GDP and the compensation of employees

Found myself looking at Compensation of Employees at FRED. It has a series name that's easy to remember: COE. "C" is for "compensation" ... "O" is for "of" ...

The everyday view: As Percent of GDP:

Graph #1: Compensation of Employees as a Percent of GDP
COE includes "wages and salaries" and "supplements to wages and salaries"
(which I think means "benefits")

Lots of sharp edges on that blue line.

The first big increase, from First Quarter 1950 to Fourth Quarter 1953, that's gotta be because of the Korean war.

The second big increase, from Q1 1966 to Q1 1970: That would be part of "the Great Inflation".

From that peak, downtrend to mid-'84: That also occurs during the Great Inflation. But it shows a downtrend. Compensation of employees didn't fall in those years. But it fell behind. GDP increased at a faster rate. In other words, after 1970 it wasn't wages driving prices up. Wages were failing to catch up.

It still irks me that the inflation was described as "wage-push".

Another oddity: The blue line goes downhill from 2001 to 2012, except for an upward burp before and during the 2009 recession, the one they call "the Great Recession." 

Plenty of "Great" stuff on this graph: Everything but the economy.

//

That "burp", where employee compensation appears to be going up at a time when the economy is on the brink of collapse, that can't be compensation going up. It's gotta be GDP falling faster than payroll.

Maybe we can see it. I can put compensation of employees and GDP on separate lines, and compare rates of growth:

Graph #2: Percent Change from Year Ago of Compensation of Employees (blue)
and Percent Change from Year Ago of GDP (red)

Yes: Just to the left of the 2010 date, the vertical gray bar marks the "great" recession. Just to the left of that, the blue line is above the red. And just a pixel or two to the left of that, the red line starts above the blue and drops below it. Red is falling faster than blue.

Something similar must happen within the vertical gray bar -- because on Graph #1 the line goes up twice in the gray bar -- but I can't see it on this graph.

Take a moment to look at the whole graph: Isn't it interesting that the two lines vary so much but still run so close together? That's how it struck me. It shouldn't be surprising, though. Okun's law says we should expect that kind of behavior.

//

I can take those two lines and subtract one from the other. Because the lines vary similarly, a lot of the variation will go away.

If I start with GDP growth and subtract compensation growth, the resulting line will be above zero when GDP is growing faster than compensation, and below zero when GDP is growing slower than compensation. So then we will be able to see what happened during the Great Recession when both lines on Graph #2 were falling.

The blue line is the result of the subtraction:

Graph #3: GDP Growth Rate minus Compensation Growth Rate

Willikers! It's jiggier now than before! I said a lot of the jigginess would go away.

Actually, much of the variation did go away. Yes, the line is still crazy up-and-down, probably more than before. But almost all of it is contained between 3 and -3 on the vertical axis. On Graph #2 the lines were mostly contained between 15 and zero or maybe 15 and -5. There was a lot more distance between highs and lows on Graph #2. On #3, a lot of that variation is gone.

Graph #3 shows only the gaps between the red and blue lines of Graph #2: the difference between the lines. The distance between red and blue on #2, that's exactly what Graph #3 shows.

//

So... We were going to look at what happens during and just before the Great Recession.

Just before the gray bar of the Great Recession, there is a sharp "V" as the blue line falls below the -1 level and then bounces back up to zero just around the start of the recession. The line goes below zero because compensation was growing more -- or shrinking less -- than GDP. This corresponds to the spot on Graph #2 where the blue line was above the red, just to the left of the Great Recession.

Now, within the gray bar: On Graph #2 we couldn't see what happened. On Graph #3 we see a very sharp, very narrow drop toward the -2 level in the middle of the gray bar. For that brief time GDP was falling faster than compensation.

The line then shoots up above the 2 level as compensation falls faster than GDP.

Immediately then, another drop toward zero, just at the end of the recession.

In sum: During the recession there were two brief times when GDP growth fell faster than compensation growth, plus the one brief time just before the recession. This accounts for the "burp" on Graph #1 -- three upticks before and during the Great Recession, where it looks like employee compensation was increasing when it should have been falling. It was falling. But so was GDP.

Well, I probably put y'all to sleep by now. Unfortunately, things are just about to get interesting. (But remember, it's me saying this.)

//

Looking at Graph #3 overall, I see a spike after every recession. Reminds me of productivity, "output per hour", where you also get a big spike after every recession.

Being a hobbyist of the economy, and one who learns mostly by looking at graphs that put other people to sleep, I have to wonder if there is any relation between the spikes in employee compensation and the spikes in output per hour. And of course I have to make another graph.

Graph #4: GDP versus Employee Compensation (Difference in
Growth Rates) (blue), and Output-per-Hour (Productivity) (red)

Well that's hard to look at. And there doesn't seem to be much similarity between red and blue. They start out alike, but then the red line drifts down and runs low for most of the graph. Then too, the jiggies seem to match until around 1970, but after 1970 they go off in different directions. At the start red and blue go up and down together. But in the 1990s (for example) red goes up while blue goes down.

Despite all these differences, if you look you'll see a good big spike after every recession, in both the red line and the blue. That similarity is interesting -- and all the more interesting because the red and blue are otherwise so different.

//

Coming out of every recession we have spikes in output per hour and in the growth rate difference, both.

The meaning of the growth rate difference is... GDP growth minus compensation growth... an upward spike means GDP is suddenly growing faster than compensation. In other words, employers are getting a bargain: more nominal output per dollar of labor.

And the meaning of an upward spike in the output-per-hour line? More real output per hour.

So those post-recession spikes mean more output per dollar of labor, and more output per hour of labor. Now, what can we learn from this? 

I have to think about that.