Energy and Productivity - Part 3: The Relationship Between GDP and Electricity Consumption

In the first post of this series, I mentioned that the relationship between GDP per capita (output of the economy) and electricity consumption has been explored in far greater depth than productivity’s relationship with electricity consumption. GDP, as explored in the second blog post, is a component of productivity and hence influences productivity. Therefore, it forms a good starting point for the discussion on the relationship between productivity and electricity consumption.

Intuitively, it is easy to understand that a relationship between electricity consumption and the output of an economy (GDP per capita) exists. Just about every business in a developed country will rely on electricity to some extent. Whether that be for lighting, heating or communication, we have gotten used to assuming that electricity will just get certain tasks done for us.

Therefore, if electricity suddenly becomes more expensive, as seen in the Australian states of South Australia and Victoria recently when on the 24th of January wholesale prices for electricity reached the price cap of AUD$14,500 per MWh, the business may just decide to stop work for the day. In the case of Australia on the 24th of January, wholesale prices were over 100 times their 2017 weekly average for South Australia and over 200 times for Victoria.[1] This is, therefore, directly reducing the output of the country.

The trends in electricity consumption and GDP per capita are illustrated in Figures 1 & 2 for the UK and US. Here, a clear break in the trend can be seen around the turn of the millennia. Before then, there is a clear positive correlation between the two, as one grows so does the other.

In fact, basic correlation estimates show that over the period of 1970 to 2000 the correlation was close to 100% (UK: 98%, US; 98.5%). On the onset of the 2007 recession, however, the correlation drops to 17% for the UK and 5% for the US (Author’s Calculations based on World Bank data). Of course, the UK and US experienced some of the most drastic changes in the relationship, Italy and Spain’s electricity and output correlations haven’t changed much over time and neither has it for China or India.

Figure 1: World Bank, 2018

Figure 1: World Bank, 2018

Figure 2: World Bank, 2018

Figure 2: World Bank, 2018

This is where things get interesting. Although an underlying correlation between electricity consumption and output is generally agreed upon, economists are quick to point out that correlation does not equal causation. Even with a strong correlation, causation could run from electricity consumption to output, which would indicate that increasing the amount of electricity people use directly causes the output of the economy to go up.

If this were the case, then policies or events that restrict the amount of electricity consumed would cause significant impacts on GDP. This would mean politicians should very much be concerned with the trends in flattening electricity demand.

The opposite could also be true, the output of the economy going up may cause people to demand more electricity. This would imply that limiting electricity consumption would have minimal effects on output. Then again, causation could be running both ways, creating a far more complicated relationship – or there might not be a causal relationship at all.

The academic literature on this is split, Omri (2014) compiled all the research completed in this area and found the following: 40% support the idea that causality runs from electricity consumption to output, 27% support the idea that causality runs from output to electricity consumption, and 33% support the idea that causality runs both ways.

This is quite an even split, which shouldn’t be surprising given the wide range of countries examined and how the different countries’ energy policies can vary greatly. Ozturk (2010) & Payne (2010) have also summarised the literature. Similarly, they found an even split between the different pathways, even if the exact figures are distributed a bit differently.

The one observation that these papers seem to support, however, is that the relationship has changed over time. Figures 1 & 2 illustrate this change that occurred somewhere around the turn of the millennia, where the electricity consumption-growth turned flat before going negative, even though GDP has continued to trend upwards.

On the other hand, countries that are still developing see a very different trend in electricity usage. Take China and India for example, the relationship between electricity consumption per capita and GDP per capita has only retained its strong correlation and continues to follow an exponential trend, as highlighted by Figures 3 & 4.

Figure 3: World Bank, 2018

Figure 3: World Bank, 2018

Figure 4: World Bank, 2018

Figure 4: World Bank, 2018

Understanding how this relationship between electricity consumption and output functions is of key importance for policymakers, businesses and consumers alike. Policy makers, for example, will need to be considerate of the impact on the output of their country if they decide to restrain electricity generation to meet carbon emission targets – particularly in developing countries where electricity consumption is still growing.

Traditional utilities, on the other hand, have relied on growth in consumption to drive their profits. In a future of steady or declining electricity usage, traditional utilities will have to reconsider their business model. This is already becoming very apparent across the globe where electricity utilities are becoming less and less profitable.

Bloomberg New Energy Finance research on US electricity utilities found that the ratio of rate base – which is essentially the profit utilities make on their investments – to sales has doubled over the last decade. This would be fine if they also expected sales to grow significantly in the future to fund these investments, however, if the growth in demand for electricity continues to be flat, these utilities may find themselves struggling to keep themselves going. Consumers too can only reduce electricity usage so much before you are making dinner in torch light.

In the end, even if the literature cannot agree upon the direction of causality, the relationship between electricity usage and the output of the country is an important one. However, as discussed in the first post, it is the way this feeds into the longer term productivity what we should all really be concerned about - this will be covered in the next post of the series.

Bibliography

AEMC. 2018. “Spot Market.” [website] URL; https://www.aemc.gov.au/energy-system/electricity/electricity-market/spot-market Last accessed 1st of February, 2019.

Bloomberg New Energy Finance [BNEF]. 2018. Bloomberg Professional. [Online]. Available at: Subscription Service (Accessed: 1 February 2019).

Omri, A. 2014. An international literature survey on energy-economic growth nexus: Evidence from country-specific studies. Renewable & Sustainable Energy Reviews, 38, 951–959.

Ozturk, I. 2010. A literature survey on energy-growth nexus. Energy Policy, 38 (2010), pp. 340-349

Payne, P. E. 2009. A survey of the electricity consumption-growth literature. Applied Energy. doi; 10.1016/j.apenergy.2009.06.034

[1] Author’s calculations based on AEMC data.