Wage Inflation Isn’t Coming. It’s Already Here.


George Soros has a great line that we’ve used a lot in Epsilon Theory notes. When he was asked how he could possibly have predicted what would happen when he famously “broke the Bank of England” in 1992, he replied in that growly Hungarian voice: I’m not predicting. I’m observing.

It’s the perfect catchphrase for the modern, narrative-savvy investor. It’s the perfect catchphrase for the Three-Body Problem market, where there is no algorithm for predicting markets (no closed-form solution, in the lingo), but only tools for calculating markets. Where there is no Answer for successful investing. But there is a Process.

I’m not predicting. I’m observing.

The Three-Body Problem

It’s a hard concept to wrap your head around, this difference between calculating the future and predicting the future, but it will change the way you see the world.

And your place in it.  

Here’s what I am observing:

Over the past four quarters, the United States has generated more wage inflation than at any point over the past 40 years.

Seeing is believing, so here’s the chart of average weekly earnings (weekly earnings, not hourly!) for Americans in private sector jobs from 1963 through today, measured quarterly.

Bloomberg: US Average Weekly Earnings SA, Quarterly Jan 1, 1963 – Mar 31, 2021

And here’s the same data on a monthly basis over the past 14 years:

Bloomberg: US Average Weekly Earnings SA, Monthly Jan, 2007 – April, 2021

These are the facts. These are not predictions. This is what has already occurred.

  • Q1 2021 wages were 7.7% higher than Q1 2020 wages.
  • Q4 2020 wages were 7.7% higher than Q4 2019 wages.
  • Q3 2020 wages were 6.2% higher than Q3 2019 wages.
  • Q2 2020 wages were 6.5% higher than Q2 2019 wages.

These are also facts. I am not making this up.

  • Over the past 10 years prior to the past 4 quarters, the highest single quarterly year-over-year wage growth was 3.6% in Q4 2018.
  • Over the past 20 years prior to the past 4 quarters, the highest single quarterly year-over-year wage growth was 4.5% in Q4 2006.
  • Over the past 30 years prior to the past 4 quarters, the highest quarterly year-over-year wage growth was 4.8% in Q4 1997.

You have to go back 40 years – to Q3 1981 – to find a higher quarterly year-over-year wage growth number (+8.5%).

This is not an anomaly. This is not a single quarter aberration. This is not transitory.

This is four straight quarters of the highest wage growth numbers in 40 years.

For those keeping score at home, the US inflation rate in 1981 was 10.3%.

Now I know what you’re thinking. You don’t believe me. Surely, you say, if this were true we would have heard some mention of this not-in-forty-years wage growth phenomenon. Surely, you say, someone involved in the creating or proselytizing or questioning the Fed’s dominant “transitory inflation” narrative would have mentioned this little nugget. I mean, it seems … relevant.

I think I know why you’ve heard nothing about this. Also, don’t call me Shirley.

The reason no one recognizes that remarkable wage inflation has already occurred is largely because of the intentional cartoonification of unemployment and wage data.

I’m using the word ‘cartoon’ in its technical sense here, as an abstraction of an abstraction. I’ve written about these cartoons of macroeconomic data in service to political ends quite a bit (in fact, earlier this week we published a nice short note on the cartoon that is CPI), but here’s the note that discusses the cartoon of average hourly wage earnings in great detail:

The Icarus Moment

We live in a Cartoon Age, an era not of alienation per Karl Marx, but of alienation per Groucho Marx.

What’s the cause, what’s the future, and what do we do about all this? 

And here’s an extended money quote from that note:

In the beginning, there was a desire to model the employment patterns of the U.S. economy to help policymakers figure out what was actually going on. So in 1884 (!) Congress established the Bureau of Labor Statistics (BLS) to do some counting and abstracting, and since 1915 (!) the BLS has been surveying employers to estimate how many Americans are working and how much they’re being paid. On the first Friday of every month, the BLS releases its report on the real-world employment patterns in the U.S. for the prior month. This data is an abstraction, to be sure, full of seasonal adjustments and model estimations, but it is a first level abstraction. This is not the cartoon.

One of the standard calculations that the BLS reports is the percentage change on a year-over-year basis in how much workers are being paid. Usually this wage growth report takes a backseat to the more famous “jobs report” of how many jobs were added or subtracted from the U.S. economy in the prior month and the even more famous “unemployment report” (which is actually based on an entirely different survey) of the percentage of Americans who were actively looking for work but were unable to find jobs. But when everyone and his cousin is either worried about wage inflation or hoping for wage increases, then the wage growth “number” takes on enormous importance. It’s the depiction and the narrative around the BLS wage growth calculation that is the cartoon. And that cartoon is everything for markets today.

The most basic way to look at wages for a monthly report would be to count up how much all workers got paid in that prior month. But that doesn’t work for a month-to-month comparison because different months have meaningfully different numbers of days. Unless you’re getting paid on a monthly or twice-monthly basis, then you’re going to be making less in February than you are in January. So the BLS uses the work week as their basic apples-to-apples comparison basis.

As far back as I can trace the theater of BLS reports — and that’s how one should think about these market data reports, as theatrical productions consciously designed to impact behavior — the “number” that’s reported isn’t the apples-to-apples comparison of weekly wages. Instead, it’s hourly wages. Why? Because back in 1915 this is how most people got paid. The abstracted idea of hourly wages connects with people more than the abstracted idea of weekly wages. It’s a more effective tool for eliciting a behavioral response, so that’s why our theatrical effort focuses on it every month.

But here’s the problem with the hourly wage abstraction. It requires introducing a new data estimation into the mix, one that has nothing (or at least very little) to do with the real-world concept we’re trying to represent, which is whether you’re taking home more money today than you did last year. That additional layer of abstraction is the average length of the work week.

Now this data estimation changes very little from month to month. Unlike the difference in work days from month to month, which can be meaningful and is incredibly easy to measure, the difference in work hours from week to week is an immaterial and almost certainly statistically spurious estimation. Here are the average number of hours in the work week since 2012.

Since 2012, the average length of the work week has been as low as 34.3 hours and as high as 34.6 hours. For more than SIX YEARS, the maximum deviation from the mean has been less than NINE MINUTES, less than ONE-HALF OF ONE PERCENT of the total work week. This is the flattest line you will ever see in any time series, and any month-to-month deviation from the mean is almost certainly a spurious statistical estimation. Meaning that the month-to-month differences in the average work week are so far inside your margin of error for this sampling and estimation process that you can have ZERO confidence that you are abstracting anything real. This is as bogus of an abstraction as you will ever see.

And yet it makes all the difference in the world for hourly wage calculations!

Why was the February wage growth number reported on March 9th as 2.6% rather than 2.9%?

Because the average work week in February 2018 was randomly estimated as being six minutes longer than it was a year ago.

Everything you read about what the March 9th wage growth number meant for your portfolio — the entire Goldilocks narrative of a “contained” wage inflation number combined with strong job growth — is based on a statistically spurious result. Everything. It’s all made up. None of it is real.

And yet, on the basis of the Goldilocks narrative, which was the all-day headline of the Wall Street Journal and the talking point of every Missionary on CNBC that Friday, the S&P 500 was up more than 1.7% on the day. That’s $415 BILLION of market wealth created in the S&P 500 alone, in one day, from a cartoon representation of annualized wage growth in the U.S. economy.

I wrote that in 2018. Here’s a chart of what’s happened since then to that average hourly work week that changes weekly earnings to hourly earnings:

Bloomberg: US Average Weekly Hours All Employees Private Sector SA, Monthly Jan, 2006 – April, 2021

You see what’s happening? We are now at an all-time high of estimated average weekly hours worked, which artificially depresses the average hourly earnings cartoon. If you just make your percentage comparisons off hourly earnings data, wage inflation doesn’t look nearly as bad. It’s still quite noticeable, but seems more of an anomaly, more of something that is “transitory”.

Bloomberg: US Average Hourly Earnings All Employees Private Sector SA, Y-o-Y %, Monthly Jan, 2007 – April, 2021

Again, there is absolutely no fundamental reason to report an hourly earnings number instead of a weekly earnings number. The BLS itself calculates the weekly number as their primary dataset to see what is truly happening with wages, and only converts to hourly wages because THAT WAS POLITICALLY ADVANTAGEOUS BACK IN FREAKIN’ 1915.

The investment question you hear constantly today is whether or not supply-driven inflation will eventually make its way into wage and price inflation. This is the wrong question. Or rather, it was the right question to ask a year ago, but now it’s been answered.

Wage and price inflation aren’t coming. They’re already here.

The right question to ask today is how bad this wage and price inflation cycle will be. I think it’s gonna be pretty bad, in large part because it’s not yet common knowledge. It’s not yet what everyone knows that everyone knows. It’s not yet contemplated as a potential outcome by our omnipotent market missionary, the Federal Reserve, who remains trapped – not by policy but by narrative – in its insistence that this cannot possibly be the start of a wage-price inflation cycle.

Inflation is transitory” is the new “subprime is contained”.

Do I think we will continue to see wage inflation running at 7% year-over-year? Not really. I dunno. I really don’t. I’m not here to predict. I mean, these things are always overdetermined, and if you want to tell me that last spring’s wage increases were a constructed illusion based on lots of low-wage workers getting booted and higher-wage workers staying on the job, I can’t say that you’re wrong. But I can tell you that month-over-month wage increases THIS spring are running at more than 10% annualized.

More importantly, I can tell you that it doesn’t matter.

Are all of these government wage and price reports constructed artifacts of a whole host of nudging and nudgeable factors? YES. That is my point! They are all cartoons. Intentionally so. Why? Because cartoons work. We are biologically hardwired and socially trained to respond to these cartoons, both as employers AND as workers. The cartoons work in both directions, to encourage deflationary expectations AND inflationary expectations.

When I observe the narrative coming out of last Friday’s jobs reports, I see employers coming to grips with the fact that they need to lift wages even more to satisfy their labor needs in a reopening economy. I see a new ballgame when it comes to wages and prices. A new ballgame that we haven’t played in forty years. A new ballgame where we are only in the first inning.

I’m not predicting. I’m observing.


To learn more about Epsilon Theory and be notified when we release new content sign up here. You’ll receive an email every week and your information will never be shared with anyone else.

Notify of
Inline Feedbacks
View all comments
1 month ago

The interesting piece here is that in March-May 2020, the case was made (rightly so) that wages bounced because a huge chunk of lower-paid workers got axed in retail, restaurants, travel, entertainment, and other services industries. Thus, the labor force in the spring of last year was left with a bunch of workers in higher-income positions, so the jump in wages that we saw at that point could be explained by showing 30M people out of work in low wage jobs, and so average wages slid up even though it wasn’t like everyone suddenly saw their incomes jump because of COVID. Not at least with how BLS calculates it for this report. But if we’re looking at it now, the question for those businesses in those sectors that are trying to reopen and trying to rehire is that this was not a seasonal closing where you go back and find the workers from last year and bump them from $11 and hour to $11.25 an hour. This was the biggest one-year labor disruption in history. A ton of people moved. A ton of people retired. A ton of people changed industries because they saw friends getting paid more in other businesses. A ton of people are still dealing with kids remote learning with no childcare. Businesses aren’t simply hiring back the same employees at a slight raise now. They have to compete with the higher wages that some of their workers have gone out and taken in the last year.… Read more »

Guy Larri
30 days ago
Reply to  Canid

The move to $15 an hour sticks, but does it create a trend where Sally demands and gets paid $17 next yet, and $20 the year after? If it stays at $15 you don’t have an inflationary spiral, you have a one off wage increase, so the wage inflation rate (the first derivative) goes back to zero. If it creates an inflationary trend then where Sally gets a series of pay rises then you have wage inflation that is game changing, but I have yet to see convincing arguments for why Sally gets a series of pay rises.

Maybe demographics might be one driver for a series of pay rises, as ratio of the number of retired people to the number of working people keeps going up.

1 month ago

I will echo a few of the points that @Canid brought up. Initially average wages went up solely because lower earning employees were laid off. Now with those employees being rehired, average wages should go down. I would have to take a look at the actual employment numbers and hours worked to see if you wage data has captured this properly. I am not disputing your narrative I just want to make certain that you have captured the above mentioned facts.

Now I will tie in John Hussman (mentioned in the ET Forum). One of Hussman’s favorite valuation measures which he uses to predict 12 year forward earnings in the stock market is what he calls MAPE (Margin Adjusted Price Earnings Ratio). He has long stated that businesses have record high margins partly due to low earnings by workers and that this would change and his MAPE calculations call for no or negative twelve year returns partly as a result of this. So obviously higher wages will result in lower earnings and potentially lower equity prices.

I am not looking forward to another bout of inflation. Dealing with inflation usually takes ones effort off of more productive endeavors.

1 month ago
Reply to  PreCambrian

All of which suggests a very nice level of heterogeneity may develop in stocks as companies with pricing power (the way we used to invest back in the inflation day) are treated differently over companies with no ability to move prices.

James Wire
1 month ago

Risking some personal inflation, your observations are worth every penny nickel dime of the subscription price. Not that it solves anything, I just wonder how many people spreading these narratives are complicit and how many are unknowing mouthpieces repeating the spin.

1 month ago

So in terms of Rusty’s Things We Need to Be True, a change in Common Knowledge would take out at least his top3:

  1. We live in a deflationary world.
  2. We live in a flat world.
  3. We live in a world in which the Fed has your back.
  1. By definition
  2. MCD and CMG can’t prepare and serve food in local markets over Zoom from India.
  3. To Ben’s point about the Fed being trapped in their own narrative, if they admit it’s not transitory, they might have to do something about it.


1 month ago

The Fed just reported that it’s balance sheet ballooned another $32 Billion last week. That’s a 23% annualized increase over 3 months and 89% growth rate from 1 year ago.

Insane rate of increases in an environment already of rising inflation .
How they can justify printing massive quantities of money is beyond me.

1 month ago
Reply to  Peter

You should read https://mishtalk.com/economics/the-psychology-of-qe-is-far-more-important-than-the-amount-of-it including the John Hussman article linked inside. The John Hussman article is also linked in the ET Forum article “Hussman Letter”. The Fed balance sheet is not inflationary because the money goes nowhere but sits on the books in banks. It does have an effect on interest rates however. The “helicopter” money (deficit spending) is another matter especially if it doesn’t get paid back. That money goes into people’s pockets for them to spend.

Guy Larri
30 days ago
Reply to  PreCambrian

Some of the money people spend gets taxed when it ends up in someone else’s salary, so some of it goes back to the Federal Government (as well as to local and state governments).

Shane Metzke
1 month ago

I am trying to better understand Ben’s points above. I get that hourly wage growth is being mathematically depressed by longer working hours. But I assume this is only for those that are actually employed?

Can we assume that there IS still slack in the US labour force that WILL come back once all the stimulus cheques and other programmes stop or shrink? Can I assume this is the case when I look at the underemployment rate (at 10.4%)?

I don’t know the answer to this but it’s important to understand this in the context of whether or not this is “transitory”.

Lawrence Pusateri
1 month ago

The Fed and the government remind me of an alcoholic that is hitting bottom. The lies that they have used to build their house of cards is starting to get wobbly. They must know tell more ridiculous lies to cover for the lies previously told.

Lawrence Pusateri
1 month ago

Now not know —we need an edit ✍️ button.

1 month ago

I like the fact that Ben used The Great Wave off Kanagawa as the image for this one. Very appropriate in that it’s a gigantic rogue wave that’s about to crash into us. Terrifying. I don’t know if it’s been coined yet, but “hyper-stagflation” is a phrase that I have not heard a lot about… yet.

The Latest From Epsilon Theory


This commentary is being provided to you as general information only and should not be taken as investment advice. The opinions expressed in these materials represent the personal views of the author(s). It is not investment research or a research recommendation, as it does not constitute substantive research or analysis. Any action that you take as a result of information contained in this document is ultimately your responsibility. Epsilon Theory will not accept liability for any loss or damage, including without limitation to any loss of profit, which may arise directly or indirectly from use of or reliance on such information. Consult your investment advisor before making any investment decisions. It must be noted, that no one can accurately predict the future of the market with certainty or guarantee future investment performance. Past performance is not a guarantee of future results.

Statements in this communication are forward-looking statements. The forward-looking statements and other views expressed herein are as of the date of this publication. Actual future results or occurrences may differ significantly from those anticipated in any forward-looking statements, and there is no guarantee that any predictions will come to pass. The views expressed herein are subject to change at any time, due to numerous market and other factors. Epsilon Theory disclaims any obligation to update publicly or revise any forward-looking statements or views expressed herein. This information is neither an offer to sell nor a solicitation of any offer to buy any securities. This commentary has been prepared without regard to the individual financial circumstances and objectives of persons who receive it. Epsilon Theory recommends that investors independently evaluate particular investments and strategies, and encourages investors to seek the advice of a financial advisor. The appropriateness of a particular investment or strategy will depend on an investor’s individual circumstances and objectives.