UK Energy Consumption and Weak Productivity Growth
Falling energy consumption in the United Kingdom is not receiving the attention it deserves.
While similar to the norm prevailing in the EU 28, the UK pattern is very strongly at variance with the global trends, which see significant increases in all sectors. There is a clear possibility that this fundamental difference is a leading causal factor behind the weak productivity growth in the United Kingdom since 2008, yet it is hardly considered by commentators calling, perhaps correctly, for aggressive “innovation” as the answer to the “productivity puzzle”. Until they do so, their appeals will be in vain: costly energy makes it rational for innovators to be risk averse.
Figure 1 charts data taken from the Digest of United Kingdom Energy Statistics (DUKES), and shows total primary inland energy consumption (yellow line) and UK GDP from 1970 to 2018.
Figure 1: Temperature corrected total primary inland energy consumption (yellow line), and UK GDP on a chained volume measure (black line), from 1970 to 2018 (1970=100). Source: Data from DUKES Table 1.1.4. Chart by the author.
Beginning with the macroscopic patterning, one observes that while energy consumption over the period has been more or less stable, GDP has exhibited a strong rising trend. This undermines claims that “de-linkage” of energy consumption and GDP is a recent phenomenon, due, for example, to efficiency measures, the digital economy and dematerialisation. On the contrary, de-linkage is a long-standing phenomenon in this data, and thus is probably not to be explained by recent novelties but rather by factors that are simpler and more fundamental
Further doubts are cast on the validity of a naïve assertion of de-linkage by examination of the finer structure of the data, in which we observe a varied and subtle relationship between energy consumption and GDP. Firstly, there is an irregular but clear downward trend in energy consumption from 1970 to the early 1980s, after which we see a moderate but steady rising trend in energy consumption up to the later 1990s and early 2000s, after which there is a flattening off and then a marked decline from 2005 onwards, a decline which is acute compared both with the previous downward trend in the 1970s and indeed with the preceding upward trend from the early 1980s. In a little over ten years the increase in consumption evident over the period 1982 to 2001 has been reversed, and in 2018 the UK consumed just under 10% less than it did in 1970. This change would be notable in itself but is particularly so when we recall that over this period population has risen from about 56 million to 65 million and that GDP has more than doubled.
As already noted, some see evidence in this data that energy consumption and economic growth have been “de-linked” in the last decade or so. There is clearly some ground for this view in the fine structure of the data at the end of the series. But equally there is evidence for an earlier de-linkage in the 1970s to the early 1980s. This can only undermine confidence in any argument suggesting that current de-linkage results from recent societal and technological modernisation, principally the digital economy. It is at least possible, and in my view probable, that some other explanation accounts for both the divergence in the 1970s and that in recent years. For the time being the de-linkage case, never theoretically strong, should be regarded as weak in comparison with alternatives.
For example, it might be inferred that the energy consumption required to support the economic growth visible in GDP is taking place elsewhere in the world. On this view, the UK economy, taken as a system, became for a short period in the 1970s more reliant on energy conversions elsewhere in the world for the goods and services it consumed, a trend that has recurred in a stronger form in the present day. If this were correct, the de-linkage of GDP and energy in the UK would be illusory.
Furthermore, the fine structure of the data also reveals that even in the divergent curves at the beginning and end of the series there is still some degree of linkage between inland energy consumption and economic activity. For example, both in the periods 1973–1975 and in 1979–1981, and again after 2008, falls in energy consumption are paralleled by falls in GDP. Indeed, in the 1970s and the early 2000s the relationship is notable for a subtle but highly suggestive character. – GDP and energy are clearly related, rising and falling together over the short term, even as they are exhibiting divergent secular trends over the longer term, with energy consumption falling and GDP rising in both periods.
These two phases at either end of the series contrast sharply with the straightforward correlation visible in the two decades from the early 1980s up to the early 2000s, when GDP and energy consumption rose together. Indeed, one interpretation could be that the Britain of today has more in common with that of the 1970s than with that of the 1980s and 1990s, a rather shocking conclusion, but one that cannot, I think, be rejected quite out of hand. It is worth asking whether a tendency towards a healthier economic function, with a more reasonable balance between inland production and imported consumption, is represented by the 1980s and 1990s, and a less satisfactorily balanced, or even anomalous operation by the 1970s and the present day.
Analysis along these lines may also shed light on the notorious “Productivity Puzzle”, that unprecedented and so far inexplicable sluggishness in productivity growth since 2008. Figure 2 charts UK inland energy consumption data (yellow line) and output per hour worked (green line).
Figure 2: Productivity (Output per Hour, seasonally adjusted), green line, and UK Inland Energy Consumption, yellow line, 1971 to 2018, 1971=100. Source: ONS, DUKES 2019, chart by the author.
Several of the points made above in relation to GDP can be made again here. Although the 1970s see productivity growth rise and energy consumption fall, there are still signs of a positive correlation in the fine structure of that divergent trend, just as there is in the divergence from around 2008 onwards. Furthermore, as with GDP, the central body of the data is characterised by the positive correlation of rising energy consumption and rising productivity growth in the period from the early 1980s to the early 2000s.
But there is also a significant difference. While GDP resumes its previous upwards rate of change quite promptly after 2008, productivity growth does not, and steers closer, as it were, to the downward energy consumption trend. One might infer, therefore, that energy consumption plays a larger part in productivity than in GDP. That is plausible, since a change in the energy consumption of inland economic activities is almost certain to have a significant and direct effect on productivity; if a production system is under-energised it does less. – If the throttle is closed, the engine decelerates. On the other hand, any effect that falling energy consumption might have on GDP can readily be offset by other factors. For instance, GDP can be enlarged by the spending of borrowed funds on imported goods, goods that are produced with energy consumption in other territories.
It seems, therefore, that there is some ground for concluding that the unprecedented stagnation of productivity growth since 2008 could be explained at least in part by factors depressing energy consumption, such as sustained and significant increases in energy cost, making it difficult to recover from the economic shock of the crash. This is no mere theoretical possibility, and strong candidates can be found, for example, in the loading of climate policy subsidy costs on to electricity, starting in 2002, and now amounting to about £10 billion per year, and very high, longer term taxes on transport fuels, totalling £28 billion a year at present. It is important to recall that both these policy impositions were charged on top of fundamental costs that were and still are themselves rising, making an underlying difficulty much worse.
There is a widespread assumption that the productivity puzzle could be addressed by a determined government focus on the enhancement of innovation. For example, a recent paper by Richard Jones, of the physics and astronomy department at the University of Sheffield, has argued exactly this and has received a generally favourable reception, even in right leaning and Conservative Party circles (see for example the short recommendation in a piece by Dominic Cummings, and for further positive endorsements see Stian Westlake and Peter Franklin).
Jones, in essence following Mazzucato’s case for an Entrepreneurial State, suggests that major government interventions in favour of research in low carbon energy, and in health and social care are “key ingredients in turning around the productivity problem” (see p. 43). Strangely, he appears to be unaware of the long-term and exorbitantly costly market coercions already favouring low carbon energy, and is thus in no position to wonder whether those distorting energy policies may be playing a significant role in creating the productivity problem in the first place. But Jones is by no means unusual in failing to take energy seriously though discussing it at length; and as a matter of fact, hardly anyone gets beyond conventional and empty gestures towards energy as the lifeblood of the economy. However, the GDP, energy consumption, and productivity data discussed above suggests that such an attempt would be worthwhile, and that a great deal could depend on it. Who would disagree with Jones and others that innovation is essential to prosperity? Why, then, is it so difficult to deliver? Because innovation is the experimental combination and application of inventions to satisfy human requirements and is a very high risk business indeed; the vast majority of innovations are failures. Cheap energy means that those failures are not expensive, and that innovators can afford to take the risk over and over and over again. When energy is expensive it is rational for innovators to be extremely risk averse, as they generally are at present in the United Kingdom.
Dr John Constable: GWPF Energy Editor