We think of capital, the assets we use in production, as heavy: machines, buildings, infrastructure, trucks and railroads. Being composed mostly of cement and steel, we would expect their production to cause a lot of greenhouse gas emissions. In a new paper, we offer a first detailed analysis of the carbon footprint of gross fixed capital formation across countries and sectors. The picture that emerges is interesting because of some small surprises.
First of all, capital is big. Capital formation constitutes about one quarter of gross global product in monetary terms. It causes about 30% of global greenhouse gas emissions. Leaving it out is a pretty big oversight.
Second, capital formation varies across countries. The country with the highest capital formation over the past decade was China, where it constituted 44% of GDP in 2007 and caused 57% of the carbon footprint. In developed economies, capital formation constitutes only about 20% of GDP.
Third, the carbon intensity of capital formation varies depending on what capital goods countries and industries invest in. While construction and machinery are carbon intensive, software and business services are not and ICT hardware has an average carbon intensity. Globally, construction accounts for half of capital formation and 60% of its footprint. In wealthier countries, capital formation is less carbon intensive than household consumption.
Fourth, capital formation increases with GDP across countries. This is maybe not surprising: affluent people save more. However, despite purchasing more software and business services, rich countries also keep investing in buildings and roads. Across countries, the carbon footprint of capital formation increases by 83% for each doubling in GDP per capita, measured on a purchasing power parity basis. China is an outlier in this picture, having a carbon footprint of capital per person higher than France or Great Britain. This increase of the carbon footprint of capital with wealth occurs in spite of investing in cleaner types of capital and having cleaner technology.
Most of the capital goods are purchased domestically. We investigated two groups of countries, those with the highest and lowest carbon intensities of capital formation, and found that in both groups, two thirds of the value added associated with capital formation was domestically produced. However, the location of GHG emissions diverges: In countries with a high carbon intensity of capital, most of the emissions occurred domestically, while in countries with a low carbon intensity of capital, three quarters of the carbon footprint was associated with imports. That is quite a stunning divergence! In economies with a clean energy mix and a high energy efficiency of production, one would expect imports to have a disproportionate share of the carbon footprint, but the differences observed for capital are especially egregious.
In Prosperity without Growth, Tim Jackson argues that investing in assets and infrastructure was an important step in the transition to a sustainable economy (p.177 f). We find that investments are just as or even more polluting than consumption, so that this strategy, per se, is not one that will solve our environmental problems. His list of suggested investments, of course, focuses on investments that would enable a more sustainable living, such a renewable energy, public transport, and energy-efficient buildings. We agree that such investments are required, but would like to point out that capital formation as such is polluting and that hence, this pollution needs to be considered when looking at the whole picture.
Finally, I would like to close by reflecting on my own field. When considering the carbon footprint of products or services, the carbon required to produce the machinery that produces the products we consume, the carbon required to lay the roads and build the ports, warehouses and shopping malls that deliver these products to us are usually not considered in the assessments, independent of whether life cycle assessment or input-output analysis is used for the quantification. We have now provided an estimate of the potential magnitude of the resulting error. It is too large to be ignored.
Source: Södersten, C.-J., R. Wood, and E. G. Hertwich. 2017. Environmental Impacts of Capital Formation. Journal of Industrial Ecology doi:10.1111/jiec.12532