Sunday, May 24, 2015

Is the slowdown in productivity growth a result of energy costs?

Slowing productivity growth in the United States has been in the news in recent months. It has become a concern to policymakers because they believe it is one of the primary contributors to a middle-class economic squeeze according to the annual report of the White House Council of Economic Advisors.

Simply put, productivity growth refers to the growth in economic output per worker or more precisely, per hour of work. When this growth slows, the potential for real wage increases diminishes since the growth in wages typically reflects the ability of workers to create more output per unit of time.

To the obstensibly naive observer the following idea may seem a plausible explanation: Higher-cost energy inputs into the production of goods and services reduce productivity growth because the economic output per dollar of energy consumed declines. And, though energy inputs aren't the only thing to consider, they are important. The high energy prices of the last decade or so may be, in part, responsible for low productivity growth. (Conversely, low energy costs would imply more output per dollar of energy consumed.)

But strangely, almost all economic models for productivity consider only so-called "tangible" factors, that is, labor and capital. In the bizarro world of modern economics, energy and materials are not considered "tangible."

Now, the way in which that productivity growth which is attributable to "technological advances" is typically calculated is to add up contributions to productivity growth from labor and capital (machines, buildings, vehicles, tools of any kind) and then subtract this sum from the known amount of total productivity growth. What is left is the so-called "residual" which is presumed to result from "technological advances" caused by increases in human knowledge. These advances and the increases in capital per worker are assumed to be the drivers of productivity growth.

Let me explain this from a slightly different angle: Obviously, if you work more hours, you will be more productive. But your output per hour will remain the same, barring some new input such as better, more efficient machines to work with or more efficient techniques, both resulting presumably from an increase in knowledge.

Note that there is no way to measure this "knowledge factor" directly. It is merely assumed that the unknown portion of productivity growth comes from "technological advancement."

But, energy researchers asked long ago whether productivity growth might be affected by changes in the quality and cost of energy inputs. Authors of a paper entitled "Energy and the U.S. Economy: A Biophysical Perspective" which appeared in Science in August 1984 noted the tight correlation between economic growth and energy consumption. They also noted that labor productivity increased with increasing energy consumption per employee. While not dismissing the effects of technological change, they believe that energy has had a central role in the persistent rise in labor productivity witnessed for most of the last century up to the time of publication:

From an energy perspective, productivity gains are facilitated by technical advances that enable laborers to empower their efforts with greater quantities of high-quality fuel embodied in and used by capital structures.

Notice the use of the term "embodied." The researchers recognized the energy necessary to produce the capital equipment used by workers. This is called the "embodied energy." The researchers also noted the following:

We found that in the U.S. manufacturing sector, output per worker-hour is closely related to the quantity of fuel used per worker-hour. A similar relation exists in the U.S. agricultural industry.

The mining sector also fits this pattern. While productivity per worker-hour has increased or, in some cases, merely stayed flat, the energy data showed just how much more energy was needed to achieve stable or growing productivity:

Technical improvements in the extractive sectors have made available previously uneconomic deposits only at the expense of more energy-intensive forms of capital and labor inputs. Physical output per kilocalorie of direct fuel input in the U.S. metal mining industries has declined 60 percent since 1939, although a few exceptions to that trend are known. The energy cost per ton of metal at the mine mouth for industrially important metals such as copper, aluminum, and iron has risen sharply as their average grade declined. For all U.S. mining industries (including fossil fuels), output per unit input of direct fuel declined 30 percent since 1939.

These findings suggest that fuel costs, fuel quality and fuel availability can be limiting factors in productivity across the economy. The idea that energy inputs used in production are central to productivity isn't so counterintuitive after all. And yet, in a sampling of recent coverage of the productivity issue, not one piece mentioned energy. (See here, here, here and here.)

One of the authors of the research cited above, Charles A. S. Hall (now retired), says that the report's findings need to be updated to see whether the relationships his team discovered still hold. It would seem wise to follow up given the exceptionally slow productivity growth associated with the period of rising energy prices before the crash in 2008 and to a certain extent with high average daily oil prices from 2011 through late 2014 (though, as one might expect, this is not even mentioned as a possible explanation in the piece cited.)

Whether such updated research would confirm the original findings can't be known. Whether it would make any difference to mainstream productivity models is known. Such new findings will make no difference whatsoever until the economics profession recognizes the central role of energy in the productivity of the workforce.

Kurt Cobb is an author, speaker, and columnist focusing on energy and the environment. He is a regular contributor to the Energy Voices section of The Christian Science Monitor and author of the peak-oil-themed novel Prelude. In addition, he has written columns for the Paris-based science news site Scitizen, and his work has been featured on Energy Bulletin (now, The Oil Drum,, Econ Matters, Peak Oil Review, 321energy, Common Dreams, Le Monde Diplomatique and many other sites. He maintains a blog called Resource Insights and can be contacted at


Nils Peterson said...

How about a different mechanism than price to achieve similar end?

Diminishing returns.

Regardless of price of a commodity (which we have seen fluctuate a lot for oil recently) with diminishing returns, proportionally more resources (labor and capital) will have to be devoted to obtaining the commodity. This will take labor and capital away from other segments of the market.

The Arthurian said...

Via Reddit.

"But strangely, almost all economic models for productivity consider only so-called 'tangible' factors, that is, labor and capital.... [N]ew findings will make no difference whatsoever until the economics profession recognizes the central role of energy in the productivity of the workforce."

A couple links that seem relevant. Maybe you've not seen them.

1. Bill C at Twenty Cent Paradigms

Bill C links to the original 1928 article where the Cobb-Douglas production function was introduced.

From the conclusion of the Cobb-Douglas PDF:

"Finally, we should ultimately look forward toward including the third factor of natural resources in our equations ...."

(The "third factor" completing AdamSmith's trilogy of land, labor and capital.)

And, knowing that Solow began with a production function of the Cobb-Douglas type to develop his growth model,

2. Sandwichman at Econospeak, on Solow's response to The Limits to Growth

"The core of Solow's argument was an analogy between labor productivity and natural resource productivity..."

Shawn B said...

This issue of modern economic models not including resources and specifically energy as a critical factor has puzzled me. I would think it general consensus that much of past politics and war was driven by the need for land and resources. Oil was a factor at least in some strategic decisions leading up to WWI, the U.S. powered the world through WWII with our oil resources, and Germany fell in part because of the lack thereof. The Iraq wars were at least in part if not entirely about oil. It is generally well understood what happens to the world economy if Strait of Hormuz or another major energy shipping route is ever blocked via war etc. Even some current unsophisticated politicians intuitively get the importance of oil to the economy, as evidenced by their admonition to drill-baby-drill. So how is something so glaringly obviously important for an economy essentially left out of economic models? I suppose one could point to a Japan or South Korea as examples of economies that grew without benefit of major domestic oil and gas resources. But exceptions don’t make the rule, and in any case my bet is the story of oil and gas use in these countries would in the end support the idea that energy is still central to productivity and economic growth. So I am back to my puzzle again. Why is energy left out of most economic models and mainstream economic discussion on productivity? A rhetorical question I suppose, but I am curious about the answer.

casamurphy said...

The author of, Gail Tverberg, has a great deal to say on this subject...

Thierry Caminel said...


FYI, an interesting study has been published recently about "How Dependent is Growth from Primary Energy ? Dependency ratio of Energy in 33 Countries", by Gaƫl Giraud and Zeynep Kahraman



Shawn B said...

Regarding this study cited in the comment above, "How Dependent is Growth from Primary Energy ?

Interesting study indeed. Worth reading the first 4-5 pages at least.

With advance apologies to any offended, taking out the economic and statistical nomenclature, my civilian (ok, amateur) paraphrase of this study is:

- “….primary energy consumption….causes GDP Growth”
- Energy consumption drives technology use/capital formation, not the reverse
- But efficiency does matter. A lot.

Duh-huh stuff for many readers of Kurt Cobb’s blog, I would suspect. I wonder what the mainstream economists would say?