Economists who treat economic progress as a matter of bookkeeping rather than human action have a bad habit of treating "productivity" as a mysterious deus ex machina that causes economic growth to slow down or speed up.
By extrapolating the past 10 years' dismal productivity growth and labor force participation into the next 10 years, the Congressional Budget Office (CBO) concludes the U.S. economy can't possibly grow faster than 1.8% a year. Since the economy is stubbornly outgrowing its lowball forecasts, the CBO felt obliged to lower GDP "projections" to 1.5-1.6% in 2021-2023 just to keep the 2018-2028 average under their magical 1.8% lid. This apparently makes sense to one prominent economist.
"Even 2% Growth Will Be Hard to Sustain" was the headline of a Feb. 14 Wall Street Journal article by the former chairman of President Obama's Council of Economic Advisers, Jason Furman. Real GDP per hour worked, he noted, rose at a "1% annual pace over the past decade. If that average continues, overall economic growth in coming years will average only 1.5%" — assuming (as he does) the labor force grows at the slow 0.5% pace it did for the past 10 years.
To declare that the next 10 years will be like the last 10 years does not require a sophisticated economic model. It's just naive extrapolation. And it conveniently blames the disappointing Obama recovery on unlucky productivity and labor force trends.
The most optimistic alternative scenario Mr. Furman can bring himself to imagine is a growth rate of 2.1%, leaving us stuck somewhere between 1.5% and 2.1%. The CBO's 1.8% projection is right in the middle of that miserable range, for similar reasons.
Growth, CBO Style
The CBO's 10-year projection of 1.8% growth is critically important. If economic growth is even 1 percentage point higher, the CBO is grossly exaggerating long-term budget deficits. That is not only possible, it is typical of CBO forecast errors after income or capital gains tax rates were reduced. From 1983 to 2000, the CBO's two-year forecasts of real GDP growth were repeatedly 1 percentage point too low, on average.
We are not at the mercy of blind fate or past trends. If business output grows more quickly, output per hour will too. Productivity in the nonfarm business sector rose by 3% a year from 1996 to 2005 because business output rose by 3.9% a year. Economic growth raises productivity, not the other way around.
Nonfarm business productivity slowed to only 1.2% a year from 2007 to 2017 mainly because business output grew by only 1.7% a year, but also because hours worked grew by only 0.4%. It is this unusually feeble "trend" in productivity over the past 10 years that Furman and the CBO ask us to project forward for the next 10 years.
Furman expands the concept to total GDP per hour, but his suggestion that we're doomed to repeat the 2007-2017 aberration (which would require another Great Recession) still fails. The OECD calculates U.S. real GDP per hour in 2010 dollars. Real GDP per hour rose from $62.35 in 1984 to $93.83 in 2007 — a gain of 2.3% a year. Since then, as Furman notes, the increase slowed abruptly to just 1% a year.
Why should we expect that the unusually weak productivity gains since 2007 constitute an inexorable trend? That same question applies to Furman's projection that the labor force will also keep rising at the depressed 0.5% rate of 2007-2017, when there was unprecedented shrinkage in the number of working-age Americans willing to work. Furman speaks of a "steady decline in labor force participation ... since around 2000," but that is misleading.
Where Did The Workers Go?
It was no surprise that the labor force participation rate dipped slightly from a record high of 67.1% at the 2000 peak to 66 in the 2008 recession. The mystery is why it kept falling after the recession to 62.7% by 2015 despite falling unemployment and rising wages.
A new Mercatus Center study finds the rising number of young male labor force dropouts appears "connected to the rising accessibility and generosity of government benefits." Similarly, The Redistribution Recession by University of Chicago economist Casey Mulligan identified many new taxes, regulations and benefits, including ObamaCare, that "reduced incentives for people to work."
If labor force dropouts can be incentivized to return to the workforce, the labor force could rise substantially faster than the working-age population. Average hours can also grow faster than the labor force if involuntary part-timers get full-time work.
Adding the unusually low productivity growth of the past 10 years to the unusually low labor force growth is a fine way to predict that past, but it's not a serious way to predict the future.
The defective reasoning that leads Jason Furman and the Congressional Budget Office to "project" a decade of 1.5%-1.8% economic growth turns out to be an excellent reason for optimism, because it implies that the CBO is grossly underestimating long-term tax revenues and (depending on spending) future budget deficits.