Who Is Reaping the AI Dividend? U.S. Labor Compensation Share of GDP Hits Lowest Level Since 1947 in Q2

U.S. labor productivity in the second quarter grew more than expected, indicating that businesses are managing to produce more output with fewer working hours amid rising costs. However, the gains from increased productivity have not fully reached employees—the share of labor compensation in economic output has fallen to its lowest level since record-keeping began in 1947.

The U.S. Bureau of Labor Statistics (BLS) reported on Thursday (June 6) that second-quarter nonfarm labor productivity—output per employee per hour—rose at an annualized rate of 1.4%, surpassing the revised first-quarter growth of 0.8% and exceeding all but one economist's forecast in a Bloomberg survey.

The productivity increase mainly reflects a surge in business output, the largest since the third quarter of 2025, while work hours increased only moderately. Unit labor costs, which measure the wages businesses pay to produce each unit of output, rose at an annualized rate of 1.3% in the second quarter—below market expectations—indicating that wages are not exerting significant upward pressure on inflation.

Since labor costs are the largest expense for most businesses, improved production efficiency theoretically allows companies to raise wages without pushing up prices, which could help improve living standards over time. As such, Federal Reserve officials, investors, and economists are closely watching whether the billions of dollars companies have invested in AI are beginning to boost employee productivity.

Federal Reserve Chair Kevin Warsh said on July 15 during a Senate hearing that he believes AI will ultimately help ease price pressures, and that products and services touched by technology will become cheaper over the long term. Andrew Sacher, an economist at Bloomberg Economics, also noted that the post-pandemic productivity growth trend continues and may reflect early AI benefits. Combined with moderate unit labor costs, these data support the Fed holding steady and suggest the labor market is not fueling inflation.

However, it remains difficult to definitively confirm AI’s actual impact on productivity and employment, as official data fluctuate significantly from quarter to quarter, requiring more time to establish a clear trend. Some economists also worry that if businesses can increase output through AI and automation, they may further slow hiring or even reduce headcount.

Notably, the share of U.S. labor compensation in nominal gross domestic product (GDP) declined from 53.7% in the first quarter to 52.9% in the second quarter—the lowest level since records began in 1947. This indicator measures how much of economic output flows to workers via wages and benefits. Its long-term decline is linked to weakening union power, globalization shifting high-paying manufacturing jobs to low-cost countries, and advances in automation technology.

The rapid recent adoption of AI now enables companies to increase output without significantly expanding their workforce. This means the benefits of higher productivity are flowing more to business owners and shareholders rather than being passed on to workers as higher wages.

Real hourly earnings, adjusted for inflation, declined at an annualized rate of 3.1% in the second quarter—the largest drop since late 2022. Real weekly earnings in the first half of 2026 were largely flat. The data highlight that while U.S. economic efficiency continues to improve, workers are not necessarily sharing equally in the fruits of productivity growth.