Quantifying Deindustrialization

At least at first glance one of the odder aspects of the long-running debate over deindustrialization in America has been the lack of reference to manufacturing value added, especially since time series' for this information have long been readily available, and economists are quite accustomed to adjusting such series' for changes in prices. As it happens, there is an explanation for that, orthodox economists hasten to argue that technological change's idiosyncrasies make such an exercise meaningless. "Why, just look at your cell phone!" they say. "How do we compare that to the products of fifty years ago?"

As with virtually everything else orthodox economists tell us this is a case of drawing a veil of obscurity over obvious facts, one in this case reliant on the gross overstatement of contemporary technological dynamism that is one of their favorite apologia for neoliberalism. After all, for all the hype, what we have tended toward in this era is not the rapid evolution but the stagnation of technology. It may well be that we have products which fifty years ago had no proper analog--but the great bulk of this in one area, information technology, with the cell phone worship confusing a great deal. Yes, you can stream Netflix on your phone, but (even if we set aside the fact that this represents a modification of prior technologies rather than something truly new, or question the actual advance in utility relative to the ways people had of amusing themselves before) it is still a physical product that must reach you through a transport system and have electricity to run, which did not change so much--a truck running on gasoline delivering it to your house, a fossil fuel-powered electrical grid indispensable to your charging its battery, etc.. Indeed, those and other tangibles essential to life--food, shelter, clothing, and the physical infrastructure underlying all of them--have not only seen far slower improvement (at best), but did not become cheaper over that same time frame, and indeed in many cases steadily more expensive relative to the incomes people actually have, not least because their production was never truly revolutionized by computers, the digital age a bust compared not only with Kurzweil-like expectations but the actual accomplishments of "Fordism" from that standpoint.

The result is that the economists' objection to calculating manufacturing value added over time as a source of at least some rough insight into the matter of whether or not deindustrialization is ongoing looks less convincing than the cell phone worshippers have it, and so I endeavored to do just that, taking Bureau of Economic Analysis figures for such value added in the United States over the period 1947-2018 and adjusting that using the Consumer Price Index--and for population growth, because, quite frankly, even a rising output might be outstripped by the more than doubling of the U.S. population that occurred over that time such that the country would be deindustrializing even as totals went up. What I found was that where U.S. manufacturing value added, almost across the board by category as well as in the aggregate, boomed between the late 1940s and mid-1960s, and afterwards continued to grow at a much slower pace until the time of the Volcker shock (1979), after that point the trend was one of consistent decline. Between 1978 and 2018 overall per capita U.S. manufacturing value added fell by about a fifth--and of course the numbers are worse when we cut the "computers and electronics" category that those who dismiss such methods quibble about so tiresomely out of the picture. Without it value added fell by more like 26 percent--the country producing a quarter less per person of everything but computers and electronics by this measure. When we also exclude those two special categories--the military-aerospace-dominated "other transportation equipment" and that particular shale boom beneficiary "petroleum and coal products" we see that the manufacturing value added for everything else fell by 32 percent. And when we look at that cluster of heavy/Fordist durables-producing categories that have long constituted the heart of industrial might--primary and fabricated metals, electrical equipment and appliances, machinery, and automotive goods--things look even worse, value added here falling by almost half per person in the United States over that time frame (48 percent), as there were areas where the situation was worse still (value added in a textiles and apparel sector already ailing before the late 1970s fell harder still, by about 80-90 percent).

No reasonable person can ignore that--especially as all this has gone along with a persistent, massive manufacturing trade deficit and weak manufacturing consumption growth suggestive not of a country booming but recklessly gorging itself on imports, but of foreign imports substituting for a vanishing domestic production. In short, if one takes these figures as at all indicative of the situation then they have to acknowledge a trend toward deindustrialization, with the fact that the biggest U.S. manufactured export as measured by dollar earnings was no longer aircraft or microchips but petroleum, with its whiff of Third World commodity producer, can seem symbolic. But it is, of course, easier to just talk about how you can buy a cell phone now where you couldn't fifty years ago, which is why easy as these calculations are to make one does not see anything like them anymore, as those whose jobs it is to propagandize the public tell them that everything's just fine, as they should be able to tell from their own lives because "You've never had it so good."