I got this graph of US productivity growth from Brad de Long , and it struck me that it’s only after 2000 that there is any real action here. As Brad says, in a normal postwar recession, we would have expected a decline in productivity growth (and maybe even negative growth) arising from labor hoarding – this was the name given to the propensity of employers to keep workers on through economic downturns. Since most employers now engage in large-scale layoffs even during booms, it’s not surprising that labor hoarding is no longer an issue.
Still, given the triumphalist rhetoric that came out of the US throughout the Clinton administration, the productivity growth for this period was remarkably unimpressive. One possible explanation for the contradiction is that the graph shows changes in output per hour worked whereas most attention in the 1990s was focused on output per worker, which was rising with increasing working hours.
In any case, if productivity growth had declined in 2001 as usual, there would have been no story in this picture. For those who attribute economic outcomes to political leadership, the obvious explanation is that Bush has produced an economic miracle.
I don’t believe in miracles, and I also think there’s a problem with Brad’s analysis in which rapid productivity and slack demand produce rising unemployment. US demand for manufactured goods (which is still the most important single part of nonfarm business product) has risen since 2000, but the increase has been met almost entirely by imports. Hence, US manufacturing output has been roughly constant and hours worked have fallen by about 15 per cent. If US productivity was really rising as fast as the graph suggests, there should have been a fair bit of import displacement, especially since the dollar began depreciating a year ago.
Perhaps there are long lags in the process of adjustment to a depreciation (Australian readers of a certain age will recall the endless wait for the “J-curve”).
But I prefer some combination of the explanations I put forward in my post on productivity. In particular, there’s the problem of factor composition. The present recession is unusual because it has been characterised by massive overinvestment. With output growth weak, capital productivity has fallen.
In these circumstances, the best way to assess the underlying productivity trend is to look at multifactor productivity. Unfortunately the data is only published annually and the most recent estimates, from the Bureau of Labor Statistics are for the change from 2000 to 2001. As would be expected from the discussion above they show a rise in labor productivity and a decline in capital productivity. The net impact was that
From 2000 to 2001, multifactor productivity fell 1.0 percent in both the
private business sector and the private nonfarm business sector
. This was the first fall since 1991. I’d expect some recovery in capital productivity to have taken place since then, since investment has been weak, but it seems unlikely that MFP growth for the period since 2000 has been more than marginally positive.
I think we’ll have to put the Bush miracle in the shed with all the others.