Update February 2020: The content in this post is somewhat outdated and was in the midst of research. The Lorenz Curve is a great way to approach inequality and the three-dimensionals plots presented here not necessarily ideal. still some points made here.

Energy Inequality is the distribution of energy across people in a society. It is to large extents determined by economic inequality, the distribution of capital (either in the form of income or wealth) across people. How do we know this? Well the coupling of economic growth and energy consumption of countries is a repeatedly observed fact. If a country has an average GDP/capita of 40 000$ and another only 10 000$, it is not hard to tell which one consumes more energy. Similar things can be observed not only across countries but within countries: Rich people have more capital than poor ones and thus with all likelihood consume more energy. But how exactly does that happen? Which channels tie having more money to consuming more energy? And how does this relationship look like anyhow? Is it linear or does it exhibit a non-linear shape? Those and another questions motivate my PhD studies.

In the quest for answering such questions, one of the things I do is visualizing my data (as obviously most scientists). I try, as good as I can, to come up with visualizations that are not only interestingly looking but also deliver insight as effectively as possible. A well-known visualization of inequality is the Lorenz Curve often applied to Income Inequality. It will also constitute one of the most important analytical elements within my future work, not least because it is directly related to the Gini coefficient. Yet I find it a little boring by now. It looks like this for Germany in 2011

It does not look that spectacular and how much insight does it really provide? Ok, we can see here immediately that energy inequality in Germany is not thaaat severe because the red line is relatively close to the blue line. If you do not know right now what that means, it does not matter.

Can we provide more insight by other means of visualizations? I asked myself how can I capture the stratification and structure of energy consumption and distribution most efficiently, aesthetically and accurately at the same time? I thought a good idea would be to plot not just two variables but three this time. Even though I am in strict favour of “as simple as possible” approaches my data comes divided according to income groups. And this nearly begs for three dimensional visuals.

For example, what I tried to visualize in the next plot is how much of certain products that exhibit certain energy intensities do certain income groups consume. So we have three dimensions:

  • Income group
  • Energy Intensity of a product
  • Money spend on such a product (by overall population in that income group)

Along the quintiles’ (the income groups) dimension one can spot the fact that the 5th quintile  usually spends a lot more on products than the 1st quintile. The bars get larger if you go from one to five. These are “the richer people”. Along the kJ/USD dimension the energy intensity of products is listed. The products are sorted according to their energy intensity. So for instance, the products in yellow shades in the front are the most energy intensive ones and the products in blue in the back the least energy intensive ones. The height of the bars, in the third dimension, displays the magnitude of expenditures. This is nothing but the amount of money spend on products by income groups. For instance go to quintile one and the golden large bar in the front and you find that on a product with an energy intensity more than 9000 kJ/USD (one cannot see that value in the graph, you can only see that the value must be larger than 4377 kJ/USD because they are labelled only every 10th product), people in the first quintile already spend more than 20 billion USD. What’s that? Mostly automotive fuels. The sector accounts for several products but the large energy intensity and the large expenditures on it mostly come from buying petrol for your car. These visuals are by no means my final “results” nor do I intend to explain only Germany. What I intend with these figures is to depict the inequality of consumption and energy use across products AND across income groups. AND steer a discourse around how to analyse and visualise inequalities or any kind of distributions in society.  

My figure is a three-dimensional space that depicts “scaling” in two ways:

  • First we see that consumption of products does not relate in a uniform way to the energy intensity of products, but exhibits what I call ‘multipolarity’.
  • Second, we can observe that the big blue bars really in the back differ not as much from each other as the mentioned large golden bars in the front. That shows that the difference in consumption is less in the product in the back but larger in the “automotive fuel” consumption in the front. It shows that the fuel product with a higher energy intensity scales steeper across income groups. A pattern that has a large impact on the energy requirements of the German society.

This is just a glimpse. And it is just the beginning of a much longer road I am about to take. But I am happy about feedback any time. Soon I hope to come up with even more, hopefully even better and more systemic ways of illustrating my data. In the long-term I hope to be able to build more dynamic visualizations that are available online. If so, in the context of the LiLi Project I am part of.