A little while back, Josh asked the question ‘how do catastrophes factor into our calculations?’ He asked the question in the context of cost-benefit analysis, but it’s a critical question for almost every facet of climate change research. The journal Nature used this week’s issue to shed some light on the subject. Since you have to pay to access Nature, I recommend the short but informative rundowns of the content from Environmental Capital and RealClimate. According to these studies, if we as a global society want to avoid catastrophe climate change, then we need to cap the world’s emissions at 1 trillion tons of CO2. Considering that we emitted one-third of that in the past 9 years, we could be in for a rough ride.
Science models can only tell us (to a degree of certainty) where to expect a greater chance of catastrophe, but how can we translate that into economic analysis. There are a couple ways to do it. First, you can use a low discount rate to better internalize the possiblity of a major disaster. Discount rates are often huge sticking points for economists arguing about each other’s models (take Bill Nordhaus vs. Nick Stern, for example) and there’s still no consensus as to what is the ‘correct’ rate.
Second, you can use risk analysis to better understand how bad disasters can be. A recent paper by Carolyn Kousky and Roger Cooke of RFF (which you really should read) does an excellent job of laying out some of these risk considerations, and it definitely provides some food for thought. The three major risk considerations are:
- Micro-correlations – The events of El Nino years are a good example. If you look at events in isolation (heavy rains and mudslides in Cali, poor fishing seasons in Peru, etc), they might not be noticeable, but that would lead you to underestimate your risk of major damages . If you take the broader picture though, you can see that disasters can be correlated and you can accordingly adjust your risk assessments.
- Fat tails (of a bell curve/normal distribution) – Basically, extreme outcomes are more likely. Disasters and extreme events compound on each other to create fat tails, which increases solvency risk of insurance, meaning that you might have way more damage than you can afford. A good example of that is Florida with their Citizen’s Property Insurance Corp., which has $3 billion to cover its $450 billion worth of exposure.
- Tail dependence – Tail dependence is essentially the likelihood that bad outcomes occur together. It is distinct from micro-correlations and fat tails, and they explain it much better than I can, but it relates to the idea that insurance lines can be independent of each other until there is a disaster, at which point they become dependent.
If you don’t account for these issues, you can severely underestimate your risk related to climate change. Cost-benefit analyses seem to have a pretty difficult time incorporating these factors, but there is still much research to be done on this front. It could be a while before we have reliable methods for incorporating catastrophes into our modeling. I don’t know about you guys, but I’m still miles and miles away from fully understanding this stuff. At least I can tell the difference between weather and climate, so I got that going for me, which is nice.