Recent legislation in the United States: consequences for the US and global energy and climate innovation systems

This Perspective sketches how recent legislation in the United States may impact the domestic and global energy and climate innovation systems, drawing on systems concepts articulated by Arnulf Grubler. It also sets out risks and gaps that could lead to less favorable outcomes.


Introduction: outside the modeling box
The U.S. Congress passed four bills between December 2020 and August 2022 that together established the most ambitious greenhouse gas emissions mitigation policies in the nation's history. Several models project that they will cut U.S. emissions by about a billion metric tons of CO 2 -e annually by 2030 (roughly 15% of the 2005 level) compared to the previous trajectory (DOE 2022).
However, 'additional non-modeled policies' (Jenkins et al 2022, p 16) may lead to emissions reductions that dwarf these projections. The chief reason is the policies' potential 'catalytic impact' on low-carbon technologies, which could set in motion trajectories replicating those of wind and solar power.
This perspective sketches how such an impact might unfold in the United States, with knock-on effects globally, drawing on systems concepts articulated by Grubler (1998). It also sets out risks and gaps that could lead to less favorable outcomes.

The New U.S. policy mix
The four bills did not flow from a single plan. U.S. policymaking is fragmented, requiring multiple layers of compromise among diverse actors. Some pieces were worked on for years through formal processes, while others were slapped together secretly overnight. The result is a complex, unwieldy mix focused on public spending and incentives, far different from the cap-and-trade program considered a decade ago (Meckling and Allan 2020).
The Energy Act of 2020 (EA2020) was signed by President Trump after the national election but before the presidency changed hands. It modernized the legal basis of many U.S. Department of Energy (DOE) programs, rebalancing DOE's portfolio, adding clean manufacturing and carbon dioxide removal (CDR), and emphasizing energy storage and carbon capture, utilization, and storage (CCUS). It also strengthened DOE's potential market impact by authorizing largescale demonstration projects in advanced nuclear power, CCUS, geothermal energy, and other fields (Powell et al 2021).
The Infrastructure Investment and Jobs Act (IIJA) won an impressive bipartisan majority from an evenly-divided Senate in August 2021. It provided massive budgets for many activities authorized by EA2020. Funding for CDR research, development, and demonstration and supporting infrastructure, for instance, grew by more than an order of magnitude, while that for CCUS tripled. The law created a new multi-technology Office of Clean Energy Demonstrations (OCED) with a five-year budget of over $20 billion, including $8 billion for regional hydrogen hubs. It poured additional billions into deployment of electricity transmission lines, electric vehicle (EV) charging networks, EV battery manufacturing facilities, and more (White House 2022).
The CHIPS and Science Act passed with a Senate vote similar to the IIJA's roughly a year later. It designated advanced energy technologies as one of ten broad fields that the United States should prioritize to 'enhance [its] competitive advantage and leadership…in the global economy' (U.S. Senate 2022, p 374). To this end, it authorized increases for many DOE research and development programs, expanded DOE's technology commercialization effort, and assigned the agency to foster regional energy innovation ecosystems.
Finally, the Inflation Reduction Act (IRA) passed on a party-line vote in August 2022. The IRA's robust tax incentives and 'clean procurement' funding promised powerful, durable demand signals for an array of technologies (Jenkins et al 2022). On the supply side, it gave $5.8 billion to OCED for industrial decarbonization and added hundreds of billions to the DOE Loan Programs Office's lending authority. It also strongly favored domestic production in some industries through targeted incentives and regulations.

Potential consequences for the U.S. innovation system
The precise impact of these four bills on energy and climate innovation is impossible to forecast. 'Technological evolution' , writes Grubler (1998, p 21), is 'uncertain, dynamic, systemic, and cumulative' . So, too, are the social processes that drive it. Nonetheless, past research suggests broad insights.
These insights rest on a capacious vision of innovation. As Grubler (1998, p 75) explains, innovation merely begins with knowledge generation. It does not end until demand is saturated and all potential applications have been explored. The innovation system thus extends far beyond research activities and funding. Grubler (1998, p 75) identified two sources of technological change in addition to 'new knowledge': 'learning' and 'entrepreneurship and organization' . All three sources are 'regulated through social "techniques"' . This tripartite division yields a rough framework for analysis.
If new knowledge is narrowly equated with basic science, the new policy mix is unlikely to change the trajectory of innovation. DOE programs that support a broad range of disciplines and applications like Basic Energy Sciences and the Advanced Research Projects Agency-Energy (ARPA-E) have been growing at 3%-5% per year since the late 2010s, and they were largely ignored by the big bills of the 2020s (Chong and Hart 2022). CHIPS and Science adopts aspirational goals to accelerate this growth, but funding to achieve them is contingent on future appropriations that are far from certain. Grubler's vision of knowledge, however, is much broader. The new U.S. policy mix is likely to accelerate knowledge underlying specific targeted technologies. Fields linked to hydrogen, CCUS, CDR, and nuclear power, for instance, will benefit from roughly 25% increases in applied research funding (Shawan et al 2021). The increment can be committed over five years and spent over an even longer period, rather than in a single year. That timeline should make it easier for the supply of researchers and facilities to grow in response, rather than merely inflating costs for scarce resources.
Learning will be even more deeply impacted by the new policy mix than knowledge generation. The massive infusion of public funds into demonstration projects for industrial decarbonization, grid modernization, and energy storage, as well as hydrogen, CCUS, CDR, and nuclear power, should allow innovators in these fields to solve practical problems at scale (Hart 2018). Expanded DOE loan authority and tax incentives for manufacturing should stimulate learning from follow-on projects in these fields and others, particularly if key results and best practices are shared, de-risking them further (Grubler and Nemet 2014). Federal investment in infrastructure like power lines and pipelines should also advance learning by enabling renewables and carbon sequestration facilities to be built in attractive locations. It should enhance learning at the systems level, too, as a wide range of energy resources is integrated and operated. All of these policies aim to 'crowd in' private capital, build public and private institutional capacity, and encourage innovation in financial instruments.
Demand side policies, such as tax incentives for buyers of low-carbon technologies like renewable energy and EVs, may enhance learning as well through scale-up and diversification. Market formation and growth often reduces unit technology costs at rates correlated with cumulative production. The slopes of these curves vary widely across technologies, as do the sites of learning, which may include production, operation, or end-use (Malhotra and Schmidt 2020, Lewis and Nemet 2021).
The most profound potential impact of the new policy mix on the U.S. innovation system would be a shift in national norms that enables entrepreneurship and organizational innovation by reducing transaction costs of many kinds. For example, acceptance of the energy transition by 'frontline communities' that have historically suffered due to their proximity to industrial facilities, as well as by 'energy communities' whose fossil fuel resources are losing value, would accelerate innovation. The new federal policies devote substantial resources to both types of communities, and to date have induced investment weighted toward states with Republican majorities, which may soften their past resistance (Meckling et al 2015, McCarthy andVirginia Olano 2023).
None of these projected consequences is inevitable. Learning is difficult to sustain in complex technologies like nuclear power (Grubler 2010). Domestic content requirements for mature massproduced technologies like solar photovoltaics and lithium-ion batteries may slow learning in these fields if domestic producers are unable to draw on knowledge held by foreign producers, particularly in China (Helveston et al 2022). New power plants, industrial facilities, transmission grids, and pipelines may be stymied by state and local regulation. And, while wealthy places aligned with the Democratic party are likely to demand EVs, induction stoves, and the like, the rest of the country may balk.
These risks might have been reduced or averted by a more thoughtful policy design. Any policies in the mix could be reversed in the future. Yet, the fact that they were adopted at all, presenting substantial opportunities along with risks, marks a welcome sea change in the United States' posture.

Potential consequences for the global innovation system
Given the United States central position in the global innovation system, the impacts of its new U.S. policy mix, both positive and negative, will inevitably spill over to other nations.
Markets may be the most important channel for these spillovers. Foreign-domiciled producers may be induced to bring know-how and expertise to the United States, while federally-subsidized U.S.-based producers may accelerate their learning by exporting into growing foreign markets if they are able to use federal support to build a sustainable competitive advantage. Boston Consulting Group (2022, p 6) estimates that global market responses to the IRA will speed energy and climate technologies down their learning curves more quickly than if only domestic responses are considered. These modeled effects are strongest (doubling or more) for the least mature technologies, like CCUS and hydrogen electrolyzers, and modest for more mature ones, like onshore wind and solar PV (see also Larsen et al 2023).
Geopolitical competition will also play a key role in mediating the impacts. Concern about China was an important motivation for Congress, particularly for Republicans who are skeptical about climate change. Domestic content requirements and associated 'friendshoring' that favors trade with specific partners will add momentum to derisking and decoupling, constricting U.S. participation in Chinacentric value chains. These measures have created tensions with other nations as well. The European Union, for instance, was prompted to advance a 'Green Deal Industrial Plan' to limit capital flight to the United States. If the new U.S. policy mix contributes to a vicious cycle that fragments key value chains, the pace of learning could be set back considerably.
Signaling and symbolism represent a third channel for global impact. President Trump's withdrawal from the Paris agreement has been eclipsed by President Biden's assertion of a paradigm shift in global climate politics. Climate-tech innovators and entrepreneurs worldwide, and those who represent them, should take heart, although they may rightly continue to fear a reversal. As Grubler (1998, 366) argued, technological evolution will determine how much damage climate change ultimately inflicts on human society. '[T]o progressively liberate the environment from human interference….we will need more technology, not less…' Researchers will never have a complete understanding of this complex and dynamic evolutionary process, much less the ability to make conclusive forecasts about the impact of policies on it. We are all blind people describing our perceptions of this elephant, economic, political, cultural, or otherwise. But we can and must do much better at pooling our understanding and synthesizing our individually limited insights to create useful if still partial pictures for decision-makers, whether they are individuals buying automobiles or national governments managing power grids. Modeling climate mitigation without incorporating these insights is, as Schumpeter (1942, p 86), the pioneering economist of innovation once put it, 'like Hamlet without the Danish prince' .

Data availability statement
No new data were created or analyzed in this study.