Giving the IPCC a Pass on Errors of Omission, and Errors of Willful Omission:

The Dutch analysis of IPCC statements on regional impacts in the 2007 report

Guest post by: Indur M. Goklany

What with the numerous panel reports on Climategate and the IPCC’s veracity, warmists may have solved our global warming problems: lots of whitewash, which should increase the earth’s albedo, and — voila — we’ll have cooling!

In this post I will discuss the Dutch Environmental Assessment Agency (PBL) report assessing the 2007 IPCC WGII report.

The accompanying press release is headlined, “Key findings of IPCC on regional climate-change impacts overall considered well founded” (emphasis added). I emphasize the word “regional” because that informs us that PBL limited itself to a review of those 8 (out of 20) WGII chapters that addressed regional impacts. The PBL report also alerts us to the fact that it did not review any of these chapters in their entirety (page 12, Executive Summary, ES). Thus, the PBL review could, at best, have a given the WGII report a clean bill of health on much less than half the report’s content. Yet many of the headlines implied a much more sweeping approval of the WGII report than warranted given PBL’s limited review. Eben Harrell at Time magazine’s Ecocentric blog, for instance, headlined his report, “Dutch agency affirms IPCC findings” and quoted Martin Parry, Chairman of WGII during the preparation of the IPCC assessment, as claiming vindication.

Let’s look more closely at what precisely PBL actually did and said. The first paragraph of the press release said:

“PBL Netherlands Environmental Assessment Agency found no errors that would undermine the main conclusions in the 2007 report of the Intergovernmental Panel on Climate Change (IPCC) on possible future regional impacts of climate change. The IPCC report conclusively shows that these effects already are visible in many places around the world, and that these will become more serious under further temperature increases. However, the foundation for some of these conclusions could have been made more transparent.” [Emphasis added.]

First, few, including myself — and, I suspect, most skeptics — would take exception to the “main conclusions” of the IPCC as described by the PBL, particularly when the impacts are characterized as possible — a meaningless weasel word which should not be confused as being synonymous with “likely” or “very likely”, even if one accepts the IPCC’s devalued definitions of these terms (compared to normal scientific and statistical parlance). Of course, climate has changed, its impacts are visible, and they would increase were there to be further temperature increases. But the notion that impacts are already serious (as implied by PBL by the “more serious” in the second sentence) is not only unsubstantiated but contradicts Figure SPM.2 in the WGII Summary for Policy Makers which indicates that impacts from a further 1 degree C increase over the 1980-1990 level would not necessarily be specially “serious.”

More importantly, given PBL’s methodology, its exoneration with respect to the listed “main conclusions” is almost a foregone conclusion because PBL focused on sins of commission rather than sins of omission, as is revealed in the following passage in the Executive Summary:

“Given the constraints regarding time and capacity, it was not possible for the PBL to check a hundred per cent of all texts and references in the eight regional chapters of the Working Group II Report for errors, considering that it had taken hundreds of authors and reviewers to produce the report over the course of five years. Instead, we limited ourselves to the IPCC summary statements, and framed the central questions of this report as follows:

“Are the summary conclusions on regional impacts well founded on the underlying chapters and literature references? Are there errors in statements that have travelled from the scientific literature references and/or the main texts through to the summary conclusions? If errors are found, do they affect the validity of these conclusions? What recommendations can we derive from our investigation in order to further improve the quality of the assessment process for the Fifth Assessment Report (due in 2014)?” [PBL Report, page 12; Italics in the original; bold is added.]

But while PBL focused on errors of commission, it is sins of omissions that are more likely to be committed.

I. The PBL did not count as errors systematic silence on the benefits of climate change

There is a general tendency to overlook errors of omission because, as the sage noted, to err is human. But by the same token, errors of omission are more likely to be committed than those of commission because its perpetrators know it is hard to prove malintent when it comes to omissions. It could result from negligence and carelessness — hardly hanging offenses, even if potentially criminal under U.S. law. But more importantly, there is that time-honored and much-abused justification for willful omissions beloved by bureaucrats and biased newspapers and their reporters, namely, the exigencies of space. This excuse works especially well for packing Executive Summaries and Summaries for Policy Makers (SPMs) with cherry-picked information. In fact, a cynic might say that the cover they provide for cherry picking is a major reason why we have such summaries. [This is why in my comments to the InterAcademy Council, I recommended dispensing with SPMs, among other things.] And, unfortunately, PBL’s PR release seems quite sympathetic to this justification:

“[T]he IPCC Working Group II Report put an emphasis on projections of the more serious, negative impacts of climate change. This selection was an obvious choice, and also had been approved by the governments that constitute the IPCC. However, this meant that the less severe impacts and any positive effects did not make it into the summaries for policymakers, which made the overall tenor of the summaries more negative than that of the underlying chapters.” [PBL PR.]

But given the uncertainties associated with impacts assessments, and the inability — or is it unwillingness — of researchers to systematically quantify or otherwise characterize these uncertainties, how does anyone compare one of set of impacts with another to determine which is more serious? At best, one may have a gut feeling. But science and risk analysis (and an IPCC assessment) must be based on more than that.

The Government of New Zealand’s comment 406 on the final draft of the Summary for Policy Makers provides a methodology to proceed on these issues:

“We are concerned that this section appears to focus on the negatives (i.e. on the most vulnerable sectors and regions). It is our view that the assessment would be much stronger if it assessed both the negatives and positives, and only then, if appropriate, concluded that the negatives outweigh the positives.” [Emphasis added.]

But it was never implemented by the IPCC.

And what about the impropriety of providing a balanced and unbiased assessment?

II. The PBL did not consider it egregious that the IPCC summary tended to single out the negative impacts of climate change while maintaining silence on benefits

Despite PBL’s — for lack of a better word, forgiving — methodology, the opening paragraph of its Executive Summary (page 9) notes that “the investigated summary conclusions tend to single out the most important negative impacts of climate change” [emphasis mine]. That is, the IPCC summary conclusions reported virtually exclusively on the costs of climate change while ignoring its benefits. But these are nothing but errors of omission!

Yet, not only does the PBL avoid calling these errors, which they most certainly are, it does not find this particularly objectionable. In fact, it confers legitimacy to systematically omitting information on the benefits by gracing it with the label, “risk oriented approach”, and claiming that this deceptive approach was implicitly endorsed by governments. Specifically, it states:

“The PBL has labelled this as a ‘risk-oriented’ approach, which had been implicitly endorsed by the governments that constitute the IPCC (including that of the Netherlands).” [Executive Summary, page 10. Emphasis added]

But where is the evidence for the PBL claim that governments “implicitly endorsed” this biased approach? The Dutch may have endorsed this implicitly, but the review comments to the final draft of the WGII Summary for Policy Makers (SPM) indicate that some governments explicitly asked for greater “balance”. See, for example, New Zealand’s comment provided above, and comment 970 for the US.

The fact that despite such comments calling for a systematic accounting of benefits and costs and greater balance, the WGII SPM “single[d] out the most negative impacts” (PBL, ES, p. 9). This attests, first, to the lack of persistence on the governments’ part rather than to their implicit endorsement. Second, it also is a testament to the fact that the one that controls the pen, controls the narrative, especially if there are hard limits on the time and space allotted for producing a product. And, believe me, the pressures to finalize an IPCC report on time are tremendous (see here). Third, they indicate that these omissions were not through oversight, but willful. That is, they are truly errors of commission.

More important, however, is that it is the IPCC’s job to provide an unbiased assessment. This duty holds no matter what. As noted on its website:

“Review is an essential part of the IPCC process, to ensure an objective and complete assessment of current information… the IPCC embodies a unique opportunity to provide rigorous and balanced scientific information to decision makers.” [Emphasis added.]

But focusing on the negative impacts, while virtually ignoring benefits is neither objective, nor complete, nor rigorous, nor balanced. In fact, had governments explicitly admonished it to provide a biased report, the IPCC would have been duty-bound to reject such an admonishment; but to produce a biased report absent any such explicit admonishment is worse than a dereliction of duty. [One is reminded here of Thomas More’s bewilderment at Richard Rich’s perjury — “Why Richard, it profits a man nothing to give his soul for the whole world… but for Wales?”]

III. The PBL considered it notable, but not reprehensible, that the IPCC summary tended to highlight the upper end of the range of negative impacts even while it was silent on benefits

To compound matters, not only did the IPCC WGII summary reports virtually ignore any benefits from climate change, to quote the PBL (page 39), “often [,] the upper ends of uncertainty ranges (the worst outcomes that are projected) were highlighted.” Of course, this is generally tempered by scientifically-precise wording such as “might”, “may”, “could” and “up to” [☺]. So, strictly speaking, these statements are not untrue, but they are certainly inaccurate (as opposed to “true but inaccurate”). These are inaccuracies of commission. Unfortunately, the PBL was silent on this tendency of the IPCC to highlight the upper end of the range of bad outcomes in both its press release (PR) and Executive Summary (ES). Worse, it seems to accept such biased reporting as acceptable practice under a “risk-oriented approach” (PBL, page 39).

[As an aside, note that both the PBL’s PR and ES noted that the IPCC summaries emphasized the “main negative impacts of climate change”. But “main negative impacts” is not synonymous with the “upper end of the range”. ]

IV. The PBL condoned the IPCC’s failure to provide context to gauge the importance of climate change — more errors of omission in the IPCC summary statements

The PBL also noted that the IPCC failed to provide information that would provide policy makers to view the impacts of climate change in the wider context of other factors that could have similar impacts. “This was even the case when these other factors were much larger than the impact that was attributable to climate change” [PBL, page 39; emphasis added]. However, while the PBL comments on some of these and notes that “some policymakers may wish to see both numbers – that is, changes with and without climate change – within the same context in a summary” (page 39), it does not label these as errors and seems to find such errors acceptable. They are, in fact, errors of willful omission, similar to the systemic neglect of benefits in the summary statements.

Here too, some governments had specifically commented on the lack of context in their comments on the final draft of the WGII SPM (e.g., US in comment 363, Finland in 78) but with no greater success. For example, comment 56 was:

“Lack of Context in which Climate Change Occurs

“In addition, the U.S. Government recommends inclusion of some discussion of the role of socio-economic and other non-climate-change-related factors, as these play important roles in both reducing and increasing vulnerability to climate risks. Climate change, for the most part, exacerbates existing problems rather than creates brand new ones (although location-specific details may vary). Fortunately, the information providing this context exists in the chapters and, in many cases, in the Technical Summary. Policymakers would benefit from being provided estimates of the relative significance of non-climate-change-related factors and climate change with respect to various climate-sensitive problems.” [Comment 56, emphasis in the original.]

Ironically, the IPCC claims that it should be “policy-relevant and yet policy-neutral, never policy-prescriptive” (see here). But what could be more policy relevant than trying to establish how climate change compares in impact with other factors?

Elsewhere, I have tried to establish the importance of climate change relative to non-climate change-related factors for a variety of impacts. See, for example, the post, Setting the Record Straight on the IPCC WG II Fourth Assessment Report, which discusses and links to yet other posts which indicate that even under the warmest scenario:

(a) Climate change would reduce the net global population at risk of water shortage, and

(b) The contribution of climate change to hunger and malaria, two reasons frequently cited for reducing carbon dioxide emissions, ranges from the trivial (4% for malaria) to the small (21% for hunger), at least through the foreseeable future.  [I define “the foreseeable future” as the 2080s.]

To quote myself, in the absence of context even the smallest molehill can be mistaken for a Mount Everest. Which is why for a report to be policy-relevant, the broader context is obligatory.

V. Summary: PBL gives the IPCC’s errors of omission a pass, and its approach to risk analysis would create their own risks

To summarize, the PBL gave the IPCC’s summary statements on regional impacts a relatively clean bill of health because it only looked for errors of commission in a limited number of chapters while deeming errors of omission to be an acceptable part of a “risk-oriented approach.” Under the latter approach, it would be acceptable for executive summaries to emphasize costs and, moreover, highlight the upper end of these costs, even as they eschew information on benefits. And providing policy makers with the broader context might be nice, but optional.

PBL may label this a “risk-oriented approach”, but most rational people would label it “biased and unbalanced”.

PBL’s approach would institutionalize biased summaries and, worse, an asymmetric approach to risk analysis so that costs (negative consequences) are conveyed to policymakers but not benefits. It’s then only a short step to justifying an asymmetric precautionary principle in which we examine, for example, the costs of a technology (e.g., genetically modified crops or DDT) but not its benefits (e.g., reduced hunger and saved lives). As detailed in the book, The Precautionary Principle, these specific asymmetric applications of the principle have led to more harm than good. Although such outcomes are not necessarily inevitable, ignoring one side of the cost-benefit equation (but not the other) is an invitation to be visited by unintended consequences.

In the global warming arena, we have seen such unintended consequences for biofuels substantially increase hunger and poverty worldwide without significantly reducing either greenhouse gas emissions or, more importantly, global temperatures.

These consequences, while unintended were not inevitable (see this paper from 1999). They were enabled by selective risk analysis. Policy makers, enamored of the benefits of biofuels, compounded by hysteria about climate change, were happy to ignore a full accounting of the social, environmental and economic consequences. And their bureaucracies were happy to oblige.

PBL’s approach to risk analysis would make such snafus more likely because it would institutionalize an asymmetric precautionary principle.