The Odds of Disaster: An Economist’s Warning on Global Warming

BY Paul Solman  May 23, 2013 at 2:54 PM EDT

By Martin Weitzman

No one can say with any assurance what the dollar value of damages would be from the highly uncertain climate changes that might accompany a planet earth that is steadily warming.

Paul Solman: Are headlines trumpeting the fact that carbon dioxide levels in the earth’s atmosphere have now passed 400 parts per million for the first time in something like three million years unduly alarmist? Or are they a timely warning?

I asked noted environmental economist Martin Weitzman to address the question.

An expert on the Soviet economy in the ’70s and ’80s, Weitzman first made news in 1984 with the publication of a book called The Share Economy, an argument for profit sharing instead of fixed wages. Fourteen years later came his paper Recombinant Growth, which revolutionized how some of us understood the enormous potential of technology.

But for many years, Weitzman has also been working on environmental economics and most recently, in a series of widely cited academic papers, on the economics of global warming; the most famous, on the “Economics of Catastrophic Climate Change.”

Weitzman’s central idea is not unlike the legendary bet proposed by the 16th century Catholic French philosopher Blaise Pascal. One way to interpret Pascal’s argument: even if you think the likelihood of God’s existence is vanishingly small, the cost if you’re wrong — eternal damnation — is infinitely high. An infinite cost times even a tiny probability is still … an infinite cost.

So you make a finite investment by believing in God and acting accordingly in order to avoid an infinite cost. To put it another way, you’re obliged, mathematically, to make the investment in belief.

You might keep Pascal’s argument in mind while reading Weitzman. Or think of the “Black Swan” argument of Nassim Taleb: certain events, however unlikely you think they may be, could have such enormous consequences, you just can’t take the chance of letting them happen.


Martin Weitzman: Recently the concentration of atmospheric carbon dioxide (CO2) reached an unprecedented level of 400 parts per million. What is the significance of this “milestone”? Does it portend catastrophic climate change? The short answer is no. The long answer is a more complicated and more nuanced maybe.

The modern era of carefully measuring and recording atmospheric CO2 began with the work of famed scientist Charles Keeling. In 1958, Keeling began to accurately monitor daily CO2 levels atop Mauna Loa, the highest mountain in Hawaii. Keeling chose this location because it was so remote from manmade sources that it would accurately track average “well mixed” CO2 levels throughout the world. Thanks to Keeling’s pioneering work we now have a continuous ongoing record of CO2 levels since 1958.

In 1958, Keeling recorded an atmospheric CO2 level of 315 ppm. Every year since then the Mauna Loa station has recorded ever-higher levels of CO2 than the year before. Atmospheric CO2 concentrations have grown relentlessly over the years until they just recently blew past the well-publicized milestone of 400 ppm.

The 400 ppm milestone is basically just a round number. To see why it might (or might not) be viewed as something unusual, or even threatening, we need to examine a longer record of CO2 levels over time.

Carbon Dioxide Levels Over Time

There is a remarkable record of CO2 concentrations preserved in tiny bubbles in Antarctic ice cores going back 800,000 years. These measurements are less accurate than modern Keeling-style instrumental readings, but they are plenty accurate enough to see the big picture clearly. All throughout the past 800,000 years, which encompasses several ice ages and interglacial warming periods, CO2 levels fluctuated in a relatively narrow band between about 180 ppm (during the colder ice ages) to 280 ppm (during the warmer interglacial periods). For about the last 10,000 years we have been living in a warm interglacial period, with CO2 concentrations at about 280 ppm. Then, beginning with the industrial revolution about 1750, CO2 concentrations gradually moved up to Keeling’s accurately measured 1958 level of 315 ppm. Since then, as we have seen, CO2 concentrations have grown rapidly to the current 2013 level of 400 ppm.

So, the current CO2 concentration of 400 ppm is some 40 percent higher than anything that has been attained in the last 800,000 years. The glacial-interglacial cycles began some two and a half million years ago. Scientists estimate that a CO2 concentration of 400 ppm has not been attained for at least 3 million years. This rapid a change in CO2 concentrations has probably not occurred for tens of millions of years.

The point here is that we are undertaking a colossal planet-wide experiment of injecting CO2 into the atmosphere that goes extraordinarily further and faster than anything within the range of natural CO2 fluctuations for tens of millions of years. The result is a great deal of uncertainty about the possible outcomes of this experiment. The higher the concentrations of CO2, the further outside the range of normal fluctuations is the planet, and the more unsure are we about the consequences.

How Much Warmer Will It Get?

Carbon dioxide is a greenhouse gas. It is by now (and for some considerable time has been) beyond any reasonable doubt that increased levels of atmospheric CO2 lead to increased average temperatures. What is still uncertain and the subject of legitimate debate is the magnitude of this effect: how much CO2 leads to how much warming? Scientists do their best to give a number, but every scientist knows that his or her best number is uncertain.

Because global warming is uncertain, scientists use a formula to represent both the average degree of global warming and its likely range, as an eventual consequence of some given steady concentration of CO2. The trouble is, each scientist has his or her own favorite variant of the formula. In what follows, I use the “consensus” formula given in the last report of the Intergovernmental Panel on Climate Change (IPCC), an organization established by the United Nations Environment Programme (UNEP) and the World Meteorological Organization (WMO).

For CO2 at the current concentration of 400 ppm, the IPCC formula translates eventually into an average temperature change of 1.5 degrees Celsius (2.7 degrees Fahrenheit) with a likely range between 1 C (1.8 F) and 2.2 C (4 F). Global average temperatures have already increased by .8 C (1.4 F), so these ultimate temperature values do not look so very scary. Therefore 400 ppm of CO2 maybe does not look catastrophic by itself — if only we could stay at 400 ppm. What does look very scary, and maybe even catastrophic, is the speed at which we blew right past 400 ppm of CO2, with no visible end in sight — and what that might portend for ultimate global warming.

If we were to continue CO2 emissions up to an atmospheric concentration of 600 ppm of CO2, the IPCC formula translates into an ultimate average temperature change of 3.3 C (5.9 F) with a likely range between 1.1 C (2 F) and 5 C (8.9 F).

If we were to continue CO2 emissions to an atmospheric concentration of 800 ppm of CO2, the IPCC formula translates into an ultimate average temperature change of 4.5 C (8.2 F) with a likely range between 3 C (5.4 F) and 6.8 C (12.3 F). The world has not seen this level of CO2 concentrations for some 50 million years, when crocodiles and palm trees thrived in the Arctic Circle, Greenland and Antarctica were ice-free, and sea levels were hundreds of feet higher than today.

So, 600 ppm of CO2 looks a lot more worrisome than 400 ppm of CO2, and 800 ppm of CO2 looks a lot more worrisome than 600 ppm of CO2. The significance of just having blown past 400 ppm is that we seem to be on a business-as-usual growth trajectory that brings us to 800 ppm (or maybe even more) within a century from now.

The key links in the chain connecting increased atmospheric CO2 concentrations to global well-being are the following. Increased CO2 concentrations lead to increased global average temperatures. Increased global average temperatures lead to increased climate changes (and planetary changes, like higher sea levels). Increased climate (and planetary) changes eventually result in increased damages to humans and the planet.

It is critical to recognize that every link of this chain is full of deep uncertainty that makes it very difficult to answer the question: by how much?

We have already discussed the first uncertain link from ultimate CO2 concentrations to ultimate global temperature changes.

As for the second link, there is yet greater uncertainty. What will be the effects of higher temperatures on precipitation patterns? Will monsoon rains be greatly altered? What will happen to Indian or Bangladeshi agriculture? Will dry places in Africa become even drier? Will tropical storms intensify? When will the ice sheets covering Greenland and West Antarctica begin to melt seriously, thereby sharply raising worldwide sea levels? Will basic essential patterns of ocean circulation currents be changed? Will the Amazon rain forest dry out or die back? Will there be large-scale releases of currently contained CO2 and methane (an even more potent greenhouse gas) under melting permafrost, thereby accelerating the process of global warming itself? What about the truly stupendous amounts of methane trapped inside the offshore continental shelves by low temperatures — might they start to become unstuck by higher ocean temperatures, thereby triggering a vicious global warming circle? What will be the effects of large-scale rapid melting of ice in the Arctic Ocean? What about the unknown unknowns we have not even thought of?

The Link Between Carbon Dioxide Concentrations and Damages

The third link, connecting to damages, is even messier to deal with. The higher the temperatures, the more difficult it is to quantify the resulting damages. No one can say with any assurance what would be the dollar value of damages from the highly uncertain climate changes that might accompany a planet earth warmed by an average of more than 3 C (5.4 F). Economists do their best, but such estimates are mostly wild extrapolations from lower temperatures, or are just plain made up. And the higher the degree of global warming, the wilder and woollier are the numbers attempting to represent estimated damages.

So what is the overall relationship between CO2 concentrations and damages? This is, after all, the ultimate welfare connection we are interested in, but it consists of three highly uncertain links, where the uncertainty in each link increases dramatically with higher CO2 levels. The point is that for higher CO2 concentrations, the relationship to ultimate damages is enormously uncertain. Suppose we tried to express uncertain damages in the same language that we used to express uncertain global warming — a central average value and a likely range. Then, no matter how it were to be calculated, the likely range of damages would be enormously wide for high CO2 concentrations. For high CO2 concentrations, the upper range of climate damages would represent genuine climate catastrophe.

Relying on averages may be OK for small amounts of uncertainty. But climate change damages from high levels of greenhouse gas concentrations are enormously uncertain. In this kind of situation, for an economist, abating CO2 emissions is like buying insurance against a catastrophe. We should cut back on CO2 emissions not only to lower the average damages, but, perhaps more importantly, to lower the probability of catastrophic damages. That could imply a lot more CO2 emissions abatement than if we were concerned only about the most likely or average damages.

Discounting the Costs of Climate Change in the Future

To add to the complexities and uncertainties, there is the fact that long periods of time are involved. The really high temperatures would likely materialize, if at all, only in the course of centuries. The worse the magnitude of the climate disaster, the more likely is it to occur at a further-off future time.

One premise of modern economics is that we humans discount the future. This simply means that we value something that happens in the here-and-now — the present — more than we value it, right now, if we will only get it in the future. A dollar today is worth more than a dollar a year from now, for example. And that means that a dollar a year from now is worth less, in today’s money, than the dollar today.

We use a discount rate to compare the two — which is, in the case of money an interest rate. So if the discount or interest rate were 3 percent a year, a dollar a year from now would be worth 3 percent less — only 97 cents — than a dollar today. At a 3 percent discount rate, that is the so-called “present value” of a dollar you wait a year to get and spend. And indeed, 3 percent a year is a commonly used discount rate for rewards in the future compared to rewards today.

It’s important to notice that if an ordinary interest rate like 3 percent were used to discount the distant future, the power of compound interest is such that the present value of even very large damages could be made to appear small. A dollar today is worth 3 percent less than a dollar a year from now: 97 cents. Discount that 97 cents by another 3 percent to wait yet another year, and so on, and by the time you repeat the process for about 24 years, a dollar is worth just half what it is today. Wait 50 years and it’s worth 22 cents. Wait a hundred years and a 2113 dollar would be worth barely 3 cents to someone living in the present.

There is a vigorous debate among economists about what interest rates should be used to discount the inter-generational damages from climate change. If we value highly the climate-associated welfare of future generations then we should be using low discount rates — say 1 percent or less — which would register the present value of their catastrophic damages as if it were equivalent to a very high level of present damages — something that must be avoided by action now. If we used market interest rates, which are usually much higher, it could still be the case that catastrophic damages should be avoided by action now if the magnitude of the future catastrophic damages were high enough. So time and discounting introduce new wrinkles, but it could still be the case that what is most worrisome about climate damages is not their average or expected or most-likely mid-range value, but the extreme upper-end values associated with various sorts of catastrophe.

Once it is in the atmosphere, CO2 remains there for a very long time. Even if CO2 emissions were cut to zero at some point in the future (a very drastic assumption), about 70 percent of CO2 concentrations over the pre-industrial level of 280 ppm would remain in the atmosphere for the following one hundred years, while about 40 percent would remain in the atmosphere for the following one thousand years. This, along with the possibility of bad outcomes, is the argument for keeping CO2 concentrations from reaching very high levels.

Most people do not realize how difficult it is to stabilize CO2 concentrations. It is not nearly enough to stabilize CO2 emissions, which would cause CO2 concentrations to keep on increasing at the same rate as before. (This is because changes in concentrations are proportional to emissions.) The problem is that if you want to stabilize CO2 concentrations, you have to make drastic cuts in CO2 emissions. This is no easy feat. Yet, unless it is done, we are liable to reach very high levels of CO2 concentrations.

Global warming skeptics would dispute or minimize the link between CO2 concentrations and temperature increases. Here is yet another uncertainty — are they or the mainstream climate scientists more right than wrong? But can we afford the luxury of assuming that a small minority of climate skeptics are more correct than the vast majority of mainstream climate scientists? What is the probability of that?

Admittedly, almost all of the relevant probabilities in this kind of rough analysis are uncomfortably indeterminate. But that is the nature of the beast here and shouldn’t be an excuse for inaction. The bottom line is that if we continue on a business-as-usual trajectory, then there is some non-trivial probability of a catastrophic climate outcome materializing at some future time. Prudence would seem to dictate taking action to cut back greenhouse gas emissions significantly. If we don’t start buying into this insurance policy soon, the human race could end up being very sorry should a future climate catastrophe rear its ugly head.


Martin L. Weitzman is Professor of Economics at Harvard University. Previously he was on the faculties of MIT and Yale. He has been elected as a fellow of the Econometric Society and the American Academy of Arts and Sciences.

This entry is cross-posted on the Rundown — NewsHour’s blog of news and insight.