Over at Relentlessly Progressive Economics (no-affiliation) Erin Weir has commented on the absurdity of Neil Reynolds* proposition that the key to productivity growth in manufacturing is slashing workers. The absurdity lies in the contradiction between short term gains in productivity driven by labour shedding and work intensification and more medium to long term growth in manufacturing (and ultimately manufacturing productivity) by expanding investment. All Reynolds is really pointing to in the US is that US manufactures have done a good job at pumping and dumping such that productivity looks good for now, but in the future such a strategy will ultimately lead to the furthering hollowing out of US manufacturing.
The FT recently ran an article about the 4-year, quarter-to-quarter tear that profit growth has been on in the US. Interestingly, the article notes that unfortunately investment has been slumping. So Reynolds laudatory celebration of the US model is not only premature it is dangerous.
Over at a Worthwhile Canadian Initiative, Stephen Gordon made a passing comment that the skew of income distribution in favour of capital maybe the result of an increasing pool of retirees drawing on pension income. I find this claim suspect because labour force participation rates have been growing. So the real metric here is the number of employed divided by the number of retirees and not the merely the percent of the population over 60 as his own graph does. Why professor Gordon chooses 60 and not 65 is beyond me, perhaps that was the only readily available comparative series. Whatever the reason, if one is attempting to track the number of retirees over the last say five years the cut off ought to be the average age of retirement and not 60.
But to bring some more clarity to the issue we can look at what has gone on in US manufacturing which speaks both to Reynolds’ and Gordon’s argument. In the graph below I have plotted three different measures of labour income shares of value added (two relative measures for manufacturing).
The first plot (black line with trend), measures the labour income share of value added in manufacturing. Clearly labours’ take of the total output they produce is declining. This is Reynolds’ supposed productivity miracle in action. The next plot (the dashed red line) plots labour income for the total US economy divided by the total value added in the total economy. Here we see only a small decline in labours income share. Notice, however, that both of the time series show an upward and temporary blip in 2001. This makes sense, owners of capital were not able to shed labour fast enough to preserve their overall income shares thus causing a temporary increase in labours relative cut. Also note that our series ends in 2003 just as the massive profit party was beginning in the US, so our series actually overstates labours income share of value added for both manufacturing and the total economy.
The last plot (the green hashed line), plots labour income in manufacturing as a percent of total, economy wide, labour income. This a nice metric because it provides a compound measure of the both the decline in manufacturing employment and the decline in the ability of manufacturing labour to preserve their relative share of the output they produce.
*It must also be said that Reynolds take on the job shedding in China is rather silly. China has been unwinding inefficient state enterprises which has provided much of the supposed loss in manufacturing employment. Similarily, just as shifting workers from low productivity sectors like agriculture to higher productivity sectors like manufacturing increases overall productivity along side, ceteris paribus, declining employment growth despite increasing employment in manufacturing.