It is time to start the next chapter of the Seeking Consensus project: Externalized Costs. This promises to be an interesting chapter because the spread and uncertainty in the available data is much greater than it was for internalized costs. As always, I’ll be providing a first estimate for each given energy option together with a simple Excel model which can be used to come up with new numbers by adjusting various important parameters.
This post will discuss the methodology I plan to follow to estimate externalized costs of various energy options. The general idea will be to split up the external costs in terms of long-term global externalities (climate change) and short-term local externalities (e.g. local air pollution).
The difference between these two types of externalities is that those who benefit from energy consumption directly experience the negative impacts of short-term local externalities, but can be separated by thousands of miles and tens of years from those experiencing the worst impacts of long-term global externalities. More details on this distinction are given in two previous articles here and here.
Long-term global externalities (climate change)
Putting a representative price on greenhouse gas emissions is essentially an impossible task. Numbers of $30-40 per ton of CO2 are generally accepted as a reasonable median, but there are many other estimates which lie far above or below this range. On the high side, we have the “long-tail” or “95th percentile” estimates which account for the much higher costs in the case that self-sustaining positive feedback loops are triggered. On the low side, there are even negative estimates stemming from positive effects of increased CO2 levels such as increased crop growth rates.
Then there is also the issue of developed vs. developing world costs. Due to the low level of economic development, developing world citizens are generally much more exposed to the effects of climate change. Storms can completely destroy their homes, heat waves can be life-threatening, droughts can result in critical shortages of food and water, and increased temperatures can bring new diseases which can wreak havoc in the absence of basic medical care. Most of these threats can be greatly reduced even by moderate levels of industrialization (at the cost of more fossil fuel consumption). Thus, it makes sense for developing nations to discount these future climate costs at a high rate in order to maximize the rate of industrialization in the here and now. This, in combination with the large historical emissions of the developed world, implies that the developing world should be allowed a significantly lower CO2 price than the developed world.
For these reasons the following methodology is proposed: Firstly, the IEA CO2 price estimates for achieving the 450 ppm scenario recommended by climate scientists is fit to a function as in the graph above. Following this, the CO2 price is discounted at the projected growth rate for the developed (2%) and developing (5%) world. This practice adjusts for the fact that the developing world does not value rising CO2 costs in the future very highly because rapid growth in the near term will allow them to handle these costs much more effectively. In this way, a cost of $53/ton for the developed world and $24/ton for the developing world can be calculated. Averaging for the global economy assumed to grow at 3.5% returns a CO2 cost of $36/ton which agrees well with the median of the wide set of estimates out there.
Short-term local externalities
Although they are less prominently discussed than greenhouse gasses, estimates of other externalities such as localized air and water pollution from fossil fuel combustion can also be found over a wide range. Naturally, these numbers vary widely from one energy option to the next and must be treated separately based on a survey of the available literature in each individual case.
However, an important additional consideration will be made in this series: the economic growth benefit experienced by the local community when not internalizing short-term local externalities. This is the main reason why many developing world citizens and governments put up with terrible local pollution as long as the rapidly growing economy grants them the opportunity to better their lives.
It is self-evident that more expensive energy will retard economic development. As an example, the historical correlation between oil prices and economic growth rates both in the US and Europe are shown below (data from the USDA database and the BP Statistical Review).
This is especially applicable to rapidly growing developing economies. Before we can expect these billions of world citizens to get really serious about global sustainability, we need to see many more incredible infrastructure buildouts such as the examples given in this link (the picture of Dubai is shown below).
This transformation will require many billions of tonnes of steel, cement and other materials to make billions of modern homes with all the required water, electricity and sewage infrastructure, many billions of appliances to fill these homes, many millions of factories and distribution centres to supply these consumers and billions of cars driving on many millions of miles of paved roads to connect everything. All of this will require an absolutely stupendous amount of energy which needs to be as cheap and as practical as at all possible.
For the billion or so people who are lucky enough to have all of this in place already, accounting for such an energy intensive infrastructure buildout is not necessary, but for the other 6 billion it sure is. This infrastructure offers large increases in productivity, thus enabling economies to sustain an exponential economic growth path over several decades. If local communities choose to make energy more expensive by internalizing short-term local externalities, even a small growth slowdown compounds exponentially over time. As shown in the graph below, even a decrease of only 2% growth sustained over 20 years can cut the potential size of the economy by over a third. See the appendix at the end of this article for a simple example illustrating this principle.
To account for this factor, short-term local externalities will be multiplied by a factor between zero and one. A factor of zero implies that costs of internalizing these externalities (reduced growth rate) outweighs the benefits, while a factor of one implies that internalization of externalities will have a negligible/positive impact on growth. Naturally, this factor will be much lower in developing nations undergoing rapid energy infrastructure buildouts than developed nations where the capital stock is stagnant or even declining.
The plan is thus to estimate external costs as follows:
- Climate change: CO2 intensity of the specific energy option in question times a CO2 price of $53/ton for developed nations and $24/ton for developing nations.
- Short-term local externalities (e.g. local air pollution): Median externalized cost for new builds of the specific energy option in question times an adjustment factor for the impact on local economic development. The adjustment factor will vary between zero (economic costs of internalizing externalities outweigh the benefits) to one (internalizing externalities has a negligible/positive impact on growth).
Appendix: Externalities and growth example
As a simple illustration of how the internalization of externalities can cut growth, we can take the example of India which will probably follow China as the world’s next heavily polluted industrialization success story. Indian GDP and electricity consumption is plotted below on a log-scale where it is shown that India has maintained almost perfect exponential growth both in GDP and electricity production over the past three decades (data from the USDA database and the BP Statistical Review).
To achieve 7% growth in 2015, India needs to grow real GDP by $147 billion and its electricity production by 85 TWh. Currently, the vast majority of this additional electricity consumption will come from dirty, cheap and practical coal plants. Let’s say that India now decides to put heavy local taxes on coal combustion to internalize short-term local externalities so that this jump in electricity production must be achieved by nuclear, onshore wind and solar PV. The graph below gives the capacity and cost of increasing the electricity supply by 7% using these four options under the following assumptions: coal (70% capacity factor, $1000/kW), nuclear (80% capacity factor, $2500/kW), wind (21% capacity factor, $1200/kW) and PV (18% capacity factor, $1800/kW).
It is clear that nuclear is the best option after coal and can deliver the required capacity to sustain the Indian exponential industrialization effort for “only” $17 billion more. If this could be done with nuclear (which has consistently scaled 40 times slower than coal in India for the past two decades), the result would be a cut in the economic growth rate of close to 1% (from 7% to 6%). $17 billion of the potential $147 billion increase in real production would now be used without any real increase in output (the new nuclear capacity would produce the same electricity as the coal capacity it displaced). Compounded over 20 years, this would cost India 11% of their potential economic output over this period.
For wind and solar, the story is much bleaker. The required capacity buildout with wind or solar would cost $55 billion or $97 billion respectively (most of the planned coal capacity would need to be built anyway due to wind/solar intermittency). A buildout of half wind and half solar would consume more than half of the $147 billion GDP increase that would be possible under the standard coal driven growth path. This would slice the potential size of the economy in 20 years in half (and more thereafter).
Given that electricity accounts for only about 40% of primary energy consumption, these kinds of growth penalties resulting from the internalization of short-term local coal externalities in the electricity sector will simply not be acceptable. When seen from the point of view of the largely impoverished local populace, the economic benefits of ignoring short-term local coal externalities will outweigh the costs in most cases.