As a follow-up to an earlier blog that considered the energy efficiency performance of the South African economy from 2000 to 2012[i], this blog looks at the electricity intensity of the economy from 1995 to 2015. The data not only tells an interesting story, but also highlights the dangers when using nominal data to consider the relationship between variables.
Using nominal GDP data (i.e. GDP at current values), it seems that the electricity intensity of the economy has decreased exponentially since the 1990s, and that the improvement in electricity intensity is now tapering off.[ii]

This, however, is a result of the effect of inflation on nominal GDP, which make it look as if more and more GDP is being generated with the same (or even less) electricity. A very different picture emerges when real GDP data with inflation removed (consistently measured in 2010 Rands) is used in the calculation. The electricity intensity of the South African economy has been decreasing in recent years, but at nowhere near the exponential rate that the nominal data indicates.

Furthermore, rather than tapering off, the improvement in electricity efficiency is still ongoing. The graph above does, however, seem to show that there was a change in the electricity intensity trend after 2004, and not as is often assumed after 2007 when South Africa was hit by an electricity supply crisis that precipitated a series of steep real increases in electricity tariffs.
In fact, linear trend lines generated for the period 1995 to 2004 (green dashed line), and 2004 to 2015 (orange dashed line) in the graph below shows that the downward trend in the electricity intensity of the South African economy has been remarkably constant since 2005. In 2008 intensity did fall more than expected, but quickly recovered thereafter.

Given that the National Energy Efficiency Strategy was published in 2005, this raises the possibility that a policy intervention, and adherence to this policy by large electricity users, rather than the impacts of the 2007-2008 electricity supply crisis and the subsequent increases in electricity tariffs has led to the increasing electricity efficiency trend that is currently being observed.
Summary (approximately 150 words)
Considering the electricity intensity of the SA economy from 1995 to 2015 tells an interesting story, and highlights the dangers inherent in using nominal data when considering the relationships between variables. Using nominal GDP data (i.e. GDP at current values), it seems that the electricity intensity of the economy has decreased exponentially since the 1990s, and that the improvement in electricity intensity is now tapering off. This, however, is a result of the effect of inflation on nominal GDP, which makes it look as if more and more GDP is being generated with the same (or even less) electricity. A very different picture emerges when real GDP data (consistently measured in 2010 Rands) is used in the calculation. The electricity intensity of the South African economy has been decreasing steadily since 2005s, but at nowhere near the exponential rate the nominal data calculation indicates. This trend, however, shows no sign of abating.
[i] [LINK] https://www.dnaeconomics.com/pages/climate_change/?zDispID=NewsArtSouth_Africas_energy_efficiency_performance_since_2000
[ii] Nominal and real GDP values obtained from StatsSA Statistical release P0441, and electricity generated and available for distribution from all producers (in Gigawatt-hours) obtained from StatsSA Statistical release P4141 (February 2017).