AI Creates More Jobs than it Eliminates
Recent predictions about artificial intelligence leading to widespread unemployment were challenged this week as April data highlighted a notable uptick in worker demand.
This information came from the Bureau of Labor Statistics. The total job openings jumped by 731,000, reaching 7,618,000. This was more than even the most optimistic forecasts from Wall Street economists suggested.
A significant factor behind this rise was Professional and business services, which contributed 668,000 job openings, making up about 91% of the total increase. Job openings in this sector surged from 1.047 million to 1.715 million within just a month. This sector is particularly sensitive to economic cycles—typically seeing more demand during periods of growth and less during downturns. An increase in this area hints at accelerated growth.
This supports the notion that AI may create issues like Jevons paradox in the labor market. As AI boosts the productivity of skilled workers, the incentives for employing these workers rise, leading to a greater overall demand for them, which outpaces any reductions from direct job replacement.
Growth in Construction and Manufacturing Jobs
The AI-driven construction boom is having a direct effect. Planned construction projects grew by 25,000, totaling 259,000. While this job growth isn’t astounding compared to post-pandemic numbers (which reached over 450,000 jobs in March 2022), it looks solid over a longer period. A significant portion of the construction worker demand likely ties back to AI, as seen in spending data, particularly in data center construction, which has seen a rise of over 28% from the previous year.
Manufacturing jobs increased by 24,000, though non-durable goods saw a decrease of 1,000 jobs, while durable goods jobs grew by 25,000. Current job openings in manufacturing are at 474,000, showing a decline compared to the immediate post-pandemic hiring surge, yet they’re up from pre-pandemic levels. The number of openings in consumer durable goods stores also sits at 321,000, an impressive figure historically, and it has been climbing since the fall.
Still, Jevons’ paradox doesn’t negate the existence of job migration. The information sector added 4,000 jobs, totaling 87,000, though this is much less than last year’s figures. Meanwhile, there were consecutive months of layoffs in the information sector, totaling 60,000. Although the layoff rate dipped to 2.1% from 2.2% in March, both figures are up historically. Data since 2002 indicates that the usual layoff rate for information jobs hovers around 1.8%, with a pandemic-adjusted rate near 1.5%. The number of employees still surpasses layoffs, but the margin is tightening. Alongside retirements, it seems that the demand for information sector jobs is dwindling—something typically reserved for economic downturns.
Low Employment Reflects Supply Constraints
Those skeptical of the optimistic interpretation of the JOLTS report point to a significant drop in employment. Jobs in trade and transportation fell by 134,000, and professional and business services experienced a decline of 131,000 despite an initial surge. The adoption rate stands at 3.2%, near the lower end of post-pandemic figures. However, this likely points to supply constraints rather than a lack of demand. The unemployment rate remains around 4%, indicating a scarcity of available workers. A turnover rate of 1.9% suggests that the normal job-to-job movement in the economy is nearly stagnant. Additionally, stricter immigration policies are limiting the labor supply among low-skilled individuals, where many unfilled jobs are concentrated.
Nonetheless, the overall impact of AI on labor demand appears to be beneficial. In fact, 493,000 jobs were added in the Western region, indicating that workers are transitioning from information-related roles to business services.
Importantly, JOLTS data can be quite volatile and has seasonal patterns. The professional and business services sector, in particular, tends to fluctuate around April, likely due to tax preparation activities. Yet, even if we account for this atypical boost, the gains seen in April were noteworthy.
A clearer picture emerging from the JOLTS report reveals: increased efficiency coupled with rising aggregate labor demand. On the flip side, tasks like coding can be directly replaced by AI. The simultaneous surge and disruption caused by AI represent two perspectives on the same underlying shifts in the labor market.
