Has hourly self-employment income stayed relatively constant to hourly payroll income?

The Brookings Institute has just come out with a new paper seeking to explain the decline in labor’s share of national income. The paper is titled, The Decline of the U.S. Labor Share. It was written by Michael W. L. Elsby, Bart Hobijn, and Ayşegül Şahin. Here is a video of Justin Wolfers explaining the basics of the paper.

The Brookings Institute paper says that the decline in labor share is over-stated because self-employed people are earning less in labor income. And the decline in self-employment income explains part of the decline in labor share.

They say that one third of the decline in labor share can be accounted for by a decrease in self-employment income. The idea is part of understanding what part of self-employment income is labor income and what part is capital income… but I question their basic assumption.

In the paper, they make an assumption.

“Self-employment income in this headline measure is imputed under the assumption that average hourly compensation for the self-employed is the same as for those on payroll. That is, hourly self-employment income = hourly payroll income for all quarters in the postwar period.”

They then say that the hours of self-employed people as a share of total labor hours (self-employment hours + payroll hours) have decreased from 14% in 1948 to 8.5% in 2012. When you multiply hours by hourly wage, you must then conclude that self-employed people are earning less as a share of national income over time. Thus they conclude that the labor share of self-employed people has declined over the years and accounts for one third of the total decline in labor share.

It all rests on the assumption that the relative ratio between hourly compensation between self-employed people and those on payroll has not changed over the years.

Can that really be a true assumption when productivity from employee work has grown much faster than their hourly compensation? Hasn’t the increased income from increased hourly productivity gone to the “owners” of labor’s work? The self-employed still hire employees who are more productive but not receiving compensation equal to their increased productivity. Aren’t the self-employed reaping increasing benefits from their more productive employees?

It would seem to follow that employers, even if they are self-employed, are getting more income from increased productivity of their workers.

And if the self-employed are working less hours, wouldn’t it might be because they have more leisure time now from increased pay per hour? Are they earning more per hour and now find they can afford more leisure time off?

I think of a cleaning cooperative in Humboldt County, California, called Restif Cleaning Services. The company is 100% employee owned. I talked with them by phone. They tell me that the employee owners receive on average $20 to $24 per hour of work. Now all of their competitors employ workers at a wage rate from $8 to $12 per hour. The self-employed owners at the other companies must be making quite a profit for themselves, especially if we assume those owners are working less hours.

I know a lady who cleans houses. She charges $45 per hour. And then pays workers $10 to $12 per hour to do the work. Needless to say, she does not like doing the work herself because she makes much less per hour. If she does a job of 5 hours by herself, she makes $225 at $45 per hour. If she has a worker do 3 of those hours, and she works 2 hours, she makes $200 ($225 – $25), at a rate of $100 per hour of work. In this case, she is making 8x the hourly wage of her employee. She says it is hard to find good dependable employees. And she is a nice religious person who wants to do missionary work in Africa to help poor children.

What about employers that are not so nice?

Side thought: I think about for a moment how supply-side schools teach that lower wages would shift the LRAS curve to the right. My research into effective demand says that lower labor share would shift the LRAS curve to the left.

We can see that the cleaning lady has an incentive to work less hours, but only because she progressively receives more per hour than the employee on payroll. Thus, the more she can get workers to do the work, the more she makes per hour. Thus, we have an explanation of why self-employed people are working comparatively less hours. It is not because they earn the same or relatively the same over time as those on payroll. It is because they are increasingly earning more per hour than those on payroll. This refutes the apparently casual assumption in the Brookings Institute’s paper.

One of the keys to saying that self-employed people are earning relatively much more than employed people is based on employee productivity gains over the decades.

If I then was to assume that self-employed hourly income had risen against payroll hourly income, I would not have reached the same conclusion as the Brookings Institute paper. I would not have said that a third of the decline in labor share is attributed to a relative decrease in self-employment labor income. It would then follow that a decline in labor’s share of income is more a result of payroll income declining than the Brookings Institute would have us believe.

However, on page 13 of the paper, it is noted…

“Moreover, rises in compensation at the very top of the distribution of proprietors have been even more extreme than among employees, suggesting that the average hourly compensation of the self-employed has soared in recent decades relative to the payroll employed, violating a key assumption underlying the headline measure.”

So they recognize a problem with the assumption. I suggest they not be so sure of saying that a third of the decline is due to lower self-employment income.