Remarks given at "Developing Clean, Innovative Commercial Energy: Will Proposed Federal Subsidies Hurt or Help?" at the Capitol Hill Club.

Introduction

I would like to thank the George C. Marshall Institute and the Nonproliferation Policy Education Center for co-sponsoring this event and for their invitation to speak this morning. Jeff Kuetter of Marshall has long been interested in this subject and it is always a pleasure to work with him. I look forward to our discussion this morning and to the project's future work. 

The Importance of the Problem

It might be useful to begin our discussion this morning with a quick look at the larger context of climate policy. This year marks the 20th anniversary of the first meeting of the IPCC. The timing for an assessment seems appropriate to ask how things are different from 1988. The answer is revealing. Last year, CO2 emissions were over a third higher than they had been in 1988, and the rise in atmospheric concentrations has accelerated through the last several decades. Even in the EU, the self-appointed, self-proclaimed greenhouse conscience of the world, emissions continue to grow. After twenty years of laborious efforts, the conclusion seems inescapable. Greenhouse gas control proposals have failed.

Without new technology, it is hard to see how future results will be any better. With current technologies, the costs of deep emission cuts are very high relative to the expected benefits. Given these high costs, only countries with politically strong domestic Green movements have committed to making them, and, even in these countries, the commitment is half-hearted. The resulting level of global effort is insufficient to move the needle on total emissions. Even were the United States to adopt controls, emission growth in China and India will swamp the impact at the global level, thus continuing the record of futility.

Technology is the only real hope for eventually breaking this pattern. And it offers some prospects for long-run success. Peter Blair of the National Academy of Engineering has pointed out that targeted past investments in science may mean that many discoveries of useful knowledge are already, "in the pipeline." Jae Edmonds has shown that speculative, but plausible, progress on some key technologies could reduce the costs of stabilizing greenhouse gas concentrations by, literally, trillions of dollars.

These prospects pose what I believe is the central policy question: How can we enhance our chances of realizing the needed technological advances? The answer, I fear, suggests that we should temper our long-run hopes with a rather skeptical view of how long this process may take. This skepticism leads to a suggestion that we consider a different kind of technology, one that might offer a safe bridge between our current situation and the more distant future.

Scientific/Technological Barriers to Success

Some of the reasons for skepticism about the near and mid term prospects relate to the nature of the technological challenges that we face. At least four factors of this type are present:

First, solutions will require new scientific knowledge, not just new gadgets. The widely cited Hoffert et al 2002 Science article, observed that existing technologies and the expected extensions of them were wholly inadequate to the task of stabilizing greenhouse gas concentrations. The article also argued that nothing less than multiple large breakthroughs in basic science could create the revolutionary new technologies that were needed. However, ex ante, the outcome of R&D is notoriously uncertain. Will the progress envisioned by Blair and Edmonds materialize? If it does, when? There is far more doubt than would be the case were we considering the simple extension of existing technologies.

Second, a long lag often occurs between the discovery of new scientific knowledge and its first use in new processes or products. Another lag is common before the latter succeed in an engineering and economic sense. And the perfected innovation may take a long while to diffuse through the economy. Economist Nathan Rosenberg has explained very clearly why the process is so time consuming, but the upshot is that the full economic payoff of discoveries in basic science is often realized only after several decades. 

Third, in the case of climate, the lags are likely to be especially long because the innovations must diffuse across most of the globe. Innovations made in America or Japan, may not fit market and institutional conditions in China and India until they have been adapted to local conditions. Those conditions may differ widely from those prevailing where the invention originated. In climate technology, therefore, we might expect the diffusion process to be unusually long. An approach like carbon capture and storage, the use of which depends completely on government policy, may have an especially hard time in countries like China and India, where governments are most unlikely to foster it.

Fourth, at this point, we do not know what technologies are likely to meet the need. It may be space-based solar power. It may be nuclear with fuel recycling. It may be microbes that produce fuel. Or, to cite Jae Edmonds again, it may well be something of which we cannot conceive until a future breakthrough in basic science opens our eyes to its possibility. One implication is for the idea that the problem here is quite different from that involved in the Apollo or Manhattan Projects. There, the scientists had a relatively clear concept of what they were looking for. Here, our vision of the goal is much cloudier.

The Institutional Roots of the Problem

Technology, however, may raise less important roadblocks than institutions do. Two market failures are central to our problem. First, without a shadow price on greenhouse gas emissions, there is no demand for emission-reducing technologies. Obviously, in that case, no one has much incentive to invent new ways of cutting emission. 

A second marked failure arises because for-profit firms often believe that they cannot capture enough of the spillover benefits produced by their inventions to repay the costs of the R&D. As a result, a large gap develops between the level of private R&D investments and the level that would be optimal for society as a whole. Patents, tax credits, and subsidies are designed to remedy the resulting R&D shortfall, but apparently they are only partly successful. The gap between actual R&D investment and the optimal level appears to be large. In the U.S., for example, R&D investment is, according to some estimates, only about a quarter of the optimal level.

A key point is that the two market failures are independent. Solving either of them leaves the other untouched. Thus, putting a positive price on greenhouse gas emissions would eliminate the first market failure but not the second. Solving the second problem, the systematic under-investment in R&D, is in many ways a harder problem.

What to Do

What, then, is to be done; especially, what is to be done about the second market failure? Part of the answer has traditionally been that government must provide financial support for research in basic science and perhaps in early development. Economists have typically recommended several interlocking prescriptions for how to manage these expenditures. Four of the most important are:

• Divide the labor between the public and private sectors in an efficient manner. Public sector funding should concentrate on basic research, and the private sector should be allowed to handle development and commercialization.
• Government climate-related R&D should fund a diverse portfolio of ideas that, in combination, promise results large enough to offer a solution. We do not know whether a given approach will pan out, so we must diversify.
• Accept failures, but stop when failure becomes evident. A R&D effort that does not have failures is likely to be too timid to produce good results, but the practice of allowing obviously failed projects to stumble along simply wastes resources.
• Increase the supply of R&D inputs and modulate the resource demands of the climate R&D effort to limit transfers of scarce R&D resources out of other high priority fields. R&D resources are specialized, and it takes time to increase their supply. We do not want to curtail the risks of climate change by draining R&D resources from urgent national security or other top priority fields.

The problem with this list and others like it is that they are easier to make than to implement. To change the way that government funds R&D requires the approval of those legislators and executive branch officials who control the current system. These interests have presumably arranged things in ways that advance their political interests. They enjoy a large measure of veto power over proposed changes. By inference, for proposed "reforms' to win approval, they must be at least as favorable to the interests of the institutional gate keepers as is the status quo.    

There are good reasons for thinking that it is difficult for reforms to meet this test. Cohen and Noll in their aptly named book The Technology Pork Barrel, explain that members of Congress, while they often support R&D funding, have weak political incentives for ensuring that the R&D efforts are structured cost effectively.

For example, members of Congress have political incentives to fight to preserve and even expand unsuccessful or dubious R&D projects that employ their constituents. Then too, reelection cycles are too short for R&D success or failure to much affect the political prospects of members who have promoted them. Moreover, the both the benefits and the costs of R&D are often too diffuse to influence constituent perceptions of congressional performance. These incentives suggest that, to appeal to Congress, reforms must offer a much stronger inducement than the promise that social returns on federal R&D investments would improve in coming decades.

I am not suggesting that we abandon all efforts to reform government R&D. It seems clear, though, that attempts to make dramatic changes in the conduct of government R&D will encounter stiff resistance. Without such changes, the performance of the various federal R&D programs is likely to remain very mixed. Under these circumstances, the future cost-effectiveness of energy research is likely to remain fairly constant. In that case, the needed technologies may well appear, but perhaps not that soon.

What to Do While We Wait for Long-Term Solutions

The question for climate policy then becomes what to do in the interim. Abrupt harmful climate change could, conceivably, appear before the desired technologies arrive. A family of technologies referred to as "geoengineering" may offer the bridge for safely crossing the coming decades.

The idea behind geoengineering is simple. When sunlight strikes the Earth's surface, greenhouse gases in the atmosphere trap some of the heat that is generated. More greenhouse gas in the atmosphere causes the planet to warm, but a slight decrease in the amount of sunlight reaching the Earth's surface could, in principle, offset that warming. Scientists estimate that deflecting into space only 2 percent of the total sunlight that strikes the Earth would be enough to cancel out the warming effect of doubling the pre-industrial levels of greenhouse gases. 

Scattering this amount of sunlight may be comparatively easy. Past volcanic eruptions have shown that injecting relatively small volumes of matter into the upper atmosphere scatters enough sunlight back into space to cause discernable cooling. The 1991 eruption of Mt. Pinatubo reduced global mean temperature by about .5 degree Celsius.

Some scientists propose, therefore, to use modern technology to create a carefully engineered analogue to this effect. Several kinds of materials appear to be possible sunlight scatterers, and several means of putting them in the upper atmosphere look practical. To produce cooling, geoengineering does not appear to need the kind of radical scientific advances required to achieve drastic emissions cuts, although progress in climate science could diminish the risk of unexpected side effects.

Proposals to study geoengineering are gaining adherents among climate experts. In early 2007, the report of a workshop jointly sponsored by NASA and Carnegie Institution noted that such distinguished scientists as Ralph Cicerone, Paul Crutzen, and Tom Wigley have suggested further study of the concept. The report went on to observe, "Promi¬nent economists such as William Nordhaus and Thomas Schelling have long argued that the con¬cept warranted further exploration as well." Since this workshop, at least three other conferences have focused on the concept, and attention in the popular media has begun to build.
 
The speed with which geoengineering can take effect could be crucial to its effectiveness as a tool for halting abrupt climate change. With the Pinatubo eruption, the temperature reductions were apparent in just a few months and persisted for about three years. By analogy, once geoengineering was developed and the capacity to use it put in place, it could begin cooling the Earth within months. Geoengineering can also be stopped quickly because within a few years at most, all of the particles will have rained out of the atmosphere.

As to cost, although the analysis remains very preliminary, the costs of geoengineering appear to be low especially compared to the large expense of deep emission cuts. A National Academy of Science Panel, for example, concluded that geoengineering options might be implemented at surprisingly low costs. 

In sum, the current state of climate policy calls for new thinking. Emission controls are not going to have much impact until a new generation of vastly cheaper technologies becomes available and matures. That is likely to take many decades at best. In the interim, the promising, although untried, state of geoengineering strongly suggests that the federal government should do the R&D needed to explore the concept. To be sure, virtually all experts agree that geoengineering is a stopgap. In the long-run emissions must come down. But the prospects for reducing the expected net costs from global warming hinges on having the time to develop efficient technologies, and geoengineering may be able to buy that time. Indeed, it may be able to buy it rather inexpensively.

Lee Lane is a resident fellow at AEI and codirector of AEI's Project on Geoengineering.