The Energy Coefficient of the United Kingdom

Total primary energy consumption in the United Kingdom has fallen more or less steadily for a decade. This makes climate change targets seem more attainable, but does it bring problems of its own, or is it to be welcomed as a further stage of economic maturation?

In 1979 William S. Humphrey and Joe Stanislaw, then at the Energy Research Group of the Cavendish Laboratory in Cambridge, published a striking study of the relation between economic growth and energy consumption.[1] The paper receives regular passing references but it cannot be said to have established a new paradigm for the study of energy intensity. Since energy consumption in the United Kingdom, and perhaps other developed countries, seems to be entering a new and historically very unusual phase it may deserve a close re-reading.Central in Humphrey and Stanislaw’s methodology is the calculation of an “energy coefficient’, which is defined as the “percentage annual growth of energy consumption divided by the percentage annual growth in real national output”. Using data from the then standard sources the paper presents coefficients for the UK in various periods, roughly decadal, between 1800 and 1975. The coefficients vary considerably, with a peak of about 2 in 1840-1850 and a low of about 0.35 in 1930 to 1938. Analysis suggested to the authors that higher coefficients were associated with periods of capital investment, the high coefficient in the 1840s, for example, being the result of iron, steel and railway investment (about half of gross domestic fixed capital formation in that decade was committed to transport).By contrast, between 1880 and 1914, the energy coefficient fell “sharply” to about 0.62. The causes of this change appear to have been a “transition to a more developed and less energy intensive economic structure”, and a “major improvement in the efficiency in which coal was used”. In the period 1914 to 1950 Humphrey and Stanislaw acknowledge data quality problems, and only report coefficients with caution, commenting that the result for the period 1925 to 1938, 0.3, does appear “rather low”. However, they grant that it may be explicable, for three reasons:

• Further changes in economic structure away from energy intensive industries and continued increases in thermal efficiency still probably had a downward impact;
• It was a period which saw a strikingly small flow of resources to capital formation, and we have seen earlier that periods of high investment tend to be associated with high coefficients and
• The mass unemployment and depressed level of economic activity in the 1930s may have induced greater efforts to conserve fuel in industry and in households which could no longer afford it, causing more energy saving than would otherwise have occurred.

In the period 1950 to 1975, the paper reports that the energy coefficient was “low by historical standards”, and suggest that this is to be explained by the substitution of oil and then gas for coal, both of which were associated with the introduction of new capital equipment with higher thermal efficiencies.Overall, and after making some useful comparisons with developed and developing countries, Humphrey and Stanislaw conclude that “UK experience appears to endorse the conclusion that periods of industrialization, involving rapid structural change in the pattern of output and, more important, the capital stock, are accompanied by relatively high growth in energy consumption in relation to GNP”.What of more recent years? I do not intend to calculate energy coefficients in this post, but it appears from a glance at the energy consumption data and the GDP data that coefficients for recent years would be historically anomalous, since the economy continues to grow, in spite of exceptions such as 2009, when the economy shrank by 5.2% (see ONS data), yet energy consumption over the last decade exhibits a strong downward trend.Indeed, taking DECC’s estimates of annual inland energy consumption on a primary fuel basis, UK energy consumption in 2015 of 195 million tonnes of oil equivalent (mtoe) represents a fall of 18% on the 2005 figure of of 236 mtoe, the second highest peak in the period since 1995. Over the same period GDP has, in spite of the 2008/9 crash, risen by somewhat over 10%.What is the significance of a very low or even negative energy coefficient over a period as long as a decade? Major shifts in economic structure combined with improvements in thermal efficiencies? Low levels of capital formation? Fundamentally depressed or passive economic activity? Something unusual is happening, something without exact precedent in the 275 year period studied by Humphrey and Stanislaw. This is not necessarily alarming, but it surely deserves serious examination. Is anyone doing this work?[1] William S. Humphrey and Joe Stanislaw, “Economic growth and energy consumption in the UK, 1700–1975”, Energy Policy 7/1 (March 1979), 29–42