Greater use of robots could generate new jobs.

Photographer: Ricardo Azoury/Getty Images

Slow-Growth Pessimists Miss the Mark

A. Gary Shilling is president of A. Gary Shilling & Co., a New Jersey consultancy, and author of “The Age of Deleveraging: Investment Strategies for a Decade of Slow Growth and Deflation.” Some portfolios he manages invest in currencies and commodities.
Read More.
a | A

I don't agree with the slow-growth-indefinitely forecasts that I detailed in Monday's column

I've long said that, until global deleveraging is completed, gross domestic product will continue to grow by about 2 percent annually. I've also noted that reducing debt levels after a financial crisis, especially one caused by a borrowing binge, normally takes about a decade. This episode is eight years old, and at the rate things are going, it may take longer than 10 years. U.S. household debt relative to after-tax income has fallen to 102 percent from 130 percent, but it's still a long way from the 65 percent norm.

Nevertheless, the process will end at some point, and I continue to believe it will be followed by rapid real economic growth of at least 3.5 percent a year. That growth will no doubt be fueled by today's new technologies, including computers, the Internet, biotech, telecom, semiconductors, robotics and 3-D printers. 

Note that the Industrial Revolution and railroads started in the late 1700s and grew explosively, but from zero starting points. It was only after the Civil War that they became big enough to drive the U.S. economy. 

QuickTake Secular Stagnation

Productivity is a complex phenomenon. It comes in waves that are often unrelated to the current economic situation, so recent weakness shouldn't be extrapolated. Even in the Depression-era 1930s, output per hour rose at a healthy 2.4 percent annual rate, higher than the 2.1 percent growth of the Roaring '20s. Many of the tech advances of the 1920s -- electrification of factories and homes, increased use of telephones, the dawn of the radio era -- weren't exploited until the following decade, when they became must-have conveniences. 

In the current setting, companies may have compensated for anemic revenue growth by cutting costs so much that productivity growth, which normally would have been spread out, was concentrated in 2009-2010. Recent productivity weakness probably indicates that companies have reached the bottom of the cost-cutting barrel. The leap in profit margins since the recession has also topped out in the last two years; corporate profits have now begun to decline. So businesses will no doubt turn from reducing costs to promoting productivity. Companies will probably redouble these efforts if wage increases push up unit labor costs. 

At the same time, companies will probably increase research-and-development spending and reinstitute labor-training programs as the supply of unemployed skilled workers shrinks. The need for trained workers will be enhanced by the growing use of robots, which generates jobs in design, engineering, maintenance, marketing and logistics. 

Solutions to the current crisis in higher education may end up promoting productivity, as well. Students and their tuition-paying parents now know that a college degree no longer guarantees a job that pays well. In this regard, two education developments are encouraging. First, many colleges are emphasizing degrees in the STEM (science, technology, engineering and math) fields, where jobs are waiting. Second, German manufacturers have transplanted their apprenticeship program to plants in the southeastern U.S., where they coordinate worker training with nearby community colleges. American businesses are beginning to copy this model. 

As for the Reinhart-Rogoff argument -- that high government debt depresses GDP -- it's probably the other way around. Slow economic growth depresses tax revenue and raises government social spending, thereby causing bigger deficits and debt levels. 

The argument that growth is stymied because companies have cut capital spending may also be missing a larger trend. Much of the spending to build new tech and social-media companies, such as Google and Facebook, isn't counted as capital outlays. The brains of the entrepreneurs and developers substitute for capital spending. A high-tech startup in a garage or a college dorm doesn't register in government data the way a new auto plant does. Many successful startups didn't need much more than a laptop, and most of the money they raised went to advertising and marketing, not building factories.

Fed economists believe that the pricing of high-tech equipment may result in understated business investment. Not surprisingly, capital equipment prices have fallen 21 percent since the early 1980s. Furthermore, the correlation between capital spending and productivity is, contrary to the belief of the slow-growth advocates, weak to nonexistent. This is shown by the correlations between private, fixed capital investment and productivity with a series of leads and lags.

From the first quarter of 1948 to the fourth quarter of 1990, the strongest relationship was between capital spending and productivity growth three quarters later, which makes sense. But the statistical fit, according to my firm's research, was very poor. Even that weak relationship broke down between the fourth quarter of 1990 and the first quarter of this year. The best fit was between productivity growth now and capital spending 16 quarters later, which defies any causal explanation.

So there doesn't appear to be any meaningful statistical relationship between capital spending and enhanced productivity. Spending money on more machines doesn't do the job, suggesting that productivity flows mainly from new technology such as robotics, better management, more motivated employees, better logistics and, probably, dumb luck. 

The pessimistic argument that American corporations have been buying back stocks instead of investing in plant and equipment carries some weight, but dividend increases still leave the payout ratio (the percentage of net income paid to shareholders) for the S&P 500 at 42 percent, well below the long-term 52 percent average. It can be argued that low interest rates make low dividend yields (dividend per share divided by price per share) acceptable. In any event, the dividend yield for the S&P 500 index is now just 1.97 percent, well below the earlier 3 percent norm. 

True, government regulation is excessive, as the slow-growth advocates maintain. Since 1970, more than half of Americans have relied on government for meaningful income; in 2007, it was 58 percent, according to my firm's research. Yet voters haven't used the ballot box to accelerate their government goodies. 

Apparently, Americans still believe they can get further on their own merit than by pushing government to redistribute income in their favor. And if I'm right about a coming economic boom, they will have even less reason to rely on government largess. 

Previously: The (Inadequate) Case for Economic Pessimism

This column does not necessarily reflect the opinion of Bloomberg View's editorial board or Bloomberg LP, its owners and investors.

To contact the author on this story:
A Gary Shilling at insight@agaryshilling.com

To contact the editor on this story:
Paula Dwyer at pdwyer11@bloomberg.net