Last week’s jobs data came in solid, boosting the market as we headed into a delightful holiday weekend.
But now that the festivities are over — it’s time to come clean…
Because in reality, last week’s print was BS.
It was a massive miscalculation that misrepresented the reality of today’s economy in a very big way. And the Federal Reserve is about to run with it.
In today’s episode, we’re going to “zoom out” and look at the actual process that the government uses to create these official statistics, so you can see exactly how they’ve made this mistake, and what that means for your portfolio going ahead.
Let’s dive in:
Video transcript:
Hello and welcome to my favorite time of the week. It is Moneyball Economics Time.
So fellow Moneyball economists, let’s get going…
I want to talk today about the payroll number that just came out because contrary to what you heard, it was an incredibly weak number. It suggests we’ve got problems out there in the broader economy, but that’s contrary to what you might’ve heard. So today I thought I would walk you through the mechanics, some of the “sausage making” that goes into this number.
And by the way, not just this number, but pretty much all conventional data out there. You’re going to come to an understanding of how messy it really is.
Starting off, why do we care about job data? Again, we know that job data is important for at least two reasons. One is, as an investor, we want to know if the economy is speeding up or slowing down.
Let’s face it, jobs function in two ways. If the jobs are growing, that means the broader economy is growing. It’s pulling along people, but it also means more directly and practically — the more job demand there is out there, the more salaries are flowing through the economy. We are a consumer-based economy, so that’s important.
But the second reason … interest rate cuts or hikes.
The US Federal Reserve looks at labor data, and so we need to look at that labor data to sort of try to figure out and predict where they’re going to go.
To cut to the chase. They’re not going to make a move in July. It looks like because of what just happened, that’s not the right move. But again, the Fed doesn’t want to cut rates at all, and so they just got some ground cover, some protection.
Now having said that, let’s take a look at the mechanics. Let’s start off with the methodology by which they grabbed data. There are a hundred million plus workers out there, and once a month they’ve got to find out how many are employed, how many aren’t employed, how many new employees do we have?
A hundred million plus, 140 million. This is an incredible task. How do they do it? Well, let’s take a look…
It starts off with a different methodology. If you’re talking jobless claims versus if you’re talking payrolls, jobless claims are much more reliable because someone had to go in and say, “I don’t have a job.” It’s very clean. There’s very minimal tweaking because someone had to submit paperwork and that paperwork had to be verified.
On the other hand, when you’re looking at payrolls and another thing called JOLT, job openings, when you’re looking at these are survey, survey based, what’s the difference? Well, again, jobless claim, someone had to say, and it had to be proven, “I don’t have a job.” Very simple, very binary.
But when you’re talking about a survey, but you have a problem, how do you reach all these businesses? Well, you can’t. And so you’re going to look at a slice of what’s going on out there and you hope that the modeling that you do is going to work. So when it comes to payrolls, they’re not looking at every business out there. They do a sampling.
That sampling is about 685,000 businesses. And when you’re talking JOLTS, for example, it’s 20,000 businesses. So in essence, jolt, which tells us how many job openings there are and what the employment intention might be, 20,000 businesses dominate that … except they don’t.
Here’s the first step. So we talked about how the sample size matters, and it’s different from jobless claims, which is 100%. Here you’re talking about a subset, big subset, but a subset. But then that subsets even smaller. It shrinks because we have a problem, ever since COVID started, with response rates.
How do you think you get this feedback from businesses? Phone calls, paper, that’s right, but they’re not obligated or forced to respond. And what’s happened is over time the response rates have crashed. And so for example, if you’re talking jolts, we’re barely getting one third, we’re barely getting 30% response rate.
That means we may be asking 21,000 companies, “how you doing?” We’re getting a response from maybe six or 7,000, just still a good sample set, but it’s not that good considering the breadth and depth that we need to track.
Similarly, as you can see, it’s the same kind of story with payrolls. The participation rate, the response rate has gone down. It’s still a large number though that we’re dealing with. However, just understand that this is a model. It’s a sample now of a sample.
And so we’re getting less and less reliability. And also understand you might have lower response rates because vacations, you might have lower response rates because weather, external shocks, whatever. But the key part here is you’re really not seeing as much of a response as you might want.
And so that’s why every month for a few months after the preliminary one is put out, you get payrolls adjusted, revised, and there can be big swings as more information comes in. And then once a year, it’s even trued up once again because there is one obligatory survey that goes out once a year in the early part of the year and businesses have to respond to that one. So you get a truing up, and that’s why every year you look back and it’s like, “whoa, the story has changed dramatically!” Hell, within a couple of months it’s changed dramatically.
So quite frankly, it’s all about the sausage making.
Now you’re talking about moves that matter because you’re investing accordingly. Now, do you want to know truth, look back last year and what was happening? No, that doesn’t matter because that’s not affecting the stock market or the bond market today. And so for all intents and purposes, the sausage making, we only care about the number when it comes out. And then maybe we care about next month depending on rate cut kind of considerations. Let’s stay with what just happened.
So the number came out and it came out strong. I thought it was going to be weak. I was right, but at the same time I was wrong. The number came out that headline number, it was strong, 147,000. So we’ve had three consecutive months where the payrolls have been 140,000 or above-ish.
Those are relatively strong numbers. Those are not numbers that correlate to a recession. And so the Fed is sitting here saying, “three months in a row, that’s a trend. This economy is solid. I can now ignore the labor data and I can focus on inflation data.”
Well, here’s the thing. We’re going to get a number next month. I think it is going to be weak. But it doesn’t matter because the fed’s going to say, well, I had three strong ones, and ignore July. That means rate cuts are not going to happen until we get to September. Now, are they going to happen in September? Depends on what happens next month in July and August.
And that’s what we want to talk about today is as we get through the sausage making, let’s talk about what it meant for this month and what it means for next month and beyond that number, that juicy number that we talked about. Well, in fact, if you peel it back, it has two parts. You’ve got the private sector part and then you got the public sector.
Let’s talk private sector. I nailed private sector. It was terrible. We went from having 140,000 private sector jobs created in the month of May to 77,000 in the month of June. That is called falling off the cliff. We went from a lot of jobs to very few job creation.
In fact, within that number that 77, 58,000 came from healthcare. So in fact, if you look at the broader economy beyond healthcare, there were no jobs added. That is a problem. Now, the reason’s simple. Not only do we have layoffs, but we also have basically companies don’t need to hire more because as I’ve been saying, the amount of activity this year is essentially the same as last year. And so you don’t hire more people. You don’t need to staff up beyond because you already have the staff, right? You’re doing the same amount of customers walking in your door as you did last year. Why would you staff up beyond what you staffed up last year?
Even worse, you should get some kind of operational efficiency so you staff a few less. So if you’re not staffing up again, rip out healthcare, ignore the government hiring. There was no hiring in June. That means the economy is frozen. It could tilt up. It could tilt down.
Let’s now talk about where the miss was. I nailed the private sector. I know what’s going on around us, but the public sector, they did a curve ball, and that’s where we want to look at some of the sausage making in the public sector, which includes not just government, but public education. They found almost, I mean, it’s incredible, 60,000 more workers in the local and state public schools.
Think about that.
This is the peak layoff season for schools. Summer vacation has started. We are deep into it, and somehow 60,000 more people were hired in this school system. Meanwhile, if you look at the private sector, they’ve laid them off. They’re in layoff mode and you can see that.
So what happened here is model mechanics, they timed the school year incorrectly. I’m going to show you this in a minute. With these charts, what happened was they said, “oh, we don’t think school is out. We don’t think all these schools have laid off their groundskeepers, the admins, the bus drivers, the cafeteria workers, whatever. We don’t think they’ve laid them off yet, and so therefore when we look at next month, they’re going to be laid off.”
So if there were 60,000 people who were somehow not fired and therefore additive to the payrolls, well next month in July, they’re going to be fired. So we’re looking at a 60,000 headwind. Now, you take this non hiring in the private sector, you take a 60,000 negative headwind. It’s going to be crappy next month, but it won’t matter. It will matter if this trend continues into August.
But let’s stick with June for a second. I mentioned sausage making. I mentioned the methodology that’s kind of messy and doesn’t work. Well, when I say that they messed up with education, let me show you:
This is local education, and what you’re going to see in this first chart is, and I’ve inverted it because they lay off people, I’ve inverted it because just easier to digest the number of people laid off in the month of June, going back 20 years, and I’ve taken out by the way, 2020 and 2021, and that’s just messy. But if you go back 20 years, ignore COVID, and you look at what happens in June, the number of people nominal, you’ll notice that this June was the fewest number of people laid off in almost 20 years!
Now, why does that matter? Because our school system is 50% larger.
You see where I’m going with this? We laid off a hell of a lot fewer people this June than we have in 20 years, and there’s no reason for that because we’ve added more cafeteria workers. We’ve added more admin staff.
To show you a different glimpse of this, let me compare this to how this number, the June number, compares to the total local education payrolls. What’s the ratio, the number of people laid off in June relative to the underlying total base? Take a look:
Again, removing 2020 and 2021, what you’ll notice is this June this year had the lowest participation rate, we’ll call it the ratio. The number of people relative to the base is the lowest we’ve seen in 20 years. It’s saying, Hey, we decided to retain more people than ever.
You see the gap, the implied gap here at the very end for 2025 is we should have seen about 30,000 more or 40,000 more layoffs in the month of June for the state, excuse me, for the local government than we saw. And the same kind of thing happens in the state.
Essentially, there’s some kind of weirdness going on with the way they decided to calendar the layoffs, excuse me, the payroll cuts for summer vacation. Now, it could be the schools didn’t respond. Again, going back to that response rate. Whatever the case, we did not have a strong June payroll. Again, going back to private sector was a shit show. Almost nobody hired except for healthcare.
Public sector was just this one-off quirk. Oh, did the schools lay off this many people this month or should it be in next month? We don’t know. We’ll just assume the best and well look at that, we got a great number for this month.
It basically sets next month up for the collapse. But then, hey, the Fed gets to say, “we had three solid months. What’s one bad month?” No reason for this to impact our decision making.
And that’s where we are going forward. I am looking at what’s going on. Hiring remains solid, but remember, remember it is not solid up. It’s just holding in place. The layoffs are not strong at the moment.
And as we go into the summertime, you’re going to have a couple of companies that lay off, Microsoft and Intel both announced they’re doing layoffs, but in general, you don’t really see huge layoffs in the summertime because people are on vacation. They’ll do this coming back in August, say, so the Feds bought itself some time. There won’t be a lot of major moves in July when it comes to payrolls and layoffs.
There might not be a lot in August, so the Fed’s kind of bought itself some time.
So net, I did see a collapse in the private sector hiring in June. It was covered up by this quirk. We get to July. There’s no covering up anything. The question is though, are companies okay? And are they hiring? I won’t know that for another week. I’m looking back at the data. The data I collect and that I see doesn’t show specific weakness. It just doesn’t show any specific strength. And when you’re talking payrolls, it’s got to be strength and additive.
Alright, stick around, we’ll talk some more and catch-up more this week.
We’re in it to win it folks,
Zatlin out!
Andrew Zatlin
Editor, Moneyball Economics