The public sector is heavily involved with research and development (R&D) in most countries throughout the world. Involvement varies widely by industrial sector; however, there is extensive spending and research in virtually every level of energy markets. This includes research into theoretical future energy forms, mineral location, combustion and transportation systems, research on pollution reduction, and the health and environmental effects of particular pollutants. Economic theory suggests that government-sponsored research may be efficient since private research efforts may be easily copied once successful, making it difficult for the private entity to capture the full benefits of its investments in R&D.

While "generic" investments into R&D may be supportable on theoretical grounds, the distribution of spending for R&D purposes can make federal intervention highly distortionary. Similarly, the controls on spending may not be as well targeted as in a private entity, reducing the overall efficiency of R&D investments. Uranium enrichment is a case in point. Spending billions on new uranium enrichment techniques (as the United States has) while there is a surplus of enrichment capacity even with current infrastructure means that the funds are not going to other uses which may have a higher return. Even if successful, lower costs of enrichment create a host of nuclear proliferation issues that further reduce the benefit of the innovation.

Obviously, not every case is the same, and U.S. public funds have also led to advances in more efficient lighting. However, as with public targeting of financial assistance, public investment in applied research can often be diverted based on political, rather than economic or societal, returns. In the energy arena, this has translated into a much greater magnitude of support for large, centralized, conventional energy resources over smaller-scale, decentralized, or efficiency-based approaches.

In addition, "generic" research is sometimes not so generic. With the exception of research into the workings of nature, research is generally driven by the most demanding use for a product or technology. As knowledge of the process and technology improves over time, costs fall, production capacity grows, and technologies will diffuse through the economy into lower-value uses. For example, microchips were first used in expensive computers before cheap watches. In energy markets, similar occurrences led to fission reactor development in the military sector before they were transferred to the commercial sector. In each case, the economic justification for the research came from the high-value use; spillover was simply a bonus.

As a result, allocating R&D expenditures to particular energy types requires some judgment. For example, is government spending on naval fission reactors a subsidy to commercial fission? Although the commercial segment undoubtedly benefited, we would argue that the costs should be borne fully by the activities that drove the research. Unless spillover benefits were a central argument in the justification for the original research, we would not attribute the R&D costs to these other activities. However, where energy-related applications are a core driver of the research, a portion of the spending is rightly counted as a subsidy to energy. For example, research on such items as ceramics (for high-temperature motors) is among the most demanding uses driving the R&D, and should therefore be allocated in part to energy efficiency.

A similar argument may be used for research into the magnitude and severity of a variety of environmental problems. Without particular forms of industrial or human activity, there would not be particular forms of emissions to deal with. Without those emissions, there would be no need to study the impacts of those emissions on the planet and its inhabitants. Therefore, for example, research into problems such as global warming may be reasonably allocated to energy on the basis of the energy sector's total contribution to greenhouse gases.