The new Brookings/AEI/Breakthrough “Post-Partisan Power” study, which calls on policymakers to focus on energy innovation rather than carbon pricing, has been generating a lot of debate over the last day. I symphathize with those who have criticized the study for pretending to be more “bipartisan” than it actually is, and for overselling the potential of energy innovation absent government incentives that increase demand. But set that aside: what’s being missed in this debate is that most of the paper is actually a smart and thoughtful discussion of how to do energy innovation policy right.
The authors have seven basic proposals: boost the DOE office of science, including the new Energy Frontier Research Centers (EFRCs) in particular; fund more energy-related science education; establish a national network of energy innovation institutes; scale up ARPA-E; reform energy subsidies to focus on cost reductions; use DoD procurement to provide stable markets for young and risky technologies; and support nuclear power, including, in particular, small modular reactors.
These are all interesting and well developed ideas. I’m particularly partial to their recommendations on the EFRCs, ARPA-E, and government procurement. I’m not going to describe all of the items – I’d encourage readers to look at the compact summary on pages eight and nine of the report – but I do want to pick at a few of the ideas, in the hopes of shifting the debate to the real meat of this report.
The authors argue, correctly, that DOE doesn’t have the same procurement power that DoD had during the Cold War, when defense procurement helped many technologies through the valley of death. They thus recommend that DoD become a big buyer of cutting-edge energy technologies. But why restrict the recommendation to DoD? The U.S. government is a huge buyer of energy, and of energy-consuming technologies. Why not recommend that the USG, more broadly, use its purchasing power to encourage new technologies? This wasn’t an option for most technologies during the Cold War – the Department of Education wasn’t going to start buying fighter aircraft – but it is an option in the energy space.
The authors also call for energy subsidies to be targeted at technologies that are becoming cheaper. That’s a nice idea, but I’d like to hear more about how it would actually work in practice. How do you decide if wind technology is really becoming cheaper in a world where steel prices fluctuate enormously? What do you do about the fact that much of a subsidy is passed on to suppliers and to labor – and hence inflates the price of a given technology? How do you account for the impact that subsidies from other governments have on the price of technologies (whether decreasing prices by subsidizing production or increasing prices by subsidizing demand)? And how do you do this all in a way that’s predictable enough to not disincentivize investment?
Finally, I’ll take issue with the following claim, which underlies many of the recommendations: “Government investments succeed not when they are blanket subsidies but rather when they are narrowly targeted to specific outcomes, such as developing computers to allow for rocket systems, building a communications network to survive a nuclear attack, or creating increasingly efficient and powerful jet engines.” It’s hard to know whether that’s true. Certainly it’s easier for the government to take credit when its investments are narrowly targeted, but that’s a different issue. Broad-based government subsidies for science education, for example, have had huge payoffs over the last half century. I suspect the authors would support the R&D tax credit – a blanket subsidy – too. (If they don’t, they should.) I think the authors have a reasonable case for pushing back against those who would only support broad-based subsidies. But, just like carbon pricing versus innovation, this isn’t either-or.