Potential GDP growth is widely estimated to have fallen sharply since the late 1990s and early 2000s, as seen in Chart 1, which shows the CBO’s estimate.
▪ Note that, as we have emphasized, potential growth can deviate materially in the short run from its longer-run projected rate.
▪ For example, CBO estimates that potential growth now is about 1½%. That’s why we say that 2% growth is cyclically “good,” meaning above potential and hence consistent with diminishing slack.
▪ However, CBO projects that potential growth will edge up after 2017, reaching 1.9% in 2020 and then remaining close to 2%, at least through 2026. That’s why we refer to potential growth in the long run as 2%.
While projected potential GDP growth of 2% would already be low relative to historical experience, recent research by staff at the San Francisco Fed suggests that even 2% may be too high.1
▪ John Fernald, a leading scholar on this subject and a senior research adviser at the Federal Reserve Bank of San Francisco, projects that potential output will grow at just a 1½%-to-1¾% rate over the longer run, with a point estimate of just 1.6%.
1 See John G. Fernald. 2016. “Reassessing Longer-Run U.S. Growth: How Low?” Federal Reserve Bank of San Francisco Working Paper 2016-18. http://www.frbsf.org/economic-research/files/wp2016-18.pdf and FedViews, December 8, 2016, Economic Research Department, Federal Reserve Bank of San Francisco.
▪ Chart 2 shows average GDP growth for several postwar periods roughly corresponding to periods of different potential growth rates; the gray-shaded region shows Fernald’s range for projected potential growth over the longer run.
A slower pace of hours growth associated with the aging of the population accounts for much of the slowdown in projected potential GDP growth.
▪ Using CBO’s labor force projections, Fernald estimates that hours worked will grow at a rate of only about ½% per year over the next decade or so.
▪ This is sharply slower hours growth than in the past. Fernald points out that, if hours growth had been that low from 1973 to 1995, GDP growth would have averaged just 1¾%, not nearly 3%.
▪ So even if productivity returns to its 1973-1995 pace of 1¼%, hours growth of the only ½% would limit GDP growth to 1¾%.
But Fernald suggests that even projected potential GDP growth of 1¾% may be too high since there is good reason to expect a slowdown in the rate of improvement in labor quality (reflecting education and experience), not just in hours growth.
▪ Fernald cites estimates that improving labor quality has added about 0.4 pp to annual productivity growth since 1973.
▪ If that rate were to be sustained, and productivity growth net of labor quality were to return to its 1973-1995 pace, potential growth would be at about the upper end of his 1½%-1¾% range.
▪ However, he cites estimates suggesting that, by early in the next decade, labor quality may contribute only 0.1 to 0.2 pp to annual productivity growth.
▪ That would slow potential GDP growth by at least 0.2 pp a year, and maybe more, taking it to the lower end of his projected range.
The pace of longer-run potential growth is important to monetary policymakers because it importantly affects the level of r-star, the long-run equilibrium real funds rate, the anchor for the fund’s rate, and indeed for all interest rates.
▪ A low level of r-star increases the prospect that monetary policymakers will find themselves constrained by the effective lower bound again in future recessions.
▪ FOMC participants have therefore provided (qualified) support for supply-side or growth-oriented policies, a focus of Trump’s economic proposals, which, by raising potential growth, would also raise r star and improve the effectiveness of monetary policy in combating adverse demand shocks.
▪ However, as Fernald highlights, there are major impediments to significantly higher potential GDP growth, and the underlying sources are not easy to change.