The declining labor share of overall national income is one of the most striking trends of our time. As recently as 2007 or so, I’d have been comfortable seeing this as overwhelmingly a business cycle issue, but it more and more looks like a structural shift in which both peaks and troughs are lower than they used to be. And one obvious culprit is mechanization and computerization. Traditionally, machines and labor have been complementary goods all things considered (and indeed in Uganda capital deepening raises wages which is why unconditional cash transfers boost growth) but that’s not a law of nature and today’s sophisticated computers might be substitutes for labor instead.
But not so says Tali Kristal in the American Sociological Review, who argues that computerization stories don’t fit the sector data. She says that if you look at where the decline happened, it was in sectors with a large labor union presence:
“It was highly unionized industries — construction, manufacturing, and transportation — that saw a large decline in labor’s share of income,” Kristal said. “By contrast, in the lightly unionized industries of trade, finance, and services, workers’ share stayed relatively constant or even increased. So, what we have is a large decrease in labor’s share of income and a significant increase in capitalists’ share in industries where unionization declined, and hardly any change in industries where unions never had much of a presence. This suggests that waning unionization, which led to the erosion of rank-and file workers’ bargaining power, was the main force behind the decline in labor’s share of national income.”
I’m not totally convinced by that causal inference. A structural power shift against rank-and-file workers could cause a reduction in unionization rates as well as vice versa. To be fully persuasive what you’d want to do is look at the international picture. My casual impression is that in Germany, for example, the politics are different so the lack of labor clout in manufacturing has led unions to agree to restrain wage demands without leading to a decline in the unionization level of German manufacturing. That’s just my offhand speculation, not real research, but I think that’s where you’d want to look for evidence that union decline caused the lack of bargaining power rather than vice versa.