Do we have potential?: An analysis of US potential economic growth

Summary

Labor productivity is one of the primary determinants of an economy's potential rate of economic growth. Productivity has waxed and waned over the years, but it is considerably slower today than it was during the halcyon economic years of the 1950s and early 1960s.

Labor productivity is determined by growth in the capital stock, changes in labor “composition” (i.e., labor quality) and changes in total factor productivity (TFP, i.e., changes in technology and other processes). Growth in the capital stock and changes in TFP have largely dictated productivity growth in recent years. We focus on growth in the capital stock in this report.

Over the past expansion (2010-2019), total labor productivity growth in the business sector slowed relative to the 1990s and the early years of the 21st century. Productivity growth was slower in the service sector, and especially in the manufacturing sector, partly due to slower rates of capital accumulation.

The economy's capital stock reflects the value of capital assets (i.e., structures, equipment and intellectual property products) at a specific point in time, and its change is determined by business fixed investment spending less depreciation. Growth in the economy's capital stock, and hence its productivity growth, downshifted after the tech bubble burst in the early 2000s.

Capital stock growth picked up in the factory sector, and more notably in the service sector, coming out of the global financial crisis. Yet growth in labor productivity remained lackluster compared to the 1992-2007 period, implying weak TFP growth during the 2010s.

Looking forward, we see the build-out in capital required for the widespread adoption of automation and artificial intelligence (AI) as a major upside for labor productivity. The necessary accumulation of hardware and software to use AI effectively may cause capital accumulation growth to strengthen and could lead to stronger productivity growth.

Reliable estimates of the volume of hardware and software investment required to support the AI transition are not readily available. Using a scenario analysis, we estimate capital accumulation can account for 1.7 percentage points of the annual rate of productivity growth by the end of the decade, which represents a six-tenthsimprovement from the past decade's average contribution.

Ultimately, capital accumulation is just one part of the productivity puzzle. The lackluster trend in labor productivity growth over the past expansion, despite the acceleration in the capital stock, implies weak TFP growth. While the AI build-out across the business sector will require significant capital investment in hardware and software, it will also likely come with a rise in efficiency and improvements to other processes that manifest in total factor productivity gains, a topic to which we will turn in the next installment of this series.

Download The Full Special Commentary