Risk: Reason and Reality

Sandy and Climate Change. Two Ways of Thinking About Association or Causation

Different audiences will try to answer the question “Did climate change cause Superstorm Sandy?” in significantly different ways. I am thinking of two; the ‘thought’ community of scientists and policy makers and academics and ‘public intellectuals’/commenters, etc., who are ostensibly sorting this out with their brains and reason…and the real world of most people, who are sorting this out as most of us do most things - and certainly most risks - with an affective blend of fact and feeling, emphasis on the feeling.

For Audience One, it may be helpful to learn from another science that has struggled with just this question – association or causation - for more than a century…epidemiology. In 1965 Sir Austin Bradford Hill published a list of 9 criteria as guidelines for whether a statistical association also indicates likely causation. Most of them are widely accepted by epidemiologists who deal with this same challenge all the time…when do bits and pieces of evidence about health patterns showing an association make a convincing-enough case that it is reasonable to claim that X causes Y. (Note that this remains a huge issue of contention with lots of health conditions that have been extensively studied. The question about whether climate change caused Sandy is not alone.) By the way, Hill is credited for having come up with the fully randomized medical trial…and was one of the first to establish the link between smoking and cancer. Interesting guy for this conversation then. Note how relevant is his description of these criteria.

“Here then are nine different viewpoints from all of which we should study association before we cry causation . . . None of my nine viewpoints can bring indisputable evidence for or against the case-and-effect hypothesis and none can be required as a sine qua non. What they can do, with greater or lesser strength, is to help us make up our minds on the fundamental question - is there any other way of explaining the set of facts before us, is there any other answer equally, or more, likely than cause and effect?”

            Here are the six criteria helpful to thinking through the climate change/Sandy question (others are strictly health…like ‘specificity’ and ‘temporality’)

1. Strength of the association. Statistically large associations are more likely to be causal, although small ones can’t be ruled out just based on this one criteria. (For an epidemiology study this would be reflected in the Odds Ratio or Risk Ratio.) For climate change broadly, this is the ‘body of the evidence’ argument, the area in which doubters/deniers try to tear down the whole house by shaking even the smallest post on which the edifice rests. For the climate change/hurricane connection, this ‘strength of association’ is measurable in part by trends over time.

2. Consistency. Causation is more likely if different investigators look at the same question in different ways with different research methods and come to the same conclusions. Smart experts about the Sandy/hurricane link seem to have inconsistent interpretations, based on different ways of looking at the question. There is far less of this inconsistency about climate change in general.

3, 4. Plausibility, Coherence. Does the association make sense based on what we know about how things work from the basic sciences…chemistry, physics, etc. Again, looked at this way, different experts in the physical sciences seem to have different perspectives about the climate change/Sandy connection.

5. Experiment. Test the idea. For climate change, this would be modeling. Post-hoc analysis of Sandy …altering variable inputs to experimentally test various hypotheses…is probably already underway.

6. Analogy. How does the observed phenomenon compare or contrast with other similar phenomenon (This is where the paleo record is so relevant and convincing re: climate change in general, and why the historic record about hurricanes raises fair questions about the climate change link to Sandy.

            By these criteria, the case for climate change is a no-brainer. The case for hurricanes in general and certainly any one storm, bumps up against several of these tests.

            But remember, this way of looking at things is consciously thoughtful and analytical, and that’s not how most folks ‘do’ risk perception. That group makes sense of questions like this using a suite of mostly subconscious mental processes that help them figure out how to FEEL about a risk, a sort of Intuitive Hill’s Criteria. An influential factor is called ‘representativeness’ (one of Daniel Kahneman’s heuristics and biases.) We make sense of partial information – any one storm – in the context of the pattern it seems to represent. An example would be the spike in belief in climate change post-Katrina? Another would be the increasing belief in climate change recently, and diminishing certainty among the doubters/deniers – in the wake of a couple years of extreme weather. (Data on this are in a recent survey.)

            We are also more likely to jump to the conclusion that 2 + 2 = 4 regarding risk when things get real (local weather now v. global climate someday) and local (could happen to ME). (Again, see that recent survey for data on this.) Further, we’re likely to assume the worst (because humans are instinctively loss averse…more threatened by loss than pleased by gain), and especially about threats that bear out catastrophically rather than in more chronic sorts of ways…lower profile changes spread out over time and space. Nasty storms like Sandy fit all these psychological criteria.

            So by the intuitive Hill’s Criteria people use to make affective judgments about the question – did climate change cause Sandy – a lot of people are ready to leap to that conclusion, even though by more rigorous scientific standards the connection, though plausible, is less robust. The evidence of that leap is clear. Note how the stormy winds of Sandy have blown the whole public conversation about climate change in an entirely new direction. A week ago it was ”Why aren’t the candidates talking about it.” Now it’s about whether climate change caused this storm…and whether it’s causing extreme weather generally. And the previous conversation was more intellectual, political punditry. Not now. The conversation about climate change has more urgency and passion and emotion, given the suffering this storm has caused.

            That suggests that this awful storm and it’s far-too-many-victims will almost certainly play a role in deepening public concern about the larger question we really need to come to grips with; Is climate change real, now, anthropogenic, and if so, what should we do about it…and even more…HOW URGENTLY. So maybe, in the end, the question isn’t “did climate change cause Sandy?” Maybe a more important question is “What role will Sandy play in the public’s concern about and response to the threat of climate change?” I think there is reason to believe that this storm will play a small but real role in spurring the public to deeper concern, and action.

comments powered by Disqus