Risk, Uncertainty and Irreversibility, Decision Theory, and Global Warming

Alex Tabarrok protests that you should not do expensive and irreversible things before you know what is going on. I think that there is an important distinction to be drawn in decision theory between (a) risk on the one hand, and (b) the interaction of uncertainty and irreversibility on the other.

In general, briefly: As a baseline, you should start out thinking that you should do what it would be best to do if your central-case forecast were to come true. And then:

1. Relative to that baseline, you should do more of things that reduce risks: a prudent portfolio should have some gold or silver in it (not much!) to guard against the potential inflationary collapse of fiat money systems and other political risks.
2. Relative to that baseline, you should do less of things that increase risk: invest less in risky enterprises, and dump less carbon dioxide and methane into the atmosphere.
3. You should delay doing expensive and irreversible things until you are confident that they are needed.
4. You should prevent processes that could cause irreversible or incredibly-expensive-to-repair damage until you are confident that their effects will be harmless.

The first two principles are the appropriate ones for dealing with risk. The last two principles are appropriate for dealing with the interaction of uncertainty and irreversibility.

Thus if congress were on the verge of banning open-carbon-cycle power generation and vehicle operation in 2010, I would think that that was a bad idea because of principle 3. But congress isn't. The uttermost limit of political feasibility over the next decade is a fifty-cents-a-gallon-equivalent tax on carbon emissions. And that seems well-covered by principle 2.

Here's Alex:

Marginal Revolution: Uncertainty is the Friend of Delay: Regarding global warming and what to do about it, Brad DeLong approvingly paraphrases Tyler, "uncertainty is not the friend of doing nothing." Bearing in mind the obvious dangers of contradicting both Brad and Tyler let me counter with "uncertainty is the friend of delay."... Suppose... that we are uncertain... there is a strong argument for delay. The argument comes from option pricing theory applied to real options. A potential decision is like an option, making the decision is like exercising the option. Uncertainty raises the value of any option which means that the more uncertainty the more we should hold on to the option, i.e. not exercise or delay our decision....

Applying the theory to global warming isn't easy because our decisions involve many options and exit costs; but if we think that our knowledge of the extent, cost, cause and solutions to global warming are increasing at a faster rate than the danger of global warming then delay of any major decision is a rational policy at the present time.

And Tyler Cowen rebuts:

I see the difference between us as such: you write: "making the decision is like exercising the option...". But many costly decisions increase option value. Such decisions include more R&D, more savings, a compensation fund, investment in greater economic flexibility and response capabilities, and so on. Anything we do counts as a "decision," so we cannot postpone decisions per se. We can and should postpone relatively irrevocable decisions, and under that heading "consumption" is a prime candidate. The correct implication of your argument is that when uncertainty increases, we should (under some specific conditions outlined in my post) consume less, which is close to my point of view...

Victor argues:

I agree on the difficulty in applying option theory to global warming policy. There’s a myriad options with varying cost/lead-time trade-offs. The uncertainties are considerable, have many dimensions, and the flow of information to resolve them is stochastic. But option theory doesn’t always argue for delay. If your doctor said there was a chance you had a flesh-eating bacterial infection that would take several days for the tests to resolve, would you delay or accept treatment? It’s a no-brainer if the treatment was just taking a bottle of antibiotics, but if your arm is turning red immediate hospitalization for intravenous antibiotics might be the optimum strategy. Acting now can preserve options going forward. At a simple level for communication, an insurance framing of the problem works better for me.

Comments

[JR]

We are in a canoe moving with the current on a broad river. The weather is fine and we are enjoying our rapid progress. Suddenly a man on shore shouts out something that sounds a bit like "Rapids ahead!!" In these conditions of uncertainty, do we beach our canoe? Do we paddle closer to shore so that we can beach quickly if necessary? Or do we stay out in mid-channel, delaying any action until we have better information?

[George Bush]

Bah. Go with your gut.

That's the best way to be a decider, which is what I am.

[Bruce Wilder]

Did we buy the $1,000,000 policy at the canoe rental place? If so, I say, "paddle harder". Let's get our money's worth!

(Isn't that going to be next argument? That any act of collective prudence will be overwhelmed by the impulses of moral hazard? After all, strapping on a seat belt leads drivers to drive more recklessly, doesn't it? [rolling eyes])

[bakho]

Prior to the anti-pollution legislation of the 70s, industries and the military created all kinds of superfund sites that have proven to be far more expensive to remediate than it would have cost to prevent.

If we were spending the Iraq War for Oil dollars on energy independence, building energy efficient devices, improving energy efficiency of our infrastructure and working on alternative energy we would be much better off. Big oil is not going to release its grip on our ecomony without a fight.

[John Emerson]

Leibniz said something like "People would disagree about the multiplication table if there were enough money in it". There are still people uncertain about whether smoking tobacco causes cancer. For that reason, there will always be uncertainty.

On environmental issues, economists seem torn between two goals: first, warding off excessive environmentally-motivated government intervention in the economy, and second, mickey-mousing economics until professional economists are able to use the technical language of economics to talk about environmental questions without sounding venal or ridiculous. These two goals are not necessarily compatible.

Abstractly discussing questions about global warming and climate change, topsoil depletion, water shortages, and air and water pollution, without any concrete contexting in terms of the actual present state of affairs and direction of change in the physical world of the environment, strikes me as malicious.

In particular it would seem malicious if economists were worried about doing too much when the actual problem we were facing was doing too little.

Economic geographers, agronomists, climatologists, and related scientists studying the non-financial physical world need to be front and center in this debate.

[Roger Albin]

"Alex Tabarrok protests that you should not do expensive and irreversible things before you know what is going on"

Like more than doubling the atmospheric CO2 concentration?

[shah8]

option theory seems to be a bad handle on the problem. They assume that the uncertainty is of natural causes, and is not driven by some agency. In a market, the number of players is wide enough that any one agenda is smeared against all the others.

The issue here is that there are few sources of data, and none who do not have conflicts of interest, not when it comes to the environment. At heart, uncertainty is a source of profit for certain players who can make themselves heard over the dinn, because it *drives delay* to the point that when the inevitable occurs and action has to be taken now, only people with the solution already hashed out, and with authority in line, will get to be heard. My approach would be forcing all the major players of pollution to a kind of insurance scheme that is structured to give them an incentive not to pay out of the pool...

[jlcg]

Nothing of any importance will be done, ever. Catastrophe will do it for those that survive it.

[Jon Fernquest]

IMHO better start from scratch. Changing the system piecemeal, we're all doomed. Life on planet earth is not a video, just hit rewind if you screw up. This is a time for "command economy" with Gore.

Install a little visible screen above every citizen's head with **their time-discounted contribution to the destruction of planet earth** via environmental degradation, climate change, not thinking of their grandchildrens' grandchildren...etc

Then creaate a 51st state and populate it with citizens dedicated to saving planet earth (but still having a good time)...that probably means a lot of telecommuting, public transportation, bicycling, and walking, chopping fire wood (but replanting it of course) good healthy living...hire a couple of environmentalists as consultants to find out other ways to reduce meter readings, then publicize the aggregate lower meter readings, and the fact that one can still have a good time without an exhaust spewing SUV, provide incentives for the rest of the US to move over to this 51st state, maybe it could even be marketed as a "war on the environment"(or "war on your own bad habits") for those who like war, enlightenment for the Buddhists, Plato's Republic for the classically inclined

[Bupa]

Now you want to apply option theory to Global Warming?

Despite Stern's attempt to frame this as an economic question the fact is that economists are out of their league here.

Lot's of discussion on the cost of trying to do something to prevent catastrophe. Looking for the lost key under the lamp post.

No discussion on the benefit of avoiding catastrophe. Stern says 5 to 20 percent of GDP. What was the going price for species extinction? Is the price higher for some species than for others? What is the price for hundreds of millions of human climate refugees?

Now option theory. What price for the risk of species extinction? Is the price higher for some species than for others?

Economist will dither as the pot boils because their tools of analysis are not up to the challenge we face.

[Walt]

Roger Albin wins the thread.

[unemployed observer]

we should also note that global warming is a public good (public bad) that some kind of regulation/action is needed facing high transaction cost.

what we are facing is the risk outcome between -200 to 0, rather than between -100 and +100.

[Alex Tolley]

I don't think the option metaphor is appropriate. Making an error on an option doesn't bring down the whole financial system, which is more likely the result of global warming on the global economy under some scenarios. These false models are just as fatuous as those earlier discussed about wealth transfers from the future. They make huge implicit assumptions about growth that are unwarranted. The US is currently engaged in a war that is clearly more far costly than originally envisaged, and yet we screwed up the cost benefit analysis pretty well there.

If we're lucky, global warming may be livable. Then again, it may not be. Certainly we can predict that a rising sea level of even 10 meters is going to create a lot of coastal land loss and hence displace a lot of people, most of them unable to migrate. What is the cost of this going to be?

James Burke said it far better in his prescient documentary of the early '90s - "The Great Warming". One of his scenarios was a great wave of refugees trying to reach Australia from Indonesian islands, and the resulting war to keep them out.

How do we factor the costs of these possible scenarios?

I recently noted that the first accurate land survey of Greenland was being done. It this point it is to determine what the land looks like to help with models of glacier movements. For option lovers, how about talking the Greenland govt into starting a land option/warrant program for plots that may emerge from the ice in this century? "Granite valley, guaranteed dry for sea level rise to 20 meters, ready to terraform at owners expense". Trump, Milken, are you listening?

[Barkley Rosser]

unemployed observer,

Well, I would agree that the downside risks/uncertainties are a lot scarier and more dangerous than the possible upside outcomes. But it should not be forgotten that the main forecasts have several countries gaining economically from global warming, lower heating costs in winter being just one example of such a gain. Indeed, the two biggest CO2 emitters, the US and China, are near washes in terms of the gains and losses economically in the main forecasts.

The kicker from the economics side is that the biggest losers (on the main forecast) are likely to be poorer and less powerful countries. Aside from those like the Seychelles Islands that are in danger of ceasing to exist entirely, the biggest likely loser is Bangladesh, with an estimated 20% loss of GDP.

Of course anyone who wishes to point out that these calculations do not fully or properly measure various non-economic ecological losses is fully justified in doing so.

[Dennis]

I feel like the biggest problem with applying option theoretic intuitions to climate change is the fact that strike is nonconstant; this expands on Victor's point in Brad's post. Climate change is like flesh-eating bacteria in that the costs you pay to fix the problem increase with time in some complicated way, because it's harder to undo more emissions than fewer and because you'd rather be cured before your arm was eaten off than after, respectively. In either case, the argument that early option exercise is almost never optimal just doesn't apply; you may well wish to call at $50 today if tomorrow's strike will be $75.

[garhane]

Perhaps the general problem of AGW is more like preparations for war than whether to build a dam or a nuclear plant. If it is July 1938 do you, as Air Ministry in the UK build a new factory to crank out Spitfires fast, say 1000 in less than a year, or just order up a few dozen. They did build it, and pretty soon they wanted a thousand, more. But who could see that clearly in the summer of 1938 after Munich? They were concerned to maybe cancel out that plane because it was hard to fabricate, they were worried about how to fit a bunch of orders for that plane into the next generation of fighters and bombers....and so on. I just can't see that interest allocations would have been all that important. And what is all this stuff about crowding out private investment? Are the AGW expenditures to be rated like additions to theme parks, or like prep for vital national interest. And who ever heard of our type economy operating at anywhere near 100% of capacity, since whatever you think it is, it always turns out you can do a lot more if you have to. We had rationing during WW11 and it worked fine; they just stopped making private cars and you learned to do quite a bit on the tools. I mean some of the comment on Stern makes me wonder if maybe Brockway had it right after all.

[JonF]

Re: The weather is fine and we are enjoying our rapid progress. Suddenly a man on shore shouts out something that sounds a bit like "Rapids ahead!!" In these conditions of uncertainty, do we beach our canoe?

To make the analogy more honest, we should added that the shoreline of the stream is known to be infested with highly venemous snakes, although it is not known if there are ny close at hand at the moment.

Re: One of his scenarios was a great wave of refugees trying to reach Australia from Indonesian islands, and the resulting war to keep them out.

The Indonesian islands are very mountainous and in most places they rise quite quickly from sea level. Indonesia is not going to be submerged. Ditto for Japan, the Philipines, Madagascar, the British Isles, Iceland, Sri Lanka, and most of the Mediterranean and Caribbean islands. The areas of greatest risk are the smaller island chains in the Pacific (not including Hawaii, which is also mountainous), and some in the Indian Ocean. And also some low-lying land areas, most notably Bangladesh.

[John Emerson]

Neither warming per se, nor rise in sea levels is the most worrisome thing. Changes in the distribution of rainfall will probably do the most damage. Some will benefit, some will lose, but during the transition there will be a net loss, and there's no guarantee that there won't be a long-term net loss.

[JonF]

Re; Neither warming per se, nor rise in sea levels is the most worrisome thing. Changes in the distribution of rainfall will probably do the most damage.

I disagree. Huge numbers of people (and lots of valuable property) are located in coastal areas. Not to mention most of our trade and shipping facilities. Rainfall patterns are unlikely to change much (and overall a warmer Earth will be a wetter Earth; today's deserts are likely to shrink). Far more economic damage and dislocation will occur because of coastal flooding.

[anne]

There is so much complexity:

http://www.nytimes.com/2005/05/24/science/earth/24clim.html?ex=1274587200&en=52b4f1b89813887a&ei=5090&partner=rssuserland&emc=rss

May 24, 2005

Ocean Warmth Tied to African Drought
By ANDREW C. REVKIN

Few places are more vulnerable to drought than Africa. From the Sahel south of the Sahara to the southern lobe of the turbulent continent, there is a simple calculus, said Dr. Richard Washington, an expert on the region's climate at Oxford: "When the rains fail, people die."

So a concerted effort has been made over the last five years to understand what drives dry spells there and what will occur in a future warmed by accumulating heat-trapping emissions in the atmosphere. Most climate experts now say that significant human-caused warming is inevitable.

One new study bodes particularly poorly for southern Africa, indicating that a 50-year-long drying trend there is likely to continue and appears tightly linked to substantial warming of the Indian Ocean.

The authors of the study say that the heating of that ocean, which lacks the natural variability of the Pacific and Atlantic, is one of the clearest fingerprints pointing to human-caused climate change.

"In our models, the Indian Ocean shows very clear and dramatic warming into the future, which means more and more drought for southern Africa," said Dr. James W. Hurrell, a scientist at the National Center for Atmospheric Research in Boulder, Colo., and an author of the study. "It is consistent with what we would expect from an increase in greenhouse gases." ...

[John Emerson]

Rainfall patterns are unlikely to change much (and overall a warmer Earth will be a wetter Earth; today's deserts are likely to shrink)

As I understand, both these statements are false. A warmer earth probably won't be wetter, and rainfall patterns probably will change.

I don't claim to be an expert about climatology, etc., etc. One of my main points here is that we've been having all these Stern Report dscussions with no apparent participation by anyone who does know much abiut the earth sciences, agronomy, etc.

[John Emerson]

I have posted an extended version of my comments here on my site. T my URL, or here:

www.idiocentrism.com /warming.htm

[JonF]

Re; A warmer earth probably won't be wetter, and rainfall patterns probably will change.

While I suppose we can debate the meaning "change much", it is absolutely true (and solidly supported by geological data from prior warming eras) that warmer=wetter for the planet as a whole. Two simple reasons: 1) water from melted glaciers increases thge overall amount of liquid water in the hydrological cycle and B) warmer temperatures increase evaporation and humidity and hence also preciptation ("what goes up must come down" is also a hard and fast rule in the hydrological cycle). Anyone who claims otherwise is not engaging in science since they are ignoring both data and logic. And I don't care how many PhDs they flaunt after their name: science rests on reason and empirical evidence. Ignore both and you're in the same category with the creationists and the astrologers.

[John Emerson]

I don't have any PhDs after my name, and don't claim to be a climatologist. I've been asking for climatologists to contribute and noting that this debate has been dominated by economists so far. My 6:48 comment was pretty peripheral to the themes of this thread.

In an area of interest to me, central Asia, warmer weather was associated with dryness. But that doesn't mean that globally that could be true. So maybe I was wrong.

In what I've read, enormous changes in ocean currents, wind patterns, and distribution of rainfall (e.g. el Nino) had disastrous effects in many cases. That was my second point.

The global warming debate seems to get stuck on icreased temperatures and rising sea levels, but there will be multipl effects.

[Gerhard Fritz]

Try the Stern Review, an in-depth study on the costs and benefits of climate protection measures. The conclusion is that the costs of climate protection are much lower than the costs of doing nothing and going on as before. The study was commissioned by the British Chancellor of the Exchequer, or finance minister - not by Greenpeace.
STERN REVIEW: The Economics of Climate Change

http://www.hm-treasury.gov.uk./media/999/76/CLOSED_SHORT_executive_summary.pdf

Summary of Conclusions
There is still time to avoid the worst impacts of climate change, if we take
strong action now.
http://www.hm-treasury.gov.uk./media/8A8/C1/Summary_of_Conclusions.pdf

You will find links to all the chapters of the full version of the Stern Review on
http://www.hm-treasury.gov.uk./independent_reviews/stern_review_economics_climate_change/stern_review_report.cfm

[guthrie]

JonF- one of the problems with rainfall is that it wont necessarily fall at the right time of year. IF dry periods get wetter, rainfall may damage crops yields or encourage crop diseases.
Here in the UK, we are already experiencing stronger, heavier rainfall over shorter periods of time, as was predicted years ago. This means that many of our local drainage systems cannot cope, and fields get waterlogged. But our summers will be warmer and dryer. Which is great except that our farmers will have to buy more equipment to water the fields.

So, to sum it up so far, changes on the magnitude expected will have far reaching and not entirely clear consequences that will cause major problems for many people.

Those of you who would like to know more about predictions of rainfall etc should look here:
http://www.metoffice.gov.uk/research/hadleycentre/models/modeldata.html

Thats the webpage for the UK Met Office climate change projections. I suggest you all go and read it.