A couple of hours ago, a Cafe patron e-mailed to me the following note (first link added; second link original):
I was intrigued by your quote of the day on Café Hayek. One of the points clearly made is link between productivity and worker wages. However, a recent post at the blog Crooked Timber insinuates that US productivity has climbed by a staggeringly higher amount than wages (even when taking non-pecuniary compensation into account). I found this line of argumentation particularly damning and troubling. How do you respond to critics who point out this inconsistency?
Here’s my reply, with a few small changes added since I sent my reply to my correspondent:
If time permits in the next day or two I might blog more on this matter. But here now are some quick, tentative – I emphasize tentative – thoughts:
(1) If Krugman is correct that what’s going on is a shift of labor income from ordinary to extraordinarily skilled workers, this fact doesn’t necessarily mean that compensation (“wages,” let’s call it for short) isn’t keeping up with productivity growth. It may well be simply that the bulk of productivity growth is at the high-skilled end of the labor market and not the low and ‘average’ skilled end.
(2) If you read Cafe Hayek you know that Russ and I suspect that the CPI overstates inflation and, hence, causes an underestimation of the growth of real wages, even for the median American worker. (Terry Fitzgerald at the Minneapolis Fed, among others, has done some work on this matter. Click here, then on the link to “Where Has All the Income Gone?”)
(3) How accurate are measures of non-wage compensation? Perhaps these measures are off – and increasingly off over the past 30 or so years. In particular, if there are a higher portion of workers today who work as independent contractors out of their homes, these workers have incentives (and a great deal of practical leeway) to overestimate their tax-deductible costs of doing business (not to mention also to underreport their cash incomes). Much of what are in fact consumption expenses, or, alternatively, non-cash compensation for the at-home worker – expenses for items such as time spent driving the new Camry, the cell-phone-and-texting plan from AT&T, the Apple MacBook Pro, the newly refurbished home office – are reported as costs, rather than as compensation, for these independent contractors. I emphasize that I have no idea if the hypothesis in this paragraph is empirically valid or, if so, how significant it is. But it strikes me as being at least worth mentioning.
(4) Let’s assume that it’s true that capital is in fact taking now an historically high portion of national income, with labor getting too little. Capital’s rate of return should reflect this happy fact for capitalists. Does it? Are measured real rates of return to capital today unusually high? Have these rates of return to capital grown over the past few decades? Again, they should have grown if it’s true that total compensation paid to workers has, over the past few decades, failed to keep up with growth in workers’ productivity.
I don’t know the answer to the question I pose here in (4). But let’s suppose now that the answer is that the real rate of return to capital has indeed risen to unusual heights since the mid- or late 1970s. What would this fact signify?….
(5) A sustained, unusually high measured real rate of return to capital might indeed be evidence of some structural problem in the economy – a problem (either designed or spontaneously evolved) that yields disproportionate benefits to capitalists at the expense of laborers. But such unusually high rates of return might also be explained by tax or regulatory policies that artificially discourage investment.
One cannot say “Oh, look! Capitalists are generally now earning real rates of return higher than the historical – ‘normal’ – rates. They must be screwing workers.” Perhaps. But how? Capital isn’t fixed in quantity. If workers’ real pay hasn’t kept pace with the growth in workers’ productivity, and if the rate of return to capitalists of hiring workers has also risen, why aren’t greedy, S.O.B. capitalists investing more in order to capture more of the higher-than-normal profits that are (here by hypothesis) available from hiring these workers? Why aren’t capitalists creating new firms and competing workers away from their current employers? Why isn’t new capital – new firms, larger factories, more labor-intensive means of production – being put into place?
Is the answer to these questions a giant conspiracy among capitalists? Unlikely. A more likely explanation, in my view, is unwise government policies that make such additional investments unattractive despite the relatively low wages that must be paid to labor.
If my “more likely explanation” is, in fact, the correct one, then while the expected risk-adjusted real rates of return to existing capital might or might not in fact (when measured properly to account for all risks) be unusually high, the expected risk-adjusted real rates of return to a great deal of new investment is surely not unusually high – for otherwise more such investment would occur.
Note finally (under this point (5)) that high rates of immigration or amounts of trade with low-wage nations cannot explain any failure of wages to keep pace with increased worker productivity in the U.S. Yes, U.S. employers undoubtedly love the possibility of hiring lower-paid immigrants, or threatening their current American employees with off-shoring whenever these employees press for pay raises. Employers’ exercise of such options might keep real wages lower for a time, but employers’ successful exercise of these options raises the real return to capital. What must be explained is why this (alleged) failure of real wages to keep pace with worker productivity has not called forth sufficient new investment to bid down the returns to capital so that capital’s share of national income is once again at its historical norm.
Again, as regards my long point number (5), that point is written under the assumption that worker productivity has for years now outstripped actual inflation-adjusted total compensation paid to workers. As some of my earlier points indicate, though, I would not be at all surprised to learn that, properly measured, total worker compensation has indeed kept up with growth in worker productivity.
‘dems my thoughts for now.