Goldman Sachs: U.S. immigration plummets by 80%, reshaping the employment "breakeven point"; AI becomes the biggest variable in the labor market
Goldman Sachs’ latest analysis points out that the fundamentals of the U.S. labor market are undergoing profound changes. Due to the Trump administration's immigration restriction policies, net immigration has plummeted by 80%. This shift is redefining the level of job growth required to keep the unemployment rate stable. The bank estimates that by the end of this year, the U.S. will only need to add about 50,000 jobs per month to maintain the unemployment rate, much lower than the current 70,000.
This sharp contraction in labor supply stems from a comprehensive tightening of immigration policy. During the Biden administration, more than 10.8 million illegal immigrants entered the U.S. After 2025, net immigration has dropped from an annual level of about 1 million in the 2010s to around 500,000. Goldman predicts it will further decline to just 200,000 in 2026.
However, labor demand remains “fragile.” Goldman notes that current job growth is narrow and job openings continue to decline—now at about 7 million, lower than pre-pandemic levels. The bank believes the biggest downside risk facing the labor market comes from artificial intelligence. AI may lead to faster and more disruptive structural adjustments, potentially dampening employers' willingness to hire and resulting in unemployment numbers exceeding current expectations.
Tighter Immigration Policy Redefines Job Growth Threshold
In its latest report, Goldman Sachs breaks down the transmission path of U.S. immigration policy tightening to the labor market. The Trump administration increased deportations, tightened visa and green card approvals, paused immigration processing from dozens of countries, and canceled temporary protection status for certain groups, sharply reducing immigration flows from multiple fronts. Data show that U.S. net immigration dropped from about 1 million per year in the 2010s to 500,000 by 2025, and is expected to shrink further to 200,000 in 2026.
The sharp contraction on the supply side directly lowers the economic “break-even” pace of job growth. Goldman estimates that by year-end, the U.S. will only need to add about 50,000 jobs per month to prevent the unemployment rate from rising, far lower than the current 70,000 or so. The report notes that since fewer new workers are entering the economy, hiring no longer needs to be as robust as before to keep the unemployment rate stable. The report states:
“Only a slight rebound is enough to keep job growth at break-even levels.”
Although Goldman’s immigration forecasts differ from those of the Brookings Institution and the Congressional Budget Office, all point to a clear downward trend. Additionally, Goldman raises a potential risk: stricter immigration enforcement may be pushing more workers into informal or off-the-books employment. If this trend is real, official employment data will underestimate the true activity in the labor market, making it more complicated for the Fed to assess economic momentum.
Demand Side Signals Continue to Weaken
Although shrinking labor supply mathematically lowers the "break-even" line for job growth, this does not necessarily mean the labor market is strong. Goldman describes current demand-side performance as “fragile,” noting job growth is more narrowly focused and mainly led by the healthcare industry, while job openings continue to decline.
Data show that job openings have now dropped to about 7 million, below pre-pandemic levels and still falling. Official data from the U.S. Bureau of Labor Statistics confirm this trend, with job vacancies at the midpoint of the 6 million range by the end of last year. Goldman warns that the continued decline in job openings will increase the risk of a more pronounced rise in unemployment—even slowing labor supply growth cannot fully offset this pressure.
The report explains that as the influx of new labor into the economy slows, hiring no longer needs to be as hot as before just to keep the unemployment rate from rising. This logic means seemingly weak employment data may increasingly mask a labor market that is only maintaining the status quo, not accelerating deterioration.
This creates a paradox: the “stability” of the labor market may increasingly look like “weakness.” As immigration slows and labor growth eases, job growth levels once seen as warning signs may soon be enough to maintain market stability.
AI Becomes the Biggest Uncertainty
Goldman Sachs sees artificial intelligence as the biggest downside risk to labor prospects—not because it has already triggered mass layoffs but because it may curb hiring at the margins. So far, the substitution effect from AI has reduced monthly job growth by 5,000 to 10,000 in the industries most affected. However, faster or more disruptive deployment could put more pressure on demand.
Goldman writes in the report:
"Our main concern about downside risk in baseline forecasts is that AI may see faster and more disruptive deployment. While there is plenty of anecdotal evidence of rapid adoption and resulting unemployment, it is hard to know how these will translate into macroeconomic outcomes."
Their data show that in the sub-sectors where AI is easiest to deploy, job growth has already slowed and turned slightly negative, and company-level evidence indicates AI is reducing demand for workers. So far, this impact, although visible, remains “moderate.”
Currently, Goldman expects the unemployment rate will only rise slightly to about 4.5%. Chief economist Jan Hatzius noted in another report that the probability of recession next year is a "moderate" 20%. The bank believes the labor market is "taking early steps toward stability."
However, with rapid development and broader potential application of AI technology, this relatively optimistic forecast faces significant uncertainty. If AI is deployed faster than expected, the impact on the job market may be larger than currently estimated.
Risk Warning and DisclaimerThe market carries risks; investments must be made cautiously. This article does not constitute individual investment advice, nor does it consider the particular investment objectives, financial circumstances, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suited to their own circumstances. Investment decisions made accordingly are at one’s own risk.