Over at Equitable Growth The estimable Neil Irwin and Tyler Cowen get, I think, things wrong here.
First, Tyler, commenting on Neil:
Tyler Cowen: Facts about non-residential investment: "One simple hypothesis is that it’s not worth spending more on American workers at current wage levels. As workers, while Americans are quite good, they are just not that much better than a variety of high-IQ individuals in cheaper countries, many of whom now have acceptable infrastructure to work with. READ MOAR
Neil Irwin @ The Upshot: "Five years into the economic recovery...
...businesses still aren’t plowing much money into big-ticket investments for the future. Nonresidential fixed investment... still hasn’t bounced back to its pre-crisis share of the economy, let alone made up for lost ground from the record lows of 2009.... If firms increased their spending enough to close that gap, it would mean an extra $220 billion in annual economic activity and perhaps a couple of million more jobs. But there may be even more important and lasting consequences for this lack of spending by businesses. Capital spending improves worker productivity. And worker productivity improves living standards. Less capital spending by businesses means less investment in the kinds of equipment, software and intellectual property that will make the economy more competitive over the long haul.
But when I look at the time series I see, for the nominal/nominal equipment spending share of potential GDP:
And for the real/real share:
For the real/real measure, neither Tyler nor Neil's point is a thing: as far as equipping U.S. workers with productive tools, understood as equipment and software, even in the depressed economy of today the U. S. is, even given the enormous amount of slack in the economy, doing a better job than it did save in the decade or so that started in 2007. In longer-run perspective this is not a thing.
For the nominal/nominal measure--which assumes that the appropriate reaction to the secular decline in equipment and software prices relative to GDP in general is for companies to boost their expenditure in order to keep the nominal investment share of potential GDP the same--we still seem to be ahead of the average pace of the 1950s, 1960s, and early and mid 1970s. Whether there is a shortfall depends on whether you think that there is little economic slack left--in which case equipment and software investment is lagging substantially behind where we would expect it to be--or whether there is still an enormous amount of economic slack. If, as I think, there is still more economic slack in the economy than there was at the business-cycle troughs of 1992 and 2001, equipment investment is not doing too bad. And when you reflect on the fact that back in 1992 and in 2002 virtually everyone expected much stronger demand growth than we expect now it seems entirely reasonable to see the nominal/nominal share of equipment investment and software in potential GDP as not lower but higher than reasonably expected.
What do I think? I think that if you want to look for an explanation of why the U.S. economy is currently depressed, then look at depressed residential investment and government-spending austerity--do not look at equipment investment to infer businesses hobbled by regulations and more pessimistic about the future than the state of present and expected future demand warrants. Similarly, I think that if you want to look for indicators that U.S. workers are overpaid (or that the value of the dollar is too high) do not look at equipment investment.