Carbon capture and storage (CCS) is a promising technology that might allow for significant reductions in CO2 emissions. But at present CCS is very expensive and its performance is highly uncertain at the scale of commercial power plants. Such challenges to deployment, though, are not new to students of technological change. Several successful technologies, including energy technologies, have faced similar challenges as CCS faces now. In this paper we draw lessons for the CCS industry from the history of other energy technologies that, as with CCS today, were risky and expensive early in their commercial development. Specifically, we analyze the development of the US nuclear-power industry, the US SO2-scrubber industry, and the global LNG industry.

We focus on three major questions in the development of these analogous industries. First, we consider the creation of the initial market to prove the technology: how and by whom was the initial niche market for these industries created? Second, we look at how risk-reduction strategies for path-breaking projects allowed the technology to evolve into a form so that it could capture a wider market and diffuse broadly into service. Third, we explore the "learning curves" that describe the cost reduction as these technologies started to capture significant market share.

Our findings suggest that directly applying to CCS the conventional wisdom that is prevalent regarding the deployment and diffusion of technologies can be very misleading. The conventional wisdom may be summarized as: "Technologies are best deployed if left in the hands of private players"; "Don't pick technology winners" or "Technology forcing is wrong"; and "Technology costs reduce as its cumulative installed capacity increases". We find that none of these readily applies when thinking about deployment of CCS.

Through analyzing the development the analogous industries, we arrive at three principal observations:  

• First, government played a decisive role in the development of all of these analogous technologies. Much of the early government role was to provide direct backing for R&D work and demonstration projects that validated the technological concepts. For example, the US government directly supported for over two decades most of the basic science and engineering research in both SO2 scrubbers and nuclear power. Most of the demonstration projects were significantly underwritten by government as well; the Japanese government was the principal backer of LNG technology through its promises to buy most of the world's LNG output over many years. Direct government support created the niche opportunities for these technologies.
• Second, diffusion of these technologies beyond the early demonstration and niche projects hinged on the credibility of incentives for industry to invest in commercial-scale projects. In each of the historical cases, government made a shift in its support strategy as the technology diffused more widely. In the early phase (when commercial uncertainties were so high that businesses found it extremely risky to participate in more than small, isolated projects) success in achieving technology diffusion required a direct role for government. But as uncertainties about the technology's performance reduced and operational experience accumulated, direct financial support became less important, and indirect instruments to lower commercial risk rose in prominence. Those instruments included tax breaks, portfolio/performance standards, purchase guarantees, and low-interest-rate loans linked to specific commercial-scale investments. It is conceivable that such incentives could have been supplied by non-governmental institutions, such as large firms or industry associations, but the three analogs point strongly to a governmental role-perhaps because only government action was viewed as credible. (In the United States, many of the key decisions to support new technologies were crafted at the state level, such as through rate base decisions to allow utilities to purchase nuclear plants.)
• Third, the conventional wisdom that experience with technologies inevitably reduces costs does not necessarily hold. Risky and capital-intensive technologies may be particularly vulnerable to diffusion without accompanying reductions in cost. In fact, we find the opposite of the conventional wisdom to be true for nuclear power in the US (1960-1980) and global LNG (1960-1995). Costs increased as cumulative installed capacity increased. A very rapid expansion of nuclear power plants in the US around 1970 led to spiraling costs, as the industry had no chance to pass lessons from one generation of investment to the next-a fact evident, for example, in the failure to standardize design and regulation that would allow firms to exploit economies of scale. For natural gas liquefaction plants, costs stayed high for decades due to a market structure marked by little competition among technology suppliers and the presence of a single dominant customer (Japanese firms organized by the Japanese government) willing to pay a premium for safety and security of supply. The same attributes that allowed LNG to expand rapidly-namely, promises of assured demand made credible by the singular backing of the Japanese state-were also a special liability as the technology struggled to compete in other markets. The experience with SO2 scrubbers was more encouraging-costs declined fairly promptly once industrial-scale investment was under way. But that happened only after sufficient clarity on technological performance and capability of FGD systems had been established. What followed was a strict performance standard-in the form of a government mandate, imposed by environmental regulators-that effectively picked FGD as a technology winner. The guaranteed market for FGD led to serious investment, innovations, and learning-by-doing cost reductions. We do not argue that this technology-forcing approach was economically efficient but merely underscore that rates of diffusion of FGD technology akin to what is imagined for CCS technology today were possible only under this technology-forcing regulatory regime.

As CCS commercialization proceeds, policymakers must remain mindful that cost reduction is not automatic-it can be derailed especially by non-competitive markets, unanticipated shifts in regulation, and unexpected technological challenges. At the same time, there may be some inevitable tradeoffs, at least for a period, between providing credible mechanisms to reduce commercial risk, such as promises of assured demand for early technology providers, and stimulating market competition that can lead to lower costs. History suggests that government-backed assurances are essential to creating the market for capital-intensive technologies; yet those very assurances can also create the context that makes it difficult for investors to feel the pressure of competition that, over successive generations of technology, leads to learning and lower costs.

We are also mindful that our history here-drawn on the experience of three technologies that have been successful in obtaining a substantial market share-is a biased one. By looking at successes we are perhaps overly prone to derive lessons for success when, in fact, most visions for substantial technological change actually fail to get traction.

Effective strategies for managing the dangers of global climate change are proving very difficult to design and implement.  They require governments to undertake a portfolio of costly efforts that yield uncertain benefits far in the future.  That portfolio includes tasks such as putting a price on carbon and devising complementary regulations to encourage firms and individuals to reduce their carbon footprint.  It includes correcting for the tendency for firms to under-invest in the public good of new technologies and knowledge that will be needed for achieving cost-effective and deep cuts in emissions.  And it also includes investments to help societies prepare for a changing climate by adapting to new climates and also readying "geoengineering" systems in case they are needed.  Many of those efforts require international coordination that has proven especially difficult to mobilize and sustain because international institutions are usually weak and thus unable to force collective action.  All these dimensions of climate diplomacy are the subject of my larger book project and a host of complementary research here at the Program on Energy & Sustainable Development.  

By far, the most important yet challenging aspect of international climate policy has been to encourage developing countries to contribute to this portfolio of efforts.  Those nations, so far, have been nearly universal in their refusal to make credible commitments to reduce growth in their emissions of greenhouse gases for two reasons.  First, most put a higher priority on economic growth-even at the expense of distant, global environmental goods.  That's why the developing country governments that have signaled their intention to slow the rise in their emissions have offered policies that differ little from what they would have done anyway to promote economic growth.  Second, the governments of the largest and most rapidly developing countries-such as China and India-actually have little administrative ability to control emissions in many sectors of their economy.  Even if they adopted policies to control emissions it is not clear that firms and local governments would actually follow.