Technology Demonstration Projects pose one of the most difficult challenges in energy-innovation policy. They are necessary to build an adequate portfolio of clean-energy options that have the potential to be deployed globally on a massive scale in the coming decades. They require public investment; private investors will not fully fund them. But the federal government’s track record of selecting, funding, and managing these projects is not encouraging. The title of the leading study of the subject, The Technology Pork Barrel (which was published in 1991 and based on projects carried out in the 1970s and early 1980s), conveys its conclusion: Demonstration projects almost inevitably become “technological turkeys.”

Defying this quarter-century-old conventional wisdom, the Obama administration initiated the first major new energy technology demonstration program in decades. (In the technology maturation process, as described in more detail below, demonstration falls between R&D, which is typically done with grant support at universities or national labs, and commercialization, which may involve tax breaks or government loan guarantees.) The performance of the Department of Energy (DOE) in running this program, which is explored in this paper, is somewhat more encouraging than The Technology Pork Barrel would lead one to expect, particularly in terminating underperforming projects. But the Obama-era experience does not provide full confidence that the challenge will be met by DOE in the future. Significant reform of DOE’s approach to demonstration projects remains in order, and Congress should consider whether to set up a new agency to run some of these projects instead of DOE.

ITIF’s study of 53 energy technology demonstration projects that were initiated between 2009 and 2011 yields the following recommendations and findings:

• The United States should build and sustain a robust, diverse portfolio of technology demonstration projects as part of a comprehensive clean-energy innovation policy. Its current portfolio is not robust, and it is rapidly dwindling.
• The federal government should continue to co-invest with private partners in clean-energy demonstration projects, because such projects yield public benefits and because private investors lack adequate incentives to bear their full costs. The 2009 stimulus package, which funded the most recent round of projects, was a good investment mechanism because it imposed a sunset date on project completion and did not require annual appropriations to sustain projects.
• Private sector partners, especially technology end users, should continue to take leadership roles in implementing clean-energy demonstration projects. The operational experience gained by these partners is a key benefit that motivates their investment in the demonstration and enables full commercialization of the technology after it has been demonstrated.
• Federal policy for private co-investment in demonstration projects should become more flexible in order to accommodate varying risk profiles, resulting in a wider range of cost-sharing ratios across projects. These ratios have clustered in the past around arbitrary levels established by Congress.
• Federally funded clean-energy demonstration projects should make information sharing among all potential users of the demonstrated technology a higher priority than it has been in the past and incorporate it into project metrics and evaluation criteria. Although such information sharing may reduce the incentive for private partners to invest in demonstrations, it accelerates diffusion and enhances competition as the technology is being commercialized.
• Initial selection of clean-energy demonstration projects should avoid excessively rapid scale-up of unproven technologies and overly optimistic assumptions about the economic and policy environment for follow-on investments in them. The Obama administration’s record was uneven in both respects.
• Federal policymakers should remain prepared and willing to terminate unsuccessful demonstration projects. The experience of the past decade suggests that the “technology pork barrel” syndrome, in which unsuccessful projects lived on because they provided local economic benefits, is not inevitable, but it remains a real possibility for large projects.
• Although DOE improved its performance in designing and managing demonstration projects compared to the 1970s and 1980s, a continuing effort should nonetheless be made to explore the viability of establishing alternatives in this domain, such as the proposed Energy Technology Corporation and Regional Innovation Demonstration Funds (see box 3 below). The factors that facilitated termination of unsuccessful projects in the Obama period may be short-lived, and DOE has not put questions about its management capability to rest.
• Federal agencies other than DOE, such as the Departments of Defense and Transportation, as well as regions and states, should consider making their own co-investments in clean-energy demonstration projects.

This paper begins by very briefly reprising the case for federal clean-energy innovation policy. The following sections explain why demonstration is essential to successful innovation in many areas of energy technology and why the private sector, quite rationally, is unenthusiastic about investing in this stage, leading it to be labeled the “second valley of death” (the first lying between research and prototype). The paper then turns to the scholarly literature on demonstration projects in general as well as the history of federally funded energy technology demonstration projects in particular to develop a set of design principles. These principles are applied to the Obama-era portfolio to provide an early assessment of its strengths and weaknesses, leading to the conclusions summarized above.