Wednesday, February 22, 2017

Prof. Brad DeLong must think that Chinese robots work more cheaply than American ones


 - by New Deal democrat

Profs. Brad DeLong and Jared Bernstein continue to dispute whether the loss of American factory jobs is primarily a matter of efficiency or primarily a matter of offshoring.

Yesterday Prof. DeLong continued to beat the drum for the role of efficiency in the loss of American factory jobs:
[T]he big deal in terms of the changing shape of the American workforce--and, quite plausibly, changing life chances, the collapse of upward mobility, and wage stagnation--is technology: rampant improvement in manufacturing technology coupled with limited demand, for while nearly all of us want one few of us want too and only a minuscule proportion of us want three refrigerators ....
....  Globalization's big effect has been to enable the construction of intercontinental value chains and to create a much finer global division of labor. It has greatly weakened the bargaining power of unskilled manufacturing workers here in the United States, yes. But has it done the same to semi-skilled and skilled manufacturing workers? ....Unskilled manufacturing jobs are not good jobs. Semi-skilled and skilled manufacturing jobs are. I think that odds are at least 50-50 that Larry [Summers] has gotten the sign of the effects of globalization on bargaining power wrong for those manufacturing jobs that are worth keeping.
Note initially by the way that DeLong appears to be addressing why *wages* should have stagnated or worse. But a loss of bargaining power shouldn't necessarily mean fewer *jobs.*

More simply put: there are lots of Chinese and other asian robots that are much more efficient than American robots.  Otherwise why would you need a robot-factory supplier half the way around the world as opposed to the robot-factory around the corner?

The data do not support the claim that there’s been an acceleration in labor-replacing technology displacing US workers. To the contrary, measures of capital investment and especially and most persuasively, productivity growth, have slowed, trends that point in the opposite direction..... . . .  If automation were increasingly displacing workers, we’d be seeing more output produced in fewer labor hours, aka, faster productivity growth.  But we see the opposite.

 I think we are at the point where we can validly say to DeLong:  Who should I believe?  You or my lying eyes?  If DeLong is correct, all of those empty American factories are an illusion.  They are busily humming away, but now they are full of robots rather than humans. (Yes, that is hyperbole, but essentially valid.)

Put another way, if DeLong is correct, and American workers have simply been replaced by American robots, then factories have been made more efficient and thus producers' supply curves should have shifted to the right (I.e., they are willing to produce the same or more at lower cost). Thus they should be supplying more per capita to consumers.

And yet that is not what industrial production, or manufacturing production per capita show:



After great strides in the 1990s, except for one year or so, industrial production has declined. Eight years after the end of the Great Recession, manufacturing production per capita  (red in the graph above) in particular is still down more than 10% from its peak. 

There is simply no getting around that production has left America. Are foreign*robots* also cheaper?!?