Policies of Scale: Efficient Global Policy

Why Energy Demand Grows with the Economy

Fellow Economonitor writer Ed Dolan posted some very good questions on a piece I wrote on the relationship between energy and agriculture that I want to address. His first question pertained to a broad statement I made about how we must assume energy demand will grow so long as the economy expands. He wondered if I made this statement because energy demand is insensitive to prices or because I predict prices will not rise. I will address this issue now, and follow up with an additional post on Dolan’s remaining questions, which pertain more specifically to the relationship between energy and agriculture, in the coming weeks. On the scale of energy as a single sector, both energy demand inelasticity and price levels may well play a role, but my statement rests more on our ability to create more wealth from less energy and the link between employment and energy consumption.

In general, the price of energy is inelastic in the US. M.A. Bernstein and J. Griffen of the RAND Corporation looked at regional differences in price elasticity of energy in 2006 and commented broadly that:

“[T]he relationship between demand and price is small. That is, demand is relatively inelastic to price. We also found that in the past 20 years, this relationship has not changed significantly; analyses performed in the 1980s showed approximately the same results. These findings might imply that there are few options available to the consumer in response to changes in the price of energy, and that price does not respond much to changes in demand.”

They also note that prices have not risen:

“On the other hand, because prices were declining in real terms over most of the period we studied, the inelasticity of demand may be more of an artifact of the lack of price increases.”

I’ll come back to these statements latter in the post, but these findings are pretty amazing when one considers changes in energy demand over the last twenty years: massive increases in highly energy-intensive technology, massive increases in gasoline prices, and substantial increases in air-conditioning loads have all contributed to a steady rise in demand:

Nonetheless, we can establish some elasticity by looking at particular sectors. As one example, this chart shows how residential electricity prices and consumption growth are generally correlated:

This correlation is particularly pronounced because residential electricity includes both a baseline, constant requirement and additional “luxury” usage. Everyone must keep the refrigerator running, but not everyone has a 250 bottle wine cooler or 7,500 square feet to air condition. Plus, residential consumers see the bill directly. The cost of energy in consumer goods, however, is hidden. Not surprisingly, the higher the household income, the more electricity it uses:

We also know that energy is one of the more volatile components of consumer prices, as Dolan himself has noted recently. Energy prices can move more frequently with more dramatic volatility than just about any other CPI component , although this is in part because oil, a major part of the energy component, is beholden to the global market while other goods are relatively more isolated by the domestic market.

One consumer price for energy is hard to formulate because the components of our “energy mix” are quite varied by sector, and are consumed at different rates by different sectors:

We see more stable household and commercial consumption relative to that of industry, a sector hard hit early in the recession when construction and manufacturing took massive hits.

Thus, creating one consumer price is difficult. However, here are two estimations of an energy consumer price index:

This graph and the one before it help illustrate how energy consumption can still correlate with negative economic growth in industrialized countries, with dips in all the indexes following 2008-2009 recession.

The relationship between energy per dollar of GDP and energy prices is also influenced by population, which influences demand for travel, housing, consumer goods, and services. In per capita terms, consumption is declining due to an extended economic recovery and improving energy efficiency. According to Energy Information Agency projections, population increases by 25% between 2010 and 2035, but energy growth rises at only 10%, while per capita use declines at an annual average of 0.5% per year from 2010 to 2035. Energy use per 2005 dollar of GDP declines by 42% from 2010 to 2035 as a result of a continued shift from manufacturing to services (and even within manufacturing to less energy-intensive industries), rising energy prices, and the adoption of policies that promote energy efficiency.

Essentially, over the last sixty-plus years our ability to convert energy into wealth has improved, and we’ve used increasingly less of it, improving our energy intensity. Looking forward, our energy consumption is expected to bottom out around 2013 and start rising well out to 2035. Nevertheless, our energy intensity is expected to improve, and our electricity prices are estimated to increase over the next twenty five years, but by no more than half a cent per kilowatt hour:

I do not predict energy prices will remain low, but they will rise slowly and gradually. Every projection I have seen for the domestic natural gas market suggests that prices will rise in the next 10-25 years, though not dramatically and not until 5-10 years from now. And I have seen no reasonable prediction that oil prices will drop markedly from current levels; most predict prices will rise indefinitely.

Of course, our future consumption of natural gas depends on our consumption of coal, and vice versa. Both are used domestically for many of the same purposes, and regulation and the relationship between investment and prices will have much to do with which source supplies more of our energy needs. Shale natural gas requires significant regular investment to ensure profitability at America’s low prices relative to those of other industrial economies. Further, significant distribution infrastructure is needed to move natural gas outside the the one to two state radious of production. Improvements in technology made shale a profitable  investment, but producers are dumping far too much gas on the market, overshooting demand:

Here’s the bottom line:

Notice in the above graph that the only sources of energy pegged to meet growing demand are liquid biofuels and other renewable sources. This is in part because government regulations and subsidies are propelling the renewable sector forward, but it is also because the cost of energy from coal and natural gas are expected to rise:

Bernstein and Griffen substantiate inelastic energy demand since 1980 but caveat that this may be because prices did not increase over the same period. Looking forward, prices and consumption are expected to begin rising, so the inelasticity will be put to the test in the coming decades. I predicate my suggestion that we can assume energy demand will grow so long as the economy expands on two reasons: (1) energy’s relatively inelastic demand is a function more of our improving energy intensity than rising prices, and (2) employment growth will lead to greater energy demand.

Generally, economic growth and improvements in material standards of living are positively correlated to energy consumption; as economies develop and grow, per capita and aggregate energy consumption rise. In the US, for example, between 1890 and 1999, both GDP and energy consumption rose by factors of ten. This correlation only holds for so much economic growth, however. In low and mid-income countries, energy demand closely tracks economic growth. In high income countries, however, GDP growth can be sustained without concomitant energy demand growth. Some interesting charts from the IMF demonstrating this can be found here.

The degree to which the two are coupled is measured by energy intensity, the amount of energy it takes to produce one unit of GDP. For reference, here are the OECD energy intensity and World Bank GDP annual growth rates for the US:


1971 1990 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
EI 0.41 0.27 0.23 0.23 0.22 0.22 0.22 0.21 0.21 0.21 0.20 0.20 0.19 0.19
GDP % 3.5 1.9 4.9 4.2 1.1 1.8 2.6 3.5 3.1 2.7 1.9 -0.4 -3.5 3.0


We see here that energy intensity does not correlate with economic growth, but improves regardless of economic growth. Going forward, energy efficiency is expected to improve:

Thus, our ability to survive higher energy prices is boosted by efficiency gains that allow our marginal cost of energy to decrease. On top of this, the link between employment and energy consumption suggests that as we add jobs, our consumption will rise:

The background of my thinking that we must assume energy demand will grow so long as the economy expands is that rising consumption with improving efficiency means greater demand despite rising prices, keeping energy a relatively inelastic good.

One Response to “Why Energy Demand Grows with the Economy”

Ed Dolan EdDolanSeptember 28th, 2012 at 6:31 pm

Some good charts and data here, and even more interesting ones in the references.

For example, the referenced World Bank data show US energy per dollar of GDP falling by 13 percent just from 2002 to 2010. Over the same period the CPI for energy increased by 66 percent while the CPI all items increased by just 21 percent. On a back-of-the-envelope basis, that shows an elasticity of demand for energy that is inelastic, in the sense the term is used by economists, namely, the percentage change in quantity is less than the percentage change in price. However, they are definitely consistent with the hypothesis that price matters.

The referenced IMF charts show a stronger linkage of energy to GDP for developing than for advanced countries. The world bank data also show that developing and emerging market countries are significantly less energy efficient than advanced countries. That is in part because of their output mix, but also in part because many developing countries retain heavy subsidies for consumer use of energy.

On the whole, I think the data give ground for hope that more realistic pricing policies for energy (end of subsidies, full-cost pricing to capture externalities) will help to break the historically strong linkage between energy and GDP growth–a relationship that already appears to be substantially weakened for the United States and even more so for countries like Norway, Japan, etc.