The IPCC consensus process: lessons from the Manhattan Project, mapping the human genome, measuring GDP… and Charles Darwin

“False facts are highly injurious to the progress of science, for they often endure long; but false views, if supported by some evidence, do little harm, for everyone takes a salutary pleasure in proving their falseness.” – Charles Darwin

Call me a big fan of Climate, Etc. and Judith Curry. She’s insightful, and fresh, and interesting; her output and energy level are prodigious. And her blog lives up to Darwin’s standard as captured in this quote. To judge by her readers (a tough crowd!), she’s mostly right and always stimulating.

She also has a good eye for interesting material put out there by others. Case in point? Some material from Mike Hulme published in Future Directions for Scientific Advice in Whitehall.  [An aside: the entire volume is a treasure trove. It features articles on all sorts of topics from twenty or so authors, ranging from John Beddington (on the science and art of effective advice) to Rebekah Higgitt and James Wilsdon (the benefits of hindsight: how history can contribute to science policy). Who wouldn’t want to read every word of all of this?]

But back to Mike Hume. His entire article (only eight pages) is worth reading in its full form. Here are some quotes from Ms. Curry’s quotes of his text (my apologies for lifting such an extensive amount of material but each of the bits included here seemed to be pertinent to the small contribution I want to add):

The drive for consensus within the IPCC process, and its subsequent public marketing, has becomes a source of scientific weakness rather than of scientific strength in the turbulent social discourses on climate change…

One of the common public expectations of science is that it speaks authoritatively about the way the physical world works and thereby what the physical consequences of different human actions and policy interventions are likely to be. Science and scientists are believed to offer something different to public life compared to that offered by politicians, journalists, lawyers, priests or celebrities. But what is meant by ‘authoritative’? And how does scientific practice best earn and maintain its authority in the face of public challenge and scepticism? The question I wish to answer can be put simply: does the pronouncement of a scientific consensus on an issue such as climate change increase or weaken the authority of science? And for whom exactly are such pronouncements effective – scientists, different publics, policymakers, politicians?

In favour of consensus

The argument in favour of consensus as authoritative is that it reflects what science supposedly is uniquely disposed to be good at: applying rules of reasoning and inference which lead unambiguously and universally from evidence to conclusion. The same evidence presented to the same disciplined mind leads to precisely the same conclusion. In this view, a lack of consensus would undermine the authority of science because it might suggest either that conflicting conclusions had been reached prematurely or that personal or cultural biases and values had protruded into the reasoning process.

This is the position that seems to be implicitly assumed by many protagonists in the climate change debate, whether they be mainstream or critical voices.

It is also the view of many critics of the scientific mainstream who assert that science properly conducted – through unbiased reasoning processes – should lead to unanimous consent. By pointing out the mere existence of minority dissenting positions outside the IPCC’s statements, ipso facto they undermine the authority of science in the eyes of the public. This of course reflects a very particular (purist) view of scientific knowledge which scholars such as Bruno Latour have described as the ‘modernist illusion of science.’ And yet it is one that offers a wide variety of protagonists a useful defence against cultural relativists.

Against consensus

But the argument against consensus as authoritative, at least in the context of wicked problems like climate change and at least in the way in which the IPCC has promoted it, seems to me to be compelling.

First is an argument by analogy. Majority rule works very effectively in maintaining authority in social institutions such as parliaments and the courts, which involve voting MPs and juries. Consensus is not required for a ruling or judgement to carry authority in wider public settings. And whatever differences we might insist on between the nature of scientific enquiry and political (or jury) debate, we must recognise that scientific assessments such as the IPCC are established explicitly as social (i.e., deliberative) institutions which scrutinise evidence.6 There are many other dimensions to the making of authoritative and trustworthy institutions than unanimity amongst members; for example, fair and agreed procedure, respect for dissent, acceptance of outcomes. Maybe the IPCC’s authority – in the eyes of critics and publics, if not also in the eyes of politicians – would therefore be enhanced if it acted on its own rules for minority reporting in the Summary for Policymakers (which it never has).

Second, the requirement of consensus is pernicious – in order to protect the authority of the group it encourages agreement in a group of experts where there is none. Maybe the IPCC should more openly embrace the idea of expert elicitation, or even expert voting as has been suggested by David Guston: “A scientific body that does not partake in … a politics of transparent social choice – one that hides both its substantive disagreements and its disciplinary and sectoral interests beneath a cloak of consensus – is not a fully democratic one.” For example, such an approach to disagreement could usefully have been applied to the case of the sea-level rise controversy in the IPCC’s 4th Assessment Report. It makes disagreements explicit and better reflects the quasi-rationality of scientific deliberation. Another example of how this might strengthen authority would be the case of the IUCN’s Polar Bear Specialist Group and the embrace of expert elicitation.

And, third, the presence of officially sanctioned – even welcomed! – credible minority views, thereby revealing the extent of dissensus, actually enhances the authority of science. It shows that it is ‘OK to disagree’ and thus indicates that the deliberative procedures of a body like the IPCC are fair and accommodating to the full range of accredited views. For science to be authoritative, it should therefore welcome – indeed seek out – its critics. In the case of large international assessments like the IPCC . . ., the process should not just allow minority reporting in its rules of procedure, but ensure that minority reporting is actively facilitated. As Dan Sarewitz has argued: “Science would provide better value to politics if it articulated the broadest set of plausible interpretations, options and perspectives, imagined by the best experts, rather than forcing convergence to an allegedly unified voice.”


In praise of both/and.

Such discussions of pros and cons arise in many contexts. Here (as in many of those other instances) it seems there’s a good case to be made for both continuing the IPCC process and fostering a full exploration of alternative views (as opposed, say. to a limited expression of minority views confined within the IPCC framework). It’s useful to find out just how far consensus can take us… and it’s equally important to know where uncertainties and opportunities for improvement and advance should be taking us next.

However, it seems that the advantages of both/and are rarely discussed. They offer the biggest advantage when it comes to consequential human affairs… matters of global importance, where time is of the essence. Earth observations, science, and services, including climate science, qualify.

Just a brief mention of three cases in science and technology where both/and have been adopted to the benefit of society.

The Manhattan Project. When the Allies were racing the Axis to be the first to build the bomb,there was need for enriched uranium. Three methods were available: electromagnetic, gaseous, and thermal. Policymakers, scientists, and engineers of the time could chosen to investigate and argue the relative merits and expense of each; but this would have led to dangerous delay. Instead, they employed all three approaches simultaneously.

Mapping the Human Genome. While DoE pursued a brute-force approach that cost initially some $10/gene in the sequence, Celera began by investigating techniques that would drive down the cost of sequencing per gene before diving in. Thanks to both approaches, at the project’s successful conclusion, the scientific community possessed the means for more cheaply and rapidly sequencing other genomes… work that continues today.

Measuring GDP. The United States has one official measure, constructed monthly by the U.S. Department of Commerce. But surrounding this routine and important work, there’s a hive of academic and private-sector activity constantly refining our understanding of GDP and how to characterize it in the face of transitions from agriculture to manufacturing to service-sector economies, and more.

This latter activity has been around a long time, and suggests the direction we might like the IPCC process to go. IPCC reports tend to inflame pre-existing views, and trigger contentious debate. By contrast, the GDP figures, though important, have been internalized by society; they’re part of the routine of everyday life. We can imagine that by the 20th such IPCC report, a similar calm will prevail.

My guess is that the mere notion of a 20th IPCC report makes every climate scientist worldwide feel tired. I can hear the collective sigh! But an additional advantage of both/and is that participants in the consensus process will increasingly recognize there’s such a thing as “good enough.” Those involved will find ways to apply some version of the 80-20 Rule (fancy name, Pareto’s principle), which reminds us we can achieve something like 80% of the result we desire with 20% of the effort. When it comes to climate change, as we realize that a succession of progressively refined reports will be coming, and we learn how to accommodate both a consensus and a bevy of new research, we’ll all lighten up, the IPCC process will come less burdensome, we’ll make better progress…both with respect to the science and society’s application of it…

…and we’ll live more harmoniously with each other.

This entry was posted in Uncategorized. Bookmark the permalink.

2 Responses to The IPCC consensus process: lessons from the Manhattan Project, mapping the human genome, measuring GDP… and Charles Darwin

  1. Bill:-

    Choosing policy responses to climate change is a classic “wicked problem” (perhaps compounded by a social mess!). Too many climate scientists (take Hansen, please!) seem to see it as a reasonably well-defined problem with a complicated, but essentially linear, solution (The mean global temperature is going up and we believe this will eventually have “bad” consequences; the concentration of CO2 is going up [and we believe driving the temperature change], therefore find ways to make the concentration of CO2 go down, or at least hold steady. QED.).

    This is exactly the wrong approach to a wicked problem.

    First, we all need to recognize that while there is a scientific basis for a relation between climate change and the concentration of CO2 in the atmosphere, the degree of depth of that relationship is still not clear (e.g., how forcing is forcing), and in particular what may actually happen is far from established. I am not claiming anything about the scientific basis, per se; only that the science is being applied based on speculation about energy usage and energy mix from now into some future date (which is generally a linear extrapolation).

    Second, the consequences of warming are rather poorly established. We know there will be some, but exactly what and how severe we really don’t know. We might also speculate there will be positive as well as negative consequences (e.g., potentially less problems with influenza, reduced transport costs through utilization of the Northwest Passage). Projections of the consequences of warming have relied on climate model predictions that are themselves subject to the concern above, and have not looked at cross effects (e.g., will possible coastal storms lead to a migration away from coastal areas thereby reducing the impacts of coastal storms).

    Third, we know there are negative consequences to virtually every policy bromide that’s been proposed. Real harm to real people real soon (an expected feature of linear solutions to wicked problems).

    As I’m teaching my students in System Thinking for Emergency Managers, the answer is to be more humble in the face of the unknown. Start with potentially positive steps with minimal harm (e.g., improve energy efficiency). Couple these with good metrics on outcomes (could be controversial to say, but the concentration of CO2 in the atmosphere is not an outcome – mean global temperature might be). Monitor how things are changing in the metrics over time, and what else is changing in tandem with your metrics. Be data- not speculation- driven. And above all, be persistent, not bold; bold solutions never work for wicked problems, but wicked problems often become less severe if there is a persistent and disciplined regimen of small steps accompanied by frequent analysis of the trajectory of change.

  2. Richard Berler (Heatwave) says:

    I like the and/or approach. We do have a good idea of the range of climate states that have taken place over the course of time including snowball earth, warmth even in polar latitudes. Yet, there are also a lot of unknowns about how these climate states have evolved (why the earth was not a “snowball” >2.5 E+09 years ago when the sun was 20% dimmer) in the past, impact of untapped reservoirs such as methane in thawed permafrost in future times, and perhaps most of all, successfully observing the oceans at high resolution in real time, and modeling them with the success that we have had with the atmosphere. Initially, as more of the climate system is modeled with explicit physics as opposed to parameterizations, it could turn out that ensemble climate system model runs would produce a larger range of possible outcomes than our present modeling in the decadal to many century scales…perhaps a mostly explicit climate system model ensemble run would then converge upon a more specific solution if the forcing was sufficient at the millenia+ scale? While we can be authoritive in our observation that our climate is warming, that continued warming is quite likely, that more water vapor in the atmosphere will likely make high end rainfall events more extreme, higher seal levels will excerbate storm surge, we cannot be authoritative on the details (will there be more/windier hurricanes, more tornados) vs. time (even at the decadal-century scale) and location. The both/and allows for the consensus to be expressed about the large scale knowns about the climate system, and overall direction that it is likely to move. It also gives an appreciation for, doesn’t sweep under the rug the smaller components of the system that we cannot be authoritative about. The small scale stuff is quite fascinating, and will ultimately be what impacts us…almost everything that matters to us occurs near the boundaries of the components of the climate system…the lower tens of meters of the atmosphere, upper tens of meters of the oceans. What happens in that thin boundary area will be the weather and weather impacts that we experience, and feedbacks on a longer time scale in this narrow boundary of changes in vegetation, land use, changes in the biosphere…these details cannot be authoritively predicted. We lack authority, currently, in attributing extreme events to the climate change that we have seen. It will be very interesting to see if exceptional blocking patterns that lead to persisting extreme temperature and precipitation events can be traced to the changes that we have already seen in our climate. 2003 Europe, 2010, Russia certainly were exceptional examples.

Leave a Reply

Your email address will not be published. Required fields are marked *