Last week’s employment report offered what may be the last clear picture of the US job market before President Donald Trump’s tariff shock. Overall, it looked pretty healthy, with a 4.2% unemployment rate, 80.4% of the prime-age population employed and 1.9 million nonfarm payroll jobs added over the past 12 months.
Where did those 1.9 million new positions come from? Almost half were in health care and social assistance, with government (where the gains were all at the state and local level), leisure and hospitality, and transportation and warehousing also big contributors. Employment was down in manufacturing, professional and business services, information (publishing, software, telecommunications and internet) and mining and logging (mostly oil and gas extraction).
Apart from construction, all the growth has been in service jobs. For better or worse, this is what Trump wants to change. Whether suddenly raising tariffs on almost every imported product will shift job growth toward goods production or just put an end to job growth is one of several big questions about his approach. Another big question is whether the job-growth trajectory of the past few decades, driven to a remarkable extent by hiring in health care and social assistance, is due to structural shifts that will outlast Trump’s tariff war.
In the near-term, the dependence of US companies on cheap imported components and finished goods argues for a shock to profits that will force belt-tightening and potentially layoffs. Over the longer run, Trump’s goal to recreate past manufacturing glory probably can’t be achieved without technological advances that would cancel out a lot of the potential job growth. And redistributing labor from services to manufacturing may also be odds with the demographics of an aging America.
There are multiple drivers of the big employment gains in health care and social assistance, including federal decisions to expand health insurance coverage with the Children’s Health Insurance Program in 1997 and Affordable Care Act in 2010. But the near-doubling of the 65-and-older population from 31 million in 1990 to 59 million as of 2023 is one of the most obvious, with the number of home health care workers — who mostly care for the elderly — rising from 262,000 in 1990 to 1.9 million last month.
Meanwhile, the goods-producing sectors — manufacturing, construction, and mining and logging — have gone from 22% of nonfarm jobs to 13.6%. Farming, a goods-producing sector but not in the nonfarm payroll data, has also seen its share of full-time-equivalent employment fall from 0.8% in 1990 to 0.5% in 2023, according to the US Bureau of Economic Analysis.
The goods-producing employment decline is almost entirely about manufacturing, where automation and the shifting of production overseas have eliminated five million jobs since 1990 and cut the sector’s share of nonfarm employment in half — although almost all of that decline was before 2010. Construction employment is up since 1990 and up a lot since the housing-bust year of 2010. Mining and logging employment has declined, but it already didn’t amount to much in 1990.
The reshaping of the job market hasn’t just been about health care and goods production. Retailing accounts for a markedly smaller share of US jobs than in 1990 — with the increase in transportation and warehousing jobs created by ecommerce making up for only a fraction of that. And after gains in the 1990s, information employment tanked as telecom providers retrenched and traditional publishers lost most of their ad revenue to Google, Facebook and their ilk (which are also part of the information sector but can harvest that revenue with far fewer employees).
The North American Industry Classification System used in government statistics has 20 sectors, which the US Bureau of Labor Statistics groups into 11 supersectors in its nonfarm payrolls reporting. The selection used in these charts is mostly supersectors, with sectors broken out when the supersector feels overly broad, as with trade, transportation and utilities and education and health services. My groupings are almost the same as the set of industry sectors for which the BLS makes employment projections, which it last did in August 2024. The agency’s view for the next decade envisioned health care and social assistance once again leading the way.
Note that employment in professional and business services, projected here to be one of the biggest gainers over the next decade, actually fell over the past 12 months. This is possibly just noise, but it could also be a sign that artificial intelligence has begun to come for the lawyers, accountants, consultants and tech service providers. The impact of technological progress is always a wild card, and seems likely to mute any employment shifts from Trump’s tariffs.
For example, Commerce Secretary Howard Lutnick’s televised remark that “the army of millions and millions of human beings screwing in little screws to make iPhones — that kind of thing is going to come to America,” has become something of a meme over the past week.
Usually left out is his next sentence: “It’s going to be automated.” Lutnick argued that this will still provide work for mechanics, HVAC specialists and electricians, but how much?
In oil and gas extraction, another Trump priority, employment has fallen 64% over the past decade even as US production of oil and natural gas has risen 130% (measured in BTUs). The employment trajectory of a mining or oil and gas boom is somewhat unique — it can take a lot of people to figure out where the resources are, and far fewer to get most of them out of the ground. But the intense pressure to increase production relative to labor costs and capital expenditures is shared by all of what are called the tradable sectors, those that produce goods and services that can be traded internationally.
Health care and social assistance is mostly nontradable. Some rich foreigners come to the US for medical procedures and some not-quite-so-rich Americans travel abroad for theirs, but the vast majority of the jobs are not affected by international competition. They are affected by government spending, and big cuts to Medicaid and social programs could weigh on job growth. But with the 65-and-older population projected to grow by another 30% over the next 25 years, would you really want to bet that America is going to need fewer health care workers than it does now?
A message from Advisor Perspectives and VettaFi: To learn more about this and other topics, check out some of our webcasts.
Bloomberg News provided this article. For more articles like this please visit
bloomberg.com.
