“A wait-and-see strategy is probably optimal because immediate cutbacks in fossil fuel make economic sense only in worst-case climate scenarios.”
The recently published Superfreakonomics contains a chapter devoted to geo-engineering. Geo-engineering refers to large-scale efforts to deliberately alter the earth’s climate, usually to reverse or offset the effects of greenhouse-gas emissions. Authors Steven Levitt and Stephen Dubner came under intense criticism for touting the virtues of a geo-engineering solution to the problem of climate change while ignoring its pitfalls.1 What the critics didn’t consider were the benefits of procrastination. The option of geo-engineering makes it much safer to continue using fossil fuels and thereby pass on extra trillions of dollars of wealth to the next generation at possibly little or even no cost.

A Few Geo-engineering Proposals

Here are some of the popular ideas:

Sulfur Dioxide Hose to the Sky. In Superfreakonomics, the authors focus on a proposal that relies on the experience of Mount Pinatubo. After its 1991 eruption, the volcano spewed forth large quantities of ash that had a measurable cooling effect on the globe. Thus, the scientists whom Levitt and Dubner interviewed are developing a scheme to suspend one or more hoses by helium balloons in order to pump sulfur dioxide into the stratosphere. This increased concentration of SO2 would reflect some of the incoming sunlight back into space, offsetting the “greenhouse effect” caused by increasing carbon dioxide (CO2) concentrations in the atmosphere. Thus, although emissions from industrial activities would cause the atmosphere to trap more heat over time, the sulfur dioxide concentration could be controlled to reflect more and more of the incoming sunlight. Levitt and Dubner were thrilled by this proposal because its total cost to completely offset human-induced global warming would be a mere $250 million—less than what Al Gore’s foundation is spending just to “raise awareness” about the dangers of climate change and about one one-thousandth of one percent of a reputable estimated cost of drastically curtailing emissions.2

Fertilizing the Oceans With Iron. Providing iron to a deficient area of the upper ocean can generate a “phytoplankton bloom.” Like land-based plants, phytoplankton rely on photosynthesis and, in the process, absorb carbon dioxide from the atmosphere. If small-scale experiments extrapolate into larger efforts, widespread iron fertilization could lead to a significant “sequestration” of carbon out of the atmosphere and, ultimately, to the bottom of the ocean.

Carbon-Eating Trees. Physicist Freeman Dyson, an amateur global-warming enthusiast, offered his own vision of geo-engineering through genetic engineering:

[E]very carbon dioxide molecule in the atmosphere is incorporated in a plant within a time of the order of twelve years. Therefore, if we can control what the plants do with the carbon, the fate of the carbon in the atmosphere is in our hands…. I consider it likely that we shall have “genetically engineered carbon-eating trees” within twenty years, and almost certainly within fifty years.

Carbon-eating trees could convert most of the carbon that they absorb from the atmosphere into some chemically stable form and bury it underground. Or they could convert the carbon into liquid fuels and other useful chemicals. Biotechnology is enormously powerful, capable of burying or transforming any molecule of carbon dioxide that comes into its grasp. . . . If one quarter of the world’s forests were replanted with carbon-eating varieties of the same species, the forests would be preserved as ecological resources and as habitats for wildlife, and the carbon dioxide in the atmosphere would be reduced by half in about fifty years.3


Critics have raised several objections to geo-engineering proposals. For solutions that merely limit incoming solar radiation, the critics point out that rising temperatures are not the only problem with growing atmospheric concentrations of carbon dioxide. In particular, ocean acidification would still occur, even if, say, the garden hose to the sky could manage to stabilize global temperatures. NASA climate modeler Gavin Schmidt recently explained why he does not view geo-engineering as a solution:

[The actual climate system] has an ozone layer, winds that depend on temperature gradients that cause European winters to warm after volcanic eruptions, rainfall that depends on the solar heating at the surface of the ocean and decreases dramatically after eruptions, clouds that depend on the presence of condensation nuclei, plants that have specific preferences for direct or diffuse light, and marine life that relies on the fact that the ocean doesn’t dissolve calcium carbonate near the surface.

The point is that a planet with increased CO2 and ever-increasing levels of sulphates in the stratosphere is not going to be the same as one without either…. The issues I listed above are the ‘known unknowns’—things we know that we don’t know…. These are issues that have been raised in existing (very preliminary) simulations. There would almost certainly be ‘unknown unknowns’—things we don’t yet know that we don’t know.4

More generally, many critics underscore the “mad-scientist” flavor of the geo-engineering proposals and worry that unanticipated consequences would force our children to abandon the techniques. If the atmosphere were to become saturated with decades of further “business as usual” emissions, they argue, the crash course to de-carbonize the economy would be much more painful than if we had done the responsible thing today.

Finally, critics claim that geo-engineering solutions would cause international tensions. For example, Russians might balk if the United States controlled mirrors in space that could regulate how much light entire continents received, while the leaders of India would be upset if a specific geo-engineering proposal interfered with the annual monsoon.

The Benefits of Procrastination

Many critics of geo-engineering overlook an important fact: there is a gain from procrastination. In some of their expositions, they argue, implicitly and sometimes explicitly, that because humans will eventually have to reduce greenhouse-gas emissions anyway, we might as well do the adult thing and start the painful adjustment today.5 But this ignores the principle that a “quick fix” can allow the deferment of solving a particular problem, lowering the total cost of the long-run solution.

Although procrastination is often a sign of immaturity, in the context of climate change it may not be. In the typical debate over geo-engineering, proponents argue that it is “the” solution to global warming, while the critics worry about all the things that could go wrong. Yet this “geo-engineering: yes or no?” debate overlooks the important possibility that the most economically efficient outcome involves the postponement of carbon-abatement strategies, along with the simultaneous research and development of varied geo-engineering techniques to be deployed if they should become necessary. Back-of-the-envelope calculations suggest that this strategy could leave our descendants many trillions of dollars richer than the alternative of implementing immediate and large cuts in emissions.

A wait-and-see strategy is probably optimal because immediate cutbacks in fossil fuel make economic sense only in worst-case climate scenarios. If we rely just on the “point estimates” of the benefits and costs of aggressive legislation, then the literature suggests it is much cheaper if governments do nothing rather than impose steep emission cuts and other regulations. Richard Tol recently published a survey article on comprehensive estimates of global welfare effects from unrestricted climate change, where the effects included monetary estimates of damages to non-market areas such as human health.6

Of the thirteen studies Tol surveyed, the best-guess estimate of global GDP effects ranged from a loss of 4.8 percent to a gain of 2.5 percent. Most of these impacts were calibrated for temperature increases of 2.5 to 3.0 degrees Celsius, which are not expected to occur until the second half of the 21st century. (Currently the globe is about 0.7 degrees Celsius warmer than the preindustrial benchmark.) Of the estimates in the eleven studies published since the year 1995, the worst case is a global GDP loss of 1.9 percent.

In contrast to these surprisingly modest estimates of the dangers of unrestricted climate change, the economic costs of zealous government interventions could be far worse than the disease. For example, the Congressional Budget Office surveyed a range of studies and concluded that the cap-and-trade emissions targets in the Waxman-Markey climate bill would reduce U.S. GDP by 1.1 percent to 3.4 percent by 2050.7 Thus, the midpoint of this range, 2.3 percent, is higher than the worst estimate of unrestricted climate change (in any surveyed study published within the last fourteen years).8 In other words, the costs of Waxman-Markey exceed even the most optimistic estimates of benefits. Moreover, the damage to the economy occurs decades earlier than the full benefits of avoided climate change and the Waxman-Markey plan, even if adopted by all major governments, would not eliminate all climate damages.

Thus if we were to rely on best-guess point estimates, it would be much cheaper to allow unrestricted carbon emissions and simply adapt to the climate change, rather than unleashing governments to embark on what would surely not be a textbook implementation of a worldwide “optimal carbon tax.” Unfortunately, this strategy by itself would be very risky. Concerned climate scientists point out that it is possible, however unlikely, that the earth’s sensitivity to emissions is much higher than the point estimates, and that today’s forecasts of tolerable climate damages will turn out to be catastrophically optimistic. This is why advocates of aggressive government intervention reject the standard cost-benefit approach to carbon policies, and instead favor the “precautionary principle.”9

But it is precisely this uncertainty over the threat of climate change that yields the great benefit of geo-engineering. So long as humans have the ability to stave off catastrophic warming for years and possibly decades, geo-engineering places far more options at our disposal. Rather than take very costly actions now, it is better to proceed with “business as usual”—while the climate scientists continue to refine their models and the engineers continue to explore new techniques of carbon-free energy production—knowing that we have an extra decade or two of “breathing room”10 because of various geo-engineering techniques.

The following example is for illustrative purposes only, but, to make the numbers plausible, it is loosely based on a leading model.11 Suppose that without geo-engineering, humanity can (a) pay $30 trillion (in present discounted value) to keep the atmosphere at a “safe” level of greenhouse-gas concentrations and thereby avoid the risk of significant climate damages down the road; or (b) allow unrestricted emissions, thereby facing a (100-x) percent probability of $20 trillion in future climate damages, and an x percent probability of a catastrophic $100 trillion in future damages when the climate passes a “tipping point.”

The proponents of option (b) would point out that in all likelihood—especially the smaller x turns out to be—humanity would be wasting $10 trillion by curbing emissions, because we would be forfeiting $30 trillion in potential output in order to avoid what would probably be only $20 trillion in climate damages. On the other hand, the proponents of option (a) would invoke the analogy of fire insurance and claim that a normal level of risk aversion—especially the larger x turns out to be—makes that “actuarially excess cost” of $10 trillion well worth the peace of mind it brings.

In this setting, we can see the advantages of a third option (c): Allow unrestricted emissions and, if it turns out that humanity has entered one of the disaster scenarios, implement a geo-engineering proposal in conjunction with draconian emission cuts, which will cost (from our vantage point now, in evaluating the scenario) $15 trillion in forfeited economic output, as well as $35 trillion in collateral damage from the geo-engineering techniques during the period when the atmosphere’s greenhouse-gas concentrations returned to safe levels.

With these stipulated numbers, option (c) (involving the possibility of geo-engineering) renders the option (a) of slamming the brakes immediately on carbon emissions much less tenable. With no possibility of geo-engineering, option (a) was plausible, despite its higher expected cost, simply because option (b) of “doing nothing” carried the threat of a catastrophic $100 trillion disaster. Yet the possibility of geo-engineering gives humanity a much safer gamble, where the worst outcome is a much more tolerable $50 trillion loss ($15 trillion plus $35 trillion.) We can continue with unrestricted carbon emissions, hoping—and expecting—that the climate’s sensitivity to greenhouse gases lies within the moderate range. But in the unlikely event that our children do realize they are on a runway to disaster, geo-engineering allows them to limit the damage. They can cap the rise in global temperatures through the various techniques discussed above, while engaging in a crash course to reduce atmospheric CO2 concentrations. Even if the fears of Gavin Schmidt and others pan out, and the geo-engineering techniques are at best a temporary fix that masks the symptoms, it is still very significant that we have such time-buying strategies at our disposal. Because costs deferred are often costs lessened, temporary fixes are often valuable. Imagine the huge economic waste involved if homeowners never used “temporary” fixes on appliances or their vehicles, and bought a new washing machine or transmission every time they encountered a minor problem.

The critics of geo-engineering ignore this crucial benefit of additional time. In general, it is cheaper to achieve a goal when there are fewer constraints. It is a matter of economics, not engineering or physics, to confidently state: “Insisting on carbon-free energy production by 2080, rather than by 2050, would raise cumulative GDP during the 21st century.”

When it comes to atmospheric concentrations of greenhouse gases, matters are not so simple as a basic statement of economic logic. Economics tells us: “It would be cheaper to stabilize the atmosphere at 420 ppm of carbon dioxide-equivalent gases from 2080 forward than it would be to stabilize it from 2050 forward.”12 However, this statement is self-evidently true only if we assume that people know from the outset the constraint on atmospheric concentrations. The critics of geo-engineering worry that industries will proceed with “business as usual” and then realize—to their horror—that they need to stabilize at 420 ppm, when the atmosphere is already far above that threshold. At that point, the actual geo-engineering technologies (and their respective costs) would come into play, to determine whether the total compliance costs from the procrastination strategy would be higher or lower.

For more on global warming, see The Economics of Climate Change, by Robert P. Murphy on Econlib, and Global Warming: A Balance Sheet, by Thomas Gale Moore in the Concise Encyclopedia of Economics.

Notwithstanding this important qualification, I still maintain that the benefits of delay are well worth it in present circumstances. Even just ten additional years of data collection could allow for much greater confidence in estimates of the climate’s sensitivity to greenhouse-gas emissions.13 If Freeman Dyson’s fanciful ideas turn out to be right, then avoiding the costly (and, in retrospect, unnecessary) impairment of economic growth could leave our descendants tens of trillions of dollars wealthier.14 And if Dyson turns out to be wrong, other options would still give the next generation valuable time to recover from the misplaced optimism.


See for example physicist (and Clinton Department of Energy official) Joe Romm’s scathing critique at his popular blog Climate Progress, as well as the blog critiques of economists Paul Krugman and Brad DeLong. See also SuperFreakonomics: Global Cooling, Patriotic Prostitutes, and Why Suicide Bombers Should Buy Life Insurance, by Steven Levitt and Stephen Dubner.

See Table 5-1 (page 82) of William Nordhaus, A Question of Balance: Weighing the Options on Global Warming Policies, proofs available in .pdf. The estimated “abatement cost” of limiting atmospheric concentrations to 1.5 times their preindustrial benchmark is more than $25 trillion in present-discounted value.

Freeman Dyson, “The Question of Global Warming,” The New York Review of Books, Vol. 55, No. 10, June 12, 2008.

Gavin Schmidt, “Why Levitt and Dubner like geo-engineering and why they are wrong,” RealClimate.org, October 18, 2009.

See for example my critique of Brad DeLong’s criticisms of Levitt’s views on the timing of the decarbonization of the energy sector.

Richard Tol, “The Economic Effects of Climate Change,”Journal of Economic Perspectives, Vol. 23, No. 2, Spring 2009, pp. 29-51.

See Table 1 (page 13) of the CBO report, “The Economic Effects of Legislation to Reduce Greenhouse-Gas Emissions,” September 2009, available in .pdf.

In the text we are comparing apples to oranges, because the Tol survey looked at effects on global GDP from climate change, whereas the CBO survey looked at effects on United States GDP from de-carbonization. However the comparison is still meaningful, because unilateral American emission curbs would do little to arrest climate change. Only if the whole world adopted strict emission targets (such as the ones in Waxman-Markey) would the global damages from climate change be mitigated, but in that case the whole world’s economy would forfeit output because of the new constraints.

Economist Martin Weitzman has been one of the leaders in this line of literature. See Robert Murphy, “Martin Weitzman’s Dismal Theorem: Do ‘Fat Tails’ Destroy Cost-Benefit Analysis?” MasterResource.org, February 1, 2009.

This is the apt phrase of physicist Nathan Myhrvold, one of the primary sources Levitt and Dubner interviewed for their chapter on geo-engineering.

See Table 5-1 (page 82) of William Nordhaus, A Question of Balance: Weighing the Options on Global Warming Policies, proofs available in .pdf.

In the text I have used 420 ppm as the “safe” concentration of carbon in the atmosphere only because that is the lowest target policy modeled by Nordhaus. In reality, many concerned scientists think a much stricter goal is necessary to reduce the likelihood of catastrophic climate impacts. For example some suggest limiting concentrations to 350 ppm (“Campaign Against Emissions Picks Number,”, New York Times, October 24, 2009) (we are already at 387 ppm), while others focus on limiting total global warming to 2 degrees Celsius (“Hit the brakes hard,” RealClimate.org, April 29, 2009) warmer than the preindustrial average (we are already at 0.7 or 0.8 degrees warmer).

Climate scientist Chip Knappenberger summarizes in“What Does the Last Decade Tell Us about Global Warming? (Hint: the ‘skeptics’ have the momentum)” the problem that the lack of warming in recent years poses for some of the high-end estimates of the globe’s sensitivity to emissions. Possibly five and certainly ten more years of data could do wonders in allowing scientists to tease out the “anthropogenic signal” in the temperature data from the natural variability.

In Nordhaus’s Table 5-1 (see note 11 above), the “Low-cost backstop” option—which assumes a “miracle cure” comes along and allows humanity to sharply curtail emissions with little impact on economic output—leaves the world about $32 trillion richer than the option of limiting atmospheric concentrations to 1.5 times their preindustrial levels, i.e. limiting them to 420 ppm.


*Robert P. Murphy is a Senior Fellow in Business and Economic Studies at Pacific Research Institute, and an economist with the Institute for Energy Research where he specializes in climate change economics. He is the author of The Politically Incorrect Guide to Capitalism (Regnery, 2007).

For more articles by Robert P. Murphy, see the Archive.