August 7, 2025 • 39 mins
Donald Trump’s decision to fire the commissioner of the Bureau of Labor Statistics, Erika McEntarfer, apparently in retribution for a report that showed slower job growth, was without precedent in recent US history. This week on Everybody’s Business we explore why presidents—at least since Richard Nixon—have left the BLS alone, and we hear from former BLS commissioner Erica Groshen on how exactly that data is collected.
According to Groshen, the recent downward revisions in the rate of job growth (which Trump alleged without any evidence was a political hatchet job) were business as usual. She explains that such revisions happen because it takes months for the more than 100,000 businesses the government surveys every month to respond. Some fill in their data electronically; some send it by email or even fax. While the BLS waits, it puts out an estimate; those estimates are often revised later on.
For the past few decades, this approach has been widely seen as a huge success. BLS data, which includes employment and inflation statistics, is relied on by researchers, economists and government policy planners—as well as by businesses. They use the data to help write budgets, plan hiring and set prices. Although Groshen optimistically contends that McEntarfer’s firing won’t immediately dent that perception, it comes amid budget cuts that have already limited the ability of BLS researchers to collect granular data and could lead to questions about reliability.
Also this week, we discuss the controversy around artificial intelligence pricing, which has recently centered around the airline industry, widely seen as the undisputed leader in customer frustration. But the strategy is in fact coming for you on pretty much every type of good and service. We also debate the significance of the American Eagle “good jeans” controversy and ask how much consumers really think about culture wars when they buy dungarees. Finally: A counterintuitive approach to warding off one of America’s most feared predators.
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 (00:11):
This is Everybody's Business from Bloomberg BusinessWeek.
Speaker 3 (00:13):
I'm Max Chafkin and I'm Stacey Vannoxsmith. Max.
Speaker 4 (00:16):
I think the theme of this week is war. We've
got the newest iteration in the trade war. Trump's new
tariff deadlines this week, India, Switzerland, all this news coming
in hot.
Speaker 2 (00:27):
Oh my god, tariff. Okay. We also have the war
on government data. We're going to do a deep dive
into the unemployment data, the data that's at the center
of all these controversies with a former head of the
Bureau of Labor Statistics.
Speaker 4 (00:41):
And then AI's war on your wallet. You might be
asking chat GPT for tips on like setting better boundaries
with your family. Meanwhile, it is silently mining your data
and figuring out how much it can overcharge you for
going home to visit your family for the holidays.
Speaker 2 (00:56):
Maybe we'll talk about that. And finally, Stacy, the underrated
story of the week, which I actually I have no
idea what it is.
Speaker 3 (01:03):
I know, I'm very excited. I've kept it a secret.
Here is what I'm gonna tell you, Max, how.
Speaker 4 (01:09):
Do you fight off one of the biggest baddest deadliest
apex predators.
Speaker 3 (01:15):
On the planet. I will give you a hint, stick around.
Oh will be revealed. Oh, Max? Another week, another round
of tariff news. This week is no.
Speaker 4 (01:37):
Exception, another big tariff deadline, lots of negotiations happening India, Switzerland.
Speaker 2 (01:44):
It's a lot, yeah. I. Meanwhile, you have these like
tech companies, Apple getting out of tariffs. We're gonna have
to come back to this, I don't know, probably as
soon as next week, Stacy. But now we want to
talk about politics, right, the politics of consumption.
Speaker 4 (01:56):
Yes, because another one of the big stories this week
had to do with Sydney Sweeney and the American Eagle ad.
Speaker 3 (02:02):
Did you see this ad?
Speaker 1 (02:04):
I'm a very world weary right now. Max, Yes, I
have seen the ad. I've absorbed the discourse somewhat unwillingly.
But I guess should I summarize.
Speaker 4 (02:14):
Yes, summarize it for in case people have been substantive stories.
Speaker 2 (02:19):
Actress Star of Euphoria is in an advertisement created by
American Eagle to sell a new line of dungarees, and
the the ads.
Speaker 3 (02:29):
Not a word for genes that they used in the
thirties or something.
Speaker 2 (02:32):
Okay, so the ads play on jeans, the pants you wear,
and genes, the genetic code in your body. The tagline
is Sydney Sweeney has great genes. This has become controversial,
I think mostly because Maga loves Sydney Sweeney.
Speaker 4 (02:53):
No, it's become controversial because in the ad, she's like,
jeans are the thing that give you your skin color and
your hair color and even your eyes. And then it
says Sidney Sweeney has gray geens and she is blonde
with blue eyes, and so people are saying this is
like a nod to eugenics or white supremacy, things like that.
Speaker 2 (03:11):
Gens are passed down from parents to offspring, often determining
traits like hair color, personality, and even eye color.
Speaker 4 (03:21):
My genes are blue.
Speaker 2 (03:23):
I don't think there are that many people in the
world who interpreted it that way, and.
Speaker 3 (03:29):
In that not only that controversy that I don't.
Speaker 2 (03:32):
Think they would have if there if a bunch of
mag accounts hadn't kind of stirred this, you know, controversy
out of nowhere. For people who are sort of invested
in the culture wars, Sidney Sweeney is this symbol of
a thing that they feel is under attack. I think
they managed to provoke a response from some very very
(03:52):
narrow corner of the left, and now all of a sudden,
here we are talking about whether or not this ad
is like a eugenic this thing or not.
Speaker 3 (04:01):
Have you seen the ad?
Speaker 2 (04:02):
I have seen the very serious several versions of the AD.
Speaker 3 (04:07):
I have actually seen the AD, and it's it is notable.
Speaker 4 (04:10):
I was kind I thought it was maybe overhyped, but
there it's it's quite I was surprised when I saw it.
Speaker 2 (04:16):
I'm not saying there isn't an undertone of something here.
I just think that we are living in this moment
where it is very easy to to sort of spin
up a controversy out of essentially nothing. I don't think
American Eagle thought that this was going to happen when
they created the ad.
Speaker 3 (04:35):
Maybe not well.
Speaker 4 (04:36):
However, Sidney Sweeney, as it turned out, was registered Republicans
loves her, yes well, And now Donald Trump has weighed
in on the American Egood.
Speaker 3 (04:45):
Here's what he said. If Sidney Sweeney is a registered Republican,
I think her AD is fantastic. So things are getting
very political.
Speaker 4 (04:55):
Like you said, all these things are very supercharged right now.
So I got cre worries about if politics are weighing
into people's buying decisions. There's been some of this with
Target and Tesla and other things. Remember, yeah, people are
buying or not buying things based on politics.
Speaker 3 (05:12):
So I wanted to see.
Speaker 4 (05:13):
New York's is a big shopping capital. There are lots
of tourists here right now. It's in August, so a
lot of people shopping. I went to an American Eagle
store and soho they have ad is huge. It's all
on this big window. It says Sidney Sweeney has great gens.
She's there like not much except for jeans and her hair,
and all these people coming in and out of the store.
(05:33):
And so I asked them what they thought about the
ad and if politics was informing their shopping.
Speaker 3 (05:38):
And here's what they said.
Speaker 4 (05:39):
Do you know about the whole Sydney Sweeney controversy.
Speaker 2 (05:43):
I don't.
Speaker 3 (05:43):
I try not to look into the politics things.
Speaker 4 (05:46):
What are the things that do factor into your decision
when you're thinking about buying something?
Speaker 2 (05:50):
Style.
Speaker 5 (05:52):
I heard about the ad when I walked across the
street and saw it in gigantic letters. I was like,
I think it's pretty disgusting.
Speaker 4 (05:59):
Honestly, do like politics ever factor into what you're buying.
Speaker 3 (06:02):
Yes.
Speaker 5 (06:03):
As an African American whose family has been here for
a very long time, it's like everything seems small until
it's not anymore. Me personally, I wouldn't shop here.
Speaker 3 (06:11):
Do politics ever influence what you buy? Yes? I like
to buy local.
Speaker 4 (06:16):
My mom was a small business owner, so I see
how the struggle is when you're competing with these really
big companies. Do you ever buy things or not buy
things because of the brand's politics?
Speaker 3 (06:26):
Not at all? It gets buy it because I want it.
Speaker 2 (06:30):
Target used to be.
Speaker 5 (06:31):
My go to store, and then after like the whole
like DII initiative. I honestly haven't been in Target whole year.
I feel like the only way big companies see things
is through like dollars and cents, and so if their
sales goes down, then it's like, Okay, that was a
bad ad and maybe let's not.
Speaker 3 (06:47):
Do it again.
Speaker 2 (06:48):
All right, Well that was I mean, I feel like
clearly this this thing has broken through. Was anyone like
what are you even talking about here? Yes?
Speaker 3 (06:55):
Yeah, several people actually had no idea what I was
talking about. It was interesting, you know, it's weird.
Speaker 2 (06:59):
I saw this on Twitter because I'm like terminally addicted
to Twitter. It was only after Donald Trump weighed in
on the controversy that people in the real world, in
my real world anyway, started talking about Like. I had
not had a real life conversation with anybody about this
ad until earlier this week, and the ad has been
around for a couple of weeks. I had really okay in.
Speaker 4 (07:21):
My real world, So Max, it does seem like shopping
has gotten political. One thing that in my career covering
business and economics for a while that I have always
thought of as kind of not political is data.
Speaker 2 (07:40):
I feel like that's wishful thinking, Stacy, but I do
see your point. There's something about these numbers, whether it's
the stock market, whether it's the prices, like the numbers
don't lie. The data doesn't lie. The unemployment rate, it
doesn't lie.
Speaker 4 (07:55):
Been the numbers in some way or another, or say well,
this number means that, or they can argue over that
like the numbers themselves.
Speaker 3 (08:01):
I feel like they're sort of like the neutral party
in the middle.
Speaker 2 (08:04):
However, it turns out that actually you can argue about him,
as we learned last week on Friday, just as our
last week's show was going live. Donald Trump showed up
and said, hold my beer, Stacy Vanix Smith and went
after the Bureau of Labor Statistics, which had just released
(08:24):
a kind of me jobs report.
Speaker 3 (08:27):
He was he's pretty disappointing. He had revised down the
other job reports. He's not good.
Speaker 4 (08:31):
He fired the head of the Bureau of Labor Statistics,
Erica mcintarfur, and truthed out tweeted out on social media
that the data was rigged.
Speaker 2 (08:39):
Right, and we wanted to talk about this with somebody
who actually understood what this data is and how it's collected.
Can you even rig it? And also, if Donald Trump
succeeds in politicizing this, what does that mean for the
US economy, for businesses, for all of us.
Speaker 4 (08:57):
Yes, we're very lucky to have Erica grosshen she headed
the Bureau of Labor Statistics from twenty thirteen to twenty seventeen,
a different Erica but also had that top job. She's
now an economic advisor at Cornell High Erica.
Speaker 6 (09:10):
Hello, glad to be here.
Speaker 3 (09:12):
So what went through.
Speaker 4 (09:13):
Your head when you heard this information? What were your
thoughts about, first of.
Speaker 3 (09:19):
All, the data getting accused of being corrupt and then
this firing.
Speaker 6 (09:23):
It's so unprecedented that I had a zillion different thoughts,
and people have asked me that over and over, and
I've told them various of my locked. I was shocked
and I was sad at the same point that the
line has been breached that had never been breached before.
I mean, presidents have tried to manipulate BLS data, but
(09:44):
they've always been stopped. The closest case, certainly in modern times,
of that kind of effect was when President Nixon decided
that the bad numbers coming out of the BLS we're
due to a Jewish cabal and tried to fire the
(10:05):
Jewish leadership in the BLS. Two of them lost their jobs,
although not the commissioner.
Speaker 4 (10:10):
Whoa So this is not entirely unprecedented. In a certain way,
it's not entirely unprecedented, but the actual firing of the
commissioner has never happened before, all.
Speaker 2 (10:20):
Right, Erica. I wanted to ask why this data that
the center of all this sort of political controversy, like
why it matters correct me if I'm wrong, like you have. Basically,
employment statistics and inflation statistics, those are the main things
that come out of the BLS, and so Some of
this is obviously useful to government, like the Social Security Administration,
(10:42):
I think uses you know, the inflation numbers figure out
how much it needs to increase the benefits. But also
lots of businesses, like if you're trying to price a good,
this is data that you can use. If you're trying
to figure out if you're a software company that makes
like human resources software, employment statistics are going to be
helpful in figuring out like what your next quarters look like?
(11:04):
Like can you just talk about, like what's at stake
with this infrastructure if we start messing with it?
Speaker 6 (11:10):
You know, our country's philosophy all the way down is
to push decision making down to the most local levels.
Speaker 2 (11:20):
Right.
Speaker 6 (11:20):
Families should make as many decisions as possible for themselves.
Businesses should make decisions as much as possible for themselves.
So the BLS has a very deep, broad website. The
very busiest part of their website is actually the Occupational
Outlook Handbook, where people who are looking for jobs, people
(11:44):
who are advising people who are looking for jobs can
see what wages and employment trends are likely to be
for over three hundred occupations.
Speaker 4 (11:56):
They're looking to see if, like what about my profession
is like is my profession hiring?
Speaker 3 (12:00):
Is it firing?
Speaker 2 (12:01):
Is? Like?
Speaker 3 (12:01):
What about the healthcare sector?
Speaker 2 (12:03):
Or I'm trying to plan compensation costs for my company
for the next year, Like what rays am I going
to have to give my employees keep them from quitting?
Speaker 6 (12:13):
What qualifications should I look for for people in that occupation?
What are ten year prospects? Where are those jobs geographically?
What are adjacent occupations that maybe I might be able
to transition to? Things like that? So there's that. When
BLS was founded, it was during a time eighteen eighty
(12:34):
four there was a lot of industrial unrest because of immigration,
trade nascent unionism.
Speaker 3 (12:43):
Technology changing.
Speaker 6 (12:45):
Technology changing, right, and the policymakers at the time seeing
all this unrest I mean, people were killing each other
in the streets over this, said well, we'd be one
step closer to industrial peace if both sides had benefit
of truth that they can trust about the labor market, conditions,
(13:07):
about pay, and about the cost of living. And so
the BLS was founded to provide that information to move
negotiators and people in conflict close to resolution faster.
Speaker 3 (13:21):
So I wanted to talk about the actual data in question.
It's a couple of surveys.
Speaker 4 (13:26):
The big one I think that people tend to really
look at and trust is the current employment statistics. That
is where we get the jobs added if I'm correct,
So you know they'll say seventy three thousand jobs added
it this month.
Speaker 3 (13:41):
Can you talk.
Speaker 4 (13:41):
About, like how do you get that information and why
the revisions? Like what is the process of getting this information?
Walk us through the data?
Speaker 6 (13:51):
Sure, So this is a survey of employers, right, every month,
one hundred and twenty thousand employers are asked about all
of the work units that they have.
Speaker 3 (14:06):
Work units being people, No.
Speaker 6 (14:08):
Not people, but the establishments the places of work oh okay,
that they have. So that covers about over six hundred
thousand places of work like offices and offices that's right, spores,
manufacturing facilities like that. So you take these one hundred
and twenty thousand employers have been recruited in advance and
(14:32):
they know they're going to be tapped for the next
X number of years to provide this information on a
monthly basis. So the biggest companies are sending this stuff
electronically an automatic feed out.
Speaker 2 (14:47):
It goes right.
Speaker 6 (14:48):
The smallest ones may be emailing it They may be
faxing it, they may be calling someone on the phone.
There are many modes because our businesses are quite varied.
So we're talking all the way from Microsoft and Amazon,
all the way down to your local dentist, mom and
pops store, the you know.
Speaker 3 (15:09):
The feed store in the middle of Iowa.
Speaker 6 (15:12):
Right the car repair shop.
Speaker 4 (15:14):
So basically you're just getting this information by any means
you can. And then are there people at the Bureau
of Labor Statistics like counting this stuff, tallying it up?
Speaker 3 (15:26):
Like what happens to this data?
Speaker 6 (15:28):
Frankly, the people aren't doing the adding. The computers are
doing the ad okay, right, this is feeding into a
data collection facility, and their reporting is fairly simple. It's
how many people did you have on your payrolls during
the pay period that contains the twelfth of the month,
(15:52):
Oh and that for that pay period, how much did
you pay all of them? How many hours on average
did they work. Then there's some basic information about the
company that's already there. What's your location, what's your industry,
And that's pretty much it. And it's more like a form,
as I say, there's no opinion in.
Speaker 4 (16:11):
There, right, So why the revisions then they have like
a week to get this information in.
Speaker 6 (16:19):
And some companies, if they're paying monthly, they haven't paid
yet for the paypier that contains the twelfth of the month.
They don't know the answer yet, or they may be
in the midst of changing their rit system, or they
may be really busy with something else, hiring people or
firing people sally, maybe sick that day right that week, right,
(16:42):
so they can't all get it in on time. Only
about sixty to seventy percent of them actually get it
in time for their first deadline.
Speaker 4 (16:52):
Oh so like thirty to forty percent of the responses
are not in on time.
Speaker 3 (16:56):
They kind of extrapolate them out right.
Speaker 7 (16:59):
So the bill then has divided all of their sample
into cells, industry location, workplace size cells. And if you
don't report during that time, then you're missing data, and
they're going to calculate the average without you, okay, And
(17:22):
so that implicitly imputes to you the average percentage change
of everybody who did report.
Speaker 2 (17:30):
So I guess one thing I'm wondering is, and I
think probably what a lot of people are wondering is
sort of like how bad is this going to be?
Other countries. There are some examples, I think Greece, Argentina
where they've sort of mucked with economic statistics.
Speaker 3 (17:47):
China and is a big one.
Speaker 2 (17:48):
Yeah, yeah, and where there have been like real ramifications
where like people are paying more in interest rates on
mortgages or to borrow money than they otherwise would because
investor like just don't really know what is going on
in the country and demand a risk premium. And the
people who pay that risk premium are basically us, like
right right, regular people who are just trying to like
(18:10):
transact in the economy. My understanding, Erica, is that the
head of the Bureau of Labor Statistics doesn't actually like
it would be pretty tricky to like get in there
and muck with the survey in any given month. Like
it's it's not like just doing this firing is going
to instantly change the survey. On the other hand, you
could imagine an erosion of trust. You could also imagine
(18:33):
a long term scenario where processes are changed and where
these statistic sortcuts, yeah, that are really seen as like
sort of gold standard. Everybody trusts them, We trust them
more than like the numbers that LinkedIn puts out or something,
because like private companies put out numbers as well, but
I don't think they're seen as quite as reliable as
these numbers. And like, what's at stake right now and
(18:55):
what does the future look like over the next few years.
Are we in danger of crossing over to a point
where basically like our statistics are like Greece's statistics, or
are we kind of a long way from there?
Speaker 6 (19:08):
Well, we're still a long way from there, but we
have crossed the line that's never been crossed before. All right,
So in the next month or two, I think there's
going to be no change in the reliability of BLS
statistics because all of these processes are still in place.
The Deputy Commissioner is now the acting commissioner. His name
(19:31):
is Bill Wyatrowski. I appointed him to.
Speaker 3 (19:34):
The job, so if you feel good about him.
Speaker 6 (19:37):
I feel good about him. He has been acting commissioner
twice already. That said, BLS is down about twenty percent
of its staff, Its leadership is down by almost by
a third. Its advisory committees have been eliminated, Its contracts
that it relied on have been terminated willy nilly, so
(20:00):
so they're not on a particularly sustainable path. Generally because
of those policy changes. Also, their funding had been lousy
for years and is still problematic. So it has continued
to produce gold standard statistics in spite of all of this.
But some of those chickens are coming home towards because
(20:21):
of that. But it's not because of manipulation.
Speaker 3 (20:25):
So there's a lot of stuff going on in the
government right now.
Speaker 4 (20:28):
And I think you could make the fair point of like, well,
so what if the data.
Speaker 3 (20:32):
Is not as trustworthy, Like what is the big deal?
Speaker 2 (20:36):
Yeah?
Speaker 6 (20:37):
So think about social Security benefits which adjusted every year
to the CPI.
Speaker 3 (20:45):
The CPI being inflation numbers.
Speaker 6 (20:48):
That's right, the consumer price Index. If the CPI is
wrong by a tenth of one percent of one basis point,
the federal government will overpay or un under pay recipients
by about a billion dollars. Oh and that's just one example,
(21:08):
right the Federal Reserve. Now, remember that the Federal Reserve
follows modern monetary policy, and it has this dual mandate
of a maximum sustainable employment and stable prices. Right, Well,
what does it rely on for that is bless data
that was originally created to quell industrial peace but has
(21:33):
turned out to be useful for this other really important decision.
Speaker 3 (21:38):
Do you think trust is eroded in the data?
Speaker 6 (21:42):
Well, I think that for the people who are listening
to what the president said, it erodes their trust in
the data. For people who are not swayed by his opinion,
then the fear that his policies will inject politics. It
(22:06):
also makes people fear the data and fear that the
data is going to be manipulated, and the knowledge that
there are these real funding and operational issues is problematic.
And then there is the actual degradation of data that
has happened because of falling survey response rates and the
(22:29):
resource issues. So right now that's showing up in the
LS in eliminating some of the granularity of the data.
So the CPI is still being produced in these inflation
inflation numbers use and really the standard error on the
top line, the national number hasn't changed very much, so
that you know, that's about as reliable as it was.
(22:52):
But the granularities, so one the city and the state
estimates of inflation, the product estimates of inflation, some of
that's just been eliminated entirely, and the margin of error,
the ability to say, okay, the big headline number went up,
But you know why and how does that impact this
(23:13):
group of people or that industry or something like that.
That's what we're losing.
Speaker 2 (23:18):
All right, Well, we're gonna have to leave it there. Eric,
thank you for joining us.
Speaker 3 (23:22):
Yes, thank you so much.
Speaker 6 (23:23):
Oh it's my pleasure. Great to be here.
Speaker 2 (23:32):
Stacy Vanicksmith, you are the biggest fan of artificial intelligence
that I know. Use it sometimes, okay, But when you
think about what do you think about like the advances
in this field, like AGI, super intelligence, Like what kind
of springs to mind? What is the future that this
technology evokes for you?
Speaker 4 (23:53):
I mean I think of AI as kind of a
personal assistant in a way.
Speaker 3 (23:58):
What are the movie times tonight? Or can you help
me write this email?
Speaker 2 (24:00):
The script for Bloomberg?
Speaker 3 (24:02):
Write this script?
Speaker 4 (24:03):
Yes, write me a shopping list kind of I don't know,
like a helper, That's how I think of it, Like
a helper, right, like.
Speaker 2 (24:09):
Some kind of amazing humanoid creation. Maybe one day it'll
even be our friends. We've talked about that. There's also
like people these futurists, like the CEO of Open AI,
Sam Altman, talks about you know, super intelligence, curing cancer
or solving global.
Speaker 3 (24:24):
Warks, Zuckerberg saying it's going to be our main friends exactly.
Speaker 2 (24:27):
You're not going to need a romantic partner anymore because
you can just hang on Facebook all day.
Speaker 3 (24:31):
It never forgets anniversaries AI exactly.
Speaker 2 (24:34):
But you know what, here's the thing, Stacey, I actually
don't think any of that really reflects where this technology
is going. I think what all of these data centers
that are going up everywhere around the country with all
of this technology, these these coders who are being offered
hundreds of millions of dollars a year to jump from
(24:54):
like Anthropic to open AI to Facebook, They're all just
going to basically make our airline tickets more expensive. Explain Okay,
So for this week in Business Week, I looked into
a sort of controversy around AI pricing. So this is
the idea that instead of you know how, like you
(25:14):
go on the website for an airline and you try
to buy a plane ticket and one day it costs
one hundred and fifty dollars, and the next day you
go back and it costs two hundred dollars, and the
day after that it maybe cost one hundred and twenty
five dollars. The price is always changing. This is called
dynamic pricing. But there's this idea in the airline industry
that they're going to use AI to make it even better.
(25:35):
Last year, during a Delta Airlines investor event, the president
of the company said they were going to like, maybe
we could raise the price twenty dollars, maybe we could
raise it forty dollars, sort of suggesting that essentially what
the AI is going to do is look at you
and figure out what you, Stacey Vanock Smith are willing
to pay to fly to Idaho today and get set
(25:56):
the price at the exact highest point that they could
poss set it.
Speaker 4 (26:01):
Okay, but this has been I know this has been
going on for a long time in certain iterations with
cookies and stuff. I know this because my family lives
in Idaho, and every time I try to book a
ticket to Idaho that I have to go into I
try to like incognito mode and all this stuff, because
they all the cookies in my data they've mined of
mine over the years mean that they're serving me up
(26:22):
expensive ticket prices.
Speaker 3 (26:23):
So how's this.
Speaker 2 (26:23):
Here's the thing, It's it's gonna get worse. Because I
looked into one of the companies that is working with
airlines is a company called Fetcher, and basically found this
sales document that they put out, And this is a
sales document not aimed at consumers, but aimed at businesses
who of course just want to raise prices, want to
raise revenue. And it talks about the idea of applying
(26:44):
what they call alien super intelligence to the problem of
figuring out how much money they can charge you. And
the idea is to take the technologies that were used
by high frequency traders, you know, these these crazy strategies
that are too complex to even perceptual, but to bring
that to domains with consumers. So when you're buying an
(27:05):
airplane ticket, there's some like crazy algorithm working behind the
scenes that is figuring out how much you're gonna pay.
And I should say this idea of AI pricing is
everywhere now, it's not just in airlines. And this sense
that you have of the frustration of buying airline tickets stacy,
it is going. It's going to be everywhere very soon,
and in fact is everywhere. We've seen landlords using it
(27:27):
to figure out rents using AI software, and pretty much
every online retailer in the planet has used AI to
price to some extent. You know about ride share fares,
I assume because like Uber, when you know depending on
when you when you do it, that affects the price.
It also affects ride share wages. So so how much
the driver is going to get paid? And we've we've
seen drivers complain, incredibly complain that it's not necessarily the
(27:51):
same thing. It's like your surge price may be different
than the driver surge price. Meat packing prices. There were
allegations of AIB used to raise those. So let me
just lay out the nightmare scenario which which you alluded to,
which is like you need to fly somewhere for a
funeral or for a medical procedure. You cannot delay your
trip and an AI figures that out and jacks it
(28:14):
up like five times a normal price. Or like you're
driving for Uber and you have zero dollars in your
bank account and Uber figures out, Oh, like this guy
is desperate, Like I'm just going to pay him five
dollars for this ride instead of fifteen, which is what
I would have pay and he's going to do it
because he's that desperate. That is the scary situation. I
(28:35):
think that that is not quite there yet. And in
the course of this story and amid the outcry over
Delta and the use of AI, the company has come
out and set essentially we are not using personal data
to set AI prices. They did not say yet, but
I do think that yet. Is there that like many
(28:56):
companies are going to use personal data?
Speaker 3 (28:58):
The Amazon does well, well yeah, yeah, openly yeah.
Speaker 2 (29:01):
So And the thing is, the thing about AI is
that you don't really know what data is being used
because it's a black box. You're just putting a bunch
of information into a large language model and telling a
large language model or this model to which the company
that works at Delta calls a large market model, but
(29:22):
basically the same thing to sort of figure out. And
we've seen with large language models that personal data sometimes
leaks into these data sets. They don't mean to put
your social Security number or your phone number into open AI,
but because open AI is crawling something with some personal information,
it could find its way in there. So like that
is a possibility. The other thing is there are ways
(29:43):
to learn about you without actually accessing your bank account
or knowing ath or looking at you or more right
like like really complicated behavioral targeting and I think the
truth is that we are just more predictable than we realize,
and they are able to figure out your own personal
interests just by your behavior. I think what we're leading
(30:05):
towards is these companies are going to have so many
even if they don't say it's personal data, they're going
to be so many different fare classes. Instead of having
like ten different fare classes, there'll be like three hundred
fare classes. And if you have three hundred fair classes,
you could imagine one of those is like desperate bereave traveler.
(30:26):
One is like broke uber drive. You. Oh, like you
could get really granular or granular at the point where
like it doesn't really matter if they know you Stacy Vanixsmith,
but they know what your exact situation is and are
able to use that to their advantage.
Speaker 4 (30:40):
Is that different than like data mining and consumer profiles
and stuff like that.
Speaker 3 (30:44):
Is it like just like just a supercharge thing steroids.
Speaker 2 (30:48):
It's concerning that they could be using these like alien
super intelligence to like squeeze a few bucks out of you.
But there's also this concern over AI collusion. So this
is the idea that I have an AI to set
prices and one of my competitors has an AI to
set prices, and the two AIS work together to just raise.
Speaker 3 (31:08):
Place for price fixing.
Speaker 2 (31:09):
Indeed, yes, And the thing is this has already happened
so at least according to a complaint against Amazon. So
Amazon tried out this AI. According to this FTC complaint,
which Amazon disputed, that would essentially raise prices briefly and
then see whether competitors would raise prices in turn, and
(31:31):
if they did, keep the prices up, but if not,
drop them down. And then this led to a bunch
of research by academics. There's a paper by some Carnegie
Mellon professors were basically showing how if you had two
companies using these AI pricing algorithms, the two algorithms would
start colluding with one another to raise prices.
Speaker 4 (31:50):
Is there a way to avoid this or is this
just star like? Is this just the future we need
to prepare for.
Speaker 2 (31:57):
The cynical tech reporter in me says, we're this is
the future we need to FuMB just because this stuff
is basically unregulated right now. The only thing that is
stopping companies from doing more of this, essentially, I think,
is embarrassment. There was a point where you had a
very active FTC under Lena Khan, where a lot of
(32:18):
those examples I rattled off, we know about those because
the FTC uncovered them and put them in complaints. Trump has,
you know, obviously made clear that he doesn't want to
regulate businesses the way they were regulated under the Bid administration. Also,
there is this really powerful force within the Trump administration
led by all these guys who gave Trump, you know,
hundreds of millions of dollars who donated to his inauguration,
(32:41):
who basically want to have zero regulation in AI. So
I'm not really holding my bread. There are ways that
you can kind of protect yourself, and you alluded to
some of them, right you're clearing your cookies, you know,
looking at airfares or any kind of price on different
web browsers, just like being more cognizant of the ways
(33:01):
that your own behavior and how you are purchasing a
given item can affect the ultimate price you get, and
using that knowledge to basically shop around, even if there's
only one airline for one route, like just try try
to buy it in a few different ways and see
what happens. The last thing, and this is maybe this
is the optimistic tech reporter in me. I do think
(33:23):
we're going to see companies that attempt to apply these
same algorithms on our own behalf. So like there's an
alien intelligence working for Delta to try or any airline
to try to maximize revenue. We're going to have our
own algorithms, I hope to try to overcome those like
will be our will have to act like our own.
Speaker 3 (33:43):
And stuff like yeah online banns.
Speaker 2 (33:46):
Kayak already does this, but I'm confident that as this
stuff becomes bigger, there will be incentives for basically middlemen
to come along and try to help make our prices lower.
So I guess what I'm saying is that Silicon Valley
will save us this time, save us from it from it. Yea,
(34:12):
all right, Stacey, I know you've got a really important
story to share with me. That story there has not
gotten enough attention from the lamestream media, and I'm ready
to hear it.
Speaker 4 (34:24):
This is my favorite story of twenty twenty five. So
I don't know how much you know about cattle ranching.
Speaker 2 (34:31):
I know there are cows and cowboys.
Speaker 4 (34:33):
I think there's a big problem on cattle ranches in
the West, which is wolves.
Speaker 3 (34:39):
So wolves got reintroduced.
Speaker 4 (34:40):
They were endangered. They were reintroduced into a lot of
the American West. The problem is the wolves come onto
cattle ranches and it's like a buffet. It's like a
golden corral. Literally, it is like they walk into a
buffet and they are like picking off these cows. And
the ranchers, because the wolves are endangered, cannot kill the wolves.
Speaker 2 (34:59):
This got be a big issue where you're from.
Speaker 3 (35:01):
It is, Yeah, my parents had cattle ranch, I didn't
have wolves. But this is a bit.
Speaker 2 (35:05):
I mean, if the wolves came, they wouldn't be like,
look at these majestic beasts. Yes, they would be.
Speaker 3 (35:10):
So wonderful to see them.
Speaker 4 (35:11):
They would be afraid, yes, And so this is a
big issue and and ranchers are trying to figure out
it is illegal to kill the wolves, but you have
to think them out of But wolves are really smart
and they're really strong. How do you keep them away
from the cattle? Cows are not fast? Okay, so they're
trying all kinds of interesting things. And one of the
(35:33):
things that they are trying to do is they fly
drones around.
Speaker 3 (35:36):
Heat seeking drones.
Speaker 4 (35:37):
Okay, they can spot wolves and then they play sounds
for the wolves to scare them away.
Speaker 2 (35:44):
Like what kind of sounds like like a dog whistle type,
like a sound that's so high pitched that human ears
can't hear it, but it sends the wolves into a tizzy.
Speaker 3 (35:51):
Yeah, what is going to scare a wolf?
Speaker 2 (35:53):
Right?
Speaker 4 (35:53):
These are apex predators. So, as it turns out, there
are a couple things that they are using to scare
the wolves. One, I mean they use like thundersounds and
gunshot sounds.
Speaker 3 (36:05):
Also a CDC. The song Thunderstruck by.
Speaker 2 (36:09):
A CDC's a great song. So I don't know what
the wolves are thinking.
Speaker 4 (36:14):
Maybe they're thinking that it's actually an overrated song, but
it gets better.
Speaker 2 (36:19):
Do you think it's the name they're like, Oh, the
song is called Thunderstruck, so not just thunderstruck.
Speaker 4 (36:25):
Ad also play a scene from a movie, the movie
Marriage Story. This movie with Adam Driver and Scarlett Johansson. Right,
there is a scene where they get into a big fight.
Speaker 2 (36:36):
And the wolves don't like that.
Speaker 3 (36:38):
They play that fight scene to scare off wolves. Here
they play, Yeah, here's the.
Speaker 2 (36:43):
Clip that they I'm gonna try to connect with my
lupine brain. People used to tell me that you were
too selfish to be a great artist, and I used
to defend you.
Speaker 6 (36:53):
They were absolutely right.
Speaker 5 (36:54):
Well, your best acting is behind you.
Speaker 2 (36:56):
You're back to being a you.
Speaker 7 (36:58):
Get flighted me, you're a Really.
Speaker 3 (37:00):
You want to present yourself as a victim because it's
a good legal strategy.
Speaker 5 (37:04):
Fine, but you and I both know you.
Speaker 2 (37:06):
Chose this life. I have never seen the movie Marriage
to it.
Speaker 4 (37:10):
Can you imagine, like this is what they're you doing
to scare off wolves?
Speaker 2 (37:13):
But wow, I just.
Speaker 3 (37:16):
A couple fighting and like splitting up.
Speaker 2 (37:18):
You just lost your appetite, right, like toxic. I don't
really want to go to the Golden Corral. I just
I just want to go and cry.
Speaker 4 (37:25):
Yeah, wolves are like it's time for some self care.
I just I can't even eat right now. I've just
lost my appetite.
Speaker 2 (37:31):
Okay, So the heat seeking drones like fly around like
there's some software that's like wolf and then it blasts
and then and marriage.
Speaker 3 (37:41):
From a marriage story, Yeah, to scare off the wolves.
Apparently it's working.
Speaker 2 (37:45):
Yeah, if you and your significant other are are hiking
together and you see a mountain lion just starts fighting.
Speaker 4 (37:50):
It's time it's time to go into like all the
resentment that you have tamped down in the back of
your brain.
Speaker 2 (37:57):
And use that stay safe out there, everybody and everybody
who is listening to this show. First of all, definitely
want to know what songs you use to scare large predators.
But also if you have a pricing story, like we're
talking about our own personal experiences around AI and pricing.
If you have experienced this in some way or you
(38:17):
have thoughts on it, definitely send us an email. Everybody's
at Bloomberg dot net. Everybody with an s at Bloomberg
dot net.
Speaker 4 (38:23):
Or if you have an idea of what else could
scare off the wolves. Maybe you know markets news, you know,
actually this podcast could be useful to scare off wolves.
Speaker 2 (38:36):
Who's the head of the Department of Interior under Trump now? Anyway,
if you're listening, just give it a shot. Yeah, could
send us right to the top of the can send
us at the top.
Speaker 3 (38:45):
Yeah, we're big with wolves, I.
Speaker 2 (38:47):
Mean for now.
Speaker 5 (38:47):
Though.
Speaker 2 (38:48):
We could use your reviews to send us further up
in these rankings and also to let other people know
about the show.
Speaker 3 (38:54):
Yes, it helps them find the show. You can do
that wherever you get your podcasts.
Speaker 2 (39:03):
This show is produced by Stacy Wong. Magnus Hendrickson is
our supervising producer. Amy Keen is our editor. We get
engineering support from Blake Maples, Dave Purcell, factchecks, Sage Bauman,
Heads Bloomberg Podcast, and special thanks to Jeff Muscus, Julia
Rubin and Maria Ling. If you have a minute, if
you are a wolf or a human, please rate and
(39:24):
review this show or drown or drone. It'll mean a
lot to us. And if you have a story that
should be our business, email us at Everybody's at Bloomberg
dot net. That's everybody with an s APN at Bloomberg
dot net. Thanks for listening. We will see you next week.
Bloomberg Audio Studios, podcasts, radio news.
Speaker 2 (00:11):
This is Everybody's Business from Bloomberg BusinessWeek.
Speaker 3 (00:13):
I'm Max Chafkin and I'm Stacey Vannoxsmith. Max.
Speaker 4 (00:16):
I think the theme of this week is war. We've
got the newest iteration in the trade war. Trump's new
tariff deadlines this week, India, Switzerland, all this news coming
in hot.
Speaker 2 (00:27):
Oh my god, tariff. Okay. We also have the war
on government data. We're going to do a deep dive
into the unemployment data, the data that's at the center
of all these controversies with a former head of the
Bureau of Labor Statistics.
Speaker 4 (00:41):
And then AI's war on your wallet. You might be
asking chat GPT for tips on like setting better boundaries
with your family. Meanwhile, it is silently mining your data
and figuring out how much it can overcharge you for
going home to visit your family for the holidays.
Speaker 2 (00:56):
Maybe we'll talk about that. And finally, Stacy, the underrated
story of the week, which I actually I have no
idea what it is.
Speaker 3 (01:03):
I know, I'm very excited. I've kept it a secret.
Here is what I'm gonna tell you, Max, how.
Speaker 4 (01:09):
Do you fight off one of the biggest baddest deadliest
apex predators.
Speaker 3 (01:15):
On the planet. I will give you a hint, stick around.
Oh will be revealed. Oh, Max? Another week, another round
of tariff news. This week is no.
Speaker 4 (01:37):
Exception, another big tariff deadline, lots of negotiations happening India, Switzerland.
Speaker 2 (01:44):
It's a lot, yeah. I. Meanwhile, you have these like
tech companies, Apple getting out of tariffs. We're gonna have
to come back to this, I don't know, probably as
soon as next week, Stacy. But now we want to
talk about politics, right, the politics of consumption.
Speaker 4 (01:56):
Yes, because another one of the big stories this week
had to do with Sydney Sweeney and the American Eagle ad.
Speaker 3 (02:02):
Did you see this ad?
Speaker 1 (02:04):
I'm a very world weary right now. Max, Yes, I
have seen the ad. I've absorbed the discourse somewhat unwillingly.
But I guess should I summarize.
Speaker 4 (02:14):
Yes, summarize it for in case people have been substantive stories.
Speaker 2 (02:19):
Actress Star of Euphoria is in an advertisement created by
American Eagle to sell a new line of dungarees, and
the the ads.
Speaker 3 (02:29):
Not a word for genes that they used in the
thirties or something.
Speaker 2 (02:32):
Okay, so the ads play on jeans, the pants you wear,
and genes, the genetic code in your body. The tagline
is Sydney Sweeney has great genes. This has become controversial,
I think mostly because Maga loves Sydney Sweeney.
Speaker 4 (02:53):
No, it's become controversial because in the ad, she's like,
jeans are the thing that give you your skin color and
your hair color and even your eyes. And then it
says Sidney Sweeney has gray geens and she is blonde
with blue eyes, and so people are saying this is
like a nod to eugenics or white supremacy, things like that.
Speaker 2 (03:11):
Gens are passed down from parents to offspring, often determining
traits like hair color, personality, and even eye color.
Speaker 4 (03:21):
My genes are blue.
Speaker 2 (03:23):
I don't think there are that many people in the
world who interpreted it that way, and.
Speaker 3 (03:29):
In that not only that controversy that I don't.
Speaker 2 (03:32):
Think they would have if there if a bunch of
mag accounts hadn't kind of stirred this, you know, controversy
out of nowhere. For people who are sort of invested
in the culture wars, Sidney Sweeney is this symbol of
a thing that they feel is under attack. I think
they managed to provoke a response from some very very
(03:52):
narrow corner of the left, and now all of a sudden,
here we are talking about whether or not this ad
is like a eugenic this thing or not.
Speaker 3 (04:01):
Have you seen the ad?
Speaker 2 (04:02):
I have seen the very serious several versions of the AD.
Speaker 3 (04:07):
I have actually seen the AD, and it's it is notable.
Speaker 4 (04:10):
I was kind I thought it was maybe overhyped, but
there it's it's quite I was surprised when I saw it.
Speaker 2 (04:16):
I'm not saying there isn't an undertone of something here.
I just think that we are living in this moment
where it is very easy to to sort of spin
up a controversy out of essentially nothing. I don't think
American Eagle thought that this was going to happen when
they created the ad.
Speaker 3 (04:35):
Maybe not well.
Speaker 4 (04:36):
However, Sidney Sweeney, as it turned out, was registered Republicans
loves her, yes well, And now Donald Trump has weighed
in on the American Egood.
Speaker 3 (04:45):
Here's what he said. If Sidney Sweeney is a registered Republican,
I think her AD is fantastic. So things are getting
very political.
Speaker 4 (04:55):
Like you said, all these things are very supercharged right now.
So I got cre worries about if politics are weighing
into people's buying decisions. There's been some of this with
Target and Tesla and other things. Remember, yeah, people are
buying or not buying things based on politics.
Speaker 3 (05:12):
So I wanted to see.
Speaker 4 (05:13):
New York's is a big shopping capital. There are lots
of tourists here right now. It's in August, so a
lot of people shopping. I went to an American Eagle
store and soho they have ad is huge. It's all
on this big window. It says Sidney Sweeney has great gens.
She's there like not much except for jeans and her hair,
and all these people coming in and out of the store.
(05:33):
And so I asked them what they thought about the
ad and if politics was informing their shopping.
Speaker 3 (05:38):
And here's what they said.
Speaker 4 (05:39):
Do you know about the whole Sydney Sweeney controversy.
Speaker 2 (05:43):
I don't.
Speaker 3 (05:43):
I try not to look into the politics things.
Speaker 4 (05:46):
What are the things that do factor into your decision
when you're thinking about buying something?
Speaker 2 (05:50):
Style.
Speaker 5 (05:52):
I heard about the ad when I walked across the
street and saw it in gigantic letters. I was like,
I think it's pretty disgusting.
Speaker 4 (05:59):
Honestly, do like politics ever factor into what you're buying.
Speaker 3 (06:02):
Yes.
Speaker 5 (06:03):
As an African American whose family has been here for
a very long time, it's like everything seems small until
it's not anymore. Me personally, I wouldn't shop here.
Speaker 3 (06:11):
Do politics ever influence what you buy? Yes? I like
to buy local.
Speaker 4 (06:16):
My mom was a small business owner, so I see
how the struggle is when you're competing with these really
big companies. Do you ever buy things or not buy
things because of the brand's politics?
Speaker 3 (06:26):
Not at all? It gets buy it because I want it.
Speaker 2 (06:30):
Target used to be.
Speaker 5 (06:31):
My go to store, and then after like the whole
like DII initiative. I honestly haven't been in Target whole year.
I feel like the only way big companies see things
is through like dollars and cents, and so if their
sales goes down, then it's like, Okay, that was a
bad ad and maybe let's not.
Speaker 3 (06:47):
Do it again.
Speaker 2 (06:48):
All right, Well that was I mean, I feel like
clearly this this thing has broken through. Was anyone like
what are you even talking about here? Yes?
Speaker 3 (06:55):
Yeah, several people actually had no idea what I was
talking about. It was interesting, you know, it's weird.
Speaker 2 (06:59):
I saw this on Twitter because I'm like terminally addicted
to Twitter. It was only after Donald Trump weighed in
on the controversy that people in the real world, in
my real world anyway, started talking about Like. I had
not had a real life conversation with anybody about this
ad until earlier this week, and the ad has been
around for a couple of weeks. I had really okay in.
Speaker 4 (07:21):
My real world, So Max, it does seem like shopping
has gotten political. One thing that in my career covering
business and economics for a while that I have always
thought of as kind of not political is data.
Speaker 2 (07:40):
I feel like that's wishful thinking, Stacy, but I do
see your point. There's something about these numbers, whether it's
the stock market, whether it's the prices, like the numbers
don't lie. The data doesn't lie. The unemployment rate, it
doesn't lie.
Speaker 4 (07:55):
Been the numbers in some way or another, or say well,
this number means that, or they can argue over that
like the numbers themselves.
Speaker 3 (08:01):
I feel like they're sort of like the neutral party
in the middle.
Speaker 2 (08:04):
However, it turns out that actually you can argue about him,
as we learned last week on Friday, just as our
last week's show was going live. Donald Trump showed up
and said, hold my beer, Stacy Vanix Smith and went
after the Bureau of Labor Statistics, which had just released
(08:24):
a kind of me jobs report.
Speaker 3 (08:27):
He was he's pretty disappointing. He had revised down the
other job reports. He's not good.
Speaker 4 (08:31):
He fired the head of the Bureau of Labor Statistics,
Erica mcintarfur, and truthed out tweeted out on social media
that the data was rigged.
Speaker 2 (08:39):
Right, and we wanted to talk about this with somebody
who actually understood what this data is and how it's collected.
Can you even rig it? And also, if Donald Trump
succeeds in politicizing this, what does that mean for the
US economy, for businesses, for all of us.
Speaker 4 (08:57):
Yes, we're very lucky to have Erica grosshen she headed
the Bureau of Labor Statistics from twenty thirteen to twenty seventeen,
a different Erica but also had that top job. She's
now an economic advisor at Cornell High Erica.
Speaker 6 (09:10):
Hello, glad to be here.
Speaker 3 (09:12):
So what went through.
Speaker 4 (09:13):
Your head when you heard this information? What were your
thoughts about, first of.
Speaker 3 (09:19):
All, the data getting accused of being corrupt and then
this firing.
Speaker 6 (09:23):
It's so unprecedented that I had a zillion different thoughts,
and people have asked me that over and over, and
I've told them various of my locked. I was shocked
and I was sad at the same point that the
line has been breached that had never been breached before.
I mean, presidents have tried to manipulate BLS data, but
(09:44):
they've always been stopped. The closest case, certainly in modern times,
of that kind of effect was when President Nixon decided
that the bad numbers coming out of the BLS we're
due to a Jewish cabal and tried to fire the
(10:05):
Jewish leadership in the BLS. Two of them lost their jobs,
although not the commissioner.
Speaker 4 (10:10):
Whoa So this is not entirely unprecedented. In a certain way,
it's not entirely unprecedented, but the actual firing of the
commissioner has never happened before, all.
Speaker 2 (10:20):
Right, Erica. I wanted to ask why this data that
the center of all this sort of political controversy, like
why it matters correct me if I'm wrong, like you have. Basically,
employment statistics and inflation statistics, those are the main things
that come out of the BLS, and so Some of
this is obviously useful to government, like the Social Security Administration,
(10:42):
I think uses you know, the inflation numbers figure out
how much it needs to increase the benefits. But also
lots of businesses, like if you're trying to price a good,
this is data that you can use. If you're trying
to figure out if you're a software company that makes
like human resources software, employment statistics are going to be
helpful in figuring out like what your next quarters look like?
(11:04):
Like can you just talk about, like what's at stake
with this infrastructure if we start messing with it?
Speaker 6 (11:10):
You know, our country's philosophy all the way down is
to push decision making down to the most local levels.
Speaker 2 (11:20):
Right.
Speaker 6 (11:20):
Families should make as many decisions as possible for themselves.
Businesses should make decisions as much as possible for themselves.
So the BLS has a very deep, broad website. The
very busiest part of their website is actually the Occupational
Outlook Handbook, where people who are looking for jobs, people
(11:44):
who are advising people who are looking for jobs can
see what wages and employment trends are likely to be
for over three hundred occupations.
Speaker 4 (11:56):
They're looking to see if, like what about my profession
is like is my profession hiring?
Speaker 3 (12:00):
Is it firing?
Speaker 2 (12:01):
Is? Like?
Speaker 3 (12:01):
What about the healthcare sector?
Speaker 2 (12:03):
Or I'm trying to plan compensation costs for my company
for the next year, Like what rays am I going
to have to give my employees keep them from quitting?
Speaker 6 (12:13):
What qualifications should I look for for people in that occupation?
What are ten year prospects? Where are those jobs geographically?
What are adjacent occupations that maybe I might be able
to transition to? Things like that? So there's that. When
BLS was founded, it was during a time eighteen eighty
(12:34):
four there was a lot of industrial unrest because of immigration,
trade nascent unionism.
Speaker 3 (12:43):
Technology changing.
Speaker 6 (12:45):
Technology changing, right, and the policymakers at the time seeing
all this unrest I mean, people were killing each other
in the streets over this, said well, we'd be one
step closer to industrial peace if both sides had benefit
of truth that they can trust about the labor market, conditions,
(13:07):
about pay, and about the cost of living. And so
the BLS was founded to provide that information to move
negotiators and people in conflict close to resolution faster.
Speaker 3 (13:21):
So I wanted to talk about the actual data in question.
It's a couple of surveys.
Speaker 4 (13:26):
The big one I think that people tend to really
look at and trust is the current employment statistics. That
is where we get the jobs added if I'm correct,
So you know they'll say seventy three thousand jobs added
it this month.
Speaker 3 (13:41):
Can you talk.
Speaker 4 (13:41):
About, like how do you get that information and why
the revisions? Like what is the process of getting this information?
Walk us through the data?
Speaker 6 (13:51):
Sure, So this is a survey of employers, right, every month,
one hundred and twenty thousand employers are asked about all
of the work units that they have.
Speaker 3 (14:06):
Work units being people, No.
Speaker 6 (14:08):
Not people, but the establishments the places of work oh okay,
that they have. So that covers about over six hundred
thousand places of work like offices and offices that's right, spores,
manufacturing facilities like that. So you take these one hundred
and twenty thousand employers have been recruited in advance and
(14:32):
they know they're going to be tapped for the next
X number of years to provide this information on a
monthly basis. So the biggest companies are sending this stuff
electronically an automatic feed out.
Speaker 2 (14:47):
It goes right.
Speaker 6 (14:48):
The smallest ones may be emailing it They may be
faxing it, they may be calling someone on the phone.
There are many modes because our businesses are quite varied.
So we're talking all the way from Microsoft and Amazon,
all the way down to your local dentist, mom and
pops store, the you know.
Speaker 3 (15:09):
The feed store in the middle of Iowa.
Speaker 6 (15:12):
Right the car repair shop.
Speaker 4 (15:14):
So basically you're just getting this information by any means
you can. And then are there people at the Bureau
of Labor Statistics like counting this stuff, tallying it up?
Speaker 3 (15:26):
Like what happens to this data?
Speaker 6 (15:28):
Frankly, the people aren't doing the adding. The computers are
doing the ad okay, right, this is feeding into a
data collection facility, and their reporting is fairly simple. It's
how many people did you have on your payrolls during
the pay period that contains the twelfth of the month,
(15:52):
Oh and that for that pay period, how much did
you pay all of them? How many hours on average
did they work. Then there's some basic information about the
company that's already there. What's your location, what's your industry,
And that's pretty much it. And it's more like a form,
as I say, there's no opinion in.
Speaker 4 (16:11):
There, right, So why the revisions then they have like
a week to get this information in.
Speaker 6 (16:19):
And some companies, if they're paying monthly, they haven't paid
yet for the paypier that contains the twelfth of the month.
They don't know the answer yet, or they may be
in the midst of changing their rit system, or they
may be really busy with something else, hiring people or
firing people sally, maybe sick that day right that week, right,
(16:42):
so they can't all get it in on time. Only
about sixty to seventy percent of them actually get it
in time for their first deadline.
Speaker 4 (16:52):
Oh so like thirty to forty percent of the responses
are not in on time.
Speaker 3 (16:56):
They kind of extrapolate them out right.
Speaker 7 (16:59):
So the bill then has divided all of their sample
into cells, industry location, workplace size cells. And if you
don't report during that time, then you're missing data, and
they're going to calculate the average without you, okay, And
(17:22):
so that implicitly imputes to you the average percentage change
of everybody who did report.
Speaker 2 (17:30):
So I guess one thing I'm wondering is, and I
think probably what a lot of people are wondering is
sort of like how bad is this going to be?
Other countries. There are some examples, I think Greece, Argentina
where they've sort of mucked with economic statistics.
Speaker 3 (17:47):
China and is a big one.
Speaker 2 (17:48):
Yeah, yeah, and where there have been like real ramifications
where like people are paying more in interest rates on
mortgages or to borrow money than they otherwise would because
investor like just don't really know what is going on
in the country and demand a risk premium. And the
people who pay that risk premium are basically us, like
right right, regular people who are just trying to like
(18:10):
transact in the economy. My understanding, Erica, is that the
head of the Bureau of Labor Statistics doesn't actually like
it would be pretty tricky to like get in there
and muck with the survey in any given month. Like
it's it's not like just doing this firing is going
to instantly change the survey. On the other hand, you
could imagine an erosion of trust. You could also imagine
(18:33):
a long term scenario where processes are changed and where
these statistic sortcuts, yeah, that are really seen as like
sort of gold standard. Everybody trusts them, We trust them
more than like the numbers that LinkedIn puts out or something,
because like private companies put out numbers as well, but
I don't think they're seen as quite as reliable as
these numbers. And like, what's at stake right now and
(18:55):
what does the future look like over the next few years.
Are we in danger of crossing over to a point
where basically like our statistics are like Greece's statistics, or
are we kind of a long way from there?
Speaker 6 (19:08):
Well, we're still a long way from there, but we
have crossed the line that's never been crossed before. All right,
So in the next month or two, I think there's
going to be no change in the reliability of BLS
statistics because all of these processes are still in place.
The Deputy Commissioner is now the acting commissioner. His name
(19:31):
is Bill Wyatrowski. I appointed him to.
Speaker 3 (19:34):
The job, so if you feel good about him.
Speaker 6 (19:37):
I feel good about him. He has been acting commissioner
twice already. That said, BLS is down about twenty percent
of its staff, Its leadership is down by almost by
a third. Its advisory committees have been eliminated, Its contracts
that it relied on have been terminated willy nilly, so
(20:00):
so they're not on a particularly sustainable path. Generally because
of those policy changes. Also, their funding had been lousy
for years and is still problematic. So it has continued
to produce gold standard statistics in spite of all of this.
But some of those chickens are coming home towards because
(20:21):
of that. But it's not because of manipulation.
Speaker 3 (20:25):
So there's a lot of stuff going on in the
government right now.
Speaker 4 (20:28):
And I think you could make the fair point of like, well,
so what if the data.
Speaker 3 (20:32):
Is not as trustworthy, Like what is the big deal?
Speaker 2 (20:36):
Yeah?
Speaker 6 (20:37):
So think about social Security benefits which adjusted every year
to the CPI.
Speaker 3 (20:45):
The CPI being inflation numbers.
Speaker 6 (20:48):
That's right, the consumer price Index. If the CPI is
wrong by a tenth of one percent of one basis point,
the federal government will overpay or un under pay recipients
by about a billion dollars. Oh and that's just one example,
(21:08):
right the Federal Reserve. Now, remember that the Federal Reserve
follows modern monetary policy, and it has this dual mandate
of a maximum sustainable employment and stable prices. Right, Well,
what does it rely on for that is bless data
that was originally created to quell industrial peace but has
(21:33):
turned out to be useful for this other really important decision.
Speaker 3 (21:38):
Do you think trust is eroded in the data?
Speaker 6 (21:42):
Well, I think that for the people who are listening
to what the president said, it erodes their trust in
the data. For people who are not swayed by his opinion,
then the fear that his policies will inject politics. It
(22:06):
also makes people fear the data and fear that the
data is going to be manipulated, and the knowledge that
there are these real funding and operational issues is problematic.
And then there is the actual degradation of data that
has happened because of falling survey response rates and the
(22:29):
resource issues. So right now that's showing up in the
LS in eliminating some of the granularity of the data.
So the CPI is still being produced in these inflation
inflation numbers use and really the standard error on the
top line, the national number hasn't changed very much, so
that you know, that's about as reliable as it was.
(22:52):
But the granularities, so one the city and the state
estimates of inflation, the product estimates of inflation, some of
that's just been eliminated entirely, and the margin of error,
the ability to say, okay, the big headline number went up,
But you know why and how does that impact this
(23:13):
group of people or that industry or something like that.
That's what we're losing.
Speaker 2 (23:18):
All right, Well, we're gonna have to leave it there. Eric,
thank you for joining us.
Speaker 3 (23:22):
Yes, thank you so much.
Speaker 6 (23:23):
Oh it's my pleasure. Great to be here.
Speaker 2 (23:32):
Stacy Vanicksmith, you are the biggest fan of artificial intelligence
that I know. Use it sometimes, okay, But when you
think about what do you think about like the advances
in this field, like AGI, super intelligence, Like what kind
of springs to mind? What is the future that this
technology evokes for you?
Speaker 4 (23:53):
I mean I think of AI as kind of a
personal assistant in a way.
Speaker 3 (23:58):
What are the movie times tonight? Or can you help
me write this email?
Speaker 2 (24:00):
The script for Bloomberg?
Speaker 3 (24:02):
Write this script?
Speaker 4 (24:03):
Yes, write me a shopping list kind of I don't know,
like a helper, That's how I think of it, Like
a helper, right, like.
Speaker 2 (24:09):
Some kind of amazing humanoid creation. Maybe one day it'll
even be our friends. We've talked about that. There's also
like people these futurists, like the CEO of Open AI,
Sam Altman, talks about you know, super intelligence, curing cancer
or solving global.
Speaker 3 (24:24):
Warks, Zuckerberg saying it's going to be our main friends exactly.
Speaker 2 (24:27):
You're not going to need a romantic partner anymore because
you can just hang on Facebook all day.
Speaker 3 (24:31):
It never forgets anniversaries AI exactly.
Speaker 2 (24:34):
But you know what, here's the thing, Stacey, I actually
don't think any of that really reflects where this technology
is going. I think what all of these data centers
that are going up everywhere around the country with all
of this technology, these these coders who are being offered
hundreds of millions of dollars a year to jump from
(24:54):
like Anthropic to open AI to Facebook, They're all just
going to basically make our airline tickets more expensive. Explain Okay,
So for this week in Business Week, I looked into
a sort of controversy around AI pricing. So this is
the idea that instead of you know how, like you
(25:14):
go on the website for an airline and you try
to buy a plane ticket and one day it costs
one hundred and fifty dollars, and the next day you
go back and it costs two hundred dollars, and the
day after that it maybe cost one hundred and twenty
five dollars. The price is always changing. This is called
dynamic pricing. But there's this idea in the airline industry
that they're going to use AI to make it even better.
(25:35):
Last year, during a Delta Airlines investor event, the president
of the company said they were going to like, maybe
we could raise the price twenty dollars, maybe we could
raise it forty dollars, sort of suggesting that essentially what
the AI is going to do is look at you
and figure out what you, Stacey Vanock Smith are willing
to pay to fly to Idaho today and get set
(25:56):
the price at the exact highest point that they could
poss set it.
Speaker 4 (26:01):
Okay, but this has been I know this has been
going on for a long time in certain iterations with
cookies and stuff. I know this because my family lives
in Idaho, and every time I try to book a
ticket to Idaho that I have to go into I
try to like incognito mode and all this stuff, because
they all the cookies in my data they've mined of
mine over the years mean that they're serving me up
(26:22):
expensive ticket prices.
Speaker 3 (26:23):
So how's this.
Speaker 2 (26:23):
Here's the thing, It's it's gonna get worse. Because I
looked into one of the companies that is working with
airlines is a company called Fetcher, and basically found this
sales document that they put out, And this is a
sales document not aimed at consumers, but aimed at businesses
who of course just want to raise prices, want to
raise revenue. And it talks about the idea of applying
(26:44):
what they call alien super intelligence to the problem of
figuring out how much money they can charge you. And
the idea is to take the technologies that were used
by high frequency traders, you know, these these crazy strategies
that are too complex to even perceptual, but to bring
that to domains with consumers. So when you're buying an
(27:05):
airplane ticket, there's some like crazy algorithm working behind the
scenes that is figuring out how much you're gonna pay.
And I should say this idea of AI pricing is
everywhere now, it's not just in airlines. And this sense
that you have of the frustration of buying airline tickets stacy,
it is going. It's going to be everywhere very soon,
and in fact is everywhere. We've seen landlords using it
(27:27):
to figure out rents using AI software, and pretty much
every online retailer in the planet has used AI to
price to some extent. You know about ride share fares,
I assume because like Uber, when you know depending on
when you when you do it, that affects the price.
It also affects ride share wages. So so how much
the driver is going to get paid? And we've we've
seen drivers complain, incredibly complain that it's not necessarily the
(27:51):
same thing. It's like your surge price may be different
than the driver surge price. Meat packing prices. There were
allegations of AIB used to raise those. So let me
just lay out the nightmare scenario which which you alluded to,
which is like you need to fly somewhere for a
funeral or for a medical procedure. You cannot delay your
trip and an AI figures that out and jacks it
(28:14):
up like five times a normal price. Or like you're
driving for Uber and you have zero dollars in your
bank account and Uber figures out, Oh, like this guy
is desperate, Like I'm just going to pay him five
dollars for this ride instead of fifteen, which is what
I would have pay and he's going to do it
because he's that desperate. That is the scary situation. I
(28:35):
think that that is not quite there yet. And in
the course of this story and amid the outcry over
Delta and the use of AI, the company has come
out and set essentially we are not using personal data
to set AI prices. They did not say yet, but
I do think that yet. Is there that like many
(28:56):
companies are going to use personal data?
Speaker 3 (28:58):
The Amazon does well, well yeah, yeah, openly yeah.
Speaker 2 (29:01):
So And the thing is, the thing about AI is
that you don't really know what data is being used
because it's a black box. You're just putting a bunch
of information into a large language model and telling a
large language model or this model to which the company
that works at Delta calls a large market model, but
(29:22):
basically the same thing to sort of figure out. And
we've seen with large language models that personal data sometimes
leaks into these data sets. They don't mean to put
your social Security number or your phone number into open AI,
but because open AI is crawling something with some personal information,
it could find its way in there. So like that
is a possibility. The other thing is there are ways
(29:43):
to learn about you without actually accessing your bank account
or knowing ath or looking at you or more right
like like really complicated behavioral targeting and I think the
truth is that we are just more predictable than we realize,
and they are able to figure out your own personal
interests just by your behavior. I think what we're leading
(30:05):
towards is these companies are going to have so many
even if they don't say it's personal data, they're going
to be so many different fare classes. Instead of having
like ten different fare classes, there'll be like three hundred
fare classes. And if you have three hundred fair classes,
you could imagine one of those is like desperate bereave traveler.
(30:26):
One is like broke uber drive. You. Oh, like you
could get really granular or granular at the point where
like it doesn't really matter if they know you Stacy Vanixsmith,
but they know what your exact situation is and are
able to use that to their advantage.
Speaker 4 (30:40):
Is that different than like data mining and consumer profiles
and stuff like that.
Speaker 3 (30:44):
Is it like just like just a supercharge thing steroids.
Speaker 2 (30:48):
It's concerning that they could be using these like alien
super intelligence to like squeeze a few bucks out of you.
But there's also this concern over AI collusion. So this
is the idea that I have an AI to set
prices and one of my competitors has an AI to
set prices, and the two AIS work together to just raise.
Speaker 3 (31:08):
Place for price fixing.
Speaker 2 (31:09):
Indeed, yes, And the thing is this has already happened
so at least according to a complaint against Amazon. So
Amazon tried out this AI. According to this FTC complaint,
which Amazon disputed, that would essentially raise prices briefly and
then see whether competitors would raise prices in turn, and
(31:31):
if they did, keep the prices up, but if not,
drop them down. And then this led to a bunch
of research by academics. There's a paper by some Carnegie
Mellon professors were basically showing how if you had two
companies using these AI pricing algorithms, the two algorithms would
start colluding with one another to raise prices.
Speaker 4 (31:50):
Is there a way to avoid this or is this
just star like? Is this just the future we need
to prepare for.
Speaker 2 (31:57):
The cynical tech reporter in me says, we're this is
the future we need to FuMB just because this stuff
is basically unregulated right now. The only thing that is
stopping companies from doing more of this, essentially, I think,
is embarrassment. There was a point where you had a
very active FTC under Lena Khan, where a lot of
(32:18):
those examples I rattled off, we know about those because
the FTC uncovered them and put them in complaints. Trump has,
you know, obviously made clear that he doesn't want to
regulate businesses the way they were regulated under the Bid administration. Also,
there is this really powerful force within the Trump administration
led by all these guys who gave Trump, you know,
hundreds of millions of dollars who donated to his inauguration,
(32:41):
who basically want to have zero regulation in AI. So
I'm not really holding my bread. There are ways that
you can kind of protect yourself, and you alluded to
some of them, right you're clearing your cookies, you know,
looking at airfares or any kind of price on different
web browsers, just like being more cognizant of the ways
(33:01):
that your own behavior and how you are purchasing a
given item can affect the ultimate price you get, and
using that knowledge to basically shop around, even if there's
only one airline for one route, like just try try
to buy it in a few different ways and see
what happens. The last thing, and this is maybe this
is the optimistic tech reporter in me. I do think
(33:23):
we're going to see companies that attempt to apply these
same algorithms on our own behalf. So like there's an
alien intelligence working for Delta to try or any airline
to try to maximize revenue. We're going to have our
own algorithms, I hope to try to overcome those like
will be our will have to act like our own.
Speaker 3 (33:43):
And stuff like yeah online banns.
Speaker 2 (33:46):
Kayak already does this, but I'm confident that as this
stuff becomes bigger, there will be incentives for basically middlemen
to come along and try to help make our prices lower.
So I guess what I'm saying is that Silicon Valley
will save us this time, save us from it from it. Yea,
(34:12):
all right, Stacey, I know you've got a really important
story to share with me. That story there has not
gotten enough attention from the lamestream media, and I'm ready
to hear it.
Speaker 4 (34:24):
This is my favorite story of twenty twenty five. So
I don't know how much you know about cattle ranching.
Speaker 2 (34:31):
I know there are cows and cowboys.
Speaker 4 (34:33):
I think there's a big problem on cattle ranches in
the West, which is wolves.
Speaker 3 (34:39):
So wolves got reintroduced.
Speaker 4 (34:40):
They were endangered. They were reintroduced into a lot of
the American West. The problem is the wolves come onto
cattle ranches and it's like a buffet. It's like a
golden corral. Literally, it is like they walk into a
buffet and they are like picking off these cows. And
the ranchers, because the wolves are endangered, cannot kill the wolves.
Speaker 2 (34:59):
This got be a big issue where you're from.
Speaker 3 (35:01):
It is, Yeah, my parents had cattle ranch, I didn't
have wolves. But this is a bit.
Speaker 2 (35:05):
I mean, if the wolves came, they wouldn't be like,
look at these majestic beasts. Yes, they would be.
Speaker 3 (35:10):
So wonderful to see them.
Speaker 4 (35:11):
They would be afraid, yes, And so this is a
big issue and and ranchers are trying to figure out
it is illegal to kill the wolves, but you have
to think them out of But wolves are really smart
and they're really strong. How do you keep them away
from the cattle? Cows are not fast? Okay, so they're
trying all kinds of interesting things. And one of the
(35:33):
things that they are trying to do is they fly
drones around.
Speaker 3 (35:36):
Heat seeking drones.
Speaker 4 (35:37):
Okay, they can spot wolves and then they play sounds
for the wolves to scare them away.
Speaker 2 (35:44):
Like what kind of sounds like like a dog whistle type,
like a sound that's so high pitched that human ears
can't hear it, but it sends the wolves into a tizzy.
Speaker 3 (35:51):
Yeah, what is going to scare a wolf?
Speaker 2 (35:53):
Right?
Speaker 4 (35:53):
These are apex predators. So, as it turns out, there
are a couple things that they are using to scare
the wolves. One, I mean they use like thundersounds and
gunshot sounds.
Speaker 3 (36:05):
Also a CDC. The song Thunderstruck by.
Speaker 2 (36:09):
A CDC's a great song. So I don't know what
the wolves are thinking.
Speaker 4 (36:14):
Maybe they're thinking that it's actually an overrated song, but
it gets better.
Speaker 2 (36:19):
Do you think it's the name they're like, Oh, the
song is called Thunderstruck, so not just thunderstruck.
Speaker 4 (36:25):
Ad also play a scene from a movie, the movie
Marriage Story. This movie with Adam Driver and Scarlett Johansson. Right,
there is a scene where they get into a big fight.
Speaker 2 (36:36):
And the wolves don't like that.
Speaker 3 (36:38):
They play that fight scene to scare off wolves. Here
they play, Yeah, here's the.
Speaker 2 (36:43):
Clip that they I'm gonna try to connect with my
lupine brain. People used to tell me that you were
too selfish to be a great artist, and I used
to defend you.
Speaker 6 (36:53):
They were absolutely right.
Speaker 5 (36:54):
Well, your best acting is behind you.
Speaker 2 (36:56):
You're back to being a you.
Speaker 7 (36:58):
Get flighted me, you're a Really.
Speaker 3 (37:00):
You want to present yourself as a victim because it's
a good legal strategy.
Speaker 5 (37:04):
Fine, but you and I both know you.
Speaker 2 (37:06):
Chose this life. I have never seen the movie Marriage
to it.
Speaker 4 (37:10):
Can you imagine, like this is what they're you doing
to scare off wolves?
Speaker 2 (37:13):
But wow, I just.
Speaker 3 (37:16):
A couple fighting and like splitting up.
Speaker 2 (37:18):
You just lost your appetite, right, like toxic. I don't
really want to go to the Golden Corral. I just
I just want to go and cry.
Speaker 4 (37:25):
Yeah, wolves are like it's time for some self care.
I just I can't even eat right now. I've just
lost my appetite.
Speaker 2 (37:31):
Okay, So the heat seeking drones like fly around like
there's some software that's like wolf and then it blasts
and then and marriage.
Speaker 3 (37:41):
From a marriage story, Yeah, to scare off the wolves.
Apparently it's working.
Speaker 2 (37:45):
Yeah, if you and your significant other are are hiking
together and you see a mountain lion just starts fighting.
Speaker 4 (37:50):
It's time it's time to go into like all the
resentment that you have tamped down in the back of
your brain.
Speaker 2 (37:57):
And use that stay safe out there, everybody and everybody
who is listening to this show. First of all, definitely
want to know what songs you use to scare large predators.
But also if you have a pricing story, like we're
talking about our own personal experiences around AI and pricing.
If you have experienced this in some way or you
(38:17):
have thoughts on it, definitely send us an email. Everybody's
at Bloomberg dot net. Everybody with an s at Bloomberg
dot net.
Speaker 4 (38:23):
Or if you have an idea of what else could
scare off the wolves. Maybe you know markets news, you know,
actually this podcast could be useful to scare off wolves.
Speaker 2 (38:36):
Who's the head of the Department of Interior under Trump now? Anyway,
if you're listening, just give it a shot. Yeah, could
send us right to the top of the can send
us at the top.
Speaker 3 (38:45):
Yeah, we're big with wolves, I.
Speaker 2 (38:47):
Mean for now.
Speaker 5 (38:47):
Though.
Speaker 2 (38:48):
We could use your reviews to send us further up
in these rankings and also to let other people know
about the show.
Speaker 3 (38:54):
Yes, it helps them find the show. You can do
that wherever you get your podcasts.
Speaker 2 (39:03):
This show is produced by Stacy Wong. Magnus Hendrickson is
our supervising producer. Amy Keen is our editor. We get
engineering support from Blake Maples, Dave Purcell, factchecks, Sage Bauman,
Heads Bloomberg Podcast, and special thanks to Jeff Muscus, Julia
Rubin and Maria Ling. If you have a minute, if
you are a wolf or a human, please rate and
(39:24):
review this show or drown or drone. It'll mean a
lot to us. And if you have a story that
should be our business, email us at Everybody's at Bloomberg
dot net. That's everybody with an s APN at Bloomberg
dot net. Thanks for listening. We will see you next week.
Popular Podcasts
Fudd Around And Find Out
UConn basketball star Azzi Fudd brings her championship swag to iHeart Women’s Sports with Fudd Around and Find Out, a weekly podcast that takes fans along for the ride as Azzi spends her final year of college trying to reclaim the National Championship and prepare to be a first round WNBA draft pick. Ever wonder what it’s like to be a world-class athlete in the public spotlight while still managing schoolwork, friendships and family time? It’s time to Fudd Around and Find Out!
The Breakfast Club
The World's Most Dangerous Morning Show, The Breakfast Club, With DJ Envy, Jess Hilarious, And Charlamagne Tha God!
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com