From Energy Wish Lists to Technological Realities
Federal policymakers have long agreed on energy technology goals; now they must come together behind the policies that can succeed.
The aspiration for new technology has been at the heart of every energy policy developed since the first oil embargo in 1973. President Bush’s 2006 State of the Union address continued the quest for new technology by proposing the Advanced Energy Initiative, which once again calls for “greater use of technologies that reduce oil use” and “generating more electricity from clean coal, advanced nuclear power, and renewable resources.” However, history shows that the hard problem for energy policy is not how to craft another technological wish list but how to turn technological aspirations into reality. And the government has been slow to use its own experience to learn how to solve this problem.
The public policy goal for energy technology can be expressed simply: to induce technological innovations in the private sector that serve national energy policy. Stated thus, this goal embodies three fundamental principles about how the innovation process works that are grounded in long experience with federal energy R&D. Because these principles differ in some important respects from conventional wisdom, understanding them is the place to begin a discussion of how to achieve the goal.
First, innovation in energy technology happens almost entirely in the private sector. The process of bringing a new product to market involves the most intimate of relationships between buyer and seller. Both are entering uncharted waters, and the balancing of risks among the parties is often a delicate compromise. The Department of Energy (DOE), or any other government bureaucracy for that matter, is too a clumsy partner to enter into this relationship in a meaningful way.
A 2001 report by the National Research Council (NRC) underscores this principle. The report examined the track record of DOE research in the areas of energy efficiency and fossil energy between 1978 and 2000. As part of its work, it asked several experts to name the most important technological innovations that actually entered the energy system during this period, without regard to their source. Then they were asked how important government-funded research was to these technologies. Of the 23 technologies that were listed, in only 3 cases did the government program make a major contribution, and in 7 DOE’s research was moderately helpful. The government’s role was minimal in the other 13.
Even where DOE played a major role, the private sector carried much of the load. In one case, DOE developed diamond bits for drilling into hot dry rock as part of its geothermal energy program. That program did not flourish, but the drill bit technology found a home in the oil and gas industry. In another, DOE invested a modest $3.2 million in the development of electronic ballasts for fluorescent lights, which enabled a small firm to introduce the product in the early 1980s. The entry of this new competitor motivated the two dominant lighting companies to adopt the technology. In the case of efficient refrigerators, DOE leveraged a $1.6 million research effort with a program to impose stricter efficiency standards on refrigerator performance, thus encouraging industry to adopt the technology developed by DOE.
Although it would be a mistake to read too much into these anecdotes, the NRC report nevertheless underscores the importance of the private sector in energy technology innovation and the varied and sometimes unintended paths by which government-sponsored research influences the private sector’s actions.
The second principle is that technological innovation is more than R&D. DOE and many other government agencies typically describe the process of developing new technology as Research, Development, Demonstration, and Deployment (RDD&D). This linear model may reasonably characterize the innovation process for technologies, such as weapons systems or space probes, for which the government is a customer. Innovation in energy technology, however, deals with the more complex problem of getting new products into the hands of private-sector buyers.
Rather than being the linear process characterized as RDD&D, the private sector innovation process is incremental, cumulative, and assimilative—in a word, messy. It typically proceeds in small steps because an incremental approach helps minimize risk to buyer and seller. However, the accumulation of such increments can ultimately add up to breakthrough technologies. Finally, the innovator often reaches out to diverse sources of knowledge and technology, assimilating them in novel ways for new markets. A Resources for the Future study of technology innovation in natural resource industries summarized the process this way: ”Even technologies subsequently recognized as revolutionary went through extended periods of adaptation and adoption. In many cases, additional technological developments were required to enhance applicability of an initial innovation. It has also been the case that one innovation does not achieve its full effectiveness until complementary albeit ostensibly unrelated technologies are developed.”
Finally, the reason for government to intervene in private-sector innovation is to remove obstacles to meeting national energy policy goals. The private sector can serve energy policy without help from government, as shown by the NRC report discussed earlier. But there are cases, often important ones, when national policy requires inducing the private sector to innovate in areas that would otherwise lie fallow. Knowing when and how to intervene is thus a crucial policy judgment.
To drive home this point, consider another conclusion of the NRC report. It calculated the economic, environmental, and security benefits produced by 39 applied research projects in DOE’s fossil energy and energy efficiency programs. Overall, the report estimated that DOE generated some $40 billion in economic benefits for the roughly $13 billion it spent on these programs between 1978 and 2000. (The report also identified environmental and security benefits that are harder to quantify.) But what is most interesting for policy is the highly skewed way in which this generally positive result was achieved. A handful of programs produced most of the benefit, whereas most of the investment resulted in very little:
- A mere 0.1% of the expenditure accounted for three-quarters of the benefit. Three programs on refrigerator efficiency, electronic ballasts for fluorescent lighting, and low-emissivity windows created $30 billion in economic benefit for a total expenditure of $13 million.
- Three-quarters of the expenditure—a little over $9 billion—produced no quantifiable economic benefit. Half of this money was applied to synthetic fuel projects that turned out to be at least a couple of decades premature. Developing synfuels technology may have been a reasonable goal at the time, but as will be discussed later, it could have been approached more modestly.
No one who has run an applied research program will be surprised by a few unexpected home runs or inevitable failures. But the DOE experience does suggest that there are lessons to be learned about how the government spends taxpayer money to influence technology innovation.
So how should government go about inducing technology innovations in the private sector that serve national energy policy? Four strategies seem especially important.
Provide private-sector incentives to pursue innovations that advance energy policy goals. Because technological innovation is primarily a private-sector activity, inducing the private sector to do most of the work of developing new energy technology is an almost self-evident strategy. The challenge for government policy is to find incentives that most effectively harness the innovative drive of the private sector.
The most effective incentive is to attach an economic value to the policy goal itself. For example, a carbon tax, or a cap on carbon emissions coupled with an allowance trading system, sets a price on carbon emissions. Similarly, requiring a floor price for oil that reflects its security and environmental risks would impart a value to reduced oil dependence. Responding to these market incentives, private-sector innovators will seek out the least-cost methods for achieving the policy goal. What makes this approach so effective is that it does not limit the innovative imagination. Thus, when a cap-and-trade system was established in the early 1990s for sulfur oxide emissions, the private sector responded by increasing the use of low-sulfur Western coal. This surprised many analysts who assumed that expensive technology to scrub sulfur oxides out of power plant exhaust gases would be the innovation of choice.
A second-best incentive is regulation designed to encourage either the introduction of new technologies or the improvement of existing ones. For example, many states now set renewable portfolio standards, which require electric utilities to generate a minimum amount of power from renewable energy sources. Another approach is to impose technology standards that require more efficient versions of familiar household appliances.
Technology standards have demonstrated their effectiveness in the case of the refrigerator standards described above and at least in the early years of the Corporate Average Fuel Economy (CAFE) standards. Nevertheless, regulation suffers from two disadvantages as compared to more broadly based market incentives. First, it tends to limit the scope of innovative activity because of the focus on specific technologies or applications. Second, regulation rather than energy policy becomes the driver of innovation, and they are not the same thing. The implementation of CAFE standards, which improved the fuel efficiency of passenger cars but promoted the market for light trucks, shows how unintended consequences can result from well-intentioned regulation.
Outright subsidies for the adoption of new technology, such as production tax credits for the adoption of solar and wind power, are the least effective form of incentive, because they address only one aspect of the problem of creating value in the marketplace. To succeed, a new technology must overcome the inertia of the market, which favors existing products, and must then rapidly become cost-competitive as production increases. No matter how good the new technology, the first few units will inevitably be more expensive than existing technology.
The value of a subsidy is that it enables the producer to sell early units at a competitive price. However, after that introductory period, the product must be able to compete on its own. To do so without the market or regulatory incentives discussed above, the cost of the new technology must drop very quickly as production increases, and this is a hard condition to meet. Arguably, the reason production why tax credits for wind power seem to have worked well is because the cost of wind technology has dropped rapidly thanks to relatively straightforward engineering and production improvements. The same cannot yet be said for solar photovoltaic cells, which require more research to become cost-competitive.
Conduct basic research to produce knowledge likely to be assimilated into the innovation process. This policy follows directly from the assimilative nature of innovation. The challenge is to design a basic research program that actually sets the table for innovators in a specific field of technology. Purely curiosity-driven research certainly produces useful knowledge, but it is not optimized for solving energy problems. On the other hand, research in support of applied technology, even if it is fundamental research, lacks the breadth to produce breakthrough ideas from unsuspected sources.
Two guidelines can help strike the right balance. One is to support ideas, even very exotic ones, that would, if successful, overcome fundamental weaknesses of known technology. For example, Marilyn Brown of Oak Ridge National Laboratory has examined how novel technologies might accelerate advances in energy efficiency. She suggests that technologies that manipulate materials at the nanoscale, apply molecular biology to energy problems, and draw on advanced computing capabilities could overcome the thermodynamic performance limits of existing energy systems.
The other guideline is to support principal investigators who are driven to apply their disciplinary knowledge to energy problems. If the innovation process is to benefit from assimilation, energy technology needs to be connected to diverse disciplinary sources of knowledge. This connection has to be a two-way street. Of course, energy technologists need to be looking for new ideas. But equally important is that scientists with new ideas need to be looking for energy applications.
The careers of Richard Smalley and Craig Venter exemplify the kind of bridges that need to be built from fundamental research in a variety of fields to long-term innovation in energy systems. Smalley, a Nobel Prize winner in chemistry, came to understand that his field of nanotechnology could greatly improve the efficiency of catalysts, solar cells, and other devices important for energy production and use. Venter, who led the private-sector group that mapped the human genome, is working to find and modify microbes that could be efficient biological sources of energy.
Target applied research toward removing specific obstacles to private-sector innovation. The fact that innovation is almost entirely a private-sector activity does not mean that it is always successful there. Economists who study innovation have identified several kinds of obstacles. One is the possibility that the innovator cannot capture enough of the benefit of innovation to justify the cost and risk of bringing a new product to market. For example, other firms may be able to copy the innovation so quickly that, despite the presumed protection of patents, the advantage of being first to market is greatly diluted. Or a firm may come up with a new product idea that is so far outside its regular business that it is unable or unwilling to bring it to market. Finally, the risk of innovation may be so great that financial markets are unwilling to risk the capital needed to proceed.
When these obstacles arise, government can often step in to put the private-sector innovation back on track. Its reason for doing so is that the innovation would produce a public good such as reduced carbon emissions that justifies intervention. Two of the home-run projects identified in the NRC study of DOE research were of this type. The technology for low-emissivity windows and electronic ballasts was important, but equally important was DOE’s ability to help have a product introduced, which in turn motivated the manufacturers of existing technology to adopt the innovations and thus advance energy policy goals.
Removing such obstacles should be a main goal of DOE’s applied research programs in nuclear, fossil, and renewable energy as well as in energy efficiency. Furthermore, it appears that this applied research is most cost-effective when it is targeted with some precision. The more precisely the private-sector obstacle is defined, the more surgically it can be removed. Doing so depends on developing a close partnership between government and the private sector.
The NRC study identified several examples of close cooperation that resulted in government research projects with solid benefit/cost ratios, even if they fell short of home-run success. For example, DOE (and before it, the Bureau of Mines) sponsored early-stage research to establish the extent of the coalbed methane resource and also pilot-tested some techniques for accessing it. Thereafter, the Gas Research Institute, an industry research organization, took the lead in developing coalbed methane. In another case, DOE organized a consortium of metal-casting companies to develop a more efficient casting technology that no one company could afford to risk on its own. The resulting technology became an industry standard.
Invest with care in technologies to serve markets that do not yet exist. The private sector innovates in the hope of creating value for future markets. Often this process, although driven entirely by private benefit, is enough to produce the innovations that energy policy desires. In some cases, however, development times are so long, the policy imperative is so profound, and the future market is so uncertain that government is justified in rushing innovation ahead of its natural pace. At the time of the first energy crisis, this idea lay behind the synthetic fuels program and DOE’s funding of coal liquefaction, coal gasification, and oil shale technologies. Today, the hydrogen economy, cellulose-based ethanol production, and zero-emission coal-fired electric power plants are the breakthrough technologies that DOE wants to accelerate in anticipation of a national commitment to reduce greenhouse gas emissions and to limit the use of oil in transportation.
Because government is not especially good at predicting when and how new markets such as these will emerge, care is required in investing in opportunities to serve them. As noted earlier, over half of the nonproductive $9 billion identified in the NRC study was spent on synthetic fuels projects. At the time it made sense to hedge against the possibility of skyrocketing oil prices, and buying some insurance with government funding was justified. But for government to aim to develop a process or product that is meant to be ready for commercial adoption, as was the goal of the synthetic fuels program, is to take too ambitious a step before the real market is more clearly in sight. Instead, attention should focus on research that would accelerate a class of technologies without presuming to pick specific winning products for undeveloped markets. A recent NRC study of DOE’s hydrogen program, for example, set forth clear research goals intended to move the hydrogen economy forward in this way.
Not surprisingly, looking at the federal government’s current energy science and technology programs though the prism of the foregoing guidelines reveals both good and bad news. By far the best news is the $500 million increase proposed in the 2007 budget for basic energy sciences and biological and environmental research at DOE. If this level of funding can be sustained or increased over time, there exists at least the potential of creating the foundation of knowledge from which future innovations in energy technology are most likely to emerge. Ensuring that these resources are targeted on a diversity of ideas and passionate researchers will be essential to this result. Because many such ideas and people are to be found in the nation’s great research universities and technology companies, DOE must reach beyond its own laboratory complex to realize the full value of this new funding. Encouragingly, the national laboratory share of the relevant science budgets seems to have drifted lower over the past five years, from around 65% of the total to about 60%.
Also in the good news column is that DOE’s applied research programs are getting more sophisticated in identifying and removing specific barriers to private-sector innovation. This observation is based on several studies of the DOE program conducted by the NRC. A review of DOE’s Industrial Technologies Program, for example, concluded that the program “has evolved over time into a well-managed and effective program….[It] significantly leverages its resources through a large and growing number of partnerships with industry, industry associations, and academic institutions.” Similarly, DOE’s FutureGen advanced electric power plant program and its research support for building the next generation of nuclear plants are being conducted in close cooperation with private-sector actors who are committed to innovation in the use of coal and nuclear power. Of course, DOE can probably do more to increase the ratio of energy policy benefits to government cost, but its program managers seem to be looking in the right places.
Unfortunately, however, the appropriations process tends to divert funding from the applied research strategies that are most likely to pay dividends. A major reason is that the targeted removal of obstacles to private-sector innovation is fairly dull work. It is tempting to embrace programs that will “solve” the energy problem by creating a hydrogen economy or a zero-emissions power plant or plug-in hybrid vehicles fueled with ethanol made from agricultural wastes. Although government-sponsored research is undoubtedly justified to position the private sector to move as the market prospects improve, it is unwise to invest in programs that have the unrealistic aim of developing products that would serve markets that do not yet exist. The risk, as noted earlier, is that such programs become too ambitious and so crowd out less glamorous but more beneficial research.
Far and away the most serious shortfall, however, is that current policy has its incentives priorities backwards. For example, the Energy Policy Act of 2005 creates $2.7 billion of production tax credits for renewable energy and an equally generous package of tax credits and loan guarantees for new nuclear and clean coal plants. Regulation gets some attention in the legislation, notably additional appliance standards, but the pace and degree of required efficiency improvement have yet to be specified. Similarly, although changes in CAFE standards are in the works, the reported targets are at best modest. And federal regulation seems less ambitious than actions being taken by a number of states.
But the missing link is to reward the private sector directly for meeting energy policy goals; that is, to put a price on carbon production and oil consumption. Although this is not the only policy that should be adopted, it enhances all the rest by focusing innovation on a specific outcome. By creating a market for public goods, this policy highlights the obstacles to innovation that government can help overcome, motivates basic researchers in a variety of disciplines to apply their knowledge to important problems, and greatly mitigates the danger of anticipating markets that may never materialize. As long as this policy tool stays on the shelf, the nation’s longstanding desire to use energy policy to stimulate technological innovation will remain unfulfilled.