LL2-14 digest — Climate change: science and the precautionary principle#

Hartmut Grassl and Bert Metz, with Panel 14.1 by Malcolm MacGarvin. Late lessons from early warnings II (EEA 2013), pp. 308–346 (PDF 310–348).

Core story#

This chapter traces climate science from Fourier (1824), Tyndall (1863) and Arrhenius (1896) to the IPCC’s 2007 “very high confidence” finding. It then follows the international policy response from the Villach warnings (1985) to Durban (2011).

The paradox at its centre: framework-level action came fast, while effective action never arrived. - The UNFCCC, with its Article 3.3 precautionary clause, was agreed two years after the first IPCC report and before human influence had been detected (pp. 321–322, 338). Kyoto followed two years after the 1995 “discernible human influence” statement (pp. 324, 326). - But commitments were “not really legally binding” (p. 322). Kyoto took until 2005 to enter into force, and the US left in 2001 (p. 324). - Global emissions still rose from ~38 to ~50 Gt CO2-eq between 1990 and 2010 (p. 325). Copenhagen failed, and 2011 pledges implied 3–4 °C of warming (p. 328).

The authors conclude that the science-policy dialogue was “only … a partial success”. Twenty years of action is “inconsistent with a strict interpretation of the precautionary principle” (p. 336).

An unattributed opening box notes that the EEA left climate out of the 2001 volume because of “too much legitimate controversy” (p. 308).

Key evidence#

• Early warnings did not land. Arrhenius’s 3–5 °C estimate was not a warning to him; he feared cooling (p. 311). Callendar (1938) was disbelieved, because colleagues distrusted his CO2 data and knew little radiative transfer (p. 311). Reliable CO2 trends came only in the 1960s, and CH4/N2O trends in the late 1980s (p. 311 fn).
• The evidence was obscured. Aerosols masked warming from the 1940s to the 1970s and still mask about a third of it (p. 313). The ocean delays the signal: less than two-thirds of committed warming is visible, and policy effects appear only after decades (p. 314).
• Institutions carried the science. Coordinated research (WCRP, IGBP) and the IPCC gave science authority with governments. That authority came from UN sponsorship, intergovernmental membership, multi-round review and sentence-by-sentence approval (Box 14.2, p. 317).
• Kyoto compliance was partly windfall. UK coal closures, German reunification, the post-Soviet collapse (–35 %) and the 2008 recession all helped (p. 324). CDM credits could halve domestic cuts. Some projects were of doubtful additionality, and some companies profited from credits for destroying HFC-23, a manufacturing by-product (pp. 325–326).
• Doubt was manufactured. The authors call this one cause of slow progress, “not the only one” (p. 330). The fossil-fuel-funded Global Climate Coalition (1989–2002) ran ads before Kyoto; a “denial industry” followed, with parallels to tobacco, acid rain and ozone (p. 330). Then came “Climategate” and “Glaciergate”. The IPCC’s “delayed and defensive” response made things worse (p. 330). EU concern fell from 57 % (2007) to 34 % (2011) (p. 336).
• Uncertainty guidance was patchy (Panel 14.1). The IPCC’s uncertainty guidance evolved from TAR to AR5. Throughout, expert groups underestimated structural uncertainty (p. 333), and WG III did not adopt the guidelines (p. 332). The AR5 guidance raised type I and type II errors “for the first time” (p. 335).

Authors’ lessons#

• Precaution shaped early framework decisions but “diminished”. It now has “little, if any, effect” (p. 338).
• All five of the authors’ evidence criteria had been met by 2001. So the science is “far beyond” needing precaution, at least on the physical question (pp. 336–337).
• Risk management drove what action happened. States committed only once they judged reductions feasible and affordable (p. 338). (This sits uneasily with p. 326, where Kyoto is adopted while feasibility and costs were “still far from certain”.)
• Precaution has mostly been applied at regional scale, where stakes were lower and “those implementing policies were generally also the ones benefitting”. Climate breaks that alignment across place and time (p. 337).
• Advocacy: reframe climate action as a co-benefit of development and “green growth” (pp. 338–339).

Main mechanisms#

• Measurement infrastructure decided when warnings became credible.
• Confounding and inertia decoupled cause, evidence and remedy.
• Conservative standards of proof delayed attribution (p. 318). Yet the IPCC’s response-strategy group urged action before attribution (p. 318).
• Collective action and free-riding: exemptions, the US exit and Byrd-Hagel.
• The causers of the problem are protected; its victims are poor and in the future.
• Concentrated short-term losers lobbied against diffuse, delayed gains (p. 338).
• Incumbents manufactured doubt, and media norms amplified it.
• Concentrated institutional credibility became a single point of failure.
• Questions that look scientific turn on value judgements: “dangerous”, “tolerable” and 2 °C as “a political decision” (pp. 322, 327).

Transferable insights (selected; full list in notes)#

1. Quantitatively correct early signals may carry no alarm; a “warning” depends on the observer’s values (p. 311). Moderate.
2. Warnings based on measurements a field distrusts are discounted until measurement infrastructure establishes credibility (pp. 311–312). Moderate (one well-described episode).
3. System inertia means waiting for observed harm locks in more harm, and remedies pay off beyond decision horizons (pp. 314, 337). Strong.
4. The evidence threshold for action is a choice, and a criteria ladder makes the degree of precaution explicit (p. 336). Moderate (built in hindsight).
5. Framework-level precaution is easier than binding, effective measures; the delay lies in implementation (pp. 321–328). Strong for this case.
6. Precaution is most readily applied when cost-bearers are also beneficiaries (p. 337). Moderate (asserted comparison).
7. Action follows confidence in feasible, affordable remedies more than proof of harm (pp. 327, 338). Moderate (the chapter’s own Kyoto account partly contradicts it, p. 326).
8. Incumbents delay action by manufacturing doubt, with recurring tactics (p. 330). Moderate (secondary sources; size of effect asserted).
9. Authoritative assessment bodies are single points of failure, and defensiveness amplifies damage (pp. 317, 330, 336). Moderate.
10. Expert groups underestimate structural uncertainty, and calibrated vocabularies can create false precision (pp. 332–334). Strong that the IPCC’s own documents recognised this; moderate as a general claim (partly MacGarvin’s analysis).
11. Offset and flexibility mechanisms, and policy-unrelated windfalls, can make formal compliance diverge from real reduction (pp. 324–326). Moderate to strong.
12. “Dangerous” and “safe” thresholds are largely value judgements (p. 322; three of Box 14.4’s four questions are called value-laden). Strong.

Main caveats#

• Insider account. Metz co-chaired IPCC WG III and Grassl led the WCRP and MPI-M (general knowledge; not disclosed in the chapter). The chapter praises the IPCC, treats its errors as minor (as the inquiries found) while conceding a “defensive” response, and relays WG III cost findings uncritically.
• Narrow causal analysis. Among political-process causes, only doubt-creation is examined in depth. The Discussion adds structural causes (collective action, distribution, inertia, framing), but treaty-architecture failure is declared “beyond the scope” (p. 330). Critics of that architecture (Victor; Prins and Rayner) appear in the reference list but are never cited, though the chapter’s citations are loose generally. Priority-setting economists are placed in the doubt-creation section as “Another line of attack on addressing climate change” and not engaged (p. 330). There is no dissenting panel.
• Internal tension. The main text says uncertainty is resolved (p. 337). The panel says structural uncertainty is systematically underestimated (p. 333).
• Factual slips. Climate sensitivity ranges are misstated (pp. 311, 319–320, 339). Several dates conflict (1897/1896; Toronto 2005/2003) or look wrong: Russia’s ratification (p. 324), Australia’s ratification (December 2008 in the text, actually December 2007; p. 326, 340) and the Nobel month (p. 327). Fig. 14.4’s lower panels are mislabelled (they show emission levels, not temperatures).
• Harms and advocacy. There is no quantification of realised harms or of the costs of delay. The green growth remedy is 2012 advocacy, not tested.