Labor’s Share, Part II

Following up on my post from the other day, I’m still thinking about the recent divergence between labor productivity and the income that labor gets. In the process of thinking about this, I’ve been looking at some more data.

The table below shows labor productivity and labor compensation in several countries compared to levels in the US. In each case, productivity and compensation is measured relative to the US level, so a productivity score of 50 would mean that workers in that country produce 50% of what US workers produce in an hour, and so forth.

Source: UNCTAD, Trade and Development Report, 2002.

Next, here’s a graph showing the relationship between productivity and compensation within the US since 1985.

Source: BLS.

Both of these point to the same conclusion: typically, the income that labor receives follows the productivity of labor. There are some exceptions, of course. Internationally, the relationship between productivity and compensation is true in a rough and general way, though with some variation from country to country. And over time in the US, there have been periods in recent history when either compensation or productivity has grown faster than the other. But in general, when productivity rises, so does labor income.

The last 3 years seem to be one of those times when compensation does not track productivity, however. As the graph shows, a relatively large gap has opened up between worker productivity and compensation. This is another reflection of labor’s shrinking share of national income.

Numerous good possible explanations were offered in the comments on my earlier post on the subject. Let me address about a couple of them. First of all, international trade seems an unlikely culprit here. Imports grew most rapidly during the period 1996-2000, which corresponds exactly with the period with fastest compensation growth in the US. On the other hand, imports into the US actually shrank in 2001 and grew very slowly in 2002, when wages diverged from productivity.

What about union power? I agree that, in theory, weaker unions should cause workers to receive a lower share of a firm’s rents. In this case, however, I think that there’s something else going on. Unions have been steadily losing power for the past two decades. I would therefore have expected workers to receive a steadily decreasing share of national income since the early 80s, not just something that has happened over the past 2 or 3 years.

Finally, what about this just reflecting the normal workings of the business cycle? There’s probably something to that — we do typically see a big increase in profits at this stage of the business cycle, and in part that serves as a signal to investors that it’s time to start expanding businesses and building new businesses to take advantage of new profit opportunities. However, the fact remains that labor’s share in national income has fallen more dramatically and to lower levels than anytime in the last 30 years. It’s plausible to think that such extreme movements in labor income may be due to more than just the recovery from what was otherwise an unusually mild recession.

So I come back to my hypothesis. If anti-trust oversight has diminished in recent years (the change in Microsoft’s fate at the hands of the Justice Department provides one example of what I’m talking about) then firms have increased their ability to keep out competition. Industries move closer to oligopoly. Output is restricted by the oligopoly, leading to lower output and employment than would otherwise be the case, with stronger competition. The labor market is therefore weaker. Workers lose, and the owners of corporations gain. I’m still persuaded by this story.

Kash