That data doesn’t mean what you think it means
What happens when the data we use to run our country is wrong?
The investors I follow on Twitter, the macro podcasts I listen to, lots of incredibly intelligent people but if this cohort has a common flaw it’s that they accept the data as is. Employment figures, business creations, some even cite the CPI (Consumer Propaganda Index as Jason Burack calls it) as gospel.
My intention with this article is to illuminate how misleading these figures can be, not necessarily due to nefarious distortions by our glorious leaders but because of the flawed methods we use to collect the numbers.
Jeff Snider and Mike Green are the two investors I follow that have been most consistently bringing attention to our wacky data collection and interpretation techniques. I will consider Jeff’s point of view in a future article, but today we’ll concern ourselves with what MG has to say.
One business dies, another is born
Like every heroic story of yore, our journey begins with a chart. This picture, as well as the following quote from Mike Green, is taken from 39:30 in this video (an excellent interview with Lacy Hunt).
The chart shows the “births and deaths” of American businesses. In other words, data measuring the rate at which businesses are being created versus closing up shop. Here’s what Mike has to say about the chart.
This is actually looking at what’s at the heart of the birth death change. The dynamics of the birth death adjustment reference something that’s called business formations data. Business formations data is an attempt to use filings for employee identification numbers (the tax treatment of a corporation) they attempt to use that to gauge the level of entrepreneurship, or new business formation, that’s occurring in the United States.
I just want to highlight the pattern of what has actually played out here, as it speaks to the extreme component of that difference between household and payroll data. 100% of that, in other words half of the jobs that have been net created in the aftermath of the pandemic, can be tied directly to business formation assumptions in the birth death adjustments.
This is actually [the chart] the EIN applications and you can see the seasonal dynamics we’re all talking about, but there are a couple of things that are really interesting. For one thing, on the Trump corporate tax cut we see a jump in business applications. On the PPP loans, which requires you to have an EIN in order to participate, unsurprisingly we see a jump in applications.
In 2021 we changed the tax reporting rules so that those who were participants in the gig economy, receiving as little as $600 in compensation from a company like Uber or DoorDash, had to have an employee identification number to receive a 1099. Guess what happens?
We’ll cover what happens in the next section… But the broader point is that the techniques we use to collect and analyze data create a distorted view of reality. Look at the so called spike in business formations when the PPP loans were announced.
Did tens of thousands of Americans set up hair salons and grocery stores at the same time? Or did a bunch of people game the system to get a favorable loan…
Let’s make these figures fly
Our next chart goes deeper into that unprecedented spike in business formations in the food and accommodation sector.
What’s up with this parabolic chart, did the government put some entrepreneurial encouraging chemicals in the water? No… Doordashers and Uber Eats drivers making more than $600 a year had to have EINs to get paid. What’s an EIN? I’ll let ChatGPT explain.
An EIN is similar to a Social Security number for an individual, but it is used to identify a business entity for tax purposes. It is required for certain types of businesses, such as corporations, partnerships, and LLCs, and it may also be required for other types of entities, such as trusts and estates.
So a bunch of Uber drivers got EINs which went on the books as business formation, skewing the data set. The problem is that our financial leaders, in their all their wisdom and benevolence, use this shoddy data as proof that we have an outstanding economy. I’ll leave it to Mike to tell you how crazy this is (from the 41:30 mark in the video).
This is accommodations and food services not elsewhere classified. In other words, Doordash and Airbnb… This is what’s sitting at the heart of the data sets! That we’re relying on to tell us what’s happening with the economy. This is insane!
Conclusions
Anecdotal to be sure, but my sister (who much to her credit is not a trained economist) is applying for jobs right now and her diagnosis is that, “the economy is shit.” Yet there is Jay Pow Pow and Yolo Yellen, telling us all how great things are…
We can create better methods to analyze the economy but first we have to concede that what we’re doing right now isn’t working. A tough sell, and Powell and Yellen are probably in too deep to ever admit they’re wrong, but maybe the next generation can make the necessary changes. Let’s hope that they do.
If you liked this article there’s more where that came from. I post exclusive content on Substack and it’s all free to read 👇
