|An example of error bars, shown in red|
Friday, January 4, 2013
Climate change and the art of airplane maintenance
An interested reader sent me a thoughtful email in response to last week’s post on global warming. The reader asked a number of questions and expressed some concerns. I welcome feedback of this sort because it helps me see how my thoughts and attempted explanations are received, and gives me the opportunity to learn from others. In this week’s post I will attempt to clarify and elaborate on some important points by answering questions paraphrased from that correspondence.
Question 1: Even if only a small percentage of models now being used predict very serious consequences of global warming for public health, or drought, or forest fires, or other types of ecosystem damage, or extinctions, etc., doesn't that imply we would be wise to take action on these models just in case they might be right?
Response 1: This question gets at a key attribute of how people respond to uncertainty: what to do about high impact but low probability events? As I have stated before, we are wired to have one of two responses to uncertainty: to ignore it or to overreact. But what would a thoughtful, rational response be? It would depend on the degree of impact and probability of occurrence, but also on the cost of actions to reduce the risk.
As an example, consider airplane design and maintenance. Airplane crashes are rare. Per mile, airline passengers are 50 times less likely to be killed than people traveling by car. When traveling by plane, people tend to ignore the risk of a crash or to be overly troubled by it, e.g., showing far more anxiety then they would if traveling in a car and facing equivalent or greater dangers (since a plane travels more than 50 times faster than a car, the chances of dying at any given moment in each travel mode are more similar). Nevertheless, there is a risk of crashing, and airlines, airplane manufacturers, and government agencies have the opportunity to make rational decisions about how to handle that risk. Why not make an indestructible plane, or as comedian Steven Wright put it, “Why don't they make the whole plane out of that black box stuff?”
Engineers will give you a straight answer to this question, emphasizing impracticality because of the weight of “black box stuff.” I, however, am skeptical that the weight itself is an insurmountable obstacle. After all, a 747 is designed to takeoff at a maximum weight of nearly 1 million pounds, and an Airbus A380 at nearly 1.3 million pounds. It may be impossible to design the entire plane using black box technology, but I imagine that the interior of a 747 or A380 could be gutted and refitted with a small black box-like cabin suitable for just a few passengers who would be protected from nearly every form of harm. The reason we dismiss this possibility is the cost. Operating costs for such a large plane are enormous and usually shared by the hundreds of paying passengers onboard. Not many people would pay 100 times the going price of airfare to reduce already low chances of dying in a plane crash to near zero. Airplane maintenance schedules raise the same issues. More frequent and more extensive inspections would improve safety but would drive up costs. In the end, airplane maintenance schedules represent a rational balance between the two.
This sort of analysis doesn’t encourage the elimination of every risk. Instead it encourages balance. This balance is what we should be striving for with respect to global warming. The real question about global warming is whether the benefits of reduced risks outweigh the costs of taking action today. In order to answer that question, we must highlight scientific uncertainty and examine it carefully with public costs and benefits in mind, which is the approach I advocated last week. I personally believe that we should be enacting policies to address the potential damage from global warming. If we perform and emphasize analyses on climate change like the airplane design and maintenance ones described above, we will disempower unwavering global warming skeptics and foster the development of rational policies.
Question 2: To the scientists I know and admire, uncertainty is a challenge, a focus, and something to highlight. Isn't this what scientists are especially proud to do? What are you implying: that scientists are uncharacteristically uncomfortable with and downplaying of uncertainty when it comes to global warming; that science is easily undermined by self-interest and discomfort with uncertainty in general; or that there is something unusual about climate change that makes it more difficult to study in objective scientific ways?
Response 2: Last week, I focused on an explicit strategy chosen by climate scientists to downplay disagreements among themselves when addressing the public about global warming. In no way does this strategy apply to debates among scientists in scientific arenas. They chose this strategy with the good intentions of breaking out of the dueling expert media formula described last week, and of conveying their genuine concerns to the public. This strategy had a cost, though, in terms of framing the debate about the certainty of global warming’s existence, when I believe a debate focused on the range of possible climate change outcomes and potential costs of avoiding the more extreme possibilities would have been more productive.
However, climate scientists are not alone in their struggle with how to portray uncertainty to non-scientists. In fisheries, scientists are often asked to recommend fishing quotas for the next year. This is a daunting task because our current understanding of the status of the fish population is always uncertain, our ability to estimate how many new recruits will be added to the population is even more unpredictable and, in many cases, we are asked to give this advice without a clear idea of how the following year’s quota might be adjusted in response to new information (e.g., not at all versus the rocket science approach). Furthermore, various sectors of society, whether they be different fishing fleets or non-fishing interests, will have different opinions as to what quota, or quota system, will be best.
Typically, fishery scientists will address uncertainties in two ways. First, they will recommend a quota but will bound it with error bars, a graphical technique that shows a range of values likely to contain the correct answer. Error bars are an honest attempt to convey uncertainty. It is my experience, though, that managers often view them skeptically as both an admission that fishery scientists don’t know the right answer, and as latitude to choose any quota value within the range of the bars.
Second, fishery scientists, along with climate scientists and every other scientist I’ve ever met, will talk at length about the uncertainties in their field, but from a scientific perspective focused on the frontiers of discovery. This framing of uncertainty does not translate directly into currencies of relevance to interest groups and policy makers. Whereas scientists often express the need to dumb down the science for policy, in reality it must be translated from the nuanced and complex scientific world to the equally but differently nuanced and complex policy world.
In sum, it’s not that scientists are self-interested or uncomfortable with uncertainty. It’s that they do not have the expertise or do not make the effort to express scientific uncertainty in useful ways for policy makers. In order to craft smarter policies, we need more emphasis in the policy process on bridging the gaps among scientific disciplines and especially the gap between the scientific and policy worlds. Addressing uncertainty more explicitly is a key step in doing so.
With many thanks for these good questions,