Toyota has recalled several million vehicles and even suspended sales in order to correct a pedal problem that rarely causes the accelerator to stick. I have not been able to find precise information about the level of risk, which concerns me personally as a Toyota owner. According to this report in the LA Times, the recall was based on a single, tragic incident in San Diego. That seems doubtful to me. Nevertheless, Toyota has described the problem as "rare." Meanwhile, this article in the Wall Street Journal estimates the direct costs at $1 billion, plus indirect costs relating to reduced consumer confidence in the quality of Toyota automobiles. So, Toyota is willing to spend $1 billion to correct a problem that rarely results in great harm. Kudos to them.
Contrast Toyota's response to low-probability disasters with the lack of attention by the US government and international community to low-probability climate disasters. In that case, scientific research has provided us some information about the probabilities of high mean-temperature changes, but much less information about the socio-economic consequences. According to Marty Weitzman's analysis of climate-sensitivity studies that informed the IPCC's 4th Assessment Report, there is an aggregate 5% chance of global mean-temperature increases of 7 degrees (Celsius) or higher over the next 150 years under "business as usual" (i.e., no policy to mitigate greenhouse gas emissions). As Marty notes, such significant temperature changes over such a short period of time are unprecedented in the history of human civilization on earth. Thus, he recommends investing in a "climate insurance" to minimize the risk of potentially catastrophic climate change.
The Toyota recall illustrates how both humans and their organizations typically and rationally respond to low-probabilities of high-magnitude harms. The disconnect when it comes to climate change is both interesting and somewhat frustrating. It might be explained to a large extent by the differing natures of the goods involved. While automobile safety certainly is, in part, a public good, there are enough private goods associated with it that Toyota and consumers both have a substantial interest in minimizing risks of potentially catastrophic harm. Climate stability, by contrast, is much more of a pure public good. It is what Scott Barrett calls (in his brilliant book, Why Cooperate?) an "aggregate effort" public good, where the supply depends on the efforts of all countries. The US cannot mitigate climate change by itself, although it may be able unilaterally to provide some geo-engineering measures. Toyota, alone, can fix the problem associated with Toyota automobiles. Neither the US nor any other country, acting alone, can solve climate change (given existing technologies). In addition, climate change remains, in large measure, a future problem with a nearly invisible present momentum. Thus, it lacks the salience of an immediate risk of death that defective automobiles carry. Finally, Toyota's recall, as noted above, was spurred by a "catalytic event" - a tragic accident in San Diego, which is so far lacking in the case of climate change. Unfortunately, policy makers often seem to require such events prior to taking remedial action. That is why some climate activists tried to argue that Hurrican Katrina was a climate-related event, although climate scientists could not substantiate those claims.
The key to investing in "climate insurance" over the next several years lies in resolving some of the collective-action problems that have hindered climate policy to date. Whether or not that can be accomplished without the extinction of the polar bears or some other such catalytic event remains to be seen.