Interlude: Why the Recent Economic Data Doesn’t Match Up With What Everyone Is Saying
Is consumer spending as weak as it seems, or is it strong like the data suggests? There’s only one way to find out – data diving!
Dear Readers,
Normally I wouldn’t be back in your inbox until next week, but something’s been bothering me about the recent consumer retail spending report, as it has with those I’ve been chatting with the last few days. And on a weekend, there’s nothing people like more than putting your feet up with an ice cold beverage and delving into economic data, amirite?
On Thursday, the Census Bureau released the monthly retail sales data for July. The top line number was that retail sales were up 1.0% in July versus June – a number more than three times the rate analysts were expecting. The market loved it, as the Dow Jones jumped over 1,000 points and the S&P 500 was up almost four percent on the week.
But something seemed off to me, as it did with other colleagues I spoke with. “What country are they living in,” was one comment I heard. “They’re just making this data up,” said another. If consumer spending is supposedly so strong at the retail level, why do retailers feel like they’re hanging on by a thread?
I felt the same way, but I’m also more trusting of government data than others may be. The answer isn’t necessarily in the top line number, but rather the details that are hidden within each report. Generally, you just have to look hard enough. I started my search in one of the most important quarterly reports we see every few months: Walmart.
The largest retailer and private employer in the world, Walmart is a good gauge for the United States as a whole. According to the company, 70 percent of Americans live within five miles of a Walmart-brand store (which includes Sam's Club); 90 percent live within 10 miles; and 96 percent live within 20 miles. That essentially means that almost every American has access to a Walmart if they choose to shop there.
And most do, as the data shows. Walmart’s domestic revenue this quarter was $110.9 billion, which averages to around $36.97 billion per month. That means that, if you exclude gas and motor vehicle sales, Walmart accounts for nearly eight percent of this country’s total retail sales. That’s an enormous share of the pie, considering how much spending there is each year.
Walmart’s results suggested the same as the economic data – consumer spending is strong. “We have not seen any incremental fraying of consumer health,” said CFO John David Rainey during the earnings call. “We aren’t experiencing a weaker consumer,” said CEO Doug McMillon. That all sounds great, if the data supported that. But again, if you look past the top-line number, there’s more to the story.
Sales of groceries were up mid-single-digits for Walmart, but general merchandise sales were flat. And they specifically mentioned that their health-and-wellness segment grew by double-digit percentages, strengthened by demand for Ozempic. In addition, their global advertising business grew 26 percent, while their Walmart+ membership dollars rose by double digits as well. The company said that more than half of its operating income growth was attributable to advertising and membership revenue – not retail sales. In fact, they even mentioned that they’ve lowered prices on over 7,000 items in order to prevent a slowing in spending. Does that sound like a strong retail market to you?
Now’s a great time to return to the economic data. Remember that 1.0% growth from June to July? Well, it turns out, that isn’t telling the full story either. June’s retail sales number was negatively impacted by an artificial drop in auto sales at dealerships due to a cyberattack that crippled dealerships across the country. If you remove those deferred motor vehicle and vehicle parts sales from this month’s data, retail spending was only up 0.4% from June. Calculate in the inflation rate from July (0.2%), and that makes for a real increase of 0.2% in retail spending. Again, not the strongest showing from the world’s largest economy.
Allow me a quick digression before we dig into this data. Most people don’t realize, but the economic data that comes out is what’s known as a “sample”. A sample is a statistical test of a small population, used to generalize to the larger population. Polling is a large industry that uses samples as well. As a whole, samples are great ways to get an idea of what a certain population is doing or thinking. But, unlike when Walmart reports a quarterly revenue, these numbers are not factual or provable.1 It is meant to give an idea of what is going on, and that’s it.
These economic numbers are a sample of about 4,800 firms around the country. It also uses what’s called a 90 percent confidence interval. Without getting more in the weeds, essentially that means that the data analysts are 90 percent sure that the data will fall within their stated range. So let’s take a look at the margin of error of this data.
Plus or minus 0.5 percent??
Sorry, my statistics PTSD is kicking in. If you take that margin of error with our 0.2% “real” retail growth, that means that the analysts are 90 percent certain that the “real” number is between -0.3% and 0.7%.2 That’s a pretty large range to state confidently that consumer spending is strong. If you saw a political poll this season that said Candidate A was up 0.2% on Candidate B, and there was a margin of error of 0.5%, that literally means that it’s a statistical tie (e.g. zero percent). It means the data cannot confidently say that either candidate is ahead. The same should be said with this economic data. It’s not wrong, by any means – in fact, the Department of Commerce’s data is some of the best we have in this country, and they are experts at gathering and reporting it. But there’s a difference between excellence in data gathering and excellence in data analysis. The Bureau just reports the raw numbers, with no opinion – as well they should. The opinions come from Wall Street analysts and economists, who should know better when it comes to analyzing basic statistical data.
So, if we’re making an argument that the retail growth for July is statistically zero, let’s now take a look at the raw data by sector in the Census Bureau’s chart, which I’ve conveniently put into a spreadsheet:3
To translate, what this data is showing is that certain sectors of retail have seen their spending drop precipitously, while spending on non-discretionary items (e.g. groceries, health and wellness, food, clothing) have increased. That makes perfect sense. As prices have skyrocketed over the last few years, consumers, who have nominally more money to spend than they did in 2021 (the average salary has increased about $10,000 per year since 2020), have to put most of that wage increase toward the essentials. That would cause overall spending to increase, certainly. But that doesn’t mean the consumer is strong.
Look at the industries that are at the bottom of that chart. Building materials are down – consumers don’t have enough money to afford a new home or mortgage, because they need to buy groceries. Sporting goods, hobby, toy, music, and bookstores are down – consumers can’t put as much money toward the leisure activities they love. Furniture – who’s re-decorating right now? We pretty much did that during the pandemic while everyone was stuck looking at their outdated belongings 24 hours a day.
Price-sensitive consumers are pulling money away from local, independent stores, because they simply can’t afford to support a small business at a moment when big box stores can dump their prices to corner the market. Spending is up across the economy. Spending is up at places like Walmart. But consumers aren’t spending it on what they want.4 That’s what I read from this month’s data. That, to me, says consumers are hanging on by a thread.
One other thing on this chart. You see that second to last line? That’s my industry.5 The reason everyone I speak to says consumer spending is weak is because we’re in a leisure industry and people are spending their money on non-discretionary goods. There’s a cognitive bias called the availability heuristic. It states that our brains are biased to assume that what is familiar to us is generalized by our brain to be true everywhere. This is why social bubbles are so dangerous. Everyone I speak with says the data should show that spending is down. But spending isn’t down – it’s simply shifted. We believe it should show a negative number because that’s what we see every day at the moment, and we’re all talking to one another.
When all is said and done, you have to simply dig into the data and draw your own conclusion. It tells a story that you can’t get from reading the headlines or even from trusting your own experiences.
Corporations would probably love it if you could take data from a few days, then extrapolate it into your quarterly results – it would save a ton of time and money. I feel like the SEC would have a problem with that, though.
It’s actually even larger than that if you want to get into standard deviations, but let’s just leave it here for simplicity’s sake.
They call me a freak in the [spread]sheets.
Other than Ozempic, apparently.
Where are all my fellow 451120 NAICS people at??