Growth Projections: Whatever Happened to Stanford's John Taylor?
First his declaration that it would be better for the U.S. economy if the U.S. were to default on its debt than to increase the debt ceiling without massive, immediate cuts in public spending...
And now this:
hy Not Go For 5% Growth?: Some skeptics have complained about the 5% national economic growth target [over 10 years] put forth by former Minnesota Governor Tim Pawlenty in his speech this week about his economic plan. They say it can’t be done. But I think the goal makes a great deal of sense. It would focus policymakers like a laser beam on the great benefits that come from higher growth and on the pro-growth policies needed to achieve it...
Note, first, Taylor's transformation of Pawlenty's 5% from a forecast of what Pawlenty's policies will accomplish to an aspirational goal.
And note that in the end Taylor doesn't get there: he only gets to 4.7%--and then rounds up:
Now if we add the 2.7 percent productivity growth to the 2 percent employment growth, we get 4.7 percent economic growth, which is within reaching distance of—or simply rounds up to—the 5 percent target set by Governor Pawlenty...
It matters: the difference between a 5.0% and a 4.7% growth rate over ten years is $700 billion in year-2021 GDP--and is a difference of a total of $1.4 trillion in tqx revenue over the next decade. Even that rounding already creates a $1.4 trillion rosy scenario.
And note how heavily Taylor has to put his thumb heavily on the scale to get to 4.7%.
Watch how he gets there.
[E]conomic growth equals employment growth plus productivity growth.... Currently the percentage of the working-age population (age 16 and over) that is actually working is very low at 58.4 percent. In the year 2000 it reached 64.7 percent, so that is at least a feasible number...
But it isn't a feasible number.
The American working-age adult population in 2020 will on average be older than it was in 2000--and older people are less likely to want to work. Americans will spend more time in school in 2020 than they did in 2000--and adults who are in school aren't working. We won't see 64.7% again--not without big changes like raising the Social Security retirement age a lot, which aren't on Pawlenty's menu. We might see 63%--which would give us 0.75% per year, not 1%, from a rising adult employment-to-population ratio.
Taylor goes on:
Adding in about 1 percent for population growth (from Census projections), gives employment growth of 2 percent per year...
But that that is not what the Census projections say.
The Census projection sees the 18-and-over population growing at 0.88% per year, and the 18-to-64 population growing at 0.23% per year.
So far we are about 0.6% per year below the Taylor-Pawlenty 5% per year target.
But there are more thumbs on the scale. Taylor goes on:
Since the productivity resurgence began around 1996, productivity growth in the United States has averaged 2.7 percent according to the Bureau of Labor Statistics. So numbers in that range are not pie in the sky...
Now I am a productivity optimist--I think that it is reasonable to say that we had a structural productivity break in the mid-1990s and thus that the 1995-present data (with its average labor productivity growth rate not of 2.7% but 2.5% per year) is a reasonable guide to what we might expect for the future if things go reasonably well.[1]
However, is that if you look at the productivity growth series you see that the bursts of productivity growth since 1980 come in the early stages of recoveries, when firms are reconfiguring their activities as demand grows without yet filling nearly as many vacancies as rising demand would lead one to expect. We saw this in 1983-85, and again in the early 1990s, the early 2000s, and at the end of the 2000s. This is a new part of the productivity growth pattern, and one that puzzles economists.
We just had such a transient early-recovery spike: 5 quarters during which productivity growth averaged 6.1% rather than 2.7%. We are unlikely to have such a spike again in the next ten years. Subtract out the excess productivity growth above trend of this spike and we have a forecast of labor productivity growth over the next ten years of not 2.5% per year but 2.2% per year. Now that is significantly better than the 1.7% per year of 1981-2011--or the 0.9% per year of 1981-1995. But it doesn't get you to 5% per year.
It gets you to 0.75% + 0.88% + 2.2% = 3.83% per year.
And that is an aspirational goal--not a central tendency projection or even a forecast of what would come to pass under any set of politically-feasible policies.
Now don't get me wrong--we might get 5% real GDP growth over the next decade. I hope that we do. But budget planning should not be based on hope, for hope is not a plan.
And the differences between the Pawlenty-Taylor 5% growth rate and my optimistic-aspirational 3.8% growth rate and my realistic 3.5% growth rate are huge. The swing is between a year-2011 real GDP number of $24.4 trillion and of $21.1 trillion. The corresponding difference in the government's tax revenue in 2021 is fully $1 trillion.
The Pawlenty-Taylor scenario thus bakes a $1.2 trillion rosy scenario into its forecasts of the total government deficit in 2021, and a $6 trillion rosy scenario into its forecasts of the total government debt in 2011.
To assume an extra $6 trillion of resources that the government can use over the next decade--that's not budgeting we can believe in.
[1] Do note, however, that most Republicans would claim that the structural break came not in 1995 but in 1981 with the inauguration of Ronald Reagan--which gives us a productivity growth rate average not of 2.5% per year but rather of 1.7% per year. In which case if you add up 0.75% + 0.88% + 1.7% = 3.33% you get not 5% per year, but only 2/3 of that.
