By Lambert Strether of Corrente
As readers may know, I was a model railroader when young, and I still buy Model Railroader. The picture of America I have is very much like the sort of town that one might model in the steam-to-diesel era, and much like the town I grew up in. My model town, besides the main line and the yard, would have spurs for a lot of small industries, like the local brewery, or the local lumberyard, or the local meatpacking plant. All that is now gone; today one would model giant unit trains of coal, or miles-long trains of containers cars loaded with consumer goods, and they’d roll through the town without stopping. Gone from my model town too, after deindustrialization, would be two- or-three story brick factories: machine shops, machine tool manufacturers, and foundries. The manufacturing base.
If I lived in the past, I might assume that re-industrialization would be as easy as building a new plant and plopping it down in my model town; “build it and they will come.” But this America is not that America. Things aren’t that frictionless. They are not, because of a concept that comes with the seventy five-cent word hysteresis attached, covered here in 2015. Martin Wolf wrote:
“Hysteresis” — the impact of past experience on subsequent performance — is very powerful. Possible causes of hysteresis include: the effect of prolonged joblessness on employability; slowdowns in investment; declines in the capacity of the financial sector to support innovation; and a pervasive loss of animal spirits.
(To “loss of animal spirits” in the entrepreneurial classes we might add “deaths of despair” in the working class.) And if there were a lot of people like me, living in the past — in a world of illusion — that too would would cause hysteresis, because we would make good choices, whether for individual careers, at the investment level, or at the policy level, only at random.
Our current discourse on a manufacturing renaissance is marked by a failure to take hysteresis into account. First, I’m select some representative voices from the discourse. Then, I will present a bracing article from Industry Week, “Is US Manufacturing Losing Its Toolbox?” I’ll conclude by merely alluding to some remedies. (I’m sure there’s a post to be written comparing the policy positions of all the candidate on manufacturing in detail, but this is not that post.)
The first voice: Donald Trump. From “‘We’re Finally Rebuilding Our Country’: President Trump Addresses National Electrical Contractors Association Convention” (2018):
“We’re in the midst of — something which nobody thought you’d hear,” Trump said. “We’re finally rebuilding our country, and we are doing it with American aluminum, American steel and with our great electrical contractors,” said Trump, adding that the original NAFTA deal “stole our dignity as a country.”
The second voice: Elizabeth Warren. From “The Coming Economic Crash — And How to Stop It” (2019):
Despite Trump’s promises of a manufacturing “renaissance,” the country is now in a manufacturing recession. The Federal Reserve just reported that the manufacturing sector had a second straight quarter of decline, falling below Wall Street’s expectations. And for the first time ever, the average hourly wage for manufacturing workers has dropped below the national average.
(One might quibble that a manufacturing renaissance is not immune from the business cycle.) A fourth voice: Trump campaign surrogate David Urban, “Trump has kept his promise to revive manufacturing” (2019):
Amazingly, under Trump, America has experienced a 2½-year manufacturing jobs boom. More Americans are now employed in well-paying manufacturing positions than before the Great Recession. The miracle hasn’t slowed. The latest jobs report continues to show robust manufacturing growth, with manufacturing job creation beating economists’ expectations, adding the most jobs since January.
Obviously, the rebound in American manufacturing didn’t happen magically; it came from Trump following through on his campaign promises — paring back job-killing regulations, cutting taxes on businesses and middle-class taxpayers, and implementing trade policies that protect American workers from foreign trade cheaters.
Then again, from the New York Times, “Trump Promised a Manufacturing Renaissance. What Happens in 2020 in Places That Lost Those Jobs?” (2019):
But nothing has reversed the decline of the county’s manufacturing base. From January 2017 to December 2018, it lost nearly 9 percent of its manufacturing jobs, and 17 other counties in Michigan that Mr. Trump carried have experienced similar losses, according to a newly updated analysis of employment data by the Brookings Institution.
Perhaps the best reality check — beyond looking at our operational capacity, as we are about to do — is to check what the people who will be called upon to do the work might think. From Industry Week, “Many Parents Undervalue Manufacturing as a Career for Their Children” (2018):
A mere 20% of parents associate desirable pay with a career in manufacturing, while research shows manufacturing workers actually earn 13%more than comparable workers in other industries.
If there were a manufacturing renaissance, then parents’ expectations salaries would be more in line with reality (in other words, they exhibit hysteresis).
[TIM COOK;] “The products we do require really advanced tooling, and the precision that you have to have, the tooling and working with the materials that we do are state of the art. And the tooling skill is very deep here.
“The vocational expertise is very very deep here, and I give the education system a lot of credit for continuing to push on that even when others were de-emphasizing vocational. Now I think many countries in the world have woke up and said this is a key thing and we’ve got to correct that. China called that right from the beginning.”
With Cook’s views in mind, let’s turn to the slap of cold water administered by Michael Collins in Industry Week, “Is US Manufacturing Losing Its Toolbox?“:
So are we really in the long-hoped-for manufacturing renaissance? The agency with the most accurate predictions on the future of jobs is the Bureau of Labor Statistics. Their projection to 2026 shows that US manufacturing sector will lose 736,000 manufacturing jobs. I spoke with BLS economists James Franklin and Kathleen Greene, who made the projections, and they were unwavering in their conclusion for a decline of manufacturing jobs.
This prompted me to look deeper into the renaissance idea, so I investigated the changes in employment and establishments in 38 manufacturing North American Industry Classification System (NAICS) industries from 2002 to 2018. I really hoped that the optimists were right about the manufacturing renaissance, but the data I collected in Table 1 (see link) shows some inconvenient truths—that 37 out of the 38 manufacturing industries are declining in terms of both number of plants and employees.
So, yeah. Mirage.
I reformatted Table I so I could mess about with it in LibreOffice. Here it is, in its original form:
And here are the changes in the NAICS manufacturing categories sorted by losses to firms:
(“1,047” in the first data cell means that number of establishments were closed, hence “-57.5%” in the second data cell is negative.) There are many fascinating stories — tragedies, really — behind these statistics. We can see the collapse of the apparel industry in “Cut and sew apparel,” (-55.6%), “textile furnishings mills” (-45.2%), and “textile fabric mills” (‘44.7%). We can see the decline of Maine’s pulp paper industry in “Paper (except newsprint)” (-19.9%). What’s doing well? Wood containers/pallets (firms down 18.0% but employment up 1%). Gotta pack those consumer goods coming in from China somehow!
However, I’ve highlighted the categories that really concern Collins in blue; they all have to do with, as Cook calls it, tooling. Collins writes:
But the most perplexing of these declining industries are the ones that are fundamental to making other manufactured products. These are industries like machining, machine tools, mold making, tool and die, semiconductors, forging, and foundries. It is difficult to see how we can achieve a manufacturing renaissance while these critical industries continue to decline… Machine tools. These are the master machines that make other machines and products…. Machine Shops. Machining is essential to hundreds of industries and thousands of products from products as tiny as a machine screw and as large as a turbine bearing for a dam. Machining is absolutely essential for all industrial products but is also used in consumer products. Machining. Two classes of machinists that are critical to manufacturing are tool and die makers and mold makers…. Industrial mold companies lost 1,241 shops (42.7%) and 5,968 workers (12%) between 2002 and 2018. The big question is, since a good deal of machining is now done overseas, is it possible to support all of the industries and companies who need machined products in the U.S by only using foreign suppliers? Foundries. The process of making parts by pouring metal to make a casting is ubiquitous and is used in the machinery, automotive, pipe, fitting, railroad equipment, valve, and pump industries. Castings are also used in everything from heart valves to aircraft carrier propellers and in every home for bathtubs, sinks, fixtures, and furnaces… Forging and Stamping. The demand for forging and stampings has been declining by the industry’s major downstream markets, which include aerospace, agricultural machinery and oil and gas machinery. Additionally, the world prices of steel, and nonferrous metals have been volatile, making it hard for the industry to secure stable purchasing contracts… Semiconductors. Semiconductors are absolutely fundamental to the electronics and computer industries… One of the big problems is that when the manufacture of semiconductors moves overseas, research and development goes with it. If the decline continues the U.S. is in danger of losing its innovative edge in electronics and computers.
A strong part of the new optimism that US manufacturing is in a renaissance is driven by digital revolution that includes progress in robotics, artificial intelligence, 3D printing, data analytics, and other digital advances. The digital revolution has great potential, but economic numbers show that it is not yet a reality. In fact, economic data from the BLS shows that since 2007, productivity growth has averaged only 1.3% growth. …[T]he biggest problem as shown in BLS data shows that 38 manufacturing industries continue to decline and nine of 10 of these industries are the critical industries that are fundamental to the manufacturing process. The digital revolution is not going to help if we eventually lose the industry.
A tooling firm closes, and a complex organism withers. The machinery is sold, sent to the scrapyard, or rusts in place. The manuals are tossed. The managers retire and the workers disperse, taking their skills and knowledge with them. The bowling alley closes. The houses sell at a loss, or won’t see at all. Others, no doubt offshore, get the contracts, the customers, and the knowledge flow that goes with all that. All this causes hysteresis. “The impact of past experience on subsequent performance” cannot be undone simply by helicoptering a new plant in place and offering some tax incentives! To begin with, why would the workers come back?
So, when I see no doubt well-meant plans like Warren’s “Economic Patriotism” — and not to pick on Warren — I’m skeptical. I’m not sure it’s enough. Here are her bullet points:
- More actively managing our currency value to promote exports and domestic manufacturing.
- Leveraging federal R&D to create domestic jobs and sustainable investments in the future.
- Production stemming from federally funded research should take place in the United States
- Taxpayers should be able to capture the upside of their research investments if they result in profitable enterprises.
- R&D investments must be spread across every region of the country, not focused on only a few coastal cities.
- Increasing export promotion to match the efforts of our competitors.
- Deploying the massive purchasing power of the federal government to create markets for American-made products.
- Restructuring worker training programs to deliver real results for American workers and American companies.
- Dramatically scale up apprenticeship programs.
- Institute new sectoral training programs.
There’s a lot to like here, but will these efforts really solve the hysteresis that’s causing our tooling problem? Just spit-balling here, but I’d think about doing more. Start with the perspective that our tooling must be, as much as possible, domestic. (“If your business depends on a platform, you don’t have a business.” Similarly, if your industrial base depends on the tooling of others, it’s not an industrial base.) As tooling ramps up, our costs will be higher. Therefore, consider tariff walls, as used by other developing nations when they industrialized. Apprenticeships and training are good, but why not consider skills-based immigration that brings in the worker we’d otherwise have to wait to train? Further, simply “training” workers and then having MBAs run the firms is a recipe for disaster; management needs to be provided, too. Finally, something needs to be done to bring the best and brightest into manufacturing, as opposed to having them work on Wall Street, or devise software that cheats customers with dark patterns. It’s simply not clear to me that a market-based solution — again, not to pick on her — like Warren’s (“sustainable investments,” “research investments,” “R&D investments,” “export promotion,” and “purchasing power”) meets the case. It is true that Warren also advocates a Department of Economic Development “that will have a single goal: creating and defending good American jobs.” I’m not sure that’s meaningful absent an actual industrial policy, democratically arrived at, and a mobilized population (which is what the Green New Deal ought to do).