It is time to stop quivering in our boots in pointless fear of the future and just roll up our sleeves and build it.
- Ray Pierrehumbert

Thursday, October 14, 2010

IPCC Good News, Bad News.

Good news: Pachauri is reappointed for another term as head of IPCC. Whether or not this would have been a good idea under ordinary circumstances, it was pretty much necessary. Anything else would have just encouraged the jackals squealing for blood.

Bad news: Andrew Sullivan quotes Jim Manzi quoting IPCC:
This is the crux of the problem with McKibben’s argument: According to the IPCC, the expected economic costs of global warming are about 3 percent of GDP more than 100 years from now. This is pretty far from the rhetoric of global devastation that McKibben, and so many others, use.
The question raised obviously is the authority of this claim. Did the IPCC make such a claim? In what context? (Manzi does not provide a reference.) Is there any basis for this claim? What is the confidence interval?

For myself, I have always thought of WG I as authoritative, and WG II and WG III as speculative. Much of the damage to the reputation of climate science has emerged from attributing WG II speculation to "climate science".

I will probably change everything I'm doing in my professional life if anyone can convince me that the 3% prediction is reliable. I think there's very little chance of that, but Manzi is right on this point: "One of the great strengths of TNR, in my opinion as an outsider, is that it has made a habit of facing up the strongest arguments of its ideological opponents." Well, I stopped following TNR fifteen years ago on the grounds of its exasperating blindness on middle east policy and a divergence of interest in general, but I support the principle. So, is there really some way to limit the damage to an unmeasurably small modest reduction of growth?

I have to say my BS detectors are pegged on this one. It will be interesting if blame for this really attaches to IPCC.

Amusingly, Sullivan takes the bait:
But since when did conservatives only care about "economic costs"? I respect Manzi's cost-benefit argument, and his policy pragmatism. But there is a moral dimension to real conservatism, even a spiritual one, that does not treat the planet as something to be used, but as something to be a sensible steward of.
Maybe so. But if we really believe that "growth" remains possible for a century, it necessarily follows that the probability of real environmental crisis is small. This is not despite, but because of, the fact that the economy is part of the environment.

In future, only actual consensus positions should be reported as such by IPCC, please and thanks. Any clues where Manzi got his idea and what we can do about it?

That all said, I like Sullivan's concluding paragraphs:

The earth is something none of us can own or control. It is something far older than our limited minds can even imagine. Our task is therefore a modest one: of stewardship, the quintessential conservative occupation.

Conservatives do not seek to remake the world anew. We do not hope to impose upon it some abstract ideological “truth” or bring about some new age for humanity. We seek as conservatives merely to live up to our generational responsibility and to care for the inheritance we have in turn been given. This ecological vision is a Burkean one, which is why Toryism’s natural colour is as much green as blue.

Confirming that the present day Republican party is not conservative.

Update: Plenty of sensible caveats (see comments) notwithstanding, I'd be enormously relieved to see that we are talking about a cost of 3% per century. This works out to a hair less than 0.03% per year.

This year's cost would be 18.3 billion dollars at that rate. Since costs are expected to accelerate, we can refute the estimate if we can find 18 billion in losses this year.

Estimates from the Pakistan flood losses alone range from 9 billion to 43 billion.

Can we put this down to climate change? I think history will do so. But for present purposes it doesn't matter. All we need to do is find excess weather/hydrosphere/cryosphere related costs. and Pakistan surely qualifies. This means that we are already bearing the brunt of business-as-usual climate change.

That is, if the estimate is correct, this year would be worse than average in terms of climate related costs averaged over the next century. I think that implies changes much smaller than the WG I consensus, which expects accelerating change. Therefore it would appear to be inconsistent with the rest of IPCC.

Note also that in comments Brian Schmidt points out that this appears in AR4 only as a quotation from AR3 upon which some doubt is subsequently cast.

In any case the idea that a problem like climate change separates out linearly from other problems itself seems highly dubious. I retain my opinion that Manzi's reasoning is based on a very flimsy premise.

Update: The MIT newspaper makes a similar case. The economists really seem to be getting this wrong; that being the case it makes a lot of sense that a lot of hostility is coming our way. Really, if the premise is correct we are totally barking up the wrong tree, but I just can't imagine what sort of reasoning could support the premise. It certainly makes the resistance to action far more understandable.

Again, my impression that we are paying attention to the wrong part of the puzzle is reinforced. The physical science doesn't matter if the economics is so far from what many of us expect. Nor does the environmental science.

42 comments:

Jim Manzi said...

Michael,

In the post that Sullivan links to, the quoted sentence is preceded by the sourcing (links to IPCC source documents in the post):

"

According to the currently governing Fourth Assessment Report produced by the United Nations Intergovernmental Panel on Climate Change (IPCC), under a reasonable set of assumptions for global economic and population growth (Scenario A1B), the world should expect to warm by about 3°C over roughly the next century (Table SPM.3). Even in the most extreme IPCC marker scenario (A1F1), the best estimate is that we should expect warming of about 4°C over roughly the next century. How bad would that be? Also according to the IPCC (page 17), a global increase in temperature of 4°C should cause the world to have about a 1 to 5 percent lower economic output than it would otherwise have. So, if we do not take measures to ameliorate global warming, the world should expect sometime in the twenty-second century to be about 3 percent poorer than it otherwise would be (though still vastly richer per capita than today).

This is the crux of problem with McKibben’s argument: According to the IPCC, the expected economic costs of global warming are about 3 percent of GDP more than 100 years from now. This is pretty far from the rhetoric of global devastation that McKibben, and so many others, use.

"

Best,
Jim Manzi

Brian said...

1. Stern Review says 20% loss from business as usual: http://en.wikipedia.org/wiki/Stern_Review

2. Eli Rabett and Don Rumsfield collaborated telepathically on the problem of known unknowns. None of the known unknowns have potential to make things better. The best we can hope for in each case is that it turns out to be insignificant (don't count on it). Obviously, these haven't been included in the impact analysis.

3. From the IPCC page Jim cites, it appears to simply restate the TAR assessment rather than assert it as a FAR assessment. It also adds this:

"It is very likely that globally aggregated figures
underestimate the damage costs because they cannot include
many non-quantifiable impacts. Taken as a whole, the range of
published evidence indicates that the net damage costs of climate
change are likely to be significant and to increase over time
[T20.3, 20.6, F20.4].
It is virtually certain that aggregate estimates of costs mask
significant differences in impacts across sectors, regions,
countries and populations. In some locations and among some
groups of people with high exposure, high sensitivity and/or low
adaptive capacity, net costs will be significantly larger than the
global aggregate [20.6, 20.ES, 7.4]."

Steve Scolnik said...

Can't place the reference from memory, but that 3%/100 yr. has a distinctly Lomborgian aroma.

Hank Roberts said...

I disagree with this conclusion, Michael:

"... if ... "growth" remains possible for a century ... the probability of real environmental crisis is small."

Growth can be fed for a long time yet by burning the seed corn and the topsoil. That's "Overshoot" --
http://www.greatchange.org/footnotes-overshoot-graphs.html

Hank Roberts said...

Oh, I suspect the trick is that GDP doesn't account for ecological services, and does increase each time something gets destroyed and replaced by new stuff.

Amusingly put here:
http://www.paecon.net/PAEReview/issue20/Vaury20.htm

"'Burn Paris and you will make GDP grow!'. Another weird remark, another problem with GDP : GDP only includes positive values: if something is destroyed, then rebuilt by a private company, GDP goes up while economic well-being is unchanged. Those who use GDP as a good measure of economic progress forget that production is closely linked to destruction. In two ways ...."

Michael Tobis said...

Jim, Brian, thanks for the elaborations.

Hank, agreed.

I remain astonished that all the public debate is about WG I issues, never mind relatively peripheral ones. Understanding how to even think about economics on a 100 year time scale is problematic.

I have a lot to say about this, but most people versed in economics seem to lack the patience to engage.

I believe this is because the only regime of economics that is explored is on the one-year to ten-year time scale, which is analogous to a dynamic predictability limit in geophysics. There is no economic discipline of the ensemble as opposed to the dynamics, comparable to how climatology relates to operation meteorology, and we desperately need one...

Hank Roberts said...

see also:
http://e360.yale.edu/content/feature.msp?id=2200

"Why the wide range of cost estimates?
At first glance, there is a bewildering range of estimates of the costs of climate protection. .... How much will it cost to reach this more ambitious target? ...."

Hat tip to:
http://rwer.wordpress.com/category/the-economy-and-the-planet/economics-of-climate-change/

jstults said...

Brian, I think you should take a closer look at the Stern Review (funny given MT's recent dirty BP pizza, that old one even mentions oil money).

Michael Tobis said...

I think "ecological services" is close, but wrong. I propose that the right concept is "natural capital". Much of what economists call production is just a transfer from natural capital to human capital. You don't get to overdraw natural capital.

On this view, aggregate human wealth has increased far less than on the view where extractive industries are perceived as "creating" wealth.

I agree that the the GDP of replacement vs the GDP of innovation is also an issue. If the air becomes unbreathable and we all have to buy oxygen tanks, the oxygen industry will prosper, compared to its present state as a very minor player...

There's also the "basket of goods" problem. A horse now is worth much more than a horse a hundred years ago, yet it serves fewer purposes. An iPad would have been worth a million dollars ten years ago, and not a penny fifty years ago. A tank of oxygen has little value now...

Even so, I'd be enormously relieved to see that we are talking about a cost of 3% per century. This works out to a hair less than 0.03% per year.

This year's cost would be 18.3 billion dollars at that rate. Since costs are expected to accelerate, we can refute the estimate if we can find 18 billion in losses this year.

Estimates from the Pakistan flood losses alone range from 9 billion to 43 billion.

Can we put this down to climate change? I think history will do so. So we're already past that 3% per century point just in conventional economic arguments not even counting Russia, or the excess flooding elsewhere.

Again, 3 % per century is consistent with this year being worse than average in terms of climate change costs over the next century. Good luck with that one.

Oh, and don't forget to work in ocean acidification, which is outside the usual IPCC turf. How much do you figure the ocean is worth?

thingsbreak said...

There's a great deal of oversimplification in Manzi's attribution of 3%/100 years.

http://thingsbreak.wordpress.com/2010/07/08/debunking-jim-manzi-in-5-easy-steps/

http://www.ordinary-gentlemen.com/2009/12/the-problem-of-denial/

Though it's hard to fault Manzi for exploiting the asinine accounting in some of the DICE modeling. Lord knows I would if I were opposed to pricing emissions.

Still, a more honest version of his 3% is more like 1-11%

Dean said...

I'm not sure that keeping Pachauri in is good news. It's true that his replacement would have encouraged deniers, but in acting on the recommendations of the IAC report, the process must be through and has to be forward-looking, by which I mean how it affects the next report, and not based on what those folks would think one way or the other. Maybe they decided that it is too late in the process to replace him now, but now the recommendation that each report has a new director is not going to be implemented for quite some time. So this one is mixed for me.

Tom Fiddaman said...

The 3% isn't really representative of the full spectrum of opinion in IPCC scenarios, but it's not a bad assessment of the central tendency. The problem is that the sample of models generating scenarios is itself highly biased toward general equilibrium and intertemporal optimization. The assumptions about markets and behavior embedded in that dominant approach are far from reality.

Steve Scolnik said...

MT,
You might want to glance at:
When Bad Economics and Climate Science Collide. Some references also at:
Discounting and intergenerational ethics.
Given that economists have so little clue even at the short range, it's not surprising that they have even less at the ultra-long. John Maynard Keynes: "In the long run, we're all dead."

NewYork said...

TB covers this in his rebuttal of Manzi, but one of the flaws in Manzi's argument (in addition to ignoring other analysis) is that it brushes aside worst-case scenarios. He simple takes the mid-point of the GDP range of a 4 C warming scenario and neglects the much higher costs if warming is on the high range (getting into double digit GDP loss). Expected value of such a distribution would be higher, as would the risk. Manzi does not appear to understand risk management.

A fallacy is also implying that damages from global warming are limited to GDP. A few hundred million might perish in a 3rd-world nation and it might not have much of an impact on GDP. Mass migration also has social/cultural costs not factored into GDP. Do all costs of increased military conflict over resources manifest itself in GDP? Where do mass extinctions come into play?

Finally, comparing GDP decline from mitigation with GDP decline from global warming has further problems. Most mitigation studies don't measure much of the benefits, such as environmental benefits (cleaner air/water), health benefits (an NRC study estimated burning energy, mainly fossil fuels, in the U.S. costs $120 billion per year), and less reliance on foreign sources of energy.

So I'd have to say Manzi's commentary is rather shallow.

Tom Curtis said...

I have generally ignored studies of future economic impacts of global warming as so much voodoo. Rather than use dollar figures, I tend to look at expected extinction levels instead, in that if, for example we are facing a Permian level extinction event from Business As Usual, that is something we want to avoid, even if we only consider likely economic flow ons. As it happens, a Permian level extinction event from Global warming is certainly possible, if not probable. Even if the net impact of BAU was only an economic loss of 3% in 2100, an approximately 5% (ball park figure) of an Permian Level extinction over the next 2 centuries makes BAU not a viable option.

However, I believe that Manzi has misread the claim. The claim comes from AR4 WG 2 Summary for Policy Makers, which in turn refferences a Chart from the Stern Review. Having skimmed the relevant sections, it appears to me that the chart reffers to annual losses, not total losses to the period. The loss is not just 3% in 2100, but also 3% in 2099, with the reduced growth from that further reducing the GDP in 2100 (and so on back).

Reading further, we see that Stern believes the total cost of BAU if you discount losses in 2100 to 40% of current values will be 20% of GDP. The discounting is important because annual 3% GDP losses from Global Warming means that by the end of the century GDP growth will be negative (falling) (expected annual growth of 1.3% without global warming minus expected 3% annual costs = -1.7% annual growth).

If you don't devalue the future, the impact of BAU global warming is, therefore, approximately the same as annual global financial crisis for the forseeable future by the end of the century. Ignoring ecological costs.
http://www.ipcc.ch/publications_and_data/ar4/wg2/en/spmsspm-c-15-magnitudes-of.html
http://webarchive.nationalarchives.gov.uk/+/http://www.hm-treasury.gov.uk/d/Chapter_6_Economic_modelling_of_climate-change_impacts.pdf

Michael Tobis said...

The difference between 3 % lost GDP at 2100 and 3% lost growth per year is sufficiently spectacular that I find it hard to believe that there is room for misinterpretation.

That said, I will go along with the voodoo angle.

I see Richard Tol is responsible for some of these numbers on the low side. I find it hard to believe that anyone as crude and impulsive as Tol, who accused me personally of favoring Maoist-style re-education camps because I mentioned the word "education" and refused to back off, is doing anything resembling serious work. In doing so, Tol essentially reduced himself to "worthless crank" in my estimation, and if WG III and the related community takes him seriously that doesn't help their reputation in my book.

I am not sure where your 1.7% baseline growth rate comes from. In my understanding the global growth rate assumed by BAU economists is much higher.

There is some circularity of argument, though. Doesn't the discount rate essentially amount to the growth rate?

As I understand it, economics as a discipline says almost nothing about negative growth regimes. A negative discount rate pretty much makes any long term impact infinite and any present expenditure counterindicated. Basically the whole structure breaks down, right?

Tom Curtis said...

The 3% cost figure is drawn from a chart (Stern's figure 6.6) which plots economic cost against temperature. As it does not include the duration at which the temperature holds, it cannot logically indicate costs over any duration beyond a year. That data is then used to calculate a utility function for the effect of global warming (Stern's box 6.3).

It is not clear to me that Stern allows the costs of global warming to compound in his calculation; but it is clear to me that he should. If GDP grows by 1.3% per annum globally with no effect from global warming (Stern's figure), and global warming introduces a 3% of GDP cost, then net growth is -1.7%; and growth (positive or negative) compounds. Presumably the negative growth can be overcome by shifting production from consumption to investment, but then the percentage of GDP devoted to consumption should fall over time (Stern keeps it at a flat 80%).

The discount rate is used in economics to reflect both asset depreciation and the uncertainty of outcomes. As Stern's calculations are pegged against per capita GDP, asset depreciation is irrelevant, so essentially Stern's discount rate is a measure of uncertainty. (This is why it is lower than standard.) Essentially, as he puts it, his discount rate reflects a probability of human extinction for reasons not related to Global Warming before the full impacts of Global Warming are upon us. His discount rate reflects a probality of 9.5% that the human race will be extinct by 2100. In his own words, "The value of δ is taken to be 0.1% per annum, so that the probability of surviving beyond time T is described by a Poisson process e-δT, where δ is the annual risk of catastrophe eliminating society, here 0.1%. So the probability of surviving beyond, say, 2106 is e-0.001*100, which is 90.5%." (Stern Review box 6.3 page 161)

Personally I think applying a discount on this basis is nonsense, not because I think the probability assigned is particularly pessimistic, but because you don't plan for the future by assuming you won't have one.

Jim Manzi said...

Thanks for the thoughtful reactions.

The Stern Review does make different projections than the IPCC. I quoted IPCC projections, designed to be and generally recognized as the best available summation of existing evidence on the topic, as my source and indicated this clearly and repeatedly. I have elsewhere -along with many, many others -
criticized the methodology of the Stern Review.

I have replied to the ThingsBreak criticism, which I think is not serious. (You can find my reply at TNR).

I have addressed (and linked to this in the TNR post indicated) the uncertainty question, whihc as I have said repeatedly, is the central issue.

In the end, you have a problem with the IPCC, not me.

Best,
Jim Manzi

Hank Roberts said...

You're ignoring ocean pH change, Jim Manzi.

This is like arguing that an avalanche isn't a problem because you're wearing warm socks and so your feet won't get cold, no matter how deep you're buried.

You ignore the big problems and pick just the GDP number out of context as though it were adequate to sum up the costs.

Hank Roberts said...

> vastly richer per capita
> than today

"A man is rich in proportion to the number of things he can afford to leave alone." -- Thoreau

"August 21st marks an unfortunate milestone: the day in which we exhaust our ecological budget for the year.... From that point until the end of the year, we meet our ecological demand by liquidating resource stocks and accumulating carbon dioxide in the atmosphere."

http://www.footprintnetwork.org/en/index.php/GFN/page/earth_overshoot_day/

Hank Roberts said...

Or if you just want to count people:

"Let's see how much of the world we can destroy before getting to 5% of global GDP.....: 81 countries, which collectively have ... about 43 percent of the world's population."

http://www.fivethirtyeight.com/2009/06/how-to-destroy-almost-half-planet-for.html

Michael Tobis said...

Jim: "In the end, you have a problem with the IPCC, not me."

Well, yeah, I suppose so. But that still leaves your position desperately wrong. Perhaps you could take uncertainty into account.

Many people think that people in 2100 will NOT be vastly richer than people today, and that growth is not sustainable. What probability should we assign to that? And how should we cope with that branch of the calculation? Because arguments from discount rates collapse into a meaningless heap if those are true.

You are not just taking IPCC for gospel, which was never the point. You are taking it as deterministic, which no sensible person would claim. Try taking a risk management point of view. If ever there were a risk management situation, this one is it.

thingsbreak said...

Jim Manzi: I have replied to the ThingsBreak criticism, which I think is not serious. (You can find my reply at TNR).

Where, exactly? In your TNR piece you still claim that we can expect 4°C of warming for A1FI and thus 3% loss of GDP. The more honest accounting is 2.4-6.4°C, yielding a range/estimate of 1-11%/5.5% GDP, not 1-5%/3%. You don't address this at all.

You don't address that there are CBAs for mitigation putting the cost lower than even the flawed 3% you use.

I would comment at TNR, but I apparently would have to subscribe to it to do so.

In what way did you reply to my "not serious" criticism, Jim?

Hank Roberts said...

> Where, exactly? In your TNR piece

I can't find it either.

Jim Manzi said...

Michael,

In the TNR piece under discussion, I linked to my earlier response to Gore in which I tried to address the issue of uncertainty at great length here:

http://www.tnr.com/blog/critics/75757/why-the-decision-tackle-climate-change-isn%E2%80%99t-simple-al-gore-says

As I argue at great length here (and at even greater length in numerous other posts), the central issue - exactly as you say - is uncertainty. But I believe if you think it through carefully, you'll see that it does not lead directly to the conclusion that we should implement aggressive emissions mitigation.


The link to the TNR piece replying to the TB argument is here:

http://www.tnr.com/blog/critics/76250/one-more-reply-bradford-plumer

Best,
Jim Manzi

Michael Tobis said...

Oh lord save us from WG II.

Jim, I do not know why you feel compelled to go to the middle of the range. If the number is 5% of GDP accumulated after the 4C increase, it is purely and simply negligible as you say.

I just don't buy it for ten seconds.

Of course, adding sea level rise and ocean acidification it gets worse, but those are neglected in AR4.

Any idea what the [F20.3] citation means?

Michael Tobis said...

It sure doesn't look like AR4 WG II Section 20.3 which reads like a precursor to the global boundaries statement of Rockstrom et al. and says nothing about GDP.

This is a bigger deal than the Himalayas glitch. Where did this bizarre 1-5% range come from?

thingsbreak said...

The link to the TNR piece replying to the TB argument is here:

http://www.tnr.com/blog/critics/76250/one-more-reply-bradford-plumer


Perhaps I'm blind, but I don't see anything in that that addresses this:

The more honest accounting [for A1FI] is 2.4-6.4°C, yielding a range/estimate of 1-11%/5.5% GDP, not 1-5%/3%.

If you want to back off using A1FI, you're free to do so. But if you're going to use A1FI as your "most extreme" scenario, why not give the more honest range/estimate? By focusing on the central value alone rather than the likely range, you're introducing an element of false precision that conveniently cuts the estimated GDP loss average virtually in half.

Now, I'm not accusing you of deliberately doing so in bad faith, but the fact that you continue to do so despite this curious accounting being brought to your attention (and your claim that you've addressed it when you've sidestepped it completely) is starting to give me pause.

thingsbreak said...

MT: Where did this bizarre 1-5% range come from?

Check your email.

Hank Roberts said...

There's no uncertainty about ocean pH change. Forget all about global warming and radiation physics.

What else can be done but drastically reduce fossil fuel use as quickly as possible?

What's the GDP of the oceans?
What was it a century ago?

Michael Tobis said...

Aha F20.3 means Figure 20.3. D'oh.

Thanks TB!

Michael Tobis said...

Interesting. These costs are obtained as follows:

The DICE model, developed by William Nordhaus, is a dynamic integrated model of climate change in which a single world producer-consumer makes choices between current consumption, investing in productive capital, and reducing emissions to slow climate change.

... Current carbon emissions add to atmospheric concentrations via a fixed retention ratio, and realized temperature change is modeled by a three-box model representing the atmosphere, mixed-layer upper ocean, and deep ocean. Damage from climate change is a quadratic function of realized temperature change with a 3-degree change calibrated to cause a 1.3 percent world GNP loss.

Emphasis added.

Begging the question, wouldn't you say?

<a href='http://sedac.ciesin.org/mva/iamcc.tg/TGsec4-2-15.html">http://sedac.ciesin.org/mva/iamcc.tg/TGsec4-2-15.html</a>

If the summary said "using a model calibrated to give the result that 3 C yields a 1.3 % GDP loss, we find that total losses are expected to be on the order of 1.3 % of GDP", as seems to be very nearly the case, would Jim Manzi bet the farm on it?

Is this really what WG II says about the matter? Well, let's check the executive summary.

"It is likely that the gloabblyaggregated figures from integrated assessment models underestimate climate costs because they do not include significant impacts that have not yet been monetised. It is virtually certain that aggregate estimates mask significant differences in impacts across sectors and across regions, countries, and locally. It is virtually certain that the real social cost of carbon and other greenhouse gases will rise over time; it is very likely that the rate of increase will be 2% to 4% per year. By 2080 it is likely that 1.1 to 3.2 billion people will be experiencing water scarcity; 200 to 600 million, hunger; 2 to 7 million more per year, coastal flooding."

So that is what IPCC actually says. Essentially what I just noticed. The Nordhaus type models are barely models at all. They are assumptions.

Michael Tobis said...

Sorry for busted link above. here.

James Annan said...

Our climatic change paper has a brief discussion based on IPCC. I don't think the 3% figure is an unreasonable summary of the literature. Note that we did not just take a mid-point in our paper but explicitly integrated across the probability distribution.

Stern's 20% was an upper end value assuming high sensitivity and adding in some extra effects to make sure he got the answer he was looking for. I think we can expect to see a lot more of that in the future.

Michael Tobis said...

James, thanks. I think even a one time 20% hit amortized over a century is noise. If I didn't think the matter were about actual decline, I wouldn't think the policy question was interesting.

I believe that conventional economics cannot prescribe policy in conditions of persistent decline. Therefore it assumes away the possibility. Thus the aggregate growth over a century is constrained to be substantial.

If, for example, 5 million people per year and their associated public and corporate wealth are wiped out by irreversible flooding (this is presumably based on the low sea level rise numbers of AR4), it is reasonable to extrapolate that to the GDP associated with 500 million people, or about 5% of the total. That's a 5% hit compounded annually, not even counting other effects or nonlinearities.

Suppose we go with fifty years of no impact, then gradually climbing to 5% / year. I get that as equivalent to a one time 78.3% hit. Now that is starting to be worth noticing, and it's just accounting reasonably for the modest coastal flooding in AR4.

Michael Tobis said...

Oh, for those with less intuition for numbers, suppose I I put in 5% a year from the beginning.

Then the total decline in GDP relative to a case without the 5% hit would be 99.4 % over a century; 99.996 % over two centuries.

Jim Manzi said...

ThingsBreak:

Here is the paragraph which I published:

[Open Quote}

According to the currently governing Fourth Assessment Report produced by the United Nations Intergovernmental Panel on Climate Change (IPCC), under a reasonable set of assumptions for global economic and population growth (Scenario A1B), the world should expect to warm by about 3°C over roughly the next century (Table SPM.3). Even in the most extreme IPCC marker scenario (A1F1), the best estimate is that we should expect warming of about 4°C over roughly the next century. How bad would that be? Also according to the IPCC (page 17), a global increase in temperature of 4°C should cause the world to have about a 1 to 5 percent lower economic output than it would otherwise have. So, if we do not take measures to ameliorate global warming, the world should expect sometime in the twenty-second century to be about 3 percent poorer than it otherwise would be (though still vastly richer per capita than today).

[Close Quote]

Please identify the erroneous statement.

Thanks,
Jim Manzi

thingsbreak said...

Jim Manzi: Please identify the erroneous statement.

Let's say that I'm running a company that had manufactured a faulty product. We've received some law suits and expect to receive more. I want a CBA comparing the relative merits of recalling the product or settling/paying out the suits.

You tell me that in the most extreme projection, we can expect a 3% loss in profit, and an optimistic consulting firm has pegged the cost of a recall at 5%, suggesting that from a simple CBA it's not worth recalling.

Then Min Jamzi comes in and tells me that you haven't been giving me the whole story. The "most extreme" scenario actually suggested losses of 1-11%. And additionally, several other consultants have pegged the cost of a recall at 2.5%.

Would I characterize what you did as "lying" or "erroneous"? I can tell you that I would have felt like you misled me on critical details, and that in your oversimplification you dramatically changed the outcome of that superficial CBA.

The point of what you're doing is to inform people, yes? You're not trying to persuade them that mitigation should not be pursued, you're just trying to tell them about the relative costs, correct?

If that's the case, don't you think your audience is better served by giving them a more accurate accounting rather than the false precision of your estimate? Saying that by not being explicitly erroneous your argument is not misleading is a little disingenous, from my perspective.

All that aside, you claimed to address my "not serious" critique. I am not the one who forced you to talk about A1FI. Your complaint of me talking about that completely side-stepped the issue of the likely range of damages under that scenario.

Or was there a different response that I have overlooked?

Aaron said...

I suggest that ALL of the climate models under estimate the rate of climate change by ~2 orders of magnitude, and thus nobody has really done a complete fault tree analysis of what could go wrong in a globally linked economy. The costs are higher, and sooner than any respected writer has hinted.

In 2007, GCM were saying the Arctic Sea Ice would be around for another 80 years, while in fact a quarter of it was gone in 0.8 years – way ahead of any timeframe, any peer reviewed GCM had suggested. As a result, we are seeing warmer and wetter air masses come out of the polar regions and these are affecting global circulation, way ahead of the timeframe suggested by the those GCM. For example, Russia gets hot, Pakistan gets wet, and warm moist air is flowing across the Greenland Ice sheet. However, everybody is still ignoring Feynman’s calculations on ice, and their implications on how fast ice sheets on land can breakup (and sea level rise).

The respected writers respect the peer reviewed literature without looking at the social dynamics within respected academic and governmental organizations that might constrain the tail(s) of the published probability curves, and their summary by IPCC. Moreover, any product coming out of US DOE (& related national labs) is very much affected by such social dynamics.) In a conceptual world that accepts the results of GCM, very rapid sea level rise is impossible. I understand that. I also understand that there is a lot of good physics that are not in GCM. For example, GCM do not understand moulins in ice sheets.

Just as two pathogens damage bees more than either pathogen alone, loss of multiple industrial facilities around the world would be much worse than the loss of just those facilities in one industry, or on one continent. We have assumed that the global industrial system has enough redundancy that no single weather event could cause major damage to either industry and certainly not to both industries simultaneously. Rapid sea level change is one exception to that assumption. Major crop failures are a second exception. Climate change does not care about our assumptions.

We have assumed that Sea Level Rise will be much slower than the plan/permit/finance/engineer/build cycle for major industrial facilities. We have assumed that weather patterns will change slower then agriculture can adapt to the new climate. If these two assumptions are correct, my points above are not problems. This is the conclusion based on the traditional GCM models, i.e., loss of Arctic Sea Ice is not a big deal because we have 80 years to adapt, and our plan/permit/finance/engineer/build cycle is only 30 years; and sea level rise is not an urgent problem because we have hundreds of years to adapt. With the timeframes suggested by old (circa 2007) GCM still in our heads, we procrastinate. In fact, time is running out 100 times faster than we thought.

Michael Tobis said...

Aaron, I don't think you can get to 2 orders of magnitude based on Arctic sea ice alone. That is a rotten-cherry pick.

If you could, though, it would makes no sense to be commenting on blogs as we would be altogether doomed, and soon.

Michael Tobis said...

Manzi: Also according to the IPCC (page 17), a global increase in temperature of 4°C should cause the world to have about a 1 to 5 percent lower economic output than it would otherwise have.

This is p 17 of WG II AR 4, which addresses these issues. It also says "it is very likely that globally aggregated figures underestimate the damage" ON THE SAME PAGE. I do not know what to make of this beyond confirming my original impression that AR 4 WG II is a mess. Manzi seems to prefer picking the convenient statement and ignoring the other one, BUT THEY ARE EQUALLY AUTHORITATIVE on the exercise of taking IPCC WG II as completely authoritative. So I think Jim should reconsider.

I usually admire TB's reasoning but I am not with him in this case. These impacts amortized over a century are small, no matter how you slice it. The question at hand is whether we believe it.

The WG II SPM tries to have it both ways. Jim seems to take that to mean he can pick and choose.

But this isn't a debating club. This is the world. Getting it right matters.

James Annan said...

mt,

I think you are conflating annual and one-off costs.

Wiping out 5% of the population (and associated wealth) is a 5% one-off cost. (In your scenario, an *additional* 0.05% per year for 100y.)

(of course there will be knock-on effects, but let's get the essentials fixed first).

The fact that the economy will be 5% smaller indefinitely into the future (compared to if the 5% was not wiped out) does not compound to anything. Or at least, it only compounds to 5% of the *compound* output of the world economy. Apples and apples comparison. There is not an additional 5% loss every subsequent year. The lost land is only lost once.

We can debate whether 5% is a reasonable figure, but that is another issue. Standard ec modelling really does say that the impacts of AGW will make us 9.8 times richer (9.5 if you prefer) rather than 10 times richer in 100y.