Breaking down the decline in the U.S. labor share of income
Understanding why the share of income going to labor is on the decline—a phenomenon stretching back several decades—is an increasingly popular area of economic research. There is some debate as to whether that labor share has declined at all, but in so much as there is agreement about the decline, the particular reason for it is very much debated. A new paper enters a different hypothesis into the debate by looking at how the labor share has changed within two large and different sectors of the U.S. economy, and how changes in technology may be the root cause of the shift.
The paper, by Francisco Alvarez-Cuadrado, Ngo Van Long, and Markus Poschke, all of McGill University, focuses on how differences between the manufacturing and service sectors might be responsible for the decline in the labor share in United States from 1960 to 2005. The U.S. economy, of course, underwent several major structural changes over this period of time amid the decline of labor’s share of income—one of the larger changes being the shift of employment from the manufacturing sector to the services sector. But why did that shift that happen?
The simple answer might be that economic activity shifted toward the service sector because this sector features a lower share of income going to labor, which would result in the aggregate labor share declining. But the authors find that the decline is more due to changes within sectorial labor shares rather than changes between the two sectors. In fact, the manufacturing sector started with a higher share of income going to labor in the 1960s, but the subsequent decline was much larger than the decline in the share of income going to labor in the service sector. So what accounts for the declining labor share if the cause is happening within the two sectors?
According to Alvarez-Cuadrado, Van Long, and Poschke the main factor is a difference in productivity between the two sectors. What they find is the importance of productivity growth depends on what specific kind of growth it is. The key factor, according to this analysis, is that the kind of productivity that the three economists call “labor-augmenting,”or the kind of productivity that does more to enhance the role of labor more than capital.
What the paper shows is that labor-augmenting productivity growth has been higher in the manufacturing sector than in the service sector. Workers with more education or skills are an example of labor-augmenting technological change. At the same time, the three economists find that the manufacturing sector more readily switches between the use of labor and capital than the service sector.
In economics speak, this phenomenon is explained by the elasticity of substitution, which is higher in manufacturing than in services industries, though both are lower than 1. This is particularly interesting given that any elasticity lower than 1 usually isn’t in line with a declining labor share.
What’s even more interesting about this new hypothesis is that technological growth is at the heart of the declining labor share of income. But the specifics show that it’s not the popular “robots” story that is often mentioned in these debates, where capital is displacing labor. Rather, in this new paper by Alvarez-Cuadrado, Van Long and Poschke, capital and labor are complements to each other, not substitutes. Their findings indicate that, technological growth has played a role in the declining share of income going to labor, but not in the way most would expect.