Technology and Learning by Doing

The Timing of Climate Change Policy

Technology and Learning by Doing

The idea that technological change will underlie any long-term solution to the climate change problem is universally shared. Analyses have shown quite convincingly that, absent technological progress, the costs of any level of long-term GHG abatement are exceptionally high.¹³ 

However, technological progress takes a great deal of time. The production of automobiles powered by hydrogen, for example, would require developing new ways to produce, store, and distribute that fuel; new training for repair and maintenance of those vehicles; and a variety of other parallel changes. Research and development were once seen as part of a “pipeline,” in which basic research at one end led to development efforts in the middle, which led to commercialization and diffusion at the other end. But a broad range of analyses of specific inventions reveals that the process is far more complex and interactive than this simple metaphor suggests. A variety of other factors appear to drive inventions and their diffusion, including procurement incentives, regulatory requirements, government research or incubation, feedback from users, and, often most importantly, the phenomenon of learning by doing. Technological progress does not happen because basic research is performed in one place, then carried to another for development, then subsequently brought to market.

Instead, a variety of analyses suggest that most of the nation’s technological progress occurs in a swarm of minor increments related to ongoing experimentation, or “learning by doing.” Leading inventions, such as the automobile, telephone, or airplane, usually have been invented by several different researchers at roughly the same time: The second application for a patent for the telephone, for example, arrived only hours after that of Alexander Graham Bell. This occurs because these epochal inventions were not truly “new,” but instead took pre-existing components, parts or products and rearranged them in new and novel ways. This “component-oriented” view of innovation shows that technological progress is more directly related to experimentation and tinkering than to a pipeline of basic research results.

This view is borne out by the persistent phenomenon of “learning curves,” which relate cumulative experience with a particular technology with declines in its unit costs. A survey of energy-related technologies such as photovoltaics, windmills, and gas turbines reveals that every doubling of cumulative experience with each technology led to a decline of about 20 percent in its cost of operation.¹⁴ Thus, early experience with new technologies appears to lead to future cost reductions.

Other economic analyses note that the level of effort devoted to technology is not fixed — rather, it responds to conditions in the economy. Technological change is not simply an autonomous process, but results from complex factors including prices, consumer values, taxes and regulations, and technology policies.¹⁵ The level of research and development, therefore, would respond to the announcement of a climate change policy. This effect is sometimes referred to as “induced technological change.” When firms come to understand that a regulatory, tax, or other regime puts a premium on restraining GHG emissions, they will shift research and development efforts towards that goal. As one workshop paper points out, “We need active training not relaxation to get into shape to run a marathon.”¹⁶ This lowers the cost of complying with a GHG reduction mandate, but also means that society’s R&D resources must be shifted from other purposes so long as science and engineering workers are in scarce supply.

A number of leading corporations have set targets to reduce their own GHG emissions, finding that acting now gives them a competitive advantage through learning by doing. As they set and meet their targets, these companies are finding low-cost emissions reduction opportunities; ancillary safety, efficiency, and environmental benefits; and competitive advantage in becoming more climate-friendly.¹⁷ Any climate policy must lead firms to begin this process in earnest — to induce technological change — if it is to lower the cost of abating GHG emissions. The very long times required to bring technologies on line, and the prospect for rising atmospheric GHG concentrations in the interim, suggest strongly that action be taken now to begin developing and diffusing those technologies.

NEXT: "No Regrets" Options

Download PDF