Nuclear energy occupies a strange place in the American psyche—representing at once a dream of endless emissions-free power and a nightmare of catastrophic meltdowns and radioactive waste. The more prosaic downside is that new plants are extremely expensive: America’s most recent attempt to build a nuclear facility, in Georgia, was supposed to be completed in four years for $14 billion. Instead it took more than 10 years and had a final price tag of $35 billion—about 10 times the cost of a natural-gas plant with the same energy output.

But the United States might not have the luxury of treating nuclear energy as a lost cause: The Department of Energy estimates that the country must triple its nuclear-power output by 2050 to be on track for its climate targets. For all the recent progress in wind and solar energy, renewables on their own almost certainly won’t be enough. Arguably, then, we have no choice but to figure out how to build nuclear plants affordably again.

Half a century ago, nuclear energy seemed destined to become the power source of the future. The first commercial-reactor designs were approved in the 1950s, and by the late ’60s, America was pumping them out at a fraction of what they cost today. In 1970, the Atomic Energy Commission predicted that more than 1,000 reactors would be operating in the United States by the year 2000.

In the popular history of atomic energy in America, the turning point was the infamous meltdown at the Three Mile Island plant in 1979. In the aftermath of the accident, environmentalists pressured regulators to impose additional safety requirements on new and existing plants. Nuclear-energy advocates argue that these regulations were mostly unnecessary. All they did, in this telling, was make plants so expensive and slow to build that utility companies turned back to coal and gas. Activists and regulators had overreacted and killed America’s best shot at carbon-free energy.

This story contains some kernels of truth. The safety risk of nuclear energy is often wildly overblown. No one died at Three Mile Island, and later studies found that it didn’t have any adverse health effects on the local community. Even including the deadly meltdowns at Chernobyl and Fukushima, nuclear power has most likely caused only a few hundred deaths, putting its safety record on par with wind turbines and solar panels, which occasionally catch fire or cause workers to fall. (The immediate areas around the sites of the Chernobyl and Fukushima disasters have, however, been rendered uninhabitable for decades because of the potential dangers of radiation.) Nuclear waste can be harmful if mishandled, but isn’t difficult to store safely. Air pollution from fossil fuels, meanwhile, is estimated to kill anywhere from 5 million to 9 million people every year.

[Read: Nuclear is hot, for the moment]

The claim that excessive regulation single-handedly ruined the American nuclear industry, however, doesn’t hold up. The cost of building new nuclear plants was already rising before Three Mile Island. Several nuclear-energy experts told me that a major driver of those cost increases was actually a lack of industry standards. According to Jessica Lovering, the executive director of Good Energy Collective and a co-author of a widely cited study on the cost of nuclear energy, throughout the ’60s and ’70s, utilities kept trying to build bigger, more ambitious reactors for every new project instead of just sticking with a single model. (Lovering used to be the head of nuclear policy at the Breakthrough Institute—a think tank that tends to warn against excessive regulation.) “It’s like if Boeing went through all the trouble to build one 737, then immediately threw out the design and started again from scratch,” she told me. “That’s a recipe for high costs.” The 94 nuclear reactors operating in the United States today are based on more than 50 different designs. In countries such as France and South Korea, by contrast, public utilities coalesced around a handful of reactor types and subsequently saw costs remain steady or fall.

Lovering also noted that the overregulation story leaves out a crucial fact: Because of a slowing economy, electricity demand flatlined in the early 1980s, causing American utilities to stop building basically every electricity-generating resource, not just nuclear plants. By the time the U.S. finally did try to build them again, in 2013, the American nuclear industry had all but withered away. “In the 1970s, we had a whole ecosystem of unionized workers and contractors and developers and utilities who knew how to build this stuff,” Josh Freed, who leads the climate and energy program at Third Way, a center-left think tank, told me. “But when we stopped building, that ecosystem died off.” This became obvious during the disastrous Vogtle project, in Georgia—the one that ended up costing $35 billion. Expensive changes had to be made to the reactor design midway through construction. Parts arrived late. Workers made all kinds of rookie mistakes. In one case, an incorrect rebar installation triggered a seven-and-a-half-month regulatory delay. Experts estimate that by the time it was finished, the project was four to six times more expensive per unit of energy produced than plants built in the early ’70s.

Given the impracticality of nuclear energy, some environmentalists argue that we should focus on wind and solar. These technologies can’t power the entire grid today, because the sun doesn’t always shine and the wind doesn’t always blow. With enough advances in battery-storage technology, however, they could in theory provide 24/7 power at a far lower price than building nuclear plants. “The nuclear industry has been promising cheap, clean energy for decades at this point,” David Schlissel, a director at the Institute for Energy Economics and Financial Analysis, told me. “Why waste our money on false hopes when we could be putting it towards technologies that have a real chance of working?”

He may be right about the technology. But just because it might one day be technically feasible to power the entire grid with renewables doesn’t mean it will ever be politically feasible. That’s because wind and solar require land—a lot of land. According to Princeton University’s “Net-Zero America” study, reaching net-zero emissions with renewables alone would involve placing solar panels on land equivalent to the area of Virginia and setting up wind farms spanning an area equivalent to Arkansas, Iowa, Kansas, Missouri, Nebraska, and Oklahoma combined. The more land you need, the more you run into the meat grinder of American NIMBYism. Efforts to build renewables are already getting bogged down by local opposition, costly lawsuits, and permitting delays. These challenges will only intensify as the easiest sites come off the board.

Transmission lines, which are needed to transport renewable energy from where it’s generated to where it’s used, may present an even bigger challenge. Some lines have taken nearly two decades just to receive their full suite of approvals. “There’s a chance we will suddenly get our act together and overcome the many, many constraints to deploying renewables,” Jesse Jenkins, who leads the Princeton Zero-Carbon Energy Systems Research and Optimization Lab, told me. “But I’m certainly not willing to bet the fate of the planet on that happening.”

The case for nuclear, then, is less about technological possibilities than it is about political realities. Nuclear can generate the same amount of power while using 1/30th as much land as solar and about 1/200th as much as wind. Reactors can be built anywhere, not just in areas with lots of natural wind and sunshine, eliminating the need for huge transmission lines and making it easier to select sites without as much local opposition. And nuclear plants happen to generate the greatest number of high-paying jobs of any energy source, by far. (On average, they employ six times as many workers as an equivalent wind or solar project does and pay those workers 50 percent more.) That helps explain why four different towns in Wyoming recently fought over the right to host a nuclear project. Nuclear power is also the only energy source with overwhelming bipartisan support in Washington, which makes Congress more likely to address future bottlenecks and hurdles as they arise.

[Brian Deese: The next front in the war against climate change]

As for how to make the economics work, there are two schools of thought. One holds that if America forgot how to build nuclear because we stopped doing it, we just need to start back up. Pick a design, build lots of plants, and we’ll eventually get better. Other countries have done this with great success; South Korea, for instance, slashed the cost of constructing nuclear plants in half from 1971 to 2008. Here, the Vogtle project carries a silver lining: The second of the plant’s two reactors was about 30 percent cheaper to build than the first, because workers and project managers learned from their mistakes the first time around. “I consider Vogtle a success,” Mike Goff, acting assistant secretary for the Department of Energy’s Office of Nuclear Energy, told me. “We learned all kinds of hard lessons. Now we just need to apply them to future projects.”

The second school of thought is that we’ve been building nuclear reactors the wrong way all along. This camp points out that over the past half century, basically every kind of major infrastructure project—highways, skyscrapers, subways—has gotten more expensive, whereas manufactured goods—TVs, solar panels, electric-vehicle batteries—have gotten cheaper. Lowering costs turns out to be much easier when a product is mass-produced on an assembly line than when it has to be built from scratch in the real world every single time. That’s why dozens of companies are now racing to build nuclear reactors that are, in a phrase I heard from multiple sources, “more like airplanes and less like airports.” Some are simply smaller versions of the reactors the U.S. used to build; others involve brand-new designs that are less likely to melt down and therefore don’t require nearly as much big, expensive equipment to operate safely. What unites them is a belief that the secret to making nuclear cheap is making it smaller, less complicated, and easier to mass-produce.

Both paths remain unproven—so the Biden administration is placing bets on each of them. The president’s signature climate bill, the Inflation Reduction Act, included generous tax credits that could reduce the cost of a nuclear project by 30 to 50 percent, and the Bipartisan Infrastructure Law included $2.5 billion to fund the construction of two new reactors using original designs. The Department of Energy, meanwhile, is exploring different options for permanent nuclear-waste storage, investing in building a domestic supply chain for uranium, and helping companies navigate the process of getting reactor designs approved.

There’s no guarantee that the U.S. will ever relearn the art of building nuclear energy efficiently. Betting on the future of atomic power requires a leap of faith. But America may have to take that leap, because the alternative is so much worse. “We just have to be successful,” Mike Goff told me. “Failure is not an option.”

Dozens of used electric-vehicle batteries were stacked like cordwood on pallets in a warehouse-style building about 30 miles east of Reno, Nevada, when I visited the site last week.

The batteries were bound for an assembly line that would begin the chemical process of recycling up to 95 percent of the lithium, cobalt, and nickel they contain. Eventually, after treatment in two more buildings on the site, the metals will become a high-value, fine black powder called cathode active material that is shipped in vacuum-sealed containers to Toyota and Panasonic for the manufacture of new EV batteries.

This cutting-edge recycling process has been developed by a company named Redwood Materials. Founded by J. B. Straubel, a former chief technology officer at Tesla, Redwood has invested about $2 billion in this 300-acre facility located in an industrial park in rural Storey County, not far from where Tesla has built a massive “Gigafactory” manufacturing complex.

With about 900 employees now on-site, and a workforce of 1,500 expected when the plant is operating at capacity, Redwood’s Nevada facility embodies the economic opportunities spinning off from the environmental imperative to reduce carbon emissions and combat the risk of global climate change.

[Read: Biden’s blue-collar bet]

The facility is also a testament to the incongruous political dynamics forming around the emerging electric-vehicle industry and the broader transition toward a clean-energy economy.

Electric vehicles are being adopted at the fastest rate inside blue-leaning major metropolitan areas. In polls, self-identified Democrats now express much more openness to purchasing an EV than Republicans.

Yet counties that Donald Trump won in 2020, such as Nevada’s tiny Storey County, are receiving the most private investment, and the jobs associated with it, in new EV-production facilities, according to a Brookings Metro analysis provided exclusively to The Atlantic.

The paradox is that even as those red-leaning places are receiving the greatest direct economic benefits from the EV transition, they have mostly elected Republican House members who voted last year to repeal the new federal tax incentives that have encouraged these investments. These places are also likely to provide most of their votes this fall to Trump, who has pledged to repeal “on day one” all of President Joe Biden’s efforts to accelerate the EV transition.

All of this raises doubts about whether it’s sustainable for the emerging EV industry to rely preponderantly on Democratic-leaning places for both its sales and political backing, while providing the greatest economic lift to Republican-leaning places electing political figures hostile to government support for the industry. Put another way: Is red makers for blue takers a viable model for the green economy?

“Every industrial order requires policy support, and so you can certainly imagine all sorts of problems if you have a complete disconnect of the production side from the consumption side,” Mark Muro, a senior fellow at Brookings Metro, told me.

The risk of losing federal support has come as the electric-vehicle industry faces a noticeable slowdown in its previously rapid sales growth. That means the industry could experience even greater disruption if Trump wins and succeeds in repealing the incentives for EV adoption that Biden signed into the Inflation Reduction Act and the bipartisan infrastructure bill.

If the incentives are rescinded, U.S. companies across the emerging EV industry will find it much more difficult to survive the rising competitive challenge from China, Albert Gore III, the executive director of the Zero Emission Transportation Association, told me.

“The role of public policy in achieving the objective of eventually out-competing China in manufacturing batteries, battery components, and EVs themselves is really significant,” said Gore, the son of the former vice president and environmental advocate Al Gore. “Those two bills have taken existing momentum [in the industry] and accelerated it and magnified it.”

At this early stage in the industry’s development, the mismatch between the geography of EV production and consumption—between the makers and takers—could hardly be greater.

In its tabulation, Brookings Metro identifies more than $123 billion in U.S. investments in EV plants since Biden took office. Almost exactly 70 percent of that spending has flowed into counties Trump won in 2020, Brookings found.

The environmental group Climate Power tracks all private-sector investment in clean energy, including facilities that manufacture components for generating solar and wind electric power, and plants that provide semiconductors for clean-energy products and improvements to the electric grid. In its latest report, the group found that since passage of the IRA, Republican-held House districts had received three-fourths of the total $352 billion in clean-energy investment under that broader definition; the GOP districts had also received 53 percent of all the jobs associated with those investments.

In contrast, the places where EVs comprise the largest share of new vehicle registrations are entirely large blue-leaning metropolitan areas, according to a recent New York Times analysis using data from S&P Global Mobility. All six of the metro areas where EVs exceed 20 percent of new registrations are on the West Coast, including five in California and Seattle; other places where EVs have made the most inroads, the Times found, include Portland, Oregon; Denver; Las Vegas; Phoenix; and Washington, D.C. Deep-blue California alone accounts for more than one-third of all U.S. EV registrations.

Polling shows that uncertainty about EV costs, reliability, and charging infrastructure is evident among a broad range of Americans. But the partisan gap over EVs remains striking.

In Gallup polling this spring, about one-fourth of Democrats said they either own or are seriously considering purchasing an EV; nearly another half of Democrats said they were somewhat open to buying one. But Republican voters have become deeply resistant to EVs;  in both the 2023 and 2024 Gallup surveys, about seven in 10 say they would never buy one. Other polling this year has found that while Democrats, by a ratio of about 10 to one, believe that EVs are better for the environment and more energy efficient than gas-powered cars,      a plurality of Republicans say that traditional internal-combustion-engine vehicles are better on both counts. Republicans are also far more likely than Democrats to say that gas-powered cars are safer, more reliable, and more affordable to operate. And of course, many more Republicans than Democrats to begin with reject the scientific consensus that carbon emissions are dangerously transforming the environment.

Brian Deese, who helped devise Biden’s clean-energy strategy as his first director of the National Economic Council, told me that economics, not politics, explained the geography of EV production. In choosing where to locate their plants, Deese said, companies are not focusing on a community’s political inclinations but rather are looking for places with lots of space, as well as nearby manufacturing and construction capacity. All of those factors, he notes, tend to be most available in communities outside major metro areas that are now preponderantly represented by Republicans. Labor officials would add one other factor: In many cases, companies are locating their new EV facilities in Republican-leaning states with right-to-work laws that impede union organizing.

Deese also thinks it’s too soon to assume that Democratic voters will remain the prime market for EVs. Although polls today show such political polarization around EVs, “it’s pretty hard to think of a technology where there was a cheaper, better technology to solve a consumer need and consumers prioritized a cultural or political patina over lower costs and higher quality,” said Deese, who is now a fellow at the Massachusetts Institute of Technology’s Center for Energy and Environmental Policy Research.

The challenge for the EV industry is that the current mismatch between its makers and takers could doom the public policies promoting its growth before it can prove its value to consumers across the political spectrum.

The industry’s key firms, as Gore told me, began investing in EV facilities long before Biden signed the IRA and infrastructure bills, or finalized the Environmental Protection Agency regulations requiring auto manufacturers to sell a growing number of zero-emission vehicles over the next decade. But companies, he added, are undoubtedly counting on both those carrots and sticks when calculating how much consumer demand will grow to support the EV-production facilities they are building.

Redwood itself typifies that dynamic. Alexis Georgeson, Redwood’s vice president of government relations and communications, pointed out that the company was formed in 2017, while Trump was president; even then, she told me, Straubel recognized that as the number of EVs on the road increased, recycling their batteries could provide an economic opportunity while also reducing U.S. reliance on foreign sources of lithium and other critical minerals the batteries require. “Our mission, our business proposition, has not changed as a result of what we have seen happening with this administration,” Georgeson said.

But Georgeson is quick to add that the policies approved under Biden have “been tremendously helpful to us.” Redwood has been approved for a $2 billion federal loan to fund further expansion of its Nevada site (although it has not received any of the money yet). It also will benefit from the IRA provisions that provide tax credits for producers and consumers in order to encourage domestic manufacture of EV batteries. Within a few years, Redwood expects to be producing enough of the recycled minerals, as well as superthin foil from recycled copper, to manufacture at least 1 million EV batteries a year.

Devon Reese, a Reno city-council member, told me that the EV industry’s rapid expansion in the area has come with “growing pains” mostly relating to ensuring reliable transportation to the isolated Storey County. But overall, he said, there’s no question that the industry’s growth “absolutely has been a net positive” for the community. There are “probably nearly 11,000 jobs that have been created in this region by the energy projects,” Reese told me. “That represents a lot of families, and homes that are owned, and apartments that are rented, and stores that are shopped in.”

Despite this enormous flow of investment into the Reno region from Redwood and Tesla’s “Gigafactory,” Mark Amodei, the area’s Republican U.S. representative, joined every other House Republican last year in voting to repeal all of the IRA’s incentives for EV production and clean energy. Every House and Senate Republican also voted against the initial passage of the IRA in 2022.

After his repeal vote, Amodei told me he questioned how many of the construction and production jobs would really go to Nevada residents. “The benefits I don’t think outweighed the negative stuff in terms of debt, inflation, and—oh, by the way—how much really came to Nevada,” Amodei said. He also said that the pace at which Biden is seeking to encourage a transition away from internal-combustion engines toward EVs “borders on suicidal” and “just makes no sense unless it’s all about a political agenda and not much about solving problems” (Amodei’s office did not respond to a question about whether his views have changed since then).

Environmentalists had hoped that the spread of clean-energy investments into Republican-held seats would politically safeguard the IRA the same way the diffusion of military projects across virtually every state and district ensures broad bipartisan congressional support for defense-spending bills. Groups, including Climate Power, are running ads in swing states this year touting the new jobs that EV manufacturers are creating, for instance in building fully electric school buses. “This is really about jobs in the United States,” Alex Glass, Climate Power’s managing director of communications, told me. “What Donald Trump has been saying is he would rather have these jobs—the jobs of the future—happen in China.”

But the willingness of all House Republicans to vote to repeal Biden’s EV incentives, even while their districts are receiving most of the investment flowing from them, challenged the traditional assumption that politicians fear voting against policies that are providing direct economic benefits to their voters.

[Read: Biden’s ‘big build’]

Now the hope among clean-energy advocates is that some Republicans whose districts are benefiting from these incentives voted to rescind them last year only because they knew that repeal could never become law with a Democratic Senate and president. If Trump wins, and Republicans seize unified control of Congress next year, a vote to repeal the IRA incentives would transform from a symbolic gesture into an actual threat to jobs in these districts. Even under a scenario of unified Republican control in Washington next year, “our perspective is that it would be quite challenging” for the GOP to assemble enough votes for repeal, Georgeson said.

As evidence, she pointed to the disconnect between Republican opposition to the IRA in Washington and the cooperation the company has received from GOP governors and other state officials in Nevada and South Carolina, where it is beginning work on an even larger recycling plant that will involve $3.5 billion in investment. “Both states have been incredibly supportive of us,” she told me, providing assistance not only in infrastructure but in forging partnerships with local colleges to train workers for the new jobs.

Gore told me that the electric-vehicle industry will mobilize to defend the federal tax and spending programs promoting its growth if Republicans try again to repeal them next year. The industry, he said, must do a better job of demonstrating how it is benefiting the Republican-leaning communities where it is primarily investing. But, he added, it can now marshal powerful evidence for that case in the form of new manufacturing plants and jobs. “No politician sees a vote against a popular growing factory in his or her district as a winning issue,” Gore told me.

As Trump and congressional Republicans escalate their threats against Biden’s environmental agenda, the best defense for the emerging clean-energy industry may be the growing number of red communities benefiting from the green of new paychecks.