If a dinner party is beginning to fizzle out, I always find it helpful to try one or two post-prandial parlour games. For example, you might want to try separating your guests into two groups. Each group is given a list of the same numbers and asked to estimate the list’s average. The only difference is that the members of the first group are given a list of numbers in ascending order whilst the second group is presented with the same numbers in descending order. And this is the fun bit: The group who averaged the ascending list will tend to underestimate the average, but the group who averaged the descending list will overestimate.
Okay, so it’s not as much fun as everyone throwing their car keys into a dish, but it does illustrate an important mental heuristic employed by all healthy minds. As one takes on additional information to help form an opinion, one is unduly influenced by the order in which the information is introduced. In this case, those who estimated the ascending list will start out thinking ‘this is a list of small numbers’ and then think ‘well, maybe not so small after all’. The other group will start out with ‘this is a list of large numbers’ before modifying that view. Both groups will converge as they re-evaluate, but each starts from a different anchorage; hence the name for this particular cognitive bias – ‘anchoring’.
Anchoring is one of a broader group of cognitive biases known as ‘Continued Influence Effect’ (CIE). The causes of CIE are numerous, but the main point is that none of us is a blank slate when confronted with new evidence; we all labour under preconceived ideas and are conservative when it comes to re-evaluating them. Nothing remarkable here, you might think.
Except, anchoring and its brethren biases do feature rather prominently in the psychoanalysis of climate change scepticism. It turns out that such cognitive biases are not the hallmark of a healthy mind after all – just the indulgence of your average, cognitively impaired climate change denier. Take, for example, this article written by Brigitte Nerlich, Professor of Science, Language and Society at the University of Nottingham. It paints a sorry picture of hapless conspiracists, mentally enfeebled by exposure to that dastardly Climategate nonsense (portraying, as it did, a totally unfair image of climatologists plotting and hoaxing their way to the top of their professional ladder). No amount of counter-evidence can help those duly exposed; they remain hopelessly anchored to a prejudice formed by what was surely a storm in a teacup. At least, that is the understanding that Professor Nerlich is anchored to.
Meanwhile, at the dog end of knowledge, climatologists are beavering away, seemingly exempted from accusations of anchoring (at least according to the likes of the dewy-eyed Professor Nerlich) because they have consensus to guide them toward the light. As Barack Obama will tell you, scientific consensus is a truthy thing; and if consensus doesn’t work for scientists, they can always try looking at the data.
Fortunately, however, not everyone is as naïve when it comes to evaluating climate science; there are those who are prepared to take seriously the possibility that anchoring plays a fundamental and central role, framing our scientific understanding of how the climate could evolve under anthropogenic influences. Take, for example, Jeroen P. Van der Sluijs, Professor, Theory of Science & Ethics of the Natural Sciences, Bergen University, whose curiosity was piqued way back in 1997 by making the following simple observation:
“The consensus-estimate of 1.5°C to 4.5°C for climate sensitivity has remained unchanged for two decades in international assessments of the climate issue. Nevertheless, during these years, climate research has expanded enormously, and scientific knowledge and complexity of climate models changed accordingly.”
In a thesis titled ‘Anchoring amid Uncertainty’, he investigated what was seen as ‘a remarkable stability’ in the consensus-estimate, and he identified a ‘repertoire of sources from which the experts managed to acquire flexibility in maintaining the same numbers for climate sensitivity whilst not ignoring changing scientific ideas’. Such sources of flexibility included:
“…(1) the modes of reasoning, (2) the types of uncertainty accounted for, (3) estimates of the best guess rather than changes in the range; (4) the connotation of the range, (5) the definition of climate sensitivity and (6) the implication of the range.”
The paper moves on to say:
“The remarkable, ostensible stability of the climate sensitivity range may play a significant role in holding together a variety of different social worlds in a situation where the state of scientific knowledge does not grant the 1.5 to 4.5°C range a higher scientific status than an ’educated guess’. But the stability can also be seen as a function of an implicit social contract amongst the various scientists and policy specialists involved, which allows ’the same’ concept to accommodate tacitly different local meanings.”
The point to take away is this: Climate science is a post-normal science in which high-stake decisions are to be taken under conditions of deep uncertainty, beset by disputed values. And, provided there is enough uncertainty, ambiguity and ambivalence in the formulation of a particular proposition, it is surprising how much wiggle room there will be available to contrive adherence to it, notwithstanding emerging evidence. One only has to exploit the uncertainty1 and endlessly re-frame the current understanding to suit one’s purposes. By such chicanery, one can even simultaneously accommodate a plurality of understanding, as long as the individuals concerned are anchored to the same social contract. Anchoring is more than just a trap that climatologists can fall into. By dint of the post-normal character of climate science, anchoring can actually become the climatologists’ scientific objective. As the Van der Sluijs paper puts it:
“…the maintained consensus about the quantitative estimate concerning a central scientific concept in the anthropogenic climate change field, namely climate sensitivity, operates as an anchoring device in science for policy”.
Two more decades have passed since Professor Van der Sluijs made his initial observation regarding the remarkably persistent level of uncertainty ascribed to climate sensitivity; two decades in which there has been a lot more scientific advancement but still little movement in the consensus estimates.2 The anchorage is still holding firm and no one seems inclined to break their social contracts.3 Indeed, even if there were to be a narrowing of estimation – say, towards a conservatively low figure – that still wouldn’t be a cause to change direction. There is still ample uncertainty within impact analysis to accommodate such a development and yet stick to a preconceived notion of risk. You say it isn’t going to get so hot after all? So what? We have now decided that the environment is much more sensitive to heat. Wiggle, wiggle, wiggle.
Anchoring is not an inherently bad thing. On the road to enlightenment we are all on a Bayesian journey, and we all set off from a house of prior beliefs. How one feels about someone’s anchoring very much depends upon how one feels about their beliefs and how much they accord with currently accepted social contracts. As far as society is concerned, the anchoring of the in-group is the settlement of science upon the moral high ground; the anchoring of the out-group is denial. Not that Professor Nerlich can appreciate the ubiquity of anchoring. In her view, climate science is unanchored, sailing majestically on its way through a sea of knowledge, heckled by climate change deniers who remain pathetically shackled to the sea floor by a cognitive bias that is uniquely their own. Not for the first time, those who are the most affected by a cognitive bias are the last to appreciate it.
 Let’s not make any bones about this. When Professor Van der Sluijs says the initial estimates for climate sensitivity were no more than an ‘educated guess’, he wasn’t making it up. To quote IPCC Lead Author, Stephen Schneider: “The range was never established by a firm decision-analytic protocol in the first place, but rather was a heuristic from a responsible, but somewhat sloppy, community in the 1970s.” It is sobering to think that this is what climatology has become anchored to.
 Professor Van der Sluijs wryly observes that, “IPCC experts felt a great need for unambiguous scientific evidence to change the range. Apparently, however, there is no equally great need for evidence to maintain the range.”
 Once ‘signed’, such social contracts have a way of self-sustaining down the years. For example, Professor C.J.E. Schuurmans said of the Toronto 1988 Conference on the Changing Atmosphere: ‘…the conference was not willing to drop these [climate sensitivity] numbers, as they were adopted from the Villach report, which forms the scientific basis of conferences such as this one. In general, questioning scientific judgments at this conference was not popular.”