Is Climate Really On the Ballot?

Why Projections of US Emissions Under Harris Versus Trump Should Be Taken With A Grain of Salt

By Ted Nordhaus

One of the long-standing tropes of climate politics in the United States is that as every presidential election comes into view, the future of the planet hangs in the balance. It does not matter what the climate movement has said about the Democratic candidate in the past. Bill McKibben at various times branded both Barack Obama and Hilary Clinton as climate deniers and the Sunrise Movement gave Joe Biden an F on his climate agenda during the 2020 primaries. Nor does it much matter that it is hard to distinguish any difference in US performance on emissions under Democratic versus Republican administrations. US emissions actually rose faster under Bill Clinton’s presidency than they did under either Bush I or II and they fell faster under Donald Trump than under Joe Biden even after excluding the deep decline in emissions in the last year of the Trump administration due to the COVID-19 pandemic. 

That’s not because either of the Bushes or Trump necessarily had better climate or energy policies than Clinton or Biden. But the reality is that economic growth rates, sectoral shifts in the composition of advanced developed economies, the business cycle, interest rates, and a range of other factors will almost always dominate emissions trajectories over the four or eight years that any administration will hold office. 

Over the longer term, technological change has a huge effect. The inflection point that tipped the US economy from rising to falling emissions at the end of the 2000s, for instance, was the financial crisis and great recession. But the thing that sustained it was the shale gas revolution, which flooded the US electricity market with cheap natural gas over the decade that followed and drove coal fired generation in the US electricity sector to record low levels. 

As we at Breakthrough Institute were the first to document, the shale gas revolution was not just a private sector technological revolution. It was made possible by decades of federal policy, including research and development, demonstration projects, public/private partnerships between the Department of Energy and the oil and gas industry, and a generous federal production tax credit for unconventional gas. But the important work that those policies did unfolded decades before shale gas remade the global energy landscape over the last decade and a half. The Carter, Reagan, Bush, and Clinton administrations can all take some credit for the shale gas revolution, even though the environmental benefits largely accrued to the emissions accounts of the Obama, Trump, and now Biden administrations.

Nonetheless, with the presidential election looming, a number of prominent energy system modelers have produced analyses purporting to quantify the emissions consequences of a Harris versus Trump victory in November. These analyses largely build upon analysis produced after the passage of the Inflation Reduction Act in 2022, which projected deep reductions in US emissions over the next decade due to those new policies, suggesting that US emissions were likely to fall by 40% or more by the end of this decade. 

As we demonstrated earlier this year, those analyses have already proven to be highly unreliable. US emissions, clean energy deployment, and electric vehicle adoption are all lagging well behind initial projections. Moreover, those projections all calculated the emissions benefits that the IRA would bring against a no-policy baseline that assumed that the US economy would decarbonize much more slowly in the absence of the IRA than it did over the thirty years prior to the passage of the IRA. Models projecting deep emissions reductions from the IRA over the next decade, in other words, did so in comparison to a counterfactual scenario that assumed that the US economy would re-carbonize over the next decade. 

The safe bet, of course, is that the US decarbonization rate over the next decade or two will not look terribly different from what it has looked like over the last three decades, featuring a slow and fairly consistent long term decline in the carbon intensity of the US economy, with short term outcomes in absolute annual emissions largely driven by macroeconomic factors that are to a significant degree out of control of federal policymakers and more particularly almost entirely independent of federal climate and energy policy. And indeed, since those initial models were released, what has become clear is that US emissions are trending toward the historic decarbonization rate. 

For this reason, the new projections showing dramatic differences in outcomes between a Trump and Harris administration should probably be taken with a grain of salt. The energy policy think tank Energy Innovation, for instance, has released new national modeling suggesting a second Trump administration could cause the U.S. to emit 1760 million metric tons of CO2-equivalents (MMT CO2e) more annually in 2030 than a Harris administration continuing Biden’s policies—a difference equivalent to annual emissions from Russia. A few days later, the Princeton REPEAT Project published a somewhat different analysis comparing modeled future post-IRA U.S. emissions with a hypothetical “frozen policies” counterfactual in which Biden had not succeeded in implementing any new policies after assuming office.  REPEAT’s analysis projected a difference of up to 600 MMT CO2e emitted in 2030. REPEAT finds a smaller disparity than Energy Innovation partly because REPEAT’s “frozen policies” assumes that policy that existed prior to Biden taking office would continue through the end of the decade, whereas Energy Innovation assumes that a Republican administration would actively attempt to repeal policies that contributed to decarbonization that preexisted Biden’s election. 

Similar to earlier post-IRA projections from both groups, these new projections forecast emissions under a Trump administration to substantially underperform the historic decarbonization rate while emissions under a Harris administration substantially overperforms that rate.

Figure 1: Modeled U.S. future emissions, adapted from 2023 and 2024 Energy Innovation analyses of climate policies under a future Harris administration (Current Policies), a future Trump administration (Project 2025), and a previous 2023 no-IRA scenario (Previous Reference). The lighter blue line marks the historic 1990-2022 trend of improving U.S. CO2e/GDP (-4.48% year-on-year), extrapolated to 2050. Note that this CO2e/GDP-based trend produces effectively identical emissions as a naive extrapolation of the U.S. gross emissions trend from 2005-2022 or 2008-2022. Source: The Breakthrough Institute

But one need not interrogate these findings, and the assumptions behind them, for long to see that there are plenty of ways that the opposite might be the case. A counterfactual world where the Biden administration did not pass the IRA and the IIJA, for instance, might have been one in which interest rates and inflation would have been at least somewhat lower, which might have kept a lot of clean energy projects that have been canceled over the last three years afloat. 

Looking forward, some financial analysts are forecasting higher economic growth in the early years of a Harris administration versus a Trump administration, due to Trump’s sweeping proposal to impose heavy tariffs on Chinese imports and higher deficits resulting from another round of tax cuts. Slower economic growth under a second Trump administration could easily result in lower emissions, even if Trump were to succeed in repealing the IRA in full. 

Recent modeling by Rocky Mountain Institute of the proposed Manchin-Barrasso permitting reform bill, to take another example, finds that the expanded transmission capacity that it projects would produce virtually the same magnitude of emissions benefit as REPEAT projects for a Harris versus Trump administration. Should Manchin-Barraso not pass in the closing days of the present Congress, it is hard to say whether prospects for permitting reform will be better under a Trump or Harris administration. 

Similar uncertainties abound about the future of nuclear energy. How will a second Trump administration view Biden era policies to support nuclear energy under provisions of both the IRA and the IIJA? How will Harris appointees to the NRC proceed on regulatory modernization and reform? Now fold in many further uncertainties. Will Harris’ newfound love of natural gas and permitting reform last beyond the election? What commitment, if any, does Trump really have to Project 2025 beyond repealing the IRA? Does it extend to repealing policies established during the first Trump administration, as assumed by Energy Innovation? 

Insofar as events, not campaign promises, more often than not shape presidencies, how will a Trump administration versus a Harris administration respond if AI drives a massive increase in electricity demand from data centers over the next four years? Or if China invades Taiwan? Even assuming that both of these initiating events occurred under either administration, the choices that they would make in response might be quite different and might have quite different consequences in terms of the macroeconomic and geopolitical context in which both emissions and broader energy system evolution would unfold. 

Decarbonization and meaningful climate mitigation is a long game. Policy, governance, and leadership matter. As I argued earlier this year, the historic decarbonization rate is, in significant part, the long term signal of past energy policy. But attempting to discern a policy signal from present policy proposals over the comparatively short duration of a single presidential term is a fool's errand. Doing so may serve various partisan and electoral purposes. But it tells us nothing about what matters and doesn’t matter in efforts to cut emissions and transition to cleaner energy. 

Indeed, if the history of US and global decarbonization tells us anything, it is that the things that matter, with the benefit of hindsight, are rarely the things that most political actors in the present imagine. Shale gas, to return to one example, was not taken seriously as a potential breakthrough technology through the 70’s, 80’s, and 90’s. At the time, far more resources, expectation, and hype were directed at oil shale, synfuels, solar, wind, and nuclear. Twenty years ago, to take a different example, various schemes to develop coal-fired electricity generation with carbon capture frequently dominated climate policy discussions while nuclear energy was, at best, an afterthought. 

Thirty or fifty years from now, there will almost certainly be a range of policy choices made by a Trump or Harris administration that will have proven to be decisive, perhaps for the better, perhaps not. IRA investments in electrified steel production or carbon removal or hydrogen production might open up new technological possibilities. Perhaps will Trump ignite a trade war that will tank the global economy, drive major expansion of US mining and processing of critical minerals, or both. 

The future is unwritten. As Vaclav Smil has reminded us for many years now, an energy transition is a remarkably slow moving thing, deeply enmeshed with the macroeconomy and constrained by the massive inertia of built infrastructure designed to last for many decades. Insofar as a Trump or Harris administration is characterized by major energy policy achievements or failures, they will likely not manifest until well after the next president has left office and the efficacy of those policies will only be apparent in hindsight. 

I write that in full recognition that Harris effectively endorsed much of the Breakthrough Institute’s energy policy agenda last week, recognizing the importance of gas for the global energy transition and security, the importance of advanced nuclear to US climate ambitions, calling for across the board investment in clean technology, and for permitting reform and other steps to clear regulatory hurdles to a clean energy economy. My own policy preferences on these questions are no more immune to the basic truth that policy agendas almost never play out the way that their advocates hope they will than are anyone else’s.

That is especially true when it comes to carbon emissions, technology, and policy. We have lots of strong evidence that technology policy can help drive important technological change. But that change is, by its very nature, shot through with uncertainty while its benefits almost always substantially lag the election cycles in which those policies are established, amended, or repealed. 

That doesn’t mean that modeling different energy futures and the potential impact of technology policies can’t tell us anything useful. At the very least, understanding what assumptions modeled projections are sensitive to can usefully help us understand which policy levers might matter, and under what circumstances. All else equal, emissions in 2030 with permitting reform and, hence, a lot more transmission probably will be lower than without, for instance. But there are all sorts of ways that the first step in that chain of events—permitting reform—might not lead to more transmission, which in turn might not result in lower emissions.

The notion that such a causal chain exists, from permitting reform to lower emissions, is based upon assumptions that are input into the models, not conclusions that come out of them. 2030 with permitting reform but also high inflation, high interest rates, high tariffs, and an economic recession, for instance, might not see a lot of new transmission or clean energy deployed even if it does see lower emissions because the economy tanks. 2030 in which the IRA has been preserved and AI-driven energy demand mushrooms might be one in which emissions were substantially higher but decarbonizing technological change proceeded faster. 

Ultimately, future U.S. emissions outcomes hinge upon economic conditions and technological factors and costs. The IRA and other Biden policies could certainly influence those factors, but only alongside a vast host of other variables. Those tumultuous political and economic uncertainties make it more unlikely than not that the future of U.S. decarbonization is narrowly synonymous with the survival of the IRA in its present form or that the trajectory of US emissions over the next four or six years are any more likely than in the past to validate Democratic or environmental policy preferences. While there are plenty of good reasons to prefer a Harris presidency over a Trump presidency, short term emissions projections are not among them.

Some Odds and Ends:

• Grist’s Kate Yoder has written a great profile of Breakthrough’s Patrick Brown and the controversy he kicked off when he criticized Nature and the peer review process that incentivizes scientists to focus overwhelmingly on climate change while ignoring other factors driving impacts associated with wildfires and other weather related climate impacts. I’ve had a long and complicated relationship with Grist. Back when it started, in the early 2000s, as a startup, web-based environmental journalism outlet, Grist put itself on the map by extensively covering the Death of Environmentalism controversy rather than trying to suppress it. In subsequent years, it published a lot of execrable ad hominem attacks on me, Breakthrough, and folks like Roger Pielke Jr. from the likes of David Roberts and Joe Romm. But over the last decade, Grist has produced some of the best and most thoughtful environmental journalism in the business, from folks like Nathanael Johnson, Shannon Osaka, Zoya Teirstein, and Kate. Kate’s story does an admirable job of playing it straight, covering a highly charged controversy and letting her readers decide what to think. A lot of big name environmental journalists could learn a lot from her reporting. 

• The vote on Matthew Marzano’s confirmation to the NRC at the Senate Environment and Public Works Committee was postponed until after the election last week, as the majority didn’t have the votes to move the confirmation out of committee. It is now scheduled for mid-November. If the Democratic majority succeeds in moving Marzano’s confirmation out of committee, it will then need to find floor time in the lame duck session amidst a lot of other high priority items. So the prospects for Marzano’s confirmation, while not dead yet, have considerably dimmed. It’s unfortunate that the White House did not choose to move forward with one of several well vetted Democratic candidates with long-standing and public records of supporting regulatory modernization. Any of these candidates would almost certainly have received broad and bipartisan support in both the committee and the Senate floor. Should the Senate, correctly in my view, refuse to confirm Marzano, that opportunity will need to wait until the next Congress. 

• By my count, Genevieve Guenther has now quietly acknowledged at least 6 of the 8 major errors that George Morrison identified in his very abbreviated fact check of her new book. These include her fundamental misunderstanding of the carbon cycle, her exaggerating the GDP impact of extreme weather in the United States by a factor of 100, her bizarre claim that bioenergy crops produce ionizing radiation, her claim that global agricultural yields have fallen since 1960, her claim that removing carbon from the atmosphere only reduces global warming by half as much as adding carbon increases it, and her claim that health care and labor productivity savings from limiting warming to 2 degrees would pay for the cost of the energy transition. She has not addressed, anywhere as far as I can tell, the seventh issue that Morrison identified, the claim that with 3C of warming, much of the United States would be too hot for people to go outside for 3 months of the year. And she has simply repeated her clearly erroneous claim that Robert Solow had admitted in a footnote in his seminal “A Contribution to the Theory of Economic Growth,” that long term economic growth was limited by the availability of agricultural land, either continuing to mistake his criticism of leading Keynesian growth theories of the time for his own theory or, more likely, simply prevaricating in order to avoid acknowledging that the central claim that underpins her entire chapter on economic growth is complete nonsense. Mostly, she has just attacked the messenger, believing, probably correctly, that as long as her supporters in the climate movement believe that our factcheck was produced by bad people in bad faith, they won’t much care that her argument is built upon claims that are demonstrably wrong.