You want a $40-a-ton carbon tax? My country is already at $55!

The Niskanen Center has published a blog post arguing that, if one regards William Nordhaus’ work on climate and the economy as correct, then one should act by taking ‘immediate and forceful climate action.’

One of the oddest things about articles on climate change, and the above blog post is no exception, is that they often urge action but rarely mention who is supposed to take it. Okay, if the measure one is talking about is a carbon tax, then of course that can only be done by governments… but which are those governments?

If you argue that all governments should take climate action, then you must at least acknowledge that many of them have already done so. The current situation is not one in which governments are doing nothing to mitigate carbon dioxide emissions, but rather one in which they are doing a lot of things at least ostensibly with that purpose (although those measures seem to be mostly ineffective).

In this article I will demonstrate that at least one country, Spain, is already above the CO2 tax recommended by Nordhaus. In other words, we are taxing CO2 more than is economically optimal. Presumably, the immediate and forceful action advocated by Niskanen consists of cutting fuel taxes!

How much should we tax CO2 (according to Nordhaus)?

Here we begin the number-crunching. First, this 2017 paper by Nordhaus himself states:

“The present study presents updated estimates based on a revised DICE model (DICE-2016R). The study estimates that the [social cost of carbon] is $31 per ton of CO2 in 2010 US$ for the current period (2015).”

In the US, inflation from 2010 to 2018 has been 15.8%. (Should I use a ‘global’ inflation index instead? I’ll leave that to the readers). On top of that, at a ‘real’ (i.e. inflation-adjusted) rate of growth of 3% a year, the real carbon tax by 2018 should be 9.3% higher than in 2015.

In other words: in 2018 dollars, the optimal CO2 tax by now should be 31 * 1.158 * 1.093 = $39.2

Of course one can argue that Nordhaus’s estimates are too optimistic, or too pessimistic, or just unreliable because it’s impossible to know exactly what will happen with climate change. I, for one, find this paragraph most troubling:

“The climate module has been revised to reflect recent Earth system models. We have set the equilibrium climate sensitivity (ECS) using the analysis in ref. 9. Ref. 9 uses a Bayesian approach, with a prior based on previous studies and a likelihood based on observational or modeled data. The reasons for using this approach are provided in ref. 5. The final estimate is mean warming of 3.1 °C for an equilibrium CO2 doubling. We adjust the transient climate sensitivity (TCS) (sometimes called the transient climate response) to correspond to models with an ECS of 3.1 °C, which produces a TCS of 1.7 °C.”

The reference he cites is a paper published in early 2012. There has been a lot of work on climate sensitivity since then; papers actually looking at the evolution of temperatures, ocean heat content and forcing in the instrumental record always find much lower sensitivity values — between 1.5ºC and 2ºC. There has also been a lot of back and forth, with researchers arguing against the instrumental papers have issues. But so far no published paper has convincingly showed the instrumental estimates to be biased low.

(I think I’m quite up to speed in that field, but if you believe there is a paper that shows why instrumental estimates of climate sensitivity are biased low, please let me know).

You don’t need to speculate about what the results would look like if one used lower climate sensitivity values to estimate the social cost of carbon, as someone has already done that. Long story short: the “optimal” CO2 tax falls by 40–50%, as seen in table 2 of the just-linked paper. (The DICE model is the one developed by Nordhaus).

So the $40-a-ton CO2 tax may make economic sense if one guesses the warming from a doubling of CO2 is 3.1ºC. If one instead takes a value of 1.7ºC or so, the “optimal” tax would be about $20 per ton.

I’m not going to spend more time discussing which is the “right” value of a CO2 tax (I don’t think an exact figure could be given even if one knew climate sensitivity with total certainty). Instead, I’ll show that even if one takes the $40-a-ton figure as valid, my country is already paying more.

How much tax we pay

In 2017 the Spanish government officially made €10,881 million off hydrocarbon taxes. However, part of the value-added tax (aka the sales tax) is derived from the hydrocarbon tax. If a liter of gasoline costs €0.50 before tax, and €1.00 after the hydrocarbon tax, it ends up as €1.21 at the pump because the VAT is 21%. In other words: to obtain the true revenue provided by the hydrocarbon tax, you have to increase it by 21%. That makes it €13,166 million.

On top of that, manufacturers and electric utilities have to obtain permits in the European Union’s Emissions Trading System for each ton of CO2 they release. Over 2017, the average price was about 5€ per ton. Since emissions from the electric system last year were 74.8 million tons of CO2, that means €374 million. I’m excluding here the permits obtained by manufacturers which burn fossil fuels onsite, as I don’t know of any easy way to estimate them; in any case they are smaller than emissions from electricity generation.

Put together the hydrocarbon tax and the emission permits and you get €13,540 million. At a 2017 average exchange rate of about 1.13 dollars per euro, that’s $15.3 billion.

Since BP pegs our emissions from fossil fuel combustion at 301.9 million tonnes (see page 49 of their Statistical Review), that means we effectively spent $50.68 per ton of CO2 released.

But aren’t fuel taxes a tax on air pollution?

There is a fundamental difference between long-lived greenhouse gases, chiefly CO2, and ‘pollution’ as has been traditionally understood. I’m going to focus on nitrogen dioxide as it’s usually considered the main component of man-made outdoor air pollution.

Long-lived greenhouse gases are also sometimes called well-mixed greenhouse gases: precisely because they stay in the atmosphere for so long, they tend to spread around the Earth’s surface very uniformly. It doesn’t matter whether a ton of CO2 is released from a coal-burning stove in Xinjiang or a pickup truck in Oklakhoma: both tons have the same effect. If the net global effects of CO2 emissions are negative, then they are negative regardless of where and how you emit the gas. The cost to (global) society is the same.

Man-made, outdoor air pollution is obviously nothing like that. Burning gasoline may be damaging in a big, crowded city but it’s innocuous in the middle of the Sahara. Plainly speaking, pollution is mostly a problem in big, dense cities or metro areas. In a dense but small population center, NO2 and other pollutants disperse into the surrounding sea or countryside; in a big (6–7 million people) but sprawling metro area like Dallas-Fort Worth, pollution levels will be lower than in similarly big but denser areas like Madrid. Just look at this NO2 map; the images are at the same scale:

In Spain’s case, automotive fuel taxes are similar all around the country; taxes on heating oil and other fuels are much lower, so when someone talks about ‘fuel taxes’ they mostly mean automotive fuels.

Taxing fuels in order to discourage pollution may make sense if the whole country were densely built up, but it’s not. In terms of population density, Madrid and Barcelona are in a league of their own; they pack 28% of our population into 3% of our area. Should someone driving in a rural area pay a pollution tax because Madrid and Barcelona are polluted? The question answers itself.

Now, an objection may be that perhaps Madrid and Barcelona account for something like 28% of automotive fuel consumption, and together with a few other big cities this may mean a major share of automotive fuels are burned in dense, built-up areas. But this is wrong; while data on how many litres are burned here and there is unavailable, it’s known that there are about 11,500 gas stations in the country; on the other hand, using this gas station finder, and looking at all fuels and brands, I find 732 in the province of Madrid and 778 in Barcelona province. That means about 13% of gas stations are in these two areas. In short, 80–90% of automotive fuel is not consumed in these provinces.

Are there other big cities with pollution problems? Maybe, but then again much of the fuel consumed in Madrid and Barcelona is burned in not-so-dense, low-pollution areas. Actually, in Madrid it’s not ‘much’ but ‘the bulk of’. The regional government keeps a network of 24 air quality stations, generally in the biggest towns of the province other than Madrid itself. Looking at the European Union’s limit on NO2, 40 micrograms per square meter averaged over a year, only one (Coslada, population 80,000) is exceeding it as of 2018. This means every other town and village, accounting for half of the province’s population and the vast majority of transportation fuel use, fell under the EU’s limit.

(An analysis of Barcelona is harder both because some of the ‘suburbs’ are in fact continuous with the city proper and because, well, I don’t live there and so I know relatively little about the province and city).

Notice also that some towns are much smaller than Madrid and Barcelona but may have relatively polluted air because of:

a) Heavy industry

b) Ports

It goes without saying that taxing diesel and gasoline cannot reduce either the pollution coming from steel mills or that coming from ships. Even if the government specifically sets out to tax ship fuel, it will be ineffective because fuel burnt near our coasts may as well have been loaded in Qatar. Shipping fuel can only be taxed or regulated internationally… as is already done by the International Maritime Organization.

But shouldn’t there still be a tax on pollution in the handful of cities that arguably may have issues with outdoor air quality? Well, the answer is that those cities already take their own measures. In Madrid, the parking meter will ask you to type your license plate number — and you will be charged more if your auto runs on diesel, predates the Euro 5 regulation, etc. Similarly, if your automobile doesn’t comply with Euro X or Euro Y, it may be barred from the city altogether on certain days. And so on.

There’s still another reason the automotive-fuel-tax-as-a-pollution-tax argument falls apart. Diesel is more polluting than gasoline, and yet it’s taxed less. Heating oil is in turn more polluting than diesel, and yet it’s taxed far less! Heating oil is also prevalent mainly in old homes… which of course tend to be the ones in dense, downtown areas. If one wanted to tax pollution one would target homes that haven’t converted to a cleaner fuel, and one would tax most heavily the fuel delivered to homes in Madrid and Barcelona proper. As you might imagine such an idea is never even floated; the closest local governments have come is subsidizing the replacement of old boilers.

There’s a tax on coal as well, but I didn’t include it in the accounting of CO2 taxes precisely for this reason: one can argue it’s a tax on particulate matter and nitrous oxides, not on carbon dioxide. (Coal plants also buy the CO2 permits discussed previously).

But aren’t fuel taxes a fee for using the roads?

That’s a better argument. Spanish roads are mostly free at the point of use — which is to say the various levels of government pay for them out of taxes. They’re free regardless of how your vehicle is fuelled, but since almost all vehicles on our streets and roads are fossil-fired it could be considered that part of the taxes on automotive fuels are a user fee that keeps the roads in shape.

But wait: every person who buys a motor vehicle in Spain pays a registration tax (separate from the VAT). And every person who owns one pays each year another tax as well. Together, these taxes contributed €3.1 billion to the public coffers in 2017. How much is that, compared to the sums of money spent on road maintenance?

The most recent data I can find refers to 2016, when €3.8 billion was invested in roads (see page 38). Now, this probably excludes some costs; for instance, I’m pretty sure it doesn’t account for the costs of policing the highways. I’m also unsure if the government’s definition of ‘investment’ includes services like snow removal, clearing vegetation close to roads, etc. But, on the other hand, around €800 million was invested by the regions, or Comunidades Autónomas as we call them; the regions manage tolled highways, so this investment cannot be considered a subsidy.

Disclosure: fully understanding how Spanish roads are paid for would take me more time than I’m willing to spend on this article, and I know my calculations are rough. The basic point on vehicle purchase and ownership probably don’t cover the whole cost of road and street maintenance, but they can only be one or two billion off. Even if one were to assume the net subsidy for roads and streets was €3 billion a year, that would mean about €10 per ton of CO2, or less than $12 per ton. In other words, the Spanish would still be effectively paying $40 per ton or more.

Why does the headline say $55 and not $50?

Spain’s effective CO2 tax was $50 or $51 per ton in 2017, but it will be higher this year for several reasons.

First, emissions in 2017 were anomalously high due to a drought, which caused a greater use of thermal power generation. Most years, emissions grow more slowly than GDP; in 2017 GDP grew by just over 3% but emissions jumped 6.9%: from 283.2 to 301.9 million tons. While it’s impossible to tell how high emissions would have been in the absence of a drought (which finished early this year), it’s reasonable to expect emissions to grow by about 2% in the face of 3% GDP growth. In other words, emissions would have been about 5% lower than they actually were.

Using more specific numbers, CO2 released from power plants grew by 17.8%, which is to say 13.3 million tons. To put this in perspective, overall emissions grew by 18.7 million tonnes (per BP’s Statistical Review). If emissions from the electric system had been stable, but non-electric emissions had grown just as they actually did (i.e. by 18.7–13.3 = 5.4 million tons), then emissions growth would have been 5.4 / 283.2 = 1.9%. Which, as mentioned above, is a typical rate for an economy expanding at a rate of 3% a year.

More importantly, in 2018 the trend is that the high-tax fuels are increasing their share of emissions. According to Cores, a good source for Spain’s consumption of hydrocarbons:

· Through August, use of oil products is up 3% year on year.

· In the same period, use of natural gas products is up only by 1.5% a year.

Oil products (e.g. gasoline, diesel) pay €150 per ton of CO2, or even more; natural gas users pay about €0.002 per kWh, which is to say about €3 per ton of CO2. As long as emissions from oil products are rising faster than those from natural gas and coal, the effective CO2 tax will rise.

Finally, even though the number of permits Spanish utilities will have to buy this year has declined (thanks to the recovery in hydro generation), the money spent on these permits will increase because the price of permits has more than doubled. It’s still too early to know what the average yearly price will be, but assuming they increase to €10 per ton (versus about €5 in 2017), there will be about €300 million in extra spending.

Who cares how much the Spanish pay ?

Well, the situation here is very similar to that of several Western and Northern European countries:

· Transportation fuels make up a big chunk of CO2 emissions — perhaps more than half

· Taxes on transportation fuels are very high — equivalent to over a hundred dollars per ton of CO2

· As a result, we pay an effective CO2 tax of $50 per ton or more, when looking at all emissions. This is not a tax on air pollution, as explained in the article.

· Even if one takes a climate sensitivity value of 3º C per doubling of CO2 as valid, a carbon tax should be lower than what we’re already paying.

I haven’t run the numbers, but the same situation broadly applies to the UK, France, Portugal, Switzerland, Sweden, and probably more. Coal-heavy countries like Germany and Poland have a smaller share of emissions coming from transportation, but still the effective CO2 tax in those nations is certainly in the tens of dollars. Or euros, if you will.

Looking across the pond, I’m not familiar with Canada’s energy and economy, so I’m not going to work through the numbers for that country. But it seems motor fuel taxes there are also substantial. Perhaps Canadian readers can chime in and compare the effective CO2 tax already in place with the (additional) tax just approved.

Furthermore, the small subsidy provided to roads and streets in Spain is likely even smaller (or non-existent, after accounting for vehicle taxes) in other European countries, where most or all highways are tolled.

Is there a bright side to this taxpalooza? You bet. I used to feel guilty about not doing anything to tackle climate change, but now it turns out we’ve been tackling climate change all along!