JPM US AI and the labor

Michael Feroli ( North America Economic Research JPMORGAN JPMorgan Chase Bank NA 04 August 2026

US: AI and the labor share Figure 2: Labor share of income %, nonfinancial corporate sector

• The labor share of national income reached an all-time 68 low in 1Q26 66 64 • This appears to reflect a genuine decline in labor’s bar- 62 gaining power rather than measurement issues. 60 • Most economic models predict that the growing use of AI 58 should further depress the labor share. 56 54 The labor share is the share of national income that workers receive. Some AI-focused economists also refer to this as the Source: BLS, J.P. Morgan “human labor share” to indicate the share of income paid to carbon-based laborers. As we will discuss later, most models Even this labor share metric is not immune to measurement that attempt to understand the effects of AI on the labor mar- concerns. To take just one example, Erick Zwick has empha- ket predict that further deployment of AI should lead to a sized that the creation of the pass-through deduction in the decline in the labor share. 2017 tax act may have changed incentives regarding corpo- rate legal forms, and hence the measurement of labor income If realized, this would come from a very low jumping-off in the corporate sector. point, as last quarter the labor share reached an all-time low. In the non-farm business sector, only 53.7 cents of every dol- A different approach to this issue is to look at the net capital lar of value added went to workers. That's down about three share of income, which in principle should move in the oppo- cents from the year before the pandemic. Given recent trends site direction of the labor share of income (modulo taxes on and the expected effects of AI, it’s possible to see the labor production and imports, and some other small adjustments). share of income fall below 50% by the end of the decade. Within the nonfarm business and nonfinancial corporate sec- tors, these measures mirror the labor share, by construction. Some data preliminaries To get an independent look at the net capital share we turn to the profit margin measures in the Census Bureau’s Quarterly The most cited measure of the labor share relates to the non- Financial Reports (QFR). While the QFR dates to WWII, the farm business sector. That measure has been trending lower longest series relate to the manufacturing sector. This may not over the last few decades, and, as mentioned above, hit an all- be as big a limitation as it sounds, because their concept of time low in 1Q26 (Figure 1). In this note we don’t propose to manufacturing includes much of tech, energy, and health care, revisit the explanations for this trend lower, but some top can- along with industrials and consumer discretionary and staples. didate explanations are automation, increasing market con- In fact, when we compare the margin measures in the nonfi- centration, and globalization and offshoring. nancial corporate sector (BEA data) to the manufacturing sec- tor in the QFR (Figure 3), we see that they share similar Figure 1: Labor share of income cyclical and secular behavior. %, nonfarm business sector

67 Figure 3: Profit margins 65 Income as % of sales 63 18 Nonfinancial 61 corporations Manufacturing Source: BLS, J.P. Morgan (QFR, 4-qtr ma) 0 One of the major measurement concerns with this measure is Source: BEA, Census, J.P. Morgan how to allocate proprietors’ income between labor and capi- tal. One proposed solution to this issue is to focus on the non- Having established that the net capital share of sales in the financial corporate sector. As seen in figure 2, that has also QFR measure is a good proxy for the aggregate economy, we been drifting lower this century and hit an all-time low in turn to size classifications. As seen in figure 4, firms with 1Q26. assets over $1 billion have very similar trends to the overall

Michael Feroli ( North America Economic Research JPMORGAN US: AI and the labor share JPMorgan Chase Bank NA 04 August 2026

sector. We think this is relevant because the concerns about Figure 5: Rate of return on capital unallocated income, or about firms reclassifying their tax sta- %, nonfinancial…