Just when I thought there wasn’t a heckuva lot more I could learn about the general, long-term (downward) direction of manufacturing employment in America, I just learned something pretty significantly new. Even better, the information bears on the big ongoing national debate about whether most of the blame for U.S. manufacturing job loss should go to trade and trade policy, or to improved productivity and the substitution of technology for labor.
I looked into the matter after watching a recording of a television show about globalization after Brexit. One of the featured guests, economist Russell Roberts, made a statement that I knew is flat wrong: “[M]anufacturing employment has been falling steadily, but long before we had free trade agreements; goes back to the end of World War II.” A few moments later, he (sort of) suggested that he was really talking about a rather different measure – manufacturing jobs as a share of total jobs. In Roberts’ words, this figure has been “falling steadily” for “about 70 years.”
Because I didn’t have the details in my head, I went back to the official statistics, kept by the Labor Department, and saw that I was indeed right on the first score. In addition, even though Roberts (as I knew) was correct on the second, the actual figures tell a much more complicated story than his standard rehash of the conventional wisdom that technology is the main culprit. Here’s what I mean.
When it comes to the absolute numbers of manufacturing jobs in the United States, they are indeed down considerably since 1945. But what’s as important as it is under-appreciated is that between V-J Day (in August, 1945) and June, 1979, American manufacturing employment rose by nearly 38 percent. That’s a period fully half as long as the 70 years Roberts mentioned. In addition, though the June, 1979 level of 19.553 million represented the absolute peak of the U.S. manufacturing workforce, even afterwards, the sector enjoyed periods of job growth. That is, its decline wasn’t all that steady.
For example, between late 1982 – when a deep recession ended – through March, 1989, the economy boosted manufacturing employment by 8.21 percent. From July, 1993 (following another, shallower, recession) to April, 1998, manufacturing increased payrolls by 5.36 percent. And of course, since the employment bottom hit soon after the latest recession ended, through last month, manufacturing employment has rebounded by 7.36 percent.
Manufacturing, as a result, has never come close to that World War II aftermath employment level. But the data also make clear that the start of manufacturing’s jobs decline (mid-1979) coincides almost exactly with the point at which trade – and especially imports – began rapidly rising as a share of the total economy, as this chart illustrates.
The same trends have held in connection with manufacturing’s share of total employment. I compared the December figures going back to 1945, and found that from then through December, 1980, the percentage of all American non-farm workers (the U.S. government’s official employment universe) employed in manufacturing dropped from 32.03 to 20.50. That’s a fall-off of just under 36 percent. Since then, the decline has been much faster – from 20.50 percent to 8.53 percent, or 58.39 percent.
Manufacturing has also managed to raise its share of the national workforce in several stretches since 1945. These periods include 1949 to 1956, and the years 1959, 1961, 1964-1966, 1971 to 1973, 1976, and 1983. But since then – and shortly after trade’s role in the economy began soaring – manufacturing has never repeated this feat.
Yet these statistics are telling us something else crucial about manufacturing employment and where it fits into the national economy. They’ve depended not only on trade and technology/productivity, but on macroeconomic factors (especially whether the economy is growing or shrinking), and on other political and policy factors – such as whether defense buildups are in progress (which gives manufacturing employment a lift) and whether all government employment is strong or weak (which affects its share of all jobs).
At the same time, better trade balances mean faster growth, manufacturing still dominates U.S. trade flows, and the sector is set to record yet another record trade deficit this year. So there can be little reasonable doubt that, although better trade policies are no manufacturing jobs cure-all, they’d make a valuable contribution.