Productivity Growth before and during the Pandemic

Featured in print Digest

Between 2010 and 2019, US productivity grew more slowly than in any other decade of the post-World War II era. The business sector grew by an average of 1.1 percent per year, less than half of the 2.5 percent average annual growth from 1950 to 2009. Analysts have struggled to explain this slowdown during the longest business cycle expansion in US history and why productivity growth in 2020 soared to 4.1 percent, a sudden surge during a year marked by the short but sharp pandemic recession.

In A New Interpretation of Productivity Growth Dynamics in the Prepandemic and Pandemic Era US Economy, 1950–2022 (NBER Working Paper 30267), Robert J. Gordon and Hassan Sayed challenge the widespread notion that deviations of productivity growth from its long-run trend result solely from autonomous “productivity shocks.” Rather, the researchers show that these deviations from trend, which they call “productivity gap changes,” are strongly procyclical. As growth in output fluctuates, due to such demand-driven components as consumer durables, fixed investment, and inventory investment, firms react by adding or cutting total hours of work — that is, total employment times weekly hours per job. But the hours response is partial and lags the change in output. Since productivity changes are by definition equal to output changes minus changes in hours of work, productivity changes surge upward in the current quarter as the counterpart of the slow response of labor hours and then fall back as labor hours gradually complete their lagged response.

Hiring recoveries following excess layoffs during the Great Recession and the pandemic led to slow productivity growth in both the 2010–19 and 2021–22 periods.

The researchers argue that the strong productivity growth experienced in 2009 and slow productivity growth during 2010–19 were both due to “excess layoffs” as firms panicked during the Great Recession and fired workers, and this was followed by a gradual pace of rehiring after 2009. Because the cuts were so large, it took nearly a decade for hiring to recover, which is a major reason productivity growth over the 2010–19 period was so weak. The researchers use a regression analysis to calculate that productivity growth in 2008–09 would have been negative 0.8 percent rather than the measured rate of positive 3.2 percent if employers had not laid off so many workers in 2009. Moreover, the 2010–16 labor recovery would have seen far more robust productivity growth of 1.7 percent rather than the actual 0.9 percent.

To understand the rebound in productivity growth in 2020, the researchers construct a quarterly database of productivity levels and changes for 17 industries that they sort into three groups: goods, work-from-home services, and contact services. They conclude that the official data substantially overstate aggregate productivity growth in the second quarter of 2020, when the recession occurred, and understate it in the third quarter because pandemic-driven lockdowns skewed the data by dramatically shifting the industry mix away from low-productivity contact services, like restaurants and hotels, and toward high-productivity work-from-home industries, such as information technology and financial services. Over the nine quarters between 2020 and early 2022, productivity growth in the goods industries was negligible, productivity in contact services declined at an average annual rate of 2.6 percent, and work-from-home industries accounted for all of the economy’s overall positive productivity growth achievement, posting a strong 3.3 percent growth rate. The researchers link the upsurge of efficiency in work-from-home industries to recent surveys showing self-assessments of greater efficiency and unmeasured extra work hours during time previously spent commuting.

The researchers develop a simulation of their 2007–19 regression analysis to estimate the effects of excess layoffs in 2020 and rebound rehiring in 2021–22. They project that rehiring will continue for several years at an annual rate that is about 2.5 percent per year faster than otherwise as industries rehire the employees they cut during the pandemic lockdowns. This rehiring phenomenon explains both why the US economy in 2022 has been characterized by positive employment growth despite negative GDP growth, and why productivity growth was negative in the first two quarters of 2022.

— Laurent Belsie