Global warming: science, policy, politics

Since my back-and-forth with John Robson three years ago, I've been continuing to argue with people about global warming. Here's my attempt to summarize the science, policy, and politics.

1. What's happening

No matter how complicated the Earth's climate is, we know that it can't violate conservation of energy.

In terms of energy, the Earth is basically a system with one input (a steady stream of solar energy) and one output (thermal radiation). A warm object radiates thermal energy into space, the same energy that you feel when you hold your hand over a hot stove. And a warmer object radiates more energy than a cooler one.

So the Sun warms the Earth to the temperature where incoming energy = outgoing energy.

By digging up and burning fossil fuels, we're raising CO2 levels in the atmosphere. Because CO2 reduces outgoing thermal radiation, input is greater than output, and so additional energy accumulates. It's like filling a bathtub; you don't need fancy models to see what's going to happen next.

We can get a sense of the scale of the process in this graph, from Wikimedia: the Earth's enormous, so it takes a huge amount of energy to increase its temperature, and we can see from post-1978 satellite measurements that the Sun's not getting any brighter.

Looking at a more detailed view, we can see extreme heat waves which would not have occurred without global warming. In 2003, for example, there was a heat wave in Europe that killed more than 70,000 people.

These maps are constructed as follows: Divide the surface of the Earth into equal areas, using a resolution of 250 km. For each area, find the average June-July-August surface temperature, from 1951 to 1980. Use this to define a probability distribution. Measure the standard deviation of the distribution. For any given year, the areas for which the average summer temperature is more than three standard deviations from the mean should cover only 0.1-0.2% of the planet. We can see that these areas - the large brown patches - are now covering between 6% and 11% of the planet!

If this is such a big problem, why haven't we done anything about it? The 1992 business-as-usual CO2 projection vs. actual CO2 levels:

The answer is that cooperation is surprisingly difficult, even in the face of a clear threat. It's an example of a collective action problem: the costs of acting are borne by the individual, while the benefits are spread across everyone, so everyone tends to free-ride and point fingers at others. It's a problem of individual incentives, not lack of awareness.

2. What we should do

Basically, we need to electrify everything, including transportation and heating; and generate much more electricity without using fossil fuels (hydro, nuclear, wind, solar).

We know that replacing the existing fossil-fuel-based infrastructure will be very costly. If you talk to economists, they're pretty much unanimous: the most cost-effective way to accelerate the transition away from fossil fuels is to use a steadily rising price on carbon, basically a sales tax on fossil fuels, with the revenue used to reduce income taxes. Since it's no longer free to dump fossil CO2 into the atmosphere, this gives individual households and businesses a direct incentive to reduce their fossil fuel usage. Joseph Heath explains the details.

British Columbia has had a carbon tax like this since 2008. It turns out the net economic cost is nearly zero: The increased drag from the carbon tax is almost exactly offset by the reduced drag from income taxes. Paul Krugman explains the basic idea.

A carbon price - as opposed to quotas - is also the most effective way to coordinate action internationally. Quotas are zero-sum: more for me means less for everyone else, so it's nearly impossible to come to agreement on how they can be allocated fairly. It should be far easier to agree on a common price. William Nordhaus explains the advantages of agreeing on a common price instead of trying to allocate emission reductions.

The IEA has worked out the carbon price path required to stabilize CO2 levels at 450 ppm: $20/t in 2020, $100/t in 2030, $140/t by 2040.

Carbon price required to stabilize CO2 at 450 ppm

Once CO2 levels are stable, it'll take about ten years for temperatures to stop rising.

Canada's actively pursuing this approach, with a minimum carbon price of C$10/ton in 2018, rising $10/t each year to $50/t in 2022. We have a self-interested reason to do so: we export oil (in fact the federal government just took over a pipeline to export oil to Asia instead of the US), and the Canadian oil sands are being targeted by the environmental movement - they're too big to stay out of the spotlight.

Canada's argument is that we're doing our part to reduce our demand for fossil fuels, using carbon pricing, and that this makes more sense than trying to withhold supply by preventing the oil sands from expanding. If we block off one source of supply without reducing demand, it's like trying to fight obesity by blockading a McDonald's. If the idea is that reducing supply will drive up the price of oil, a better approach is to use carbon pricing to raise the price directly.

3. How to get people to do it

At the national level, we can use carbon pricing to align people's incentives. But that doesn't answer the question of why a particular national government would seek to act, rather than free-riding and putting their political capital into other issues.

It's difficult to generalize, because it's going to depend on the political context in that particular country. In Canada, for example, the Harper government's approach was to do nothing, and to accept or even deepen polarization (muzzling government scientists, withdrawing from Kyoto). Harper figured that he just needed 40% of the voters on his side, and having 60% of the country be angry at him might even be helpful. Trudeau's approach, on the other hand, is to take a compromise position: use carbon pricing to reduce demand for fossil fuels, and continue shipping oil. Naturally he gets attacked from both sides, but he believes that he can get most Canadians to accept this compromise.

If a country doesn't happen to have a confluence of political interests which line up in favor of action, it may be possible to use diplomacy. Hans Morgenthau describes the three elements of diplomacy - how to get another country to do what you want - as persuasion, compromise, and threats. (Machiavelli's History of Florence: "It is customary with those who wish to obtain a favor, to make use either of prayers, presents, or threats, that pity, convenience, or fear, may induce a compliance with their requests.")

For example, William Nordhaus's most recent proposal is a climate club, giving reluctant countries a direct incentive to act based on tariffs.

In addition, my guess is that whenever there's an extreme weather event, people will put a great deal of pressure on their own government to do something.

A major non-governmental factor is the environmental movement. On the one hand, environmentalists have raised awareness of climate change, putting pressure on governments to act, and the Sierra Club has been leading a successful campaign to shut down coal-fired power plants for economic reasons. On the other hand, environmentalists are often criticized for attacking supply-side projects, like the Keystone XL pipeline, rather than trying to push for more abstract demand-side policies like carbon pricing. David Roberts explains the logic: it's simply easier to get people in the streets to protest against a pipeline. "It is true that each individual fossil fuel project is only a tiny contributor to the totality of climate change. But the industry as a whole is a high-functioning, high-earning, high-influence death machine that is driving civilization toward disaster, knowingly so. Some sand has got to be thrown in the fucking gears." In some ways that feels like giving up on policies like carbon pricing, and resorting to targeting a scapegoat for its symbolic value.

If the multilateral, cooperative, orderly approach doesn't work out, the emergency fallback is geo-engineering.