Thursday, October 22, 2009

Response to 'Economic Case for Slashing Carbon Emissions'

I'm reproducing my comment to the opinion piece The Economic Case for Slashing Carbon Emissions here because it apparently didn't make the editorial cut on Yale 360. The article is about a study done that supports capping emissions so that atmospheric C02 levels reach 350 parts per million in this century.

"At first glance, there is a bewildering range of estimates of the costs of climate protection.
The outliers ... funded by partisan lobbying groups ... An analysis by journalist Eric Pooley documents the excessive, often uncritical attention given to these studies by the media."

I think an 'uncertainty' analysis that's based on a journalist's survey of reports is pretty poor. The MIT guys have a much better approach to this, ie a no kidding uncertainty analysis using Bayes' Theorem and Monte Carlo sims. It's by no means perfect, but it is an honest analysis rather than ad hominem attack.

Also, Nordhaus' analysis with his DICE model seems to indicate that just trying to fix CO2 at a certain level could be a pretty poor strategy, and actually end up costing a significant amount more than an optimal approach while still not avoiding many of the supposed costs of environmental damage.

The costs: this is where things get *very* tenuous. The connection between climate change and solid estimates of costs and benefits seems on quite poor footing. It's hard even to get the sign of the effect right, not to mention the magnitude. Some studies even show an economic benefit of warming...

I suppose it could have been worded better (all writing is a series of drafts right?). The point I was trying to make in the first paragraph is that discounting a study because of its funding source isn't science, it's politics. If the study is so fundamentally flawed, it should be easy to point out the errors in underlying assumptions and methodology (that's science). No need to sling mud, "OMG, Teh Evil Exxon paid for that study!!1!eleventyone!" (that's politics).

The second part of the comment was just pointing out some research that seems to show that trying to control CO2 at a fixed level is not a very economical strategy for avoiding costs due to environemtal damage and at the same time avoiding tanking your economy through too much unproductive wealth redistribution (an optimal approach if you will). Further, if you read that paper about the DICE model, you'll see that the cost function (relating climate change to cost of damage) is very uncertain. So uncertain in fact that it can't even be reasonably quantified (and much to his credit, Nordhaus is quite upfront about this). Not only is the magnitude of the effect unquantifiably uncertain, there is question as to the sign of the effect (whether warming is a bad or a good). Which leads me to one of my favourite Lord Kelvin quotes:

In other words: Quantify Your Uncertainty! This is very hard to do when you can't validate your model with properly designed experiments. Here's a graphic taken from a 2006 study about costs of climate change:

Notice anything about the range on the costs of climate change from 'previous studies'? The error bars include zero! We are not even sure of the sign of the effect, much less the magnitude. It may turn out to be good public policy to subsidize carbon dioxide emissions rather than tax them.

It seems that these types of results (DICE model runs) are very sensitive to the chosen discount rate.
The discount rate is therefore the parameter that most influences the 22nd century behaviour of the modelled climate.
-- Schneider (2003)

It is interesting to note that the discount rate in the model runs supporting 'The Economics of 350' conclusions are changed from the defaults and some hand-waving justification is given. However, no quantitative sensitivity analysis is presented.
Curiouser and Curiouser
-- Winnie-the-Pooh

In fact, the discount rate is not actually stated in the narrative, the Stern Review is referenced. The discount rate used there is 1.4%, quite a low rate of return for an investment with a time horizon of a century. In fact, the only way you'll get a majority of folks to invest in something that poor would be either to lie to them or to coerce them with the machinery of state.
A dollar invested in the private sector can provide a stream of consumption at 4 per cent. Social funds are clearly limited. If we cannot invest in every desireable social activity, clearly we should begin by investing in our best social opportunities first. If climate change can earn only a 1.5 per cent return each year, there are many more deserving social activities that we must fund before we get to climate. Although climate impacts are long term, that does not justify using a different price for time.
-- Robert Mendelsohn, Perspective Paper 1.1, 'Global Crises, Global Solutions'

Talk about tweaking the knobs on the model to get the result you're looking for...


  1. Interesting article by Nathan Myhrvold that touches on the silliness of so much of the hysterics around climate science.

  2. Another problem with the Stern Review's climate change cost estimates emerges:

    In fact as much as 40% of the Stern Reivew projections for the global costs of unmitigated climate change derive from its misuse of the Muir-Wood et al. paper.
    --Roger Pielke, What a Tangled Web We Weave

  3. Muir-Wood on misuse of his paper:
    The Stern report, citing Muir-Wood, said: “New analysis based on insurance industry data has shown that weather-related catastrophe losses have increased by 2% each year since the 1970s over and above changes in wealth, inflation and population growth/movement.

    “If this trend continued or intensified with rising global temperatures, losses from extreme weather could reach 0.5%-1% of world GDP by the middle of the century.”

    Muir-Wood said his research showed no such thing and accused Stern of “going far beyond what was an acceptable extrapolation of the evidence”.