Why Is Unemployment Falling as Job Growth Weakens?

TL;DR

Falling unemployment does not necessarily signal a strong labor market, because people who stop looking for work are excluded from the unemployment count. Payroll growth is weak, labor force participation has fallen sharply, and job openings have generally declined, but low layoffs and initial claims suggest the economy has not entered a recessionary feedback loop.

Transcript

Hey everyone and thanks for jumping back into the macroverse. Today we're going to talk about the most recent labor market report and many of the mixed signals that we got. We're going to try to decipher what's actually going on and why the market is having a hard time understanding what the Fed is going to do next. If you guys like the content, ma... Read More

Key Insights

• Nonfarm payroll growth is weakening, with the latest report showing 23,000 jobs against expectations of roughly 80,000. Negative monthly readings have become more common since 2025, signaling deterioration even though one weak month alone does not establish an economic contraction.
• Year-over-year nonfarm payroll growth remains positive, which matters because negative annual changes have historically coincided with recessions in the data discussed. The annual measure came close to turning negative in February but subsequently recovered, leaving the economy without that recession signal for now.
• Net job creation is extremely limited compared with earlier growth. Payroll employment was growing roughly 4% to 5% annually in 2022, weakened through 2023 and 2024, and is now described as effectively near zero, creating little room for a policy mistake by the Federal Reserve.
• The unemployment rate can fall while hiring weakens because it counts jobless people who want and are seeking work. The rate declined from 4.5% in November to 4.1%, while labor force participation fell from 62.5% to 61.4%, removing some nonworkers from the unemployment calculation.
• Layoffs are not the main source of current labor market weakness. Although layoffs have trended upward and some large reductions have received attention, their overall level is described as roughly comparable with the prepandemic period rather than the pronounced surge normally visible during a recession.
• Initial unemployment claims remain low, with a recent reading of 189,000 described as the lowest in a long time. This indicates that companies are generally retaining workers and that the reinforcing cycle of layoffs, reduced demand, and further layoffs has not begun.
• Labor market conditions contain both strength and weakness across indicators. Job openings have generally trended downward but recently showed some resilience, while temporary help services declined for a long period before beginning to recover, making the overall signal difficult for markets and policymakers to interpret.
• Regional and industry differences can conceal the national picture. Some industries may already feel recessionary, especially risk-sensitive areas such as crypto, while others remain resilient. State unemployment changes are mixed, unlike 2008, when deterioration spread across the country and left little economic optimism.

Questions & Answers

Q: Why can unemployment fall when job growth is weakening?

The unemployment rate counts people who lack a job but want work and are actively looking for it. People who stop seeking employment are not included as unemployed. In the period discussed, the unemployment rate fell from 4.5% to 4.1%, while labor force participation dropped from 62.5% to 61.4%. That participation decline helps explain why unemployment fell despite weaker payroll conditions.

Q: Does one negative monthly payroll reading mean a recession?

One negative monthly payroll change does not automatically establish a recession because individual months can be affected by temporary or random events. The transcript points to July through August 1997 as an example of a decline that did not necessarily represent contraction. A sustained decline is more significant, and quarterly or year-over-year comparisons can reduce the false signals created by volatile monthly data.

Q: What does year-over-year payroll growth indicate about recession risk?

Year-over-year nonfarm payroll growth has historically turned negative during recessions in the data presented. It has not turned negative in the current period, although it came close in February before moving higher. This distinction explains why a negative monthly reading is not enough to declare a recession. Annual payroll growth remains positive, but its weakness leaves little margin for further deterioration.

Q: Where is the current labor market weakness coming from?

The weakness is primarily visible in limited job creation, declining labor force participation, and a general downward trend in job openings, rather than in widespread layoffs. Monthly payroll changes have become negative more often since 2025, and annual payroll growth is described as very small. Temporary help services also weakened for an extended period, although that indicator recently began trending upward again.

Q: Are layoffs currently signaling a recession?

Layoffs have moved upward, but the level described is roughly comparable with the prepandemic period rather than the sharp increases associated with recession. Initial unemployment claims also recently reached 189,000, described as the lowest reading in a long time. Together, these indicators suggest employers are generally retaining workers and that a self-reinforcing cycle of layoffs, weaker demand, and additional layoffs has not started.

Q: Why is the Federal Reserve's next move difficult to predict?

The Federal Reserve faces mixed labor market evidence. Payroll growth is weak, participation is falling, and job openings have mostly declined, yet unemployment has also fallen, layoffs remain near prepandemic levels, and initial claims are low. These conflicting signals make it difficult to determine whether the economy requires support or remains resilient. With job creation already limited, the transcript argues that a policy mistake could turn annual payroll growth negative.

Q: How do regional differences affect the national labor market picture?

Labor conditions can differ substantially among states and regions. Over the previous six months discussed, unemployment rose in some states and fell in others, creating a mixed national map. That differs from 2008, when worsening unemployment spread across the country. Continued strength in some locations can provide a pocket of optimism even while residents of weaker regions feel that recessionary conditions have already arrived.

Q: How should long-term investors interpret weaker employment data?

Weak employment data does not guarantee a stock market decline of 50%. The transcript notes that recessions have also involved declines of roughly 20% or 30%, and market drawdowns are an expected part of long-term investing. The stated approach is to average into low-expense-ratio index funds with a 20-year horizon, rather than making decisions based on uncertain short-term labor reports or market fear.

Summary & Key Takeaways

• Nonfarm payrolls missed expectations, with 23,000 jobs reported against an expected figure of roughly 80,000. Monthly and quarterly changes have weakened, while year-over-year payroll growth remains positive. The labor market is therefore close to stagnation, but the historical recession signal of negative annual payroll growth has not appeared yet.

• The unemployment rate fell from 4.5% in November to 4.1%, even as payroll conditions weakened. A major explanation is that labor force participation dropped from 62.5% to 61.4%. People without jobs are counted as unemployed only when they want work and are looking, so departures from the workforce can lower unemployment.

• Layoffs have increased only to approximately prepandemic levels, while initial unemployment claims recently reached 189,000, described as their lowest level in a long time. Weakness instead appears in sluggish job creation, declining participation, and generally falling job openings. Conditions also vary considerably among industries, risk markets, regions, and individual states.