17 July 2014

Guest Post: Kerry Emanuel Clarifies a Recent Quote in the NYT

The comment below is by Kerry Emanuel, at MIT, who is clarifying a recent quote of his in the New York Times.
I would like first to thank Roger for allowing me to post this response to the article about John Christy by Michael Wines in Tuesday's New York Times. Although I was quoted accurately, the context in which the quotation was phrased distorted its intended meaning.

Several weeks ago, I had several phone conversations with Mr. Wines about the work of John Christy. In those conversations, I emphasized the value of skepticism in science and also said that I agreed with some elements of John's point of view, in particular, that projections are still highly uncertain, that climate models leave a great deal to be desired, and that some of the decisions that have to be made about how to deal with climate change are very tough indeed. Wines asked me to explain where I differ from John. I told him that we differ primarily in our assessment of the magnitude of climate tail risk. Wines asked me to explain what I meant by "tail risk", and I offered the metaphor of advising a small girl whether she should cross a busy street to catch her bus (a metaphor I have used before).

Unfortunately, the positioning of the quotation within the article makes it seem as though I am suggesting that John is the kind if person who would let the girl take the risk. I state here that I have absolutely no reason to question John's motives; indeed, he strikes me as the sort of person who would risk his own life to save a child who wandered into a busy street. My metaphor was intended only to illustrate the nature of tail risk.


  1. The problem with this explanation is that "tail risk" is not a "substantial chance." Tail risk are events that are outside of three standard deviations from the mean or events that happen 0.03% of the time.

    1. I agree. If we stick with this (inappropriate) metaphor, a more accurate version would be advising a small girl to cross the street when the walk light is green. That's a lot closer to the "tail risk" than jaywalking a busy street.

  2. Stupid metaphor. Made worse by overstating risk. If he insists on using such a stupid construction, it would be more appropriate to frame it as how many old ladies should the girl slam into with her bicycle in trying to catch this bus instead of the one coming in another minute.

    1. Calm down. The "tail" emphasizes that it's on the tail end of a probability distribution. How can that be overstating it?

  3. There are so many analogies and metaphors running around trying to capture climate change risks... it's as legion as the number of bears living in the woods ;-)

    But, as far as the risks go, is it not a question of dual certainty/uncertainty? Meaning, it's not just about "the percentage chance of death from car impact" that's on the table, but also the question of "How busy is the street really?"

    Take smoking for example... A study within the last ten years pegs the chance of a "heavy smoker" to develop cancer at 25% ... Are we apparently orders of magnitude MORE certain that our current actions will lead to a catastrophic diagnosis on Earth..?

    ...And even then, in the face of direct knowledge of the future expected outcomes, people still choose to smoke in the present. Ergo (in my opinion), short of finding a way around due-process and/or shutting voices out completely, I don't see how exactly any of the solutions involving controlling behavior or forced acceptance of infrastructure are going to work on the scale required vis-à-vis projections in energy demand and carbon sensitivity.

  4. Thank you for that clarification Kerry, and thank you Roger for posting it. I read the piece on Christy in the NY Times and thought perhaps the "child-crossing-the-street" metaphor was not the most apt for climate risk. There's the potential for misunderstanding which I see in hindsight.

    But I think we need other analogies for communicating climate risk. The decision to cross or not cross the street at any given instant is a binary one. You go or you don't. You are either in the road or you are not. You are not saying "I will cut my street presence by 5%."

    But with GHG emissions we are not talking about a binary choice to emit or not (at least in serious policy discussion), we are talking about a continuum of possibilities, such as "we will cut carbon emissions by 5% by 2020." And then we are considering how that kind of decision would affect the physical climate system, and the tail risks as represented by heat waves, cyclones etc.

    I don't know what the best analogy would be in this case. Smoking works on some levels in the sense that as a smoker you can cut down the risk on a continuum, such as cutting back from two packs a day to one, etc. But this doesn't completely remove your health risks. Non-smokers still get cancer and cardiovascular disease due to other risk factors - environmental exposure, genetics, diet, etc.

    So with the climate system, "tail risk" would still exist even at pre-industrial GHG concentrations. There would remain a rate of occurrence of floods, droughts, cyclones, heat waves etc. Less of a risk than at higher radiative imbalances? How much less?

  5. I am pleased that Kerry Emanuel has clarified the statement referenced in that article about John Christy. It does not take too much investigation to see that when it comes to caring about people, especially the poor and the disadvantaged, John Christy doesn't talk the talk, he walks the walk. It is most unfortunate that the article gave the impression that Kerry thinks John reckless. I hope this post helps to clarify that is not the caae. Better would be a clarification in the Times.

  6. What is left out of the bus analogy is the gain from catching the bus. In the girl case the missed bus is sold as no big deal.

    In the climate case, the tail risk must be measured against the very high loss of economic activity from decarbonisation.

    In most situations people faced with a small probability of losing something big, but a definite loss will often risk the big loss. I for example do not have theft insurance on my car. The definite loss of the policy is far worse in the long run than the small risk my car will be stolen. However I have third party insurance, because the risk I cause a crash causing hundreds of thousands of dollars' damage is quite high, and I can't afford that.

    Mr Emanuel needs to specify that by missing the bus the girl is missing the opportunity of something very important. Say a flight overseas. If he does that, of course, his analogy becomes much more accurate but also looks very weak indeed. Most people will risk a road crossing if the bus is the last one to the airport.

  7. This seems like a nice way for Kerry Emanual to clear his conscious without clearing the record. Far more people read the New York Times than Roger Pielke Jr.'s blog (sorry Roger). Those people are still left with the impression that John Christy does not care about children, and, judging by the comments above, they will also have the impression that 'tail risk' is a much higher risk than it actually is.

    In the world of climate change, false impressions are like magic beans that grow into giant bean stalks. The greatest false impression of all is the one generated from the facts that almost all scientists agree that adding CO2 to the atmosphere, all else being equal, will tend to cause some warming, and that there was some warming in the late 20th Century. Those innocent little beans have grown into the giant consensus that 97% of scientists believe that humans will be the main cause of inevitable, catastrophic global warming, if we don't drastically change our ways!

    Then other little magical stalks (fallacies) sprout from the original magic vine, like the notion that global warming crisis skeptics are an ignorant little group that denies the holocaust and evolution. Or that skeptics are receiving huge amounts of money from 'Big Oil', or that they don't care about the future or even the fate of the little bitty children!

    Kerry Emanuel has watered and fertilized the ever-growing global warming fallacies at the New York times. Coming here to make a correction is like trying to chop down those massive vines with an overripe banana.

  8. It is interesting that the official guidelines for pedestrians crossing the road do not engage in statistical reasoning but urge to follow rules of thumb. Thus when Kerry Emmanuel says "In assessing risk, one has to estimate the probability distribution of the event (car colliding with girl), convolve that with an outcome function (girl likely dies if hit), and account for the cost of mitigation (take 5 minutes to walk to a traffic light)" -- this looks irrelevant to everyday decision making.

    Pedestrians' risks are not calculated but dealt with in a pragmatic way. Here is a link to the relevant sections of the UK Highway code. Note that "Children should be taught the Code and should not be allowed out alone until they can understand and use it properly. The age when they can do this is different for each child. Many children cannot judge how fast vehicles are going or how far away they are. Children learn by example, so parents and carers should always use the Code in full when out with their children. They are responsible for deciding at what age children can use it safely by themselves." Through observation and learning children become confident in making these decisions, and the two alternatives presented in Emmanuel's example are extremes (1% risk v safe passage at traffic light--mind you that a green light still poses a risk).

    General rules for pedestrians are: First find a safe place to cross; Stop just before you get to the kerb; Look all around for traffic and listen; If traffic is coming, let it pass; When it is safe, go straight across the road – do not run.

    Children learn how to become competent road users and traffic participants before they are adults, before they have a driving licence, and independent of knowing what "risk distributions" are.

    Perhaps we should apply the same heuristic to climate policy.

    I did a quick search for crossing rules in the USA but could not find a similar set of rules as the UK has (apparently these vary across states and most pertain to marked crosswalks). Maybe this cultural difference has a bearing on the choice of metaphor in the first place? Crossing the road in other locations than signed crosswalks seems common, even for children, outside the USA.

    1. Upon re-reading the post it occurs to me that maybe, just maybe, the real message advanced by Kerry Emanuel is that there are children who must learn, and adults who know. These roles are assigned to the public, and to climate scientists, respectively.

    2. Dr. Emanuel can all of the metaphorical lipstick on that pig of an assertion he wants. He misses the main point: He has severely misrepresented the nature of the risk and the ways to deal with low probability high impact events. He has even missed the idea of valid risk. A young child crossing the street is not really a low probability high impact risk. Additionally, the variables that make up the risk are well defined, tested and provable. Unlike the climate crisis.
      Well at least he clarified the impression he gave that climate skeptics want to kill kids.
      Keep it classy, Dr Emanuel. There is lots of lipstick for that pig.

    3. ReinerGrundmann,

      So climate scientists are "adults"?

      One would be hard-pressed to conclude that by reading what passes as "communication" for many of them ("mailed fist" or "dark alley" anyone?)

      But your analogy is not completely useless, as it does bring to mind "The Emperor Has No Clothes".

  9. JohnM

    I meant to say that in the analogy Emanuel seems to equate climate scientists to adults, the public to children. It is not my view. Sorry if this was not clear.