Energy & Income: a Doubling Rule

Peter Mott

February 2021

Faced with Climate Change there is a view in rich countries that consumption must be cut back and energy use curtailed in order to “tackle” this perceived threat. But is it true that as countries grow richer their tendency is to consume more energy? It might seem so - there are more cars, more travel, and more appliances in a rich country than in a poor. In fact the situation seems a lot more complicated, as the following charts and tables will hopefully illustrate.

A data point in Chart 1 shows income against energy consumption for all the years and countries in the dataset I have. The years are from 2001 to 2019. The countries are all those in both the BP energy statistics for 2020 and the World Bank income data.

The highest point is Qatar in 2005 which consumed 259MWh per person; Qatar’s income then was 38K USD per person. For every year, Qatar consumes more energy per person than anywhere else, illustrating a fact that rankings in energy consumption are not very volatile. The richest point on the chart is Norway in 2014 with a per capita income of 104K USD and an energy consumption of 101MWh. Norway, like Qatar, is so rich because of its oil but it consumes less than half the energy per capita that Qatar does. The lowest energy consumption was Bangladesh in 2001 using 1 MWh per person. It was bottom in 2018 as well though its energy use had more than doubled to 2.6 MWh per capita (and its population increased by 23%).

Chart 1 Energy & Income, All Countries 2001-19.

The fitted regression (energy on the square root of income) is not very useful: enquire its prediction for the energy consumption of a country with income of 60K USD and you will get the answer 87MWh ± 60MWh ⊕Here is another chart of this idea https://bit.ly/2Xqo9BS which you could have guessed anyway. Energy use does increase with income but not smoothly and a function that maps income onto energy within a useful error range will not, I think, be found. This is a shame because if the regression were good you could estimate how much energy a country was going to use from a projection of its GDP.

Chart 2 provides a better way of looking at the issue.

Chart 2 Energy & Income by Income Group

The chart is for 2008 and divides countries according to World Bank income group. I chose 2008 because later years will not show Low income countries because there are fewer of them and those that there are are not in the BP Energy statistics. The most notable thing about this chart is how World Bank Income Group determines energy use with higher groups using more than lower ones. The table summarises this:

The most surprising thing about the table is the regular spacing of energy use per income group: within quite a close approximation the median user in each energy group uses double the energy of the group below. This is not just for 2008: looking at Chart 3 we see it remains true when each of the years 2001-2019 are considered. Which gives us a Doubling Rule: as a country moves up a level in the World Bank income hierarchy its energy use will double.

Chart 3 Median Energy Consumption by Income Group 2001-2019

⊕The chart shows both the median energy use and adds the inter-quartile range as a ribbon. The median energy use divides the sample such that half the countries use less and half use more. The inter-quartile range covers 50% of the countries. Equivalently, 75% of cases fall below the top of the band and 25% below the bottom. From Chart 3 median energy use seems to have been declining until around 2009 after which it has remained pretty stable, except in the High Income group where the decline seems to be continuing. This is confirmed in Table 2. The second row of the table,for example, says that of the 20 countries classified in 2001 as Lower-middle income sixteen (80%) of them show an increasing trend of energy use between 2001 and 2019, while only 4 (20%) show a Reducing one. But, by contrast, with only two exceptions, high income countries show a reducing trend in energy use:

Of course if the World Bank reclassifies a country as Upper-middle income in 2015 its energy usage is not going to double in 2016. Rather the trend for developing countries is for increasing energy use as they strive to pass from one level to the next. High income countries reached a plateau of usage around 2005 and now are on a reducing trend.

I started this piece because I thought there would be a simple relationship between prosperity and energy use: but there is not. Rather the picture is more complicated: as countries pass through the steps of the World Bank income levels their energy use doubles. But at the top of the staircase, in the high income world, energy use is in fact declining. The sum of these two effects is far from clear: but one can be pretty confident that rich and poor countries will not agree as to what is to be done about Climate Chang (I mean in policy, in action, not in words, pledges and promises)