# Oil as a function of GDP

In Mankiw, N. (2012). Principles of Macroeconomics (6th ed.)., on pg 456 Mankiw says "The amount of oil used to produce a unit of real GDP has declined about 40 percent since the OPEC shocks of the 1970s."

How would one measure that?

That's unfortunately just kind of dropped at the end of a chapter as the second to last sentence in a five paragraph side note about OPEC, and I've been curious for years now if that reduction was because the mix of energy supply changed (rise in natural gas consumption?), or production is simply more energy efficient, or what.

Can someone clarify the claim, the process, or point toward primary sources (US EIA maybe?), rereading this book has reminded me how frustrating that bit was.

I don't think it's any more complicated than looking at the ratio of global oil consumption and global Gross Domestic Product. The OECD, World Bank, IMF and others produce estimates of each.

There is some fuel-switching: for example, the world burns much less electricity from oil than it used to: in 1973 (the first oil crisis started around October that year) 251 million tonnes were used for electricity; by 2000, this had dropped to 110; by 2012, to 70 (table 7A on IV.81 of the IEA Electricity Information 2014).

There's several decades of energy efficiency, kickstarted in large part by the oil shocks of the 1970s.

The share of services (relative to physical goods) within global GDP has increased; IIRC services require much lower energy inputs than goods, on average, even though transport is one of the services (I'll check this and edit later - please nudge me in comments if I haven't after a week or so). The World Bank only gives services as a share of global GDP from 1990 onwards: this share grew from 60.7% in 1990 to 70.2% in 2011.

The combination of energy efficiency and the increase in the share of services (relative to physical goods) within global GDP, means that energy consumption per unit of global GDP has fallen: J. Bradford DeLong gives global GDP at \$12 trillion for 1970 and \$41 trillion for 2000 (preferred measure, 1990 international dollars). The BP Statistical Review of World Energy 2014 gives 1970 energy consumption at 4.9 billion tonnes of oil equivalent (btoe) and 9.3 btoe for 2000. So energy per unit global GDP almost halved (44% reduction) over those thirty years 1970-2000.

Understanding an individual country's changes are harder, because we then have to unpick whether oil intensity of GDP has dropped because of energy-efficiency & fuel-switching, or because oil-intensive goods are now imported rather than produced domestically: for further reading on this, see papers by John Barrett and by Dieter Helm.

• I was kind of hoping you'd answer. :) Looking at it as world makes more sense. I had always assumed he meant the domestic (US) ratio because the excerpt is about the effects of the oil shocks on the US economy. But i guess it was left intentionally vague because global seems much easier to quantify, especially for what was essentially a throw away line. – Jason Nichols Jan 2 '15 at 15:39
• Oh, thank you! Yes, quantifying the national picture is harder because of the offshoring of the manufacture of oil-intensive / energy-intensive goods. – 410 gone Jan 2 '15 at 16:12