The WSJ assures us that productivity is holding up well as the economy weakens. Perhaps, the last few quarters haven't been bad, but the record since the second quarter of 2004 is pretty dismal. Productivity growth has averaged less than 1.7 percent annually. This is only a bit higher than the 1.5 percent annual rate during the long slowdown from 1973-1995. The article includes some bizarre assertions. It reports that the shift away from construction due to the housing crash would raise productivity. Actually, construction is a relatively high productivity sector so the shift away from construction would lower productivity, other things equal. The article also reports that the improvement in the trade balance should raise productivity because exporting sectors have higher productivity on average than industries that don't export. Actually the main improvement on the trade balance has been on the import side, as imports have fallen sharply. The replacement of imports with domestically produced goods might be expected to lower productivity, since the domestically produced goods had been more expensive than imported goods before the drop in the dollar. As a more general rule the higher prices for imported goods in general is likely to be a major drag on growth since firms will be likely to substitute labor for imports that are now more expensive. Just as the rise in the dollar might have provided a temporary boost to productivity growth, the decline in the dollar might be expected to be a drag on productivity growth. There are two other important measurement issues that should be noted in an article like this one. There has almost certainly been a much sharper decline in employment in the construction industry than the data show because many undocumented workers in this industry never showed up in the data. This means that productivity growth was likely overstated in the boom and is being understated now. The other factor to keep in mind is that these data are subject to large revisions, especially in periods near recessions. It is likely that the final data will show a very different pattern in productivity than our current data now indicate.
--Dean Baker
