that market actors with significant resources and economic interest in BLS numbers for Business and not Political reasons are not on some jihad against BLS already tells us all something - that this whole affaire is a political one and not actually about good government or anything like that.
Now the POLITICAL people and innumerate journalists who don't ... mmmm -well that's politics and journalistic story telling by probably 3/4 of journalists who maybe had at most econ 101.....
(https://en.wikipedia.org/wiki/Bayesian_statistics for those who feel like understanding his reference to this)
You mention zombie employment growth in the health sector. I wouldn't count on that. The trillion-dollar cutback in Medicaid and in ACA subsidies and changes in 340B ...will all close a lot of hospitals...putting a lot of folks out of work....and a lot of hospitals that don't close will endure service and staff constrictions, e.g., labor and delivery deserts (meaning no staff), ERs no longer operating 24 hours a day...no staff...you get the picture...Further, almost all small/rural hospitals will take a financial haircut and be unable to hire new/additional staff....and add in aging in the health sector.... Don't bet on employment growth
ah yes - the cut backs may indeed choke health sector growth - i don't know one should count on heavy heavy impacts, but that withdrawal of money from health is indeed likely to see private health operators at least slow-walk hiring.
During the Biden administration mainstream media spun that two quarters of negative growth was not necessarily a recession. The mainstream media spoke upon a soft landing throughout most of Biden’s term. The cheerleading for one POTUS vs the other in mainstream media, reversed on Fox news has caused most Americans to forget mainstream news entirely.
Yes but was that two quarters of negative economic growth experienced as a recession in any real sense (beyond the technical definition)? If not, the “spin” was correct. The singular economic problem during the Biden administration was inflation, not recession; economic growth was good.
Isn't three consecutive quarters of negative growth the benchmark for recession? There was no recession declaration at any point during Biden's term. Maybe the landing wasn't soft, but there was recovery without a recession (let alone a depression, as many feared in the worst phase of the pandemic) after the greatest worldwide shock to society in my lifetime (73 years). Pretty dam good imo.
well... there's not any official standard on this. Quaters neg gdP is rule of thumb.
IMF discussion for example (written with notihng to do with either Trump or Biden so the howling partisans shall have to bark over other things
(https://www.imf.org/external/pubs/ft/fandd/basics/recess.htm#:~:text=Most%20commentators%20and%20analysts%20use,of%20thumb%2C%20it%20has%20drawbacks.)
Calling a recession
There is no official definition of recession, but there is general recognition that the term refers to a period of decline in economic activity. Very short periods of decline are not considered recessions. Most commentators and analysts use, as a practical definition of recession, two consecutive quarters of decline in a country’s real (inflation-adjusted) gross domestic product (GDP)—the value of all goods and services a country produces. Although this definition is a useful rule of thumb, it has drawbacks. A focus on GDP alone is narrow, and it is often better to consider a wider set of measures of economic activity to determine whether a country is indeed suffering a recession. Using other indicators can also provide a timelier gauge of the state of the economy.
In the United States, the private National Bureau of Economic Research (NBER), which maintains a chronology of the beginning and ending dates of U.S. recessions, uses a broader definition and considers a number of measures of activity to determine the dates of recessions. The NBER’s Business Cycle Dating Committee defines a recession as “a significant decline in economic activity spread across the economy, lasting more than a few months, normally visible in production, employment, real income, and other indicators. A recession begins when the economy reaches a peak of activity and ends when the economy reaches its trough.” Consistent with this definition, the Committee focuses on a comprehensive set of measures—including not only GDP, but also employment, income, sales, and industrial production—to analyze the trends in economic activity.'
Note NBER doesn't use two quarters.
And here we have 2022 NPR - ('Main Stream media' I should think for our pretending totally not to be a Trumpie, totally pinky finger swear)
Economy
U.S. economy just had a 2nd quarter of negative growth. Is it in a recession?
Updated July 28, 20228:32 AM ET
https://www.npr.org/2022/07/28/1113649843/gdp-2q-economy-2022-recession-two-quarters
So essential - NBER which is not 'Government Appointtees' and not even government and not political doesn't use 2Q, and uses multi-factor standard. Which is economicaly sensible if not politically the talking point.
I'm just an educated layperson, not an economist, who lacks your erudition. Nevertheless, I'm sticking to my general point that there was no consensus judgment that a recession occurred on Biden"s watch.
If people were informed adults, they would just look at the numbers and the underlying issues and trends.
But politicians want to spin the data, and newspapers want headlines, and average people couldn't describe GDP correctly if you spotted them 10 words in the wiki entry. (Gross domestic product is a monetary measure of the total market value of ")
So instead we get politicians counting coup with data fluctuations that have nothing to do with their policies.
And for the record, you would be hard pressed to find a softer landing than what actually happened.
US Real GDP
https://fred.stlouisfed.org/graph/fredgraph.png?g=1L57L&height=490
World Real GDP (2020 = 1 for inflation)
https://fred.stlouisfed.org/graph/fredgraph.png?g=1L586&height=490
Well since in mainstream not-Trumpie-agitation -and-propaganda, 2Qs is not the NBER definition of a recession, not used by NBER and not an economic standard as such, there is no spin at all.
Of course if one's concern is in fact politics and spin, then one is upset that one's opponent isn't beaten up but that's not economics, nor statistics, it[s pure tribal politics.
Slowing from *the first* round of Tariffs and a little bit from the current round, due to the *lag* on when this report is discussing, let's make that clear.
The irony is that the administration is removing the warning siren that could save them, not silencing a propaganda tool against them.
Biden, for all his stupidity, was trying to correct the economy even as he lied through his teeth about what was happening because he *did* listen to the warning. (Wiser move would be to admit it was happening and that we were in 'conditions unseen since the time of horse and buggy' and just ride it out, but I digress...)
Trump's doubling down and is going to make his problem worse.
But what do I know? I mean, I was the person crazy enough to buy AMD at $99 and invest in European defense stocks in February...
I think better to say Elderly Old Man Stubbornness - although it was stupidity for the Biden Admin to try to Spin & Inflation-Splaining their way out of their public position. I do agree it would have been much wiser to have grappled with their largely self-generated inflation with a more 'we're seeing inflation and working to address, we know the supply chains and businesses are struggling (etc)'
Denialism was a big error.
Absolutely agree that Trumpies are engaging in self-blinding that is more likely to harm themselves than not.
But see the completely foolish and self-harming idiocy of attacking the Fed / Powell to try to get lower rates in face of inflation re-acceleration.
Orange Erdoganism.
Sorry big revisions, whether within historical norms or not, tell me your data is UNRELIABLE. Firing someone who does a bad job is reasonable. The Democrats act like government jobs are sacred. If the new appointee skews data then we can consider this a problem, but firing a bad actor is a problem. Had this data come out earlier, the excuse Powell has been making about keeping interest rates stagnant would have evaporated. Next meeting rates will go down. Businesses rely upon accurate data, not major revisions which are “still within historical norms.” She is gone and won’t be missed.
This excuse was made for polling data in 2016, 2020 and to a lesser degree 2024. Yes, data was within margins of error, but people no longer consider polls a strong indicator of who wins. Sorry, you have to get the big questions right.
Publishing timely and mostly accurate data then revising it has a lot of value… that’s why companies issue earnings projections and guidance. And why weather forecasters offer (less accurate) 5 day forecasts in addition to just telling you weather the sun is shining.
Bls projections give good, timely labor market data to everyone. If the number was only released 3 months after the fact private banks etc would have their own projections and use them to get a major advantage over smaller players. There would also be a cottage industry of forecasters doing the month by month projections (but worse ) to try to trade on the bond market.
Mostly accurate leaves a lot to be desired based on recent update. Private banks already have huge advantages. I get why you want data quickly, but not bad data. More importantly, why should anyone cry for a privileged political appointee who lost her job after a major revision? I would concede if Trump fires an appointee who begins with a bad number that will look very politically motivated.
Try reading the actual article/post by Nate. (well we know you won't in fact and are reacting purely politically0
These revisions should give you more faith in the numbers as it is clear that accuracy is their goal.
Would Trump have fired BLS head Friday if the May/June revisions were the same number but opposite sign (upward revision?)
No, he would have nominated her for the Fed if the coin flip had ended up that way.
The "big numbers are big" team has entered the room.
Nate even did the math for you on the percent error considering the total number of jobs.
You probably think the deficit is bigger if you count it in pennies.
Typical. If you do not blindly support a Biden appointee you are a Trump supporter? I would say you are a blind DNC supporter. How did that work out in 2024? If DNC wants to get back in power, stop defending bad government. Just for your information, I did not and would not ever vote for Trump. I do blame the excuse for the Democratic party for where we are today and yes, I do loathe the incapacitated Biden administration. But sorry that does not equate to being a Trumpster. 70% of America will not shed one tear for McEntarfer, the other 30% just blindly hate Trump. Trump comes across in this instance as someone holding government appointees accountable.
Utterly pathetic spin (and frankly don't believe you on your trump claim but that's neither here nor there).
no one is commenting on "Biden Appointee" (in fact a bipartisan supported technocrat who might easily have been a Bush or T-1 appointee) - people are commenting on BLS and what Nate actually wrote (but what you evidently did not read or worse, did not in any way shape or form understand).
Always nice to see made up number in response to an in depth analysis of BLS information.
Or perhaps you have a source for that 70% and 30%?
Uncertainty in=uncertainty out.
And Trumponomics pegs the needle on uncertainty.
Sorry big revisions, whether within historical norms or not, tell me your data is UNRELIABLE. Firing someone who does a bad job is reasonable. The Democrats act like government jobs are sacred. If the new appointee skews data then we can consider this a problem, but firing a bad actor is a problem. Had this data come out earlier, the excuse Powell has been making about keeping interest rates stagnant would have evaporated. Next meeting rates will go down. Businesses rely upon accurate data, not major revisions which are “still within historical norms.” She is gone and won’t be missed.
This excuse was made for polling data in 2016, 2020 and to a lesser degree 2024. Yes, data was within margins of error, but people no longer consider polls a strong indicator of who wins. Sorry, you have to get the big questions right.
You seem to be illiterate as well as innumerate. Clearly, you have not absorbed what Silver is (quite accurately) telling us.
My developing markets experience makes this rather... a nasty jolt. Very negative sign. The limp Republican responses are even more disappointing.
A reason 'third world shit-holes' to quote Trump are the way they are is elevated uncertainty from bad and above all manipulated or 'bullied' data (and the pretense this is about accuracy is transparently laughable as the Trump Tantrum came only a negative-for-him revision, not on the positive).
Not that the bullying is going to work outside the Pre-Convinced - any more than the Biden & Lefty Proggy Democrats self-deceptions on inflation worked for them on wider audiences. But it is long-term damage that is expensive to undo (if Americans even realize they need to undo it from their excessive comfort, boiling frogs...)
Question becomes is Trump going to succeed in his gambit to become the Orange Erdogan and destroy the Central Bank credibility by inserting sycophants (worked brilliantly for Turkey - what with inflation escalating towards triple digits annualized). This is of course a route to becoming Argentina - it's really a modernized version of Peronism in broad terms.
I do expect by Fall the numbers will start to look rather unhappy as the lead times of the feed-through tariff direct effects and secondary effects including the ricochets on TACO uncertainty (with this weeks return to Tariff Jihadism we can expect more realization that basically US has an economic policy that varies day-to-day with the whims of an aging tantrum thrower).
Of course it's not new, it's this behaviour pattern that saw his operating business go to bankruptcy. (Eastern Shuttle etc)
We are constantly told by people we must accept bad government. In fact bad government is within historical norms. The next appointee will want the data to be more accurate up front. Media should monitor said numbers more carefully and perhaps produce their own results to compare with government. You know: maybe the media should not just parrot government officials, but actually check them for accuracy. Instead mainstream media mostly thinks its job should be to parrot whatever government states. Kind of like state media.
It’s like you didn’t read the article.
He has feelings. He doesn't need to "read" or "understand facts".
Well - he read the headline it seems - although if he did read the article, he literally understood nothing of it.
(ETA; and one must admit he singly falsifies Nate's headline... [not literally meant we all know but...)
How is issuing revisions ‘bad government’? Some companies do not answer the initial survey question because they are unsure of the answer to provide. Maybe they are planning changes to their company structure. Eventually they give BLS the jobs number, which gets folded into the revision. BLS has followed this process for decades.
Probably better to stay quiet and have people not know if you are ignorant than writing a rant and proving you are.
In fact issuing revised data is pretty Good Government (also it is Good Business and Good Management - if one's in analytics business and one needs to revise one's estimates as initial estimates were off, that's GOOD - as otherwise managers, your data consumers are using bad numbers. Initial estimates are just that - estimates.
Of course it is both Bad Government and Bad Business Management to suppress numbers and news you dislike as not flattering.
It's very much how you get failure, bankrupcty (or division of product failure).
Trump of course has long had this habit, his bankruptcies in all his operating businesses (versis his Reality TV and his brand rental / marketing) resolve back to general operating biz incompetence in contrast with his pure genuine in marketing.
If the data is bad next month, then it will be too late to shoot the messenger, but I am not interested in defending a government appointee who put out bad data, regardless of how hard the job may be. Also, it is only until Trump draws attention to this that we even begin demanding more accuracy up front. Here is an idea: if it takes 3 months to get up to date data: then wait to release it. Don’t release bad information. Quick wrong information is not helpful long term for anyone.
Impressive the level of either total and complete misunderstanding of what Nate ably described about the economic data generation OR impressive on the sheer mindless spin.
"Demanding more accuracy' is right up there in understanding with Socialist Mamdani 'demanding' affordable housing by restricting supply.
Actual real consumers of the data know the surveys, the BLS publishes clear statistical information on the surveying and modeling - and as Nate notes, private collection on this tends to be 'slightly worse' - or better - messier.
This is the nature of a real economy where we are not Stalinist planning and without a central party enforcing parties to report information, one has to in real actual private actors (of which large numbers of small private employers) that one has to estimate on as they themselves don't have statistical departments.
That is Real World differen than Reality TV or Magical TV / movie world where data magically is generated in omniscient all knowingness. (and why Socialist planning always fails)
Edited to Add (1hr later)
However this does falsify Nate's headline as very evidently Trump's denialism will fool some people who will go Hook Line & Sinker for Trump marketing spin (of course Nate wasn't being literal -and the old adage, 'there's a sucker born every minute' comes to mind.
What did I misunderstand? My only point is I could care less that a government appointee with a cozy job was fired for reporting bad data. I understand excuses and don’t care. IF, the new head of BLS is immediately fired for posting a poor and fairly accurate report, then we can condemn Trump. Until then, let’s not shed tears.
It’s clear you’re prejudiced against government employees for some reason, and assume her job is “cozy”, when more likely than not the job is incredibly difficult. Saying things like I don’t care …., just proves that your viewpoints isn’t grounded in anything but feelings/emotions,
She is an appointed official and at 1/3 of nation riled up by her firing. That’s a cozy job. But you are right, I am sure appointed government officials are some of hardest working people in nation. Of course, I have worked for a few and never seen it, but maybe! LOL
Substance free political maundering on....
what "1/3' of the nation is doing is utterly besides the point and only apparently your concern since you're posting about your political grievances and not about the subject of Nate's analysis - which you either did not read or understood literally nothing of.
Well you evidently understood nothing of what Nate actually wrote and instead are engaging in Party Political ideological posturing about "government appointees" that has zero factual grounding on the problem.
Basically you're engaging in political shouting and whinging on about of the same quality as the Socialists side.
For anyone reading the actual content of Nate's comment the phrase "fairly accurate report" relative to the statistical work of the BLS and these revisions bascially shows you're an econometric illiterate just shouting political slogans.
Whatever. Keep your head in the sand and get back to me in 2028. I will play it fair the next few years and will acknowledge when I think Trump is acting fraudulently. Trump is acting politically and legally in firing BLS head and FEW, except angry Democrats, WILL SHED ANY TEARS FOR THIS POLITICAL APPOINTEE. Also, when numbers are consistently wrong and subject to major revisions, value is lost.
Ah the whole "Head in the Sand" rhetorical gambit.
That and repeating in Caps "Political Appointee" - so very convincing (not).
Shouting political slogans shows the response is in fact political - not business, not econometric.
Whether "few" are shedding any tears is utterly irrelevant (although in fact the Republican Senators responses are broadly negative - this in the fact of known habit of Trump violently attacking anyone criticising him - but there's your fine point of Management Incompetence and Failure as suppressing bad news and data is how you get into bankruptcy. But Trump has extensive experience in that, taking a profitable cash cow and wrecking it via operational management incompetence).
Ironically, the downward revision in jobs that Trump cried about Friday increases the likelihood that interest rate reductions will actually happen. I doubt he would have been happier if BLS had picked up the negative jobs signal faster, though.
Yeah - of course deeper irony is that public attacks on the Fed about interest rate level
A. Made Fed action on this more difficult not easier on their own institutional interest level - so impeding the very goal
B. Made the image of any rate change worse not better
C. Display a naive and child like understanding of the market rates and the degree to which a Fed cut that is not in alignment with overall market fundamentals will have any kind of 'positive' result.
(See Ergodan, Turkey)
but this isn't real policy making, it's Realy TV Showmanship.
Trump is such an immature cry baby. How he’s gone about this will not lead to accuracy. The way to do this is to investigate why the data is being revised and ask them to create systems to avoid this (eg delay returns until they’re accurate).. Give them 2-3 months to create systems and turn it around. Complaining that it’s rigged is pathetic and the sign of a bad leader. Nobody will trust the figures under Trump regardless of what he does. May as well not bother with these numbers now.
Well the way to do this is in fact learn the statistical econometrics - or get briefed on them and then come to an understanding if the reporting is "inaccurate' - and further to have engagement with the private sector market actors (investment banks, economic research consultancies, Conference Board) who are large consumers of BLS to see if they consider that there is a data problem - since they have heavy direct non-political interest in these numbers.
Inside investment circles one knows very well that there's statistical noise - the BLS is perfectly and solidly transparent on the econometrics
(now if one wants to throw more money at data collection that could be a legit thing to do - maybe)
That’s probably then a problem with how these numbers are advertised if at all. Needing to learn statistics to understand the accuracy of a model doesn’t create accessible information which the government should provided; not just for the economically literate in Wall Street.
Personally I don’t see a problem with holding onto the numbers for 3 months for accuracy.
You seem to be the kind of person who watches the house burn down and advise the inhabitants to wait and see if it spreads to the next house on the block before sending in the firemen. Are you totally clueless about the number of key government officials and experienced federal workers Trump has fired? Not to mention entire departments he has defended?
Now just watch those July number gets even worse next month.
And tbh anybody especially new grads will tell you that the job numbers are bs for awhile - this is where sentiment comes in.
Yes. That is side effect of the process related statistical bias Nate describes.
When the economy is in transition some of the late reporting companies have correlated reasons to report late.
A burst of hiring during a boom, or a layoff during a downturn can make government paperwork a second tier priority.
This is just another Trump attempt to hide the reality of his chaotic tariff policies, another in a long history of trying to hid his business mistakes, going back to the 1980's, something that earned him 34 felony convictions last year.
Trump has been relatively successful in covering his tracks in the past, but doing that on his individual construction projects is a much different scale than doing it with basic metrics for the United States economy.
Trump didn't just fire a shot across the bow of the US economy, he fired below the water line.
When businesses perceive that the numbers are being cooked to make Trump look like the sage his isn't, businesses will become much more cautious about increasing expenses to increase business, like hiring, infrastructure, capital investment, or any of a myriad of decisions for which they rely on data. They will likely do what they did in 2008, sit on their money. That turned out well, didn't it?
Tomorrow morning, market response will be "interesting" in the sense of the Chinese curse. The vast complexity of the US economy will require some time before the effect of this latest stunt reverberates through it, but a skittish market may react quickly.
There is of course a big difference between the Biden White House dismissing consumer fears and not acting on it...and firing an apolitical appointee for doing her job.
Trump is a great believer in the maxim, if you don't like the message, shoot the messenger. Trump is a pathetically incompetent administrator with a mile wide track record of business failures to prove it. As to personal failures, his shoddy tabloid record is there for all the world to see. His so-called trade "deals" are disastrous for the US. On one level, they are nothing more than a national sales tax. On another, they will drive trading partners to seek other markets and more reliable partners. Much of the "deals" call for certain levels if spending and investment in the US which a previous "deal" with China prove are totally unenforceable. Net net the American public will be the poorer. While the damage to our economy and standing in the world will be severely damaged, and while international consumers, angered by Trump's actions, will also result in far less demand for US made goods, the most pressing question now is will the US be able to recover post DJT's destructive reign of terror and, if so, how long will it take?
Does anyone have any reason to believe that some of what Trump is doing is a "prescribed burn" of the economy, to trigger a recession after a long expansion at a convenient time in his administration to stimulate a good rebound afterwards?
Personally, I don't think he's that clever.
I don't think he is that clever.
I do think he is petty enough to attack California and New York for political reasons, completely ignoring that together they are over 20% of the GDP.
Shooting the messenger, figuratively, is a dick move. Totalitarians often do it literally.
We're starting to see a pattern coming from Trumponomics. Basically, it turns the uncertainty up to 11 so companies are scared to hire or otherwise invest, at least in the short term.
Conservative and rather Trump-skeptic commentator Erick Erickson often points out that the negative effects of tariffs are often delayed, sometimes with a brief short-term gain, as the effects work their way through the supply chain.
TACO is doing two things--one good and one bad. The bad is that it makes the uncertainty even more uncertain. The good is that the additional uncertainty is in a "good" direction.
Lastly, at some point companies and investors can't remain on the sideline forever, especially if their competitors are taking the risk. I think the market has absorbed this risk well, but there's still the "what if" if none of the tariff mess had come into play and the market might have done better.
But the overall effect is depressive.
While I agree firing the BLS commissioner was foolish, I think the post overlooks the severe defects in BLS employment numbers since CoVID.
The pandemic accelerated a restructuring of US labor markets. Gig jobs were a minor employment factor prior to 2020, in 2025 they may represent half of all jobs. Not only are businesses responding to BLS surveys less frequently, the BLS is having more trouble learning what businesses are in operation in the first place.
If Trump had said he was dissatisfied with the speed of BLS adaptations and replaced the commissioner with someone with fewer ties to legacy methods and more fresh ideas for improvement, it might have been a good thing. But firing the commissioner without a replacement in mind for publishing numbers he didn't like is a bad thing.
But no economic number is a perfect reflection of what it is supposed to measure. We learn about the economy by studying lots of numbers, plus qualitative information. The problem with BLS numbers is they don't seem to align with historical patterns, not that they favor Democrats. And all numbers published by governments should be interpreted with the knowledge they come from governments.
That is the reason the BLS also does a broad population survey also.
100k-ish random data points is pretty decent. About 0.3%
But the trouble is the business survey and the household survey are no longer agreeing, nor are they showing historical relations to other measures.
You're minimizing the revision. Going from 144,000 to 19,000 jobs, or from 147,000 to 14,000, does not represent the deviation from truth, but the same people estimating the same thing in the same way with a bit of extra data. If the little extra you learn in a few weeks can change your estimates by 125,000 or 133,000, it seems plausible that both your old and new estimates could be off by hundreds of thousands of jobs.
I'm not sure how you get 0.3%. 133,000 is 0.08% of the 163 million jobs in the US, but that's not an appropriate denominator. It's more meaningful to compare it to the 3 million or so people who leave jobs or take jobs in a typical month. Getting the net of of those two numbers wrong by 4% or 5% seems significant to me--especially since we're not talking about the deviation from truth, but the deviation from what you thought a few weeks ago.
0.3% is the standard MoE calculation for 95% confidence.
You can slice it to whatever subsample you want, but that isn't the sample population in the survey. As usual, MoE calculations for subpopulations differ. I'll let you do the calculation.
And while you think the surveys differ, the employed labor force of 159.5M is not really all that far from 163.1M, given the different methodology.
To your original point, the survey absolutely collects data on the gig economy, which is what I was addressing.
I read your original comment too quickly. I thought you were saying a revision of 100,000 jobs amounted to 0.3% error.
Now I see you meant that 100,000 fair coin flips will result in 50,000 heads plus or minus 300 (0.3% of 100,000) about 95% of the time.
But this vastly underestimates the error in real surveys. Samples are never fully random representations of the population, and people don't always tell the truth. Answers are slotted into categories that don't represent the full complexity of reality. Lots of adjustments are made to the raw numbers.
I'm aware that the BLS tries to accommodate gig workers but in my opinion it will require fundamental changes. Ten years ago a part-time worker who wanted more hours might have worked for Walmart and wanted more shifts. Today it could be a gal who drives for Uber, delivers for Amazon and does some web design; and could get more Uber hours anytime she wants them but wants more web design work and is planning to quit Amazon. You can't put survey answers from these two people in the same categories.
As long as they answer the questions, they are included in the data.
Agreed. But the aggregated answers mean different things in 2025 versus 2015, and have different relations to other labor market measures.
The ADP numbers are historically inaccurate. I’m not even sure they’re directionally correct on average.
Not the least because they completely skip all government jobs.
I also think they have no model for companies that don't use their services, so they miss a lot of small businesses.
You can do a lot with QuickBooks.
Most people know to take the ADP numbers with a grain of salt, or at least recognize their limitations.