Is Slow Still the New Normal for GDP Growth?

new John Fernald (FRBSF) on the impact of demographics and productivity on GDP growth

"....The solid growth in GDP over the past year naturally raises the question of whether we have been understating the potential trend. Nevertheless, it is far too early to declare that trend growth is higher than current estimates of 1.6-1.7%. The most natural interpretation of the strength in hours worked is that any structural element represents an effect on the levels of employment, hours, and potential output—perhaps in response to the TCJA—not a persistent growth effect.Policy changes such as family-friendly policies common in other OECD economies might lead to trend increases in labor force participation (Daly et al. 2018). But such policies are unlikely to be implemented soon.....slow pace relative to history reflects two factors. Most importantly, demographic trends leave little doubt that the labor force will grow slowly at best over the next decade. There is more uncertainty about the other factor, productivity growth. Our forecast assumes that productivity will grow at a pace similar to recent decades. Averting this slow-growth future would require a sustained increase in productivity growth..... Demographics are the key reason future trend growth is expected to be low relative to history. Figure 1 shows that labor force growth has slowed markedly since the 1970s, when baby boomers reached working age and female participation in the labor force rose rapidly. However, stagnating growth in female participation since the 1990s combined with projections for birth rates, immigration, and mortality suggest growth of the traditional working-age population ages 16-64 is slowing to a historically low pace.....Figure 2 shows growth of real GDP per hour, a broad measure of labor productivity, broken into periods since 1973 that reflect variation in productivity growth. Figure 2 depicts how productivity growth has shifted between normal and exceptional periods (Fernald 2016). Unusually influential innovations—such as the steam engine, electric dynamo, internal combustion engine, and microprocessor—typically led to a host of complementary innovations that boosted productivity growth broadly for a time.....This measure of productivity growth has slowed since 2004. The green dashed line in Figure 2 is productivity growth averaged over 2004-2018, which is very close to the 1973-1995 average pace. During the fast-growth period from 1995 to 2004, productivity growth averaged 2½%. During the slower periods of 1973-1995 and 2004-2018, growth averaged only 1% to 1¼%, and that pace dropped dramatically lower most recently in 2010-2018....The black dot in the figure shows 2018 growth in real GDP per hour of 1.1%. This is higher than the average growth rate over 2010-2018 and may suggest a pickup in productivity growth. However, in our view, it is best to view the years since 2004 as a whole; in this case, the 2018 pace is close to the period average. We focus on the average since 2004 because labor productivity has some persistent cyclical dynamics. The large shock of the Great Recession caused a temporary boost in productivity growth from 2007 to 2010 that has since unwound. For one thing, less-skilled workers were more likely to lose jobs in the Great Recession than more-skilled workers, who typically have higher productivity. For another, businesses came out of the Great Recession with ample productive capacity but limited demand. Over time, increases in capacity per worker are an important source of productivity growth, but for a time after the recession, businesses could meet the recovery in demand without investing. Hence, that capacity channel was temporarily dialed down. Once we average over the transitory ups and downs, it appears that we have been in a slow growth regime for the past 15 years. Regimes generally appear persistent—lasting a decade or longer. For this reason, our most likely or “modal” estimate is that the slow-growth regime will continue for at least the next few years.That said, productivity is highly uncertain in both a statistical and an economic sense. Under the regime view of productivity growth, much of the statistical uncertainty is about which regime we will be in. Neither economists nor statisticians have a good track record of forecasting changes in trend productivity growth. In addition, even within a regime, productivity is inherently volatile from year to year. Economically, major technological gains could remain narrowly focused, leaving modest and incremental productivity growth as the norm for the broader economy, as we assume in our benchmark. Or, we might see another broad-based wave of the IT revolution. Alternatively, future growth might look substantially different from the past, reflecting the innovative contributions of robots and machine learning. Despite this inherent uncertainty, our best guess is that productivity growth over the next five to six years will be in line with previous slow regimes. During those regimes, GDP per hour rose around 1-1¼% per year...."

John Fernald and Huiyu Li, "Is Slow Still the New Normal for GDP Growth?," Federal Reserve Bank of San Francisco, June 24, 2019, https://www.frbsf.org/economic-research/publications/economic-letter/2019/june/is-slow-still-new-normal-for-gdp-growth/