Employee Compensation

On the Think Markets Blog, I was looking at a comment by Pietro regarding the BLS series on Average Hourly Earnings. Pietro argued that the increases in Average Hourly Earnings during the Great Recession could be caused by greater reductions in the employment of lower wage workers than higher wage workers. It is possible that the remaining employed workers all took pay cuts, but if enough workers whose pay was lower than average were laid off, then the average hourly wage could rise.

Of course, most of the decrease in total employment is due to a sharp drop in the number of new jobs being created. But this just reinforces Pietro's point. Suppose the new jobs that aren't being created would be low wage, entry-level jobs. That these jobs have not been created tends to raise average wages, even if old workers receive no raises or even pay cuts.

If layoffs and quits and the jobs that haven't been created were representative of the jobs that exist, then changes in the average would give a more accurate estimate of how the wages of particular workers have changed. Are those continuing to work receiving more than before? Or are they receiving the same or even less? And, of course, what are they really receiving, corrected for inflation?

While interesting questions, the monetary disequilibrium approach focuses on how changes in nominal expenditures , in the face of sticky prices, results in lower real expenditures, lower real expenditures result in lower production, and lower production results in lower employment.

Of course, if surpluses of products and labor result in lower prices and wages, then the decrease in nominal expenditure can occur with a smaller, or perhaps no decrease in real expenditure. The possibility that there is no decrease in real expenditure, or that any decrease in real expenditure simply reflects a decrease in productive capacity, reflects the scenario where prices and wages are perfectly flexible.

Scott Sumner has sometimes argued that if nominal expenditure falls 10 percent below trend, all wages and salaries must immediately fall 10 percent below trend, or employment must suffer. Rather than thinking about individual wages and salaries all dropping, it seems plausible that a decrease in final sales of domestic product would need to be matched by a proportional drop in compensation of labor, if employment it to be maintained.

So rather than look at average hourly earnings, I began to look at total compensation of employees, in particular, the series National Income: Compensation to Employees, Paid.

In the fourth quarter of 2009, Employee Compensation was $7882 billion, which was an increase of about .5 percent from the previous quarter. Still it was 2.3 percent below its peak value of $8,069 billion in the third quarter of 2008. Perhaps it should be no surprise that when Final Sales of Domestic Product began falling, so did Employee compensation. The dollar volume of sales fell, and so did the amount spent on labor.

The trend growth rate of Employee compensation between the beginning of the Great Moderation in 1984 and its end in the third quarter of 2008, was approximately 5.3 percent. During the fourth quarter of 2008 it fell at at a 1.9 percent annual rate. In the first quarter of 2009, it fell at a 10.3 percent annual rate. After a slight additional decrease in the second quarter of 2009, it began to grow, but at a rate well below trend.

The gap between employee compensation and the trend growth path in the fourth quarter of 2009 was 13.9 percent.

This is substantially greater than the 10.3 percent gap between Final Sales of Domestic Product and its trend growth path. This difference is consistent with Final Sales of Domestic Product falling approximately 2 percent from its peak--roughly 2/3 of the decrease in Employee Compensation.

Employment is currently 5.5 percent below its peak in the first quarter of 2008, and 9.27 percent below its trend from the Great Moderation.

What has happened to the compensation per employee?

While compensation per worker is currently $57, 059, an all time high, it did fall between the fourth quarter of 2008, $55,795, to $55,271 in the first quarter of 2009, a drop of nearly one percent. It had risen again by the second quarter.

Compensation per worker remains 4.6 percent below the trend of the Great Moderation.

While compensation paid to workers is below its trend, the decrease is a bit less than half of the decrease in total final sales. As Sumner argues, employment suffered.

What about real wages?

After making an adjustment for changes in the GDP deflator, real compensation peaked in the first quarter of 2008, at $7452 billion in 2005 prices. It current value of $7169.8 billion, also in 2005 prices, has fallen 3.8 percent from the peak.

Real Compensation of Employees is currently 12.3 percent below trend.

The trend growth rate of real compensation during the Great Moderation was approximately three percent. The annual rate of decrease in the first quarter of 2009 was 12.2 percent. While it has grown in the third and fourth quarter, the growth rates remain below trend. Real employee compensation is not closing the gap towards its long term trend.

What about the average? Real compensation per employee fell below its long term trend in the second quarter of 2005. It's "peak" was in the fourth quarter of 2008, but this was after remaining nearly unchanged from the beginning of the recession. It then dropped 1.4 percent, (a 5.6 percent annual rate) in the first quarter of 2009. And then it began several quarters of increase, so that it is now 1.8 percent higher than at the beginning of the recession. However, it remains 3 percent below its long term trend.

The trend growth rate of real compensation per worker during the Great Moderation was 1.6 percent. The sharp drop can be seen, as well as the rapid recovery.

The rapid increase in real compensation per employee might suggest an increase in real wages, but there remains the problem that employment has continued to fall during that period. Perhaps the decrease in employment was of workers with less than average compensation.

Another factor is that compensation per worker isn't the same thing as compensation per unit of labor. The index of aggregate hours for the business sector shows a decrease of nearly 8 percent. This is 19 percent greater than the loss in employment in the business sector during the period. (Total employment is about 5 percent below its peak, but business sector employment has fallen 7 percent from its peak.) Real compensation per worker understates real compensation per hour worked.