Greg Craven published a series of Youtube videos claiming to give a stark a simple choice for any non-scientist to rise above the “shouting match” a beliefs to reach a decision, with confidence as what to do about climate change. By understanding the flaws in his argument it is possible to develop a more sophisticated grid that enables us to make a more rounded assessment of the climate change problem.

Craven’s Grid

Greg Craven is a high school science teacher in Oregon, who has produced a series of Youtube videos (and later in a book), asking:-

“Which is the more acceptable risk?

• Do nothing and accept the potential catastrophe of global warming?
  
• Take action now, potentially harming the economy, but averting potential catastrophe?”
  

He puts this into a simple grid, with whether to take action along the horizontal axis and whether the global warming hypothesis is true or not on the vertical axis. Recognising that there are costs to combating climate change he allows for the worst case scenario. The cost will be global depression. However, this is as nothing compared with the worst case scenario the scientists are telling us will happen if we do nothing. The choice is stark. We do not know whether the global warming hypothesis is true or false – that is a shouting match that we will get no sense out of – but we should not take a chance, for our own sakes, the sake of the world’s poor and the sake of our children.

Someone then pointed out a gaping hole in Craven’s argument. This grid could be applied to anything that could potentially happen, no matter how unlikely. He uses the example of an invasion of giant mutant space hamsters. The outcome of everyone on the planet being eaten by such creatures is far worse than the costs of building the necessary defence shield to repel them. We must therefore look at the credibility of those who hold the hypothesis against those who do not. If the backers of the hypothesis are the vast majority of leading scientific organisations, and the opponents are few and far between, then we can give much greater weight to the global warming hypothesis. So the grid becomes like this:-

Craven’s Grid with numbers

Craven’s argument becomes much clearer if we put in some numbers. Let us say that the cost of action is 20, but the impact of the catastrophic warming is 400, or 20 times worse. The basic grid becomes:-

The figures are all negative, as the planet is worse off if catastrophic warming occurs, and the people on this planet are worse off if action is taken. The figure of twenty times worse is the upper limit of cost-benefits from the Stern Review (Tol & Yohe 2006). I chose 20 as a base to save on too many decimals when the grid becomes more sophisticated. (Twenty times of benefits over costs is the upper limit of the Stern Review?) Note that there is now an average return, with equal weighting given to whether the hypothesis is true or false. We end up with an expected cost of not taking action of 10 times greater than if we did. We can now incorporate Craven’s argument is that we should give much greater weighting to the consensus scientists than to the sceptics. Let us say the chances of the hypothesis being true are 4 times that of it being false. The weighted grid becomes:-

So the expected outcome of not taking action is 16 times worse than of not taking action.

Major issues with Craven’s Grid

There are a number of issues with this basic decision-making grid. Most fundamentally, it relies on opinions of opinions. That is of the expert climate scientists convincing the scientists from other fields that they have a convincing case, then getting this much larger group to pass a resolution. If the vast majority of the membership can be shown to have studied this and agreed that the climate science is as strong as in their disparate fields and used for a basis of voting, then this might be a reliable, secondary, source. However, if there is evidence that the resolution is promulgated by a small number of the membership (and the majority not interested enough to oppose); or that there has been outside pressure to pass a resolution; or that there have been efforts to silence critics through abuse; or that pressure groups try to undermine the research grants on which critical scientists depend – then the resolutions should be at a minimum be viewed as worthless. This is along similar as the verdict of a jury in a court of law is viewed as worthless if there is any evidence jury tampering. Most courts would not just nullify the evidence, they may throw out the case, then prosecute those who would try to pervert the course of justice. If there is considerable evidence of this undue influence, then the balance of opinion should swing the other way. If the argument relies on secondary or tertiary opinions, then one should look very carefully at the actual scientific and policy arguments. As Craven does not offer up this argument (just mentions a paper by Richard Lindzen, to then neglect this issue entirely), I shall leave this as argument as just invalid and ignore it entirely. The alternative – to take as circumstantial evidence that the underlying case is weak – would cloud the analysis. Instead I shall concentrate on the science and policy issues. These can be divided into four areas:-

1. The climate models do not just predict huge rises in temperatures. Rather, they offer up a range of scenarios. The UNIPCC’s scenarios show a greater than 90% chance of temperatures rising between 1.1^oC and 6.4^oC this century (UNIPCC 2007.1 p8). Both extremes are fairly unlikely. However, if these scenarios were taken to their peak then the distribution of final temperatures would be skewed towards the top end. That is the cut-off of 2099 is likely to be a matter of timing
2. The consequence of such temperature rise, the models predict, will not simply be higher temperatures, but severe disruption to weather systems. Extreme weather events, like hurricanes, floods and droughts will become far more frequent. Although impossible to predict the extent of climate disruption with each rise degree in temperature, the impacts are more likely to be more towards the exponential than the linear.
3. Craven appears to assume that the policy proposed will be entirely successful in combating warming. The UNIPCC recognise that this is not realistic. They propose of target of limiting CO2 to 550ppm (twice the pre-industrial levels) warming to 2.0^oC. If this policy successfully contains the warming to 2.0^oC, then there would still be some adverse consequences to that warming. Further, this proposal must be signed up to by all the nations of the world and then translated into actual policies.
4. There are also two broad approaches to policy that need to be considered. The most frequently advocated is mitigation – stopping global warming by stopping the rise in anthropogenic greenhouse emissions, or even reducing them. The other is adaptation – adapting to the climatic changes that occur, such as improving coastal defences; changing crop production; improving irrigation systems; and changing house design.
5. The policy targets then need to be implemented, recognising that often large, and ill-defined government projects poorly managed, tend to over-run on costs and under-perform on benefits. Further, governments may sometimes get freeloaders, who generate costs but do not contribute anything towards the desired outcome.

This gives four areas that can be analysed separately, under the acronym FCPO.

Forecast – How much temperatures are predicted to rise and in what time period.

Consequences – The forecast change in the weather patterns (droughts, floods, hurricanes, etc.), along with the impacts on crop yields, forestation, sea levels etc.

Policy – The adopted remedy recommended to combat the consequences, along with actual legislation enacted and structures created.

Outturn (of the policy) – The actual policies enacted, along the results of those actions – both intended and unintended.

There is a clear dividing line between the climate science (Forecast and Consequences) and the social sciences (Policy and Outturn). The sub-divisions between the Forecast and Consequences and between Policy and Outturn maybe less distinct, but, as later analysis will demonstrate, are useful for purposes of clarity.

Extending the Grid for FCPO

The grid can now be extended include the four elements discussed in a series of stages. We start with Craven’s scenario were a mitigation policy is 100% successful in combating the adverse effects of global warming.

1. The costs of action
   

Extending Craven’s Grid by:-

• Keeping the costs of mitigation to 20.
• Extending the True / False Scenario to five categories of warming. (None, lower, median, higher & extreme).
• Including an extra option of “Adapt”. This is split into two parts. First are low cost contingency plans, given a value of 5. Then there are the reactive costs. These costs rise steeply with more extreme climate change.

This gives a revised grid below.

1. The Climate Change Impact
   

This shows the impact of taking action. Points to note

• The extreme case of not taking action is still twenty times the impact of not taking action.
• Adapting to climate change is only half as effective as mitigation.
• The likely catastrophic impacts are likely to be non-linear. I have assumed that it will double with each category rise.

  		Climate Change Impact
  Category	Extent	Mitigate	Adapt	None
  A	None	0	0	0
  B	Lower	0	-25	-50
  C	Median	0	-50	-100
  D	Higher	0	-100	-200
  E	Extreme	0	-200	-400
  	RETURN	0	-75	-150

1. Net Costs of Action
   

   Adding 1. and 2. together.

   		Net Costs of taking Action
   Category	Extent	Mitigate	Adapt	None
   A	None	-20	-5	0
   B	Lower	-20	-35	-50
   C	Median	-20	-65	-100
   D	Higher	-20	-125	-200
   E	Extreme	-20	-245	-400
   	RETURN	-20	-95	-150

    

2. Weighted Returns – UNIPCC Scenarios as a basis
   

   The UNIPCC gives a 90% chance of temperatures rising between 1.1^oC and 6.4^oC this century. Using this pattern, I assume that the most extreme cases have a 5% probability and the median case as the most likely with a 50% probability. This gives a 20% probability to the lower and higher cases.

   			Weighted Net Costs of Action
   Cat	Extent	%	Mitigate	Adapt	None
   A	None	5%	-20	-5	0
   B	Lower	20%	-20	-35	-50
   C	Median	50%	-20	-65	-100
   D	Higher	20%	-20	-125	-200
   E	Extreme	5%	-20	-245	-400
   	RETURN		-20.0	-77.0	-120.0

    

3. Weighted Returns – Based on Peak Scenario Warming
   

   As noted above, the final peak temperatures from the models are likely to be more skewed towards the upper end, as some of the peaks are likely to be well after 2010.

    

    

   			Weighted Net Costs of Action
   Cat	Extent	%	Mitigate	Adapt	None
   A	None	5%	-20.0	-5.0	0.0
   B	Lower	5%	-20.0	-35.0	-50.0
   C	Median	40%	-20.0	-65.0	-100.0
   D	Higher	40%	-20.0	-125.0	-200.0
   E	Extreme	10%	-20.0	-245.0	-400.0
   	RETURN		-20.0	-102.5	-162.5

4. Weighted returns – restricting CO2 to 550ppm
   

   The policy to mitigate climate change does not have the unrealistic aim of stopping any future warming. Rather it seeks to stabalise CO2 at 550ppm or double pre-industrial levels. This will equate to a temperature rise of around 2 ^oC. This is around the lower level impacts.

   The climate change impact grid becomes

   		Climate Change Impact
   Category	Extent	Mitigate	Adapt	None
   A	None	0	0	0
   B	Lower	-20	-25	-50
   C	Median	-20	-50	-100
   D	Higher	-20	-100	-200
   E	Extreme	-20	-200	-400
   	RETURN	-16	-75	-150

   Note that the lower levels have some benefit from the climate change. But mitigation prevents the most catastrophic scenarios.

   The final grid becomes

   			Weighted Net Costs of Action
   Cat	Extent	%	Mitigate	Adapt	None
   A	None	5%	-20.0	-5.0	0.0
   B	Lower	5%	-40.0	-35.0	-50.0
   C	Median	40%	-40.0	-65.0	-100.0
   D	Higher	40%	-40.0	-125.0	-200.0
   E	Extreme	10%	-40.0	-245.0	-400.0
   	RETURN		-39.0	-102.5	-162.5

    

The FCPO Grid

The grid derived simply, but effectively summarises the “Consensus” case.

1. There is a high degree of likelihood that we will have some serious global warming with huge consequences for the planet and the human race.
2. The best policy is to limit CO2 to double pre-industrial levels. Re-active adaptation strategies are risky and ineffective compared with mitigation.
3. The expected average costs of doing nothing are far greater than any costs of mitigation.
4. More importantly than any expected average impact, there is a very real possibility of a massive catastrophe.

What is more it is a way of effectively gauging some of the arguments, underpinned with some risk analysis. By trying to turn the uncertainties into our best-guess risks we can get a more balanced and rigorous assessment of the science than Greg Craven’s. Further, we can also bring into the same analysis the potential effectiveness of any policy, and the project management aspect of delivering the planned results.

However, the ability of this grid to evaluate the science is limited. It can only be used for most top-down or broad-brush aspects. It is a way to get the relative importance of various aspects of science and policy within the totality of the subject, when they are vastly different in degree. If used for splitting hairs, then this will just be another tool to advance the “shouting match”. To focus and clarify the evaluation, a new question is required. “What is the worst that can happen?” is too loaded towards thinking of extreme, but highly improbable, events.

The New Question

To focus and clarify the evaluation, a new question is required. “What is the worst that can happen?” is too loaded towards thinking of extreme, but highly improbable, events. The FCPO Grid looks at both the future scenarios projected by the science and at our collective ability to provide a remedy. There the new question becomes

By taking action, is there, on the balance of the evidence, an expectation that the future state will be improved?

The climate change consensus would contend that this is clear-cut – something that my FCPO grid replicates.

References

Craven, Greg 2007 Jun – Youtube video “The Most Terrifying Video You’ll Ever See”

Craven, Greg 2007 Oct – Youtube video “How it all ends”

Craven, Greg 2009 – What’s the Worst That Could Happen?

Craven, Greg Website – http://www.gregcraven.org/

Tol & Yohe 2006 – A Review of the Stern Review – WORLD ECONOMICS • Vol. 7 • No. 4 • October–December 2006

UNIPCC 2007.1 – Climate Change 2007: Synthesis Report – Summary for Policymakers (Dec 08 format)