OPINION:
The economy is growing, but jobs are harder to find.
Second-quarter growth in gross domestic product was 3.8%, thanks to robust investment in information processing equipment, software, and research and development, mostly to build out artificial intelligence.
With government scorekeepers furloughed last month, policymakers and analysts must rely on private forecasters and data for a sense of third-quarter performance and what lies beyond.
The Wall Street Journal survey of private prognosticators pegged third-quarter growth at 2.5%, in line with the trend for the eight years of the first Trump and Biden presidencies.
Wells Fargo’s economics unit and the proprietary service Action Economics estimated 3.5% and 3.0%. My estimate is 3.0%.
Longer term, most economists believe the best the economy should accomplish is about 2%, perhaps less, owing to slower labor force growth in the wake of President Trump’s limits on immigration.
Hence, they expect growth to slow next year. I don’t agree, thanks to an even faster pace of AI investment and an upshift in workplace efficiency to compensate for fewer workers.
We are not in an AI bubble as much as we’re living through a revolution.
Enthusiasm for AI extends well beyond Big Tech and the money managers and retail investors who ran up share prices for publicly traded companies.
Like the settlers during the Oklahoma land rush, venture capitalists are chasing AI startups, not the other way around. Innovators such as Decagon, which makes AI tools for customer service applications, are being showered with gifts and offers of favors by venture capitalists who hope to lead or get a good piece of their next funding round.
How would you like to receive NBA tickets and expensive meals from bankers hoping to land your next car loan or fund your second year at law school?
The lion’s share of fresh money is going into AI because the folks who sniff out the winners — the next Nvidia, Apple or Salesforce — sense that AI will pay off, but not necessarily by enabling computers to think like Nobel laureates. Rather, advanced AI will spawn specialized agents that permit computers to replace millions of middle-level workers in administrative, managerial and research functions.
Just as the tractor, moving assembly line and personal computer radically altered farms, factories and offices during the 20th century, AI will drive paradigmatic change.
Since May, job creation has been well below the 60,000 per month that may be possible from indigenous population growth and the immigration regime Mr. Trump now has in place.
That’s well below the 141,000 monthly pace during the first Trump and Biden administrations.
So how can the economy be growing so fast with so little uptake of new workers?
One factor is uncertainty about how Mr. Trump’s tariffs and trade war with China will sort out.
Still, stuff must be made, and we are far into 2025 with a lot of growth under our belts and an almost stagnant level of employment.
The answer must be that AI is paying off with elevated productivity growth, but you should be careful of economists on this.
During the 1980s, desktop computers emerged and reduced headcounts in clerical and administrative tasks. In the 1990s, the internet accelerated this process, but the bean counters couldn’t find it in the data.
Nobel Prize-winning economist Robert Solow lamented, “You can see the computer age everywhere but in the productivity statistics.”
The same thing may be happening again: Economists are equivocating about AI.
Some latch onto stories about AI creating workslop, for example, AI-generated content that contains nonexistent sources and fictitious claims. All those indicate is that what computers spit out still must be fact-checked by human beings.
Importantly, economists get tenure at universities, raise money for projects and keep corporate bosses happy doing surveys and parsing data — the stuff of “Freakonomics.”
Those tend to distort or lag what’s going on in real time.
Journalists who directly observe businesses bring back a more optimistic story.
Axios reporters recently spoke with 20 CEOs from various companies, and all said they were reducing hiring ambitions as AI comes into plain sight.
Walmart plans to keep its headcount constant as sales grow over the next three years. Along with Amazon, it is rapidly replacing workers in warehouses with robots aided by AI.
The real challenge ahead is what’s next for all those displaced workers.
Some things can’t be automated (for example, much of the more mundane work nurses do). Rural hospitals are at the leading edge of the return to vocational secondary education by training teenagers.
If the glass is half empty for anyone, it’s colleges and universities. High school graduates are shunning majors and institutions that deliver big debt but too few employment opportunities.
To be employable, Americans may have to come full circle by working with their hands as more machines become the mind of man.
• Peter Morici is an economist and emeritus business professor at the University of Maryland, and a national columnist.
Please read our comment policy before commenting.