re: #68 recusancy

For any econ nerds here’s some notes of Krugman thinking through a “toy model” on carbon pricing and investing.

Sorry, I read through this and kept saying to myself, but lambda is zero. I have to tell you where I am coming from on this. I have a Master’s degree in Statistics, and although most of my career has been spent in the area of computer science, I’ve done some Box-Jenkins style time-series analysis work, primarily to analyze bug reports on software. ;-)

First, current CO2 levels are killing people. Not by drowning in rising seas, or dying of heat stroke in heat waves. Higher CO2 levels favor sepsis in wounds. Where this is a big problem is in bed sores in the elderly or otherwise bedridden people. Is it a contributing factor in other health problems? Probably, but I haven’t seen the research.

When talking about current and even future CO2 levels, they are not high relative to historical levels on Earth, but they are about as high as they have ever been since the human race evolved. So that is an additional huge, unknown risk.

For those reasons I think that reducing carbon emissions is important. What about the effect of CO2 on global warming? We are back to lambda is zero. A subset of statisticians understand time series analysis (TSA). Probably there is a much larger population of physicists and financial analysts who understand TSA—they use it every day. What makes TSA so difficult? The assumptions of independence and normality which most parametric statistics is built on do not hold when working with time series.

If you have TSA tools available on your home computer and spend a few minutes with the long term data on CO2 and temperature, you too will conclude that CO2 levels lag global temperatures by about 800 years, and there is no forcing. Huh? In fact when volcanoes increase CO2 levels, global temperatures promptly go down—I can point to half a dozen cases in the last couple of hundred years, and some bigger drops before then. But that is almost certainly due to the sulfur dioxide that accompanies the CO2.

I don’t want to go too deep here, but make it understandable to people on both sides of the AGW argument. Hmm. Let me start with methane. You will see lots of articles saying that methane contributes 65 times as much as CO2 to global warming. But methane decomposes rapidly in the atmosphere to water and CO2. If you don’t believe me, test the outside air for methane. Unless you do it in the middle of a cow pasture, you won’t find a thousandth as much CH4 as CO2. So anyone who sells methane as a powerful contributor to global warming is selling snake oil. The only way it contributes is by becoming more CO2.

But, but…what about CO2? The most potent greenhouse grass in terms of its contribution to the temperature of the Earth is water vapor. What happens when you mix CO2 and H2O? In theory, you get carbonic acid H2CO3, in practice you get a lot of various ions as well, even when you mix CO2 and water as gasses. To make a huge, complex story simple, no one who knows how to analyze the data* is willing to accept the hypothesis that CO2 when added to the water vapor already present, is a greenhouse gas globally. Locally, in deserts and cold, dry arctic locales, CO2 does have a net greenhouse effect. In hot, humid areas, CO2 seems to increase precipitation and may have a net cooling effect. (Again, not enough data to be sure.)

Remember again, that as a statistician, I am telling you that increasing CO2 kills people. But it doesn’t seem to cause global warming. What does? Sunspots! But I’m sure you’ve heard that from the global warming deniers too. (I know that global warming is happening, I just don’t think CO2 is causing it.)

* On the other hand, if all you have is a hammer, every problem looks like a nail. :-(