August 4, 2025 • 24 mins
Late last week, Donald Trump shocked Wall Street by firing Erika McEntarfer, the head of the Bureau of Labor Statistics, the agency responsible for publishing some of America's most important economic data. The firing came after the BLS released a weaker than expected jobs report for July, with just 73,000 new jobs added for the month (compared to forecasts for 103,000). The bureau also revised jobs numbers for the prior two months down by nearly 260,000 jobs. Trump called the data "rigged." But why does the BLS make these revisions, and what does the firing of the BLS chief mean for anyone trying to gauge the direction of the US economy? In this episode, we speak to Bill Beach, a former BLS chief, about the latest drama in US economic statistics.
Read more:
Trump to Name New Fed Governor, BLS Head in Coming Days
S&P 500 Bounces 1% After Weak Jobs Data Stokes Rate-Cut Optimism
Only Bloomberg - Business News, Stock Markets, Finance, Breaking & World News subscribers can get the Odd Lots newsletter in their inbox each week, plus unlimited access to the site and app. Subscribe at bloomberg.com/subscriptions/oddlots
See omnystudio.com/listener for privacy information.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:02):
Bloomberg Audio Studios, Podcasts, Radio News.
Speaker 2 (00:18):
Hello and welcome to another episode of the All Thoughts Podcast.
Speaker 3 (00:21):
I'm Tracy Alloway and I'm Joe Wisenthal.
Speaker 2 (00:23):
Joe, I kind of missed this on Friday because I
was not feeling very well, so I was out sick.
But what was Friday morning like when you were sat
in front of your Bloomberg terminal.
Speaker 3 (00:34):
Well, the jobs report was pretty shocking. Obviously Friday was
just a crazy day. Yeah, So the jobs report, obviously
it was bad. I mean, the last two months massively
revised down, all of this evidence of labor market momentum
perhaps seems to be stalling the tariff effect. Maybe we
really are and it's sort of some sort of terriff
(00:54):
and do slow down. The only job creation is in
healthcare and social service.
Speaker 2 (00:59):
Not great, right, Okay, So I saw the initial jobs
report headline. So we had non farm payrolls coming in
at plus seventy three thousand, which I think was like
thirty percent lower than the average expectation. We had the
unemployment rate taking up to like four point three percent,
four point almost four point three percent. Oh, let me
(01:19):
let me sorry, I'm rounding up here.
Speaker 3 (01:21):
Okay, Okay, okay, although maybe I shouldn't.
Speaker 2 (01:23):
Do that on a podcast all about labor market statistics.
But the big thing that seemed to catch everyone's attention
was we also had these massive revisions to the report
for May and June, so we had a combined two
hundred and fifty eight thousand jobs basically lowered from the
initial reports. And this was like the biggest revision since
(01:43):
the depths of the pandemic. And then the headline that
really caught my eye while I was, you know, laying
in my sick bed on Friday afternoon, was Trump firing
the head of the Bureau of Labor Statistics, so the
agency responsible for putting out the non farm pay Roule
Report every month.
Speaker 3 (02:01):
Yeah, this is I would say. The key thing is
that a we've been used to these, you know, seeing
significant revisions. We've been talking you in particular, have been
writing a lot in the newsletter about deteriorating quality of
labor statistics. Response rates to a lot of surveys have
gone down over time, kind of like we're seeing in
political opinion surveys and so forth. So there's already been
(02:23):
this sort of anxiety. You have a lot of people online,
including President Trumpet himself, you know, been like stoking sort
of these conspiracy theories about these revisions, these attempts to
actually be transparent, and then you get the first sort
of like genuinely negative report under this administration. It's the
first one that was like, Okay, this was bad and boom,
(02:44):
Trump fires the person responsible for it. Now again, this
is one of those things where sort of like Doge itself,
where it's like, I like the idea of government efficiency, right, yeah, sure, Doge.
I like the idea of the BLS sort of doing
a better job in some way or addressing these responses.
But like Doge, which you know, we all have seen
how that's turned out. I think there are a lot
(03:06):
of more than a lot of questions, more than a
lot of questions about whether firing to be less and
replacing the head with someone who's you know, a Trumpet
point t will you know, add a more credible transparent
data clauding right.
Speaker 2 (03:20):
One thing that we have learned from all our episodes
on data collection in the US is that it is
actually a really really insanely labor intensive thing to do.
You have to actually call up a bunch of people
homeowners for instance, or you have to call up a
bunch of businesses and ask questions. You have to go
out into the field and gather individual prices. And we
(03:40):
can have a debate over whether or not you could
maybe use new technology or price scanning data to make
all of that more efficient, but for the time being,
that's how it's done, and so you need people to
do it. And if DOGE comes in and eliminates, you know,
a big chunk of the budget, then it becomes harder
to do.
Speaker 3 (03:57):
Or if it just gets more costly and the budget's
gone up, and then the one thing I would say
is that in our conversations, and again you have talked
to the BLS many times, they are really good. They're professionals,
and they take it very seriously.
Speaker 2 (04:09):
They're very responsive, however, very transparent. We could talk about
this a little bit, but perhaps less responsive than they
used to be in recent weeks. So anyway, we should
talk about all of this. I think it's really important.
And as you said, there are big questions like what
happens if a president ouse a BLS chief and then
introduces a new one. We have the perfect guests. Someone
we've spoken to before about the decline of America's data infrastructure.
(04:33):
We're going to be speaking with Bill Beach. He is,
of course, the former Commissioner of Labor Statistics, former head
of the BLS. So Bill, thank you so much for
coming back on all thoughts.
Speaker 4 (04:43):
Yeah, it's just really a pleasure to be back with you.
I wish we were talking about some other subject besides
this termination. I mean, it's just yeah, shocked.
Speaker 2 (04:52):
Well, on that note, Friday, Trump tweets or posts that
he wants to fire Erica mc Andterfer, the Commissioner of
Labor Statistics. What was going through your mind when you
saw that news.
Speaker 4 (05:04):
I was dumbcast. I just finished a luncheon meeting at
a very nice restaurant, and I was sitting in my
pickup truck and I just couldn't believe. And then I
sat there for probably thirty forty minutes, car running and
answering emails, people just sorting it out. It isn't that
he didn't have the authority to do so. He does
have the authority. We all, you know, everyone in the
(05:26):
executive branch serves as the pleasure of the president, except
those that have been specifically exempted by statute, and where
that statute has been signed by the President. I think
what was shocking about this is not that she was dismissed.
That's shocking enough. It was the possibility that we could
have now a sustained attack on official economic statistics, and
(05:46):
that would undermine confidence in those statistics and put us
in a path that other countries have followed. I mean
instantly people went to worst case. After a while, that
sort of settled down, and we began to look at
the reasons for the dismissal, which then subsequently said on
Twitter or x that they were they were unfounded. And
i haven't dropped that position since, so I'm happy to
(06:09):
talk about it.
Speaker 3 (06:10):
Well, why don't we just back up for a second.
We talked to you before, because you've been sounding the
alarm about the capacity constraints to collect good data. But
just for people who haven't listened to or haven't followed you,
you were a Trump appointing. I mean, you served at
the head of the BLS in twenty seventeen, and so
I do you know, and it's tart of like, okay,
your credibility. Why don't you just give us a little
(06:31):
bit of background about who you are, your position, and
the concerns that you've been raising for some time about
the constraints at the BLS.
Speaker 4 (06:40):
So I was really honored to be the Commissioner of
Labor Statistics from twenty nineteen until twenty twenty three. So
I served a little less than two years under President Trump.
He nominated me. It is a presidential nomination, Senate confirmed positions,
it has a statutory for your term. And then I
served the rest of my under President Biden. So I
(07:02):
was in both terms. I think that's kind of important
because the BLS Commissioner has been exempt from the normal
turning over of the presidency by party. And that's because
the view as well, these data are so important they
should not even be a part of the political appointment
process except when they come to. It's sort of like
(07:24):
you know, the Federal Reserve chair or something of that
nature has just seen as we want to keep it
out of politics. So I served over that period of time.
You know, that was the COVID period, so there were
all kinds of real challenges from that, but I think
the main thing that I drew from that period is
how important it is crucial, almost were the crisis level,
to modernize the way we collect data. The response rates
(07:47):
on the surveys are falling dramatically. I've made that point
on this program, and our costs arising dramatically, which makes
it really difficult to conduct those surveys in the future
and the way we have in the past, and then
the future is kind of going in the direction of
the employment survey, and I just said surveys are following
with the employment surveys a little different. It's not an
(08:09):
in person survey, it is an electronic survey, and we
need to invest more in the electronic side of statistical
creation or production, blended data, etc. I could do a
whole program on this, but your point was well made.
BLS has had some challenges in front of it, as
has the entire statistical system. Unfortunately, for this particular episode,
(08:32):
those challenges were really not I don't say relevant. They're relevant,
of course, but they're really not the reason why we
should have had a change at the top.
Speaker 2 (08:41):
Well, speaking of the employment report, talk to us about
what exactly happens when the BLS is publishing the initial
non farm payrolls and then what happens, you know, in
the couple months or so before it publishes the revision,
and why does it seem we are getting these large
gaps at least in recent months and years.
Speaker 4 (09:01):
Right, So the employment survey, the jobs numbers come from
a survey of businesses. The unemployment rate comes from a
survey of households. So there's two surveys involved in every
First Friday report. What we're talking about now is the
survey that goes out to businesses, and it goes out
to hundreds of thousands of businesses. Well, of course there's
(09:22):
still a probability sample, since there's well in excess of
twelve million businesses in our Census of business, our business register. Well,
those businesses are supposed to turn their surveys in at
the end of the month basically, but only about sixty
eight percent usually that's the average do so, and so
(09:43):
BLS keeps the window open for two more months, so
sixty eight percent or so at the end of the
first month they make the first estimate. Then at the
end of the second month we've get about eighty three
percent completion, and they revise that number of the first month,
and then by the third month we're into the nineties.
Usually end up around ninety three to ninety four percent
(10:05):
of all the sent out returns. Come back to us
with information. So it's a wonderful survey. The revisions are
done because we get more information from businesses. Oftentimes big
businesses answer first they have the capacity to do so,
and then we get smaller entities, state and local governments,
smaller businesses answer in those next two months. We always
(10:27):
have revisions. There's there's even when we say the number
did not change, it doesn't mean we didn't have revisions.
They just canceled each other out. There's just all these
revisions are coming in. The revisions have been high recently,
but this is not a typical of a period when
the economy is either going back to growth or going
(10:48):
down to subsidence. So oftentimes turning points in the economy
are accompanied by changes larger changes in the revisions. Particularly
in this case where we SUSPEC effect. Smaller businesses are
feeling the effects of the supply shock coming from the
tariff policy and from immigration, and that these supply shocks
(11:09):
are affecting smaller businesses more than they are larger businesses,
and state and local governments are being affected by another factor,
and that is the expiration of the COVID era money.
I think that was actually the root of a lot
of the changes that you saw on Friday. State and
local governments are not able to hire at the levels
they were going to hire at in previous years or
(11:30):
did hire at in previous years because they did not
have the subsidies that they had in previous years. So
these are important revisions. Actually, the research shows, and this
may shock you, that BLS is getting much better at
its first estimate than it was thirty years ago. In fact,
it's really kind of almost a steady progress towards greater accuracy.
(11:55):
So the big revisions are indicative of not air of
more information, and the bigger the revisions, the more likely
we are in a turning point.
Speaker 3 (12:06):
There's so much that was in that answer that I
found to be very helpful and clarifying, so thank you.
I mean, just this idea, you know, a that actually
over time, contrary to what people on Twitter might think,
that the equality of that first pass has actually gone
up is very striking.
Speaker 1 (12:24):
You know.
Speaker 3 (12:24):
Let's stipulate that, Okay, the BLIS could use some upgrading.
As you mentioned that there are these capacity constraints, non response,
the cost of going out. Maybe it needs to adopt
more technology in some way to solve for this. Does
the BLS currently have that capacity, either budgetary wise or
statutory authority such that a commissioner could just do that
(12:47):
whatever it is that ideal state, or would it need
either some sort of budgetary allocation or an Act of
Congress to get to the point where it's you know,
getting to the level that we are happy with.
Speaker 4 (12:58):
Again, it has the authority to make the changes, it
doesn't have the budget to do so. And it isn't
that BLS just needs more money, right. As a matter
of fact, I think Congress, if it paid more attention,
would say, well, what are you spending your money on now?
And let's make sure that the low priority is canceled
so that we can support the high priority. I think
(13:18):
BLS would be very happy with that. But we do
need more money. And why Well, if you're going to
change an official statistic, the unemployment rate, let's just take that.
You want to be really careful, right, You want to
run tests, you want to test out your new idea.
You don't want to have a count the number of
geese in the air, and that's the unemployment rate. You
(13:39):
want to make sure that that is a particularly good
indicator of the employment market, and you want to test
it in real time as well as experimental time. You
in real time, so run it parallel. We these tests
expand the costs of your unemployment statistics program because you're
running essentially two systems simultaneously. That to modify the current
(14:03):
population survey, which is the survey of households, would cost
around fifteen million to twenty million in experimental costs over
a two year period. It's not great. The program itself
costs about forty five million a year, but it does
add to the cost for those two years or so.
Congress could could easily do that. I mean, oh my gosh,
Congress is spending a lot more money than that on
(14:25):
things that we might disagree with. But they need to
allocate a little bit of money to CPS for that.
Does it have the authority? Here's a really interesting story.
During COVID, we canceled all travel, all conferences, you know,
and I was surprised at how many millions of dollars
went into travel and conferences and meetings. So this became
(14:46):
for me couch money during the COVID years, and I
was able to spend that money, you know, on important
things I thought, So I did a lot of things,
like I've built a new data center to get us
out of the basement of one hundred and four year
old building out to a wonderful above ground local location.
But I also started a number of research programs.
Speaker 2 (15:06):
You know.
Speaker 4 (15:06):
I made thirty one big modifications to do the CPI.
We started a whole new approach on looking at consumer expenditures,
and I subsidized some experimentation on the Jobs Report well
as it as COVID was over, we started to get
congressionally mandated programs that then sucked up all that counch money.
(15:30):
If we had a little bit of extra, we I
mean be only asked a little bit of extra to
do these things. Let's just think of the innovation that
they would happen. I think we would quickly be out
of the problems that we are facing from a methodological standpoint,
and that might then reduce the political pressure that we
saw that was so evident on Friday.
Speaker 2 (15:52):
So we're recording this August fourth, nine, five am in
the morning and thirty three seconds. I feel like I
need to be that specific nowadays with the newsflow, so
who knows what happens between now and when this episode
actually comes out, but we are expecting the administration to
announce a new BLS head in the coming days. What
(16:13):
happens now, you know, there's a Trump appointee for a
new statistics commissioner after the President has specifically stated that
he thinks the previous commissioner was politically motivated in some way.
Are people still going to believe the numbers that are
coming out? And how do you think that's actually going
to play internally at the BLS, at an organization whose whole,
(16:36):
you know, Raison Deutch is to find, you know, factual
numbers and statistics to portray the American economy.
Speaker 4 (16:44):
Well, I think there has been damage and it will
take time to recover from that damage. The last time,
there was a serious political effort on BLSS during the
Nixon administration and it took some time for BLS to
recover from that, though the action taken by the President
at that time visa BLS was not, I think as
(17:05):
significant as this one. So he just imagined that the
President decides to appoint Saint Peter, you know, as the
new BLS commissioner, and of course Saint Peter has a
reputation for honesty and for clarity of thought. Still it'll
be the case because Saint Peter won't have any control
over the way that the data are collected, the data
(17:27):
are assembled, or the estimates that are done. That's all
done outside of the of the knowledge of the commissioner.
Commissioner has no control over that. In fact, you're locked
out of that whole process. So it will be the
case that Saint Peter will have a month come when
the unemployment rate will be disappointingly low, could even be negative.
But I think given what has just happened, people will say, well,
(17:51):
Saint Peter probably influenced that number and it's not as
bad as it really is. So for a while, that
suspicion that the estimate that's in ount it's really not
the real estimate, will be in the minds of some people,
not in my mind, because I know that these people
who work there, the professionals, the full time staff, the
patriots that work there, the loyal Americans who are just
(18:15):
extremely diligent in doing an absolutely objective job. Those people
are still in place. But if you don't have my
level of knowledge, or a reasonable level of knowledge of
how internally BLS works, you're going to be subject to
these falsehoods and accusations, and that will be damaging, That
will reduce investment, that will reduce economic activity, That will
(18:38):
make at least that will create greater uncertainty about what's
happening inside the US economy. Policymakers will be less clear
in the direction that they take, So there will be
a time when we need to recover. No matter who's appointed,
I hope that it will be a short lived time.
And my guess is, given the reputation of BLS, if
(18:59):
someone you know, really reputable is appointed, then yeah, I
don't think the period of transition will be that a lot.
Speaker 2 (19:07):
All right, Well, Bill Beach really appreciate you coming back
on the show. Thank you so much for being on
Off Box.
Speaker 3 (19:13):
Thanks Bill, that was great.
Speaker 4 (19:15):
Thank you very much for asking me.
Speaker 2 (19:29):
Joe so good to get Bill back on to talk
about this. Really the perfect guest. There are so many,
I guess ironies involved in this whole conversation. The big one,
of course, is this idea that like, well, if Trump
wants to convince everyone that the US economy is doing
fantastic and job growth is great, then firing the head
of statistics and having everyone distrust, you know, the subsequent numbers,
(19:52):
because the subsequent head might be a political appointee who's
you know, a loyalist to Trump. It seems like you're
sort of shooting yourself in the foot.
Speaker 3 (20:00):
The other huge contradiction is that he says Powell is
too late. Yea wait, this is the other huge contradiction,
which is every day he slams Powell, he's like, oh,
you're too late, you need to cut rates. Well, if
the job numbers were weak, then I could understand this
argument to some extent, oh the job But he's also
casting expersions on the negative jobs numbers. So how is
(20:20):
Powell too late? I mean, this is I mean, it's
an irretrievable contradiction in the two criticism of both of
the Fed and the BLS.
Speaker 2 (20:28):
It's definitely a contradiction. I don't think, however, that Trump
necessarily draws a direct connection between the unemployment rate and interest.
Speaker 3 (20:36):
Rates as well.
Speaker 2 (20:37):
You know, he seems to just like low interest rates.
Speaker 3 (20:40):
And he seems to perceive that low interest rates are
basically a reward for an improving economy. But yes, but I.
Speaker 2 (20:48):
Mean, lots of other contradictions. You're absolutely right. So another
one is this idea that Okay, a lot of the
weakness came through on the revision side. The revisions are
mostly coming from small businesses who reply late. And if
you think about the Trump administration, you know, they always
sort of portray themselves as a good business environment for
smaller businesses, and it might be that that scenario that's
(21:10):
seeing some weakness.
Speaker 3 (21:12):
I thought your question about sort of internal morale, and
as Bill said, you know, an agency staffed with patriots,
I mean, I just you know, is Bilska going to
be continue to be a destination for people who are
patriots right or take their job seriously in the future
is a really interesting question. Also, the idea that like
maybe we could like get high data quality for like
(21:34):
twenty million dollars. Given how much that's worth, it's just
like drives you crazy if that's really all it took.
But also I really like this idea that it would
have to always be done in parallel with the current thing,
such that the new data is sort of provably backwards
compatible with the old data. Right, so if you're going
to adopt a new methodology, you need some allocation, run
two surveys at once, et cetera. Anyway, even in a
(21:56):
very short conversation. Learned a lot from Bill. Absolutely.
Speaker 2 (21:58):
I do think it's worth pointing out out that there
are other statistical agencies in the world that are experimenting
with new technology for data sources. The UK Statistics Agency
in particular, they're going to start using you know, price
scanning data for CPI.
Speaker 3 (22:12):
Wait, have you heard about the whole thing though, with
how they moved the Ons to Wales and no one
wanted to go out there, And have you heard about this?
Speaker 2 (22:20):
I have Wales. It's all that bad.
Speaker 3 (22:23):
Attribute the largest of the government across. Everyone loves that idea.
Let's not have it all be concentrated in London. They're like, okay,
we're going to make everyone move to Wales and now
it's a lot harder for the Ons to hire anyone. Anyway,
it's just a little interesting story out Maybe we'll cover
it at.
Speaker 2 (22:38):
Some point why people don't want to move to Wales.
Speaker 3 (22:41):
No interesting, it's interesting, right, Like there's intuitively I love
this idea of like, oh, why should all of the
government spending on public sector stuff at the federal level
in the richest city in the UK, or why should
it all be in dc et cetera. Should this idea
of diffuse anyway, it's a separate thing. It's actually some
interesting challenges arise for.
Speaker 2 (23:00):
Okay, shall we leave it there?
Speaker 3 (23:01):
Yeah's leave it there.
Speaker 2 (23:02):
This has been another episode of the oud Lots podcast.
I'm Tracy Alloway. You can follow me at Tracy Alloway.
Speaker 3 (23:07):
And I'm Joe Wisenthal. You can follow me at the Stalwart.
Follow our guest William Beach, He's at Beach WW four
five three. Follow our producers Carmen Rodriguez at Kerman Ermann,
Dashel Bennett at Dashbot, and Kale Brooks and Kale Brooks.
More odd Lots content go to Bloomberg dot com slash
odd Lots, where we have a daily newsletter and all
of our episodes, and you can chat about all of
(23:28):
these topics twenty four seven in our discord Discord dot
gg slash oud locks.
Speaker 2 (23:33):
And if you enjoy odd Lots, if you like it
when we talk about the future of economic statistic collection
in the US, then please leave us a positive review
on your or the UK, and please leave us a
positive review on your favorite podcast platform. And remember, if
you are a Bloomberg subscriber, you can listen to all
of our episodes absolutely add free. All you need to
do is find the Bloomberg channel on Apple Podcasts and
(23:55):
follow the instructions there. Thanks for listening in
Bloomberg Audio Studios, Podcasts, Radio News.
Speaker 2 (00:18):
Hello and welcome to another episode of the All Thoughts Podcast.
Speaker 3 (00:21):
I'm Tracy Alloway and I'm Joe Wisenthal.
Speaker 2 (00:23):
Joe, I kind of missed this on Friday because I
was not feeling very well, so I was out sick.
But what was Friday morning like when you were sat
in front of your Bloomberg terminal.
Speaker 3 (00:34):
Well, the jobs report was pretty shocking. Obviously Friday was
just a crazy day. Yeah, So the jobs report, obviously
it was bad. I mean, the last two months massively
revised down, all of this evidence of labor market momentum
perhaps seems to be stalling the tariff effect. Maybe we
really are and it's sort of some sort of terriff
(00:54):
and do slow down. The only job creation is in
healthcare and social service.
Speaker 2 (00:59):
Not great, right, Okay, So I saw the initial jobs
report headline. So we had non farm payrolls coming in
at plus seventy three thousand, which I think was like
thirty percent lower than the average expectation. We had the
unemployment rate taking up to like four point three percent,
four point almost four point three percent. Oh, let me
(01:19):
let me sorry, I'm rounding up here.
Speaker 3 (01:21):
Okay, Okay, okay, although maybe I shouldn't.
Speaker 2 (01:23):
Do that on a podcast all about labor market statistics.
But the big thing that seemed to catch everyone's attention
was we also had these massive revisions to the report
for May and June, so we had a combined two
hundred and fifty eight thousand jobs basically lowered from the
initial reports. And this was like the biggest revision since
(01:43):
the depths of the pandemic. And then the headline that
really caught my eye while I was, you know, laying
in my sick bed on Friday afternoon, was Trump firing
the head of the Bureau of Labor Statistics, so the
agency responsible for putting out the non farm pay Roule
Report every month.
Speaker 3 (02:01):
Yeah, this is I would say. The key thing is
that a we've been used to these, you know, seeing
significant revisions. We've been talking you in particular, have been
writing a lot in the newsletter about deteriorating quality of
labor statistics. Response rates to a lot of surveys have
gone down over time, kind of like we're seeing in
political opinion surveys and so forth. So there's already been
(02:23):
this sort of anxiety. You have a lot of people online,
including President Trumpet himself, you know, been like stoking sort
of these conspiracy theories about these revisions, these attempts to
actually be transparent, and then you get the first sort
of like genuinely negative report under this administration. It's the
first one that was like, Okay, this was bad and boom,
(02:44):
Trump fires the person responsible for it. Now again, this
is one of those things where sort of like Doge itself,
where it's like, I like the idea of government efficiency, right, yeah, sure, Doge.
I like the idea of the BLS sort of doing
a better job in some way or addressing these responses.
But like Doge, which you know, we all have seen
how that's turned out. I think there are a lot
(03:06):
of more than a lot of questions, more than a
lot of questions about whether firing to be less and
replacing the head with someone who's you know, a Trumpet
point t will you know, add a more credible transparent
data clauding right.
Speaker 2 (03:20):
One thing that we have learned from all our episodes
on data collection in the US is that it is
actually a really really insanely labor intensive thing to do.
You have to actually call up a bunch of people
homeowners for instance, or you have to call up a
bunch of businesses and ask questions. You have to go
out into the field and gather individual prices. And we
(03:40):
can have a debate over whether or not you could
maybe use new technology or price scanning data to make
all of that more efficient, but for the time being,
that's how it's done, and so you need people to
do it. And if DOGE comes in and eliminates, you know,
a big chunk of the budget, then it becomes harder
to do.
Speaker 3 (03:57):
Or if it just gets more costly and the budget's
gone up, and then the one thing I would say
is that in our conversations, and again you have talked
to the BLS many times, they are really good. They're professionals,
and they take it very seriously.
Speaker 2 (04:09):
They're very responsive, however, very transparent. We could talk about
this a little bit, but perhaps less responsive than they
used to be in recent weeks. So anyway, we should
talk about all of this. I think it's really important.
And as you said, there are big questions like what
happens if a president ouse a BLS chief and then
introduces a new one. We have the perfect guests. Someone
we've spoken to before about the decline of America's data infrastructure.
(04:33):
We're going to be speaking with Bill Beach. He is,
of course, the former Commissioner of Labor Statistics, former head
of the BLS. So Bill, thank you so much for
coming back on all thoughts.
Speaker 4 (04:43):
Yeah, it's just really a pleasure to be back with you.
I wish we were talking about some other subject besides
this termination. I mean, it's just yeah, shocked.
Speaker 2 (04:52):
Well, on that note, Friday, Trump tweets or posts that
he wants to fire Erica mc Andterfer, the Commissioner of
Labor Statistics. What was going through your mind when you
saw that news.
Speaker 4 (05:04):
I was dumbcast. I just finished a luncheon meeting at
a very nice restaurant, and I was sitting in my
pickup truck and I just couldn't believe. And then I
sat there for probably thirty forty minutes, car running and
answering emails, people just sorting it out. It isn't that
he didn't have the authority to do so. He does
have the authority. We all, you know, everyone in the
(05:26):
executive branch serves as the pleasure of the president, except
those that have been specifically exempted by statute, and where
that statute has been signed by the President. I think
what was shocking about this is not that she was dismissed.
That's shocking enough. It was the possibility that we could
have now a sustained attack on official economic statistics, and
(05:46):
that would undermine confidence in those statistics and put us
in a path that other countries have followed. I mean
instantly people went to worst case. After a while, that
sort of settled down, and we began to look at
the reasons for the dismissal, which then subsequently said on
Twitter or x that they were they were unfounded. And
i haven't dropped that position since, so I'm happy to
(06:09):
talk about it.
Speaker 3 (06:10):
Well, why don't we just back up for a second.
We talked to you before, because you've been sounding the
alarm about the capacity constraints to collect good data. But
just for people who haven't listened to or haven't followed you,
you were a Trump appointing. I mean, you served at
the head of the BLS in twenty seventeen, and so
I do you know, and it's tart of like, okay,
your credibility. Why don't you just give us a little
(06:31):
bit of background about who you are, your position, and
the concerns that you've been raising for some time about
the constraints at the BLS.
Speaker 4 (06:40):
So I was really honored to be the Commissioner of
Labor Statistics from twenty nineteen until twenty twenty three. So
I served a little less than two years under President Trump.
He nominated me. It is a presidential nomination, Senate confirmed positions,
it has a statutory for your term. And then I
served the rest of my under President Biden. So I
(07:02):
was in both terms. I think that's kind of important
because the BLS Commissioner has been exempt from the normal
turning over of the presidency by party. And that's because
the view as well, these data are so important they
should not even be a part of the political appointment
process except when they come to. It's sort of like
(07:24):
you know, the Federal Reserve chair or something of that
nature has just seen as we want to keep it
out of politics. So I served over that period of time.
You know, that was the COVID period, so there were
all kinds of real challenges from that, but I think
the main thing that I drew from that period is
how important it is crucial, almost were the crisis level,
to modernize the way we collect data. The response rates
(07:47):
on the surveys are falling dramatically. I've made that point
on this program, and our costs arising dramatically, which makes
it really difficult to conduct those surveys in the future
and the way we have in the past, and then
the future is kind of going in the direction of
the employment survey, and I just said surveys are following
with the employment surveys a little different. It's not an
(08:09):
in person survey, it is an electronic survey, and we
need to invest more in the electronic side of statistical
creation or production, blended data, etc. I could do a
whole program on this, but your point was well made.
BLS has had some challenges in front of it, as
has the entire statistical system. Unfortunately, for this particular episode,
(08:32):
those challenges were really not I don't say relevant. They're relevant,
of course, but they're really not the reason why we
should have had a change at the top.
Speaker 2 (08:41):
Well, speaking of the employment report, talk to us about
what exactly happens when the BLS is publishing the initial
non farm payrolls and then what happens, you know, in
the couple months or so before it publishes the revision,
and why does it seem we are getting these large
gaps at least in recent months and years.
Speaker 4 (09:01):
Right, So the employment survey, the jobs numbers come from
a survey of businesses. The unemployment rate comes from a
survey of households. So there's two surveys involved in every
First Friday report. What we're talking about now is the
survey that goes out to businesses, and it goes out
to hundreds of thousands of businesses. Well, of course there's
(09:22):
still a probability sample, since there's well in excess of
twelve million businesses in our Census of business, our business register. Well,
those businesses are supposed to turn their surveys in at
the end of the month basically, but only about sixty
eight percent usually that's the average do so, and so
(09:43):
BLS keeps the window open for two more months, so
sixty eight percent or so at the end of the
first month they make the first estimate. Then at the
end of the second month we've get about eighty three
percent completion, and they revise that number of the first month,
and then by the third month we're into the nineties.
Usually end up around ninety three to ninety four percent
(10:05):
of all the sent out returns. Come back to us
with information. So it's a wonderful survey. The revisions are
done because we get more information from businesses. Oftentimes big
businesses answer first they have the capacity to do so,
and then we get smaller entities, state and local governments,
smaller businesses answer in those next two months. We always
(10:27):
have revisions. There's there's even when we say the number
did not change, it doesn't mean we didn't have revisions.
They just canceled each other out. There's just all these
revisions are coming in. The revisions have been high recently,
but this is not a typical of a period when
the economy is either going back to growth or going
(10:48):
down to subsidence. So oftentimes turning points in the economy
are accompanied by changes larger changes in the revisions. Particularly
in this case where we SUSPEC effect. Smaller businesses are
feeling the effects of the supply shock coming from the
tariff policy and from immigration, and that these supply shocks
(11:09):
are affecting smaller businesses more than they are larger businesses,
and state and local governments are being affected by another factor,
and that is the expiration of the COVID era money.
I think that was actually the root of a lot
of the changes that you saw on Friday. State and
local governments are not able to hire at the levels
they were going to hire at in previous years or
(11:30):
did hire at in previous years because they did not
have the subsidies that they had in previous years. So
these are important revisions. Actually, the research shows, and this
may shock you, that BLS is getting much better at
its first estimate than it was thirty years ago. In fact,
it's really kind of almost a steady progress towards greater accuracy.
(11:55):
So the big revisions are indicative of not air of
more information, and the bigger the revisions, the more likely
we are in a turning point.
Speaker 3 (12:06):
There's so much that was in that answer that I
found to be very helpful and clarifying, so thank you.
I mean, just this idea, you know, a that actually
over time, contrary to what people on Twitter might think,
that the equality of that first pass has actually gone
up is very striking.
Speaker 1 (12:24):
You know.
Speaker 3 (12:24):
Let's stipulate that, Okay, the BLIS could use some upgrading.
As you mentioned that there are these capacity constraints, non response,
the cost of going out. Maybe it needs to adopt
more technology in some way to solve for this. Does
the BLS currently have that capacity, either budgetary wise or
statutory authority such that a commissioner could just do that
(12:47):
whatever it is that ideal state, or would it need
either some sort of budgetary allocation or an Act of
Congress to get to the point where it's you know,
getting to the level that we are happy with.
Speaker 4 (12:58):
Again, it has the authority to make the changes, it
doesn't have the budget to do so. And it isn't
that BLS just needs more money, right. As a matter
of fact, I think Congress, if it paid more attention,
would say, well, what are you spending your money on now?
And let's make sure that the low priority is canceled
so that we can support the high priority. I think
(13:18):
BLS would be very happy with that. But we do
need more money. And why Well, if you're going to
change an official statistic, the unemployment rate, let's just take that.
You want to be really careful, right, You want to
run tests, you want to test out your new idea.
You don't want to have a count the number of
geese in the air, and that's the unemployment rate. You
(13:39):
want to make sure that that is a particularly good
indicator of the employment market, and you want to test
it in real time as well as experimental time. You
in real time, so run it parallel. We these tests
expand the costs of your unemployment statistics program because you're
running essentially two systems simultaneously. That to modify the current
(14:03):
population survey, which is the survey of households, would cost
around fifteen million to twenty million in experimental costs over
a two year period. It's not great. The program itself
costs about forty five million a year, but it does
add to the cost for those two years or so.
Congress could could easily do that. I mean, oh my gosh,
Congress is spending a lot more money than that on
(14:25):
things that we might disagree with. But they need to
allocate a little bit of money to CPS for that.
Does it have the authority? Here's a really interesting story.
During COVID, we canceled all travel, all conferences, you know,
and I was surprised at how many millions of dollars
went into travel and conferences and meetings. So this became
(14:46):
for me couch money during the COVID years, and I
was able to spend that money, you know, on important
things I thought, So I did a lot of things,
like I've built a new data center to get us
out of the basement of one hundred and four year
old building out to a wonderful above ground local location.
But I also started a number of research programs.
Speaker 2 (15:06):
You know.
Speaker 4 (15:06):
I made thirty one big modifications to do the CPI.
We started a whole new approach on looking at consumer expenditures,
and I subsidized some experimentation on the Jobs Report well
as it as COVID was over, we started to get
congressionally mandated programs that then sucked up all that counch money.
(15:30):
If we had a little bit of extra, we I
mean be only asked a little bit of extra to
do these things. Let's just think of the innovation that
they would happen. I think we would quickly be out
of the problems that we are facing from a methodological standpoint,
and that might then reduce the political pressure that we
saw that was so evident on Friday.
Speaker 2 (15:52):
So we're recording this August fourth, nine, five am in
the morning and thirty three seconds. I feel like I
need to be that specific nowadays with the newsflow, so
who knows what happens between now and when this episode
actually comes out, but we are expecting the administration to
announce a new BLS head in the coming days. What
(16:13):
happens now, you know, there's a Trump appointee for a
new statistics commissioner after the President has specifically stated that
he thinks the previous commissioner was politically motivated in some way.
Are people still going to believe the numbers that are
coming out? And how do you think that's actually going
to play internally at the BLS, at an organization whose whole,
(16:36):
you know, Raison Deutch is to find, you know, factual
numbers and statistics to portray the American economy.
Speaker 4 (16:44):
Well, I think there has been damage and it will
take time to recover from that damage. The last time,
there was a serious political effort on BLSS during the
Nixon administration and it took some time for BLS to
recover from that, though the action taken by the President
at that time visa BLS was not, I think as
(17:05):
significant as this one. So he just imagined that the
President decides to appoint Saint Peter, you know, as the
new BLS commissioner, and of course Saint Peter has a
reputation for honesty and for clarity of thought. Still it'll
be the case because Saint Peter won't have any control
over the way that the data are collected, the data
(17:27):
are assembled, or the estimates that are done. That's all
done outside of the of the knowledge of the commissioner.
Commissioner has no control over that. In fact, you're locked
out of that whole process. So it will be the
case that Saint Peter will have a month come when
the unemployment rate will be disappointingly low, could even be negative.
But I think given what has just happened, people will say, well,
(17:51):
Saint Peter probably influenced that number and it's not as
bad as it really is. So for a while, that
suspicion that the estimate that's in ount it's really not
the real estimate, will be in the minds of some people,
not in my mind, because I know that these people
who work there, the professionals, the full time staff, the
patriots that work there, the loyal Americans who are just
(18:15):
extremely diligent in doing an absolutely objective job. Those people
are still in place. But if you don't have my
level of knowledge, or a reasonable level of knowledge of
how internally BLS works, you're going to be subject to
these falsehoods and accusations, and that will be damaging, That
will reduce investment, that will reduce economic activity, That will
(18:38):
make at least that will create greater uncertainty about what's
happening inside the US economy. Policymakers will be less clear
in the direction that they take, So there will be
a time when we need to recover. No matter who's appointed,
I hope that it will be a short lived time.
And my guess is, given the reputation of BLS, if
(18:59):
someone you know, really reputable is appointed, then yeah, I
don't think the period of transition will be that a lot.
Speaker 2 (19:07):
All right, Well, Bill Beach really appreciate you coming back
on the show. Thank you so much for being on
Off Box.
Speaker 3 (19:13):
Thanks Bill, that was great.
Speaker 4 (19:15):
Thank you very much for asking me.
Speaker 2 (19:29):
Joe so good to get Bill back on to talk
about this. Really the perfect guest. There are so many,
I guess ironies involved in this whole conversation. The big one,
of course, is this idea that like, well, if Trump
wants to convince everyone that the US economy is doing
fantastic and job growth is great, then firing the head
of statistics and having everyone distrust, you know, the subsequent numbers,
(19:52):
because the subsequent head might be a political appointee who's
you know, a loyalist to Trump. It seems like you're
sort of shooting yourself in the foot.
Speaker 3 (20:00):
The other huge contradiction is that he says Powell is
too late. Yea wait, this is the other huge contradiction,
which is every day he slams Powell, he's like, oh,
you're too late, you need to cut rates. Well, if
the job numbers were weak, then I could understand this
argument to some extent, oh the job But he's also
casting expersions on the negative jobs numbers. So how is
(20:20):
Powell too late? I mean, this is I mean, it's
an irretrievable contradiction in the two criticism of both of
the Fed and the BLS.
Speaker 2 (20:28):
It's definitely a contradiction. I don't think, however, that Trump
necessarily draws a direct connection between the unemployment rate and interest.
Speaker 3 (20:36):
Rates as well.
Speaker 2 (20:37):
You know, he seems to just like low interest rates.
Speaker 3 (20:40):
And he seems to perceive that low interest rates are
basically a reward for an improving economy. But yes, but I.
Speaker 2 (20:48):
Mean, lots of other contradictions. You're absolutely right. So another
one is this idea that Okay, a lot of the
weakness came through on the revision side. The revisions are
mostly coming from small businesses who reply late. And if
you think about the Trump administration, you know, they always
sort of portray themselves as a good business environment for
smaller businesses, and it might be that that scenario that's
(21:10):
seeing some weakness.
Speaker 3 (21:12):
I thought your question about sort of internal morale, and
as Bill said, you know, an agency staffed with patriots,
I mean, I just you know, is Bilska going to
be continue to be a destination for people who are
patriots right or take their job seriously in the future
is a really interesting question. Also, the idea that like
maybe we could like get high data quality for like
(21:34):
twenty million dollars. Given how much that's worth, it's just
like drives you crazy if that's really all it took.
But also I really like this idea that it would
have to always be done in parallel with the current thing,
such that the new data is sort of provably backwards
compatible with the old data. Right, so if you're going
to adopt a new methodology, you need some allocation, run
two surveys at once, et cetera. Anyway, even in a
(21:56):
very short conversation. Learned a lot from Bill. Absolutely.
Speaker 2 (21:58):
I do think it's worth pointing out out that there
are other statistical agencies in the world that are experimenting
with new technology for data sources. The UK Statistics Agency
in particular, they're going to start using you know, price
scanning data for CPI.
Speaker 3 (22:12):
Wait, have you heard about the whole thing though, with
how they moved the Ons to Wales and no one
wanted to go out there, And have you heard about this?
Speaker 2 (22:20):
I have Wales. It's all that bad.
Speaker 3 (22:23):
Attribute the largest of the government across. Everyone loves that idea.
Let's not have it all be concentrated in London. They're like, okay,
we're going to make everyone move to Wales and now
it's a lot harder for the Ons to hire anyone. Anyway,
it's just a little interesting story out Maybe we'll cover
it at.
Speaker 2 (22:38):
Some point why people don't want to move to Wales.
Speaker 3 (22:41):
No interesting, it's interesting, right, Like there's intuitively I love
this idea of like, oh, why should all of the
government spending on public sector stuff at the federal level
in the richest city in the UK, or why should
it all be in dc et cetera. Should this idea
of diffuse anyway, it's a separate thing. It's actually some
interesting challenges arise for.
Speaker 2 (23:00):
Okay, shall we leave it there?
Speaker 3 (23:01):
Yeah's leave it there.
Speaker 2 (23:02):
This has been another episode of the oud Lots podcast.
I'm Tracy Alloway. You can follow me at Tracy Alloway.
Speaker 3 (23:07):
And I'm Joe Wisenthal. You can follow me at the Stalwart.
Follow our guest William Beach, He's at Beach WW four
five three. Follow our producers Carmen Rodriguez at Kerman Ermann,
Dashel Bennett at Dashbot, and Kale Brooks and Kale Brooks.
More odd Lots content go to Bloomberg dot com slash
odd Lots, where we have a daily newsletter and all
of our episodes, and you can chat about all of
(23:28):
these topics twenty four seven in our discord Discord dot
gg slash oud locks.
Speaker 2 (23:33):
And if you enjoy odd Lots, if you like it
when we talk about the future of economic statistic collection
in the US, then please leave us a positive review
on your or the UK, and please leave us a
positive review on your favorite podcast platform. And remember, if
you are a Bloomberg subscriber, you can listen to all
of our episodes absolutely add free. All you need to
do is find the Bloomberg channel on Apple Podcasts and
(23:55):
follow the instructions there. Thanks for listening in
Hosts And Creators
Joe Weisenthal
Tracy Alloway
Popular Podcasts
On Purpose with Jay Shetty
I’m Jay Shetty host of On Purpose the worlds #1 Mental Health podcast and I’m so grateful you found us. I started this podcast 5 years ago to invite you into conversations and workshops that are designed to help make you happier, healthier and more healed. I believe that when you (yes you) feel seen, heard and understood you’re able to deal with relationship struggles, work challenges and life’s ups and downs with more ease and grace. I interview experts, celebrities, thought leaders and athletes so that we can grow our mindset, build better habits and uncover a side of them we’ve never seen before. New episodes every Monday and Friday. Your support means the world to me and I don’t take it for granted — click the follow button and leave a review to help us spread the love with On Purpose. I can’t wait for you to listen to your first or 500th episode!
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Special Summer Offer: Exclusively on Apple Podcasts, try our Dateline Premium subscription completely free for one month! With Dateline Premium, you get every episode ad-free plus exclusive bonus content.
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.