The quiet hum of economic data often obscures the shifts happening beneath the surface. For months, economists have been parsing the surprisingly robust productivity numbers, struggling to reconcile them with a labor market that, while resilient, hasn’t shown the dramatic efficiency gains one might expect. The latest research from David Jang and Jon D. Samuels at the Bureau of Economic Analysis offers a compelling explanation: it isn’t just how much work is being done, but who is doing it. Their paper, “Labor Composition and the Recent Productivity Acceleration: Evidence From a Prototype Quarterly Industry Dataset,” reveals that changes in labor composition – the skills, experience, and education levels of the workforce – are a significant, and increasingly important, driver of productivity growth.
This isn’t a new idea. Economists have long understood that a more skilled workforce can produce more output. However, the BEA’s new quarterly industry-level dataset allows for a far more granular look at this dynamic, revealing that labor composition changes accounted for approximately 14% of aggregate labor productivity growth between 2005 and 2024. This contribution increased to 10% from 2019Q4 to 2024Q4, suggesting that the effect is becoming more pronounced. The implications are significant, as they challenge the narrative that productivity is solely driven by capital investment or technological advancements.
But the headline number hides a crucial nuance. While labor composition changes have contributed to overall productivity gains, those gains aren’t evenly distributed. The research highlights a shift in the primary drivers of productivity growth beginning in 2019Q4, with sectors like computer systems design, retail trade, data processing, and management of companies emerging as key contributors. Critically, the analysis shows that this acceleration isn’t simply due to a broader improvement in worker quality across all industries. The variance in labor composition contributions increased after 2019, yet explains almost none of the cross-industry variance in overall labor productivity growth. The gains are concentrated in specific sectors, and the increasing disparity in contributions across industries is a key observation.
The Labor Force Level shows a steady incline, but this aggregate view hides the compositional shifts driving productivity gains.
Furthermore, when looking at shorter-term momentum, the picture becomes even more focused. Comparing the four-quarter moving average in 2019Q4 (pre-pandemic) to 2024Q4, the data reveals that the majority of the acceleration is concentrated in a handful of industries: retail trade, ambulatory health care services, miscellaneous professional and scientific services, and hospitals and residential care facilities. This demonstrates the vital importance of high-frequency, quarterly industry-level data for capturing short-term structural shifts. This contrasts sharply with the longer-term averages, which can obscure these crucial dynamics.
However, the BEA's dataset, while groundbreaking, has limitations. The authors themselves acknowledge that it doesn't fully capture all aspects of labor quality, nor does it account for the impact of capital deepening. The paper’s growth accounting framework focuses specifically on labor composition, grouping capital contributions and total factor productivity (TFP) together. This means the extent to which capital investment or technological advancements contribute to productivity gains remains unclear. The analysis relies on the production possibility frontier concept, which is a simplifying assumption in a complex economic system, and the prototype data is still under development.
Average Hourly Earnings remain relatively stable, suggesting wage growth isn’t the primary driver of the observed productivity gains.
Returning to the initial puzzle of robust productivity growth despite a seemingly stable labor market, the BEA’s research provides a crucial piece of the puzzle. The gains aren’t coming from across the board, or from simply paying workers more. They’re concentrated in specific sectors, driven by shifts in the composition of the workforce. This has implications for how we interpret economic data and what we expect in the future.
The recent uptick in productivity growth, while welcome, may not be a signal of broad-based economic strength, but rather a reflection of specific sectoral adjustments. As signals from the labor market indicate a slowing of growth and wage increases, alongside the current concentration of productivity gains, monitoring the composition of labor within those key sectors—retail, healthcare, and professional services—becomes paramount. The question isn't just whether productivity will continue to rise, but where those gains will come from, and whether they will be sustained.