Photo: Scott Olson/Getty Images
A recent CSIS panel teed up the question “Does shifting toward lower net carbon energy production put U.S. manufacturing at risk?” Consistent power supply, day and night, winter and summer, whether it’s windy or sunny, is crucial for industry and a baseline expectation of consumers. The next question is whether policies that promote clean energy will undermine U.S. manufacturing through higher costs. Volkswagen’s recent announcement that it plans to close multiple plans in Germany due to cost factors including energy further highlights the question of how energy costs affect manufacturing location.
Economic research shows that, since the early 2000s, electricity and natural gas prices have had little impact on U.S. manufacturing jobs or output, after controlling for other factors. Today, U.S. manufacturing should be even less sensitive to electricity and natural gas prices for two reasons. First, China and Mexico became less attractive destinations for manufacturing, especially for export to the United States, and are likely to become even less so under the second Trump administration. Second, increased corporate sensitivity to carbon emissions has made many corporations willing to pay more for greener power.
A company building a new plant decides where to locate it—what country, state, and local jurisdiction. That’s the time when energy prices can play the greatest role, in a from-scratch decision. But energy prices are weighed with many competing considerations. These questions include a suite of labor issues: Is the right labor force available? What is the local cost of the kinds of labor you expect to hire? How tight is the local labor market? Is it a “right-to-work” state? Other top considerations include local land availability and cost, and state and local incentives such as tax breaks.
Unless the manufacturing plant is in a very energy-intensive industry, other considerations are likely to dominate. Very energy-intensive industries include aluminum, fertilizer, and battery manufacturing. Tennessee has won a striking number of new battery plants—and it is among the states with the lowest electricity prices. In August 2024, Tennessee had the fourth lowest industrial electricity rates.
For an existing plant, it would take a much bigger energy price change—or other factors making the current location less attractive—for a company to relocate it. Greenfield plants can cost billions more to build. Even if a company needed to fully replace an existing plant, its site and workforce have value.
Academic research tries to disentangle plant location considerations and isolate the energy price effect. Looking across studies, the common finding is that energy prices do have a statistically significant effect on the amount of manufacturing in the United States, but that the effect is quite small.
The following three papers represent the broader research. Gray, Linn, and Morgenstern find that the 50 percent drop in natural gas prices between 2007 and 2012 raised overall manufacturing employment by 0.6 percent. In gas-intensive industries, it was three times higher, but still under 2 percent. All growth is good growth—but a half percent to 2 percent move in employment is quite modest in response to a massive 50 percent drop in natural gas prices.
Other papers look at the impact of climate policies on manufacturing, through the channel of electricity prices. Wolverton, Shadbegian, and Gray look at the 27 states that had mandatory Renewable Portfolio Standards (RPS) by 2015. RPS require a minimum share of renewables in the state power mix that grows over time. They find that RPS raised electricity prices by under 2 percent and scarcely moved manufacturing employment or output—about one-tenth the move in power prices. The results were similar for the most energy-intensive, trade-exposed industries as manufacturing overall. In sum, based on the early 2000s, studies find big moves in electricity prices could cause some loss of U.S. manufacturing, with a downside to jobs and economic output of about one-tenth of the move in electricity prices.
Several major changes are on the horizon.
Recent developments in terms of the U.S.-China relationship and increased corporate sensitivity to emissions have reduced those historical sensitivities to electricity prices. Moving production from the United States to China has become less attractive due not only to tensions between China and the United States, but also to higher labor and land costs in China than in decades past, both in absolute levels and relative to alternative destinations. China’s manufacturing wages rose over 130 percent over a decade, widening its wage gap to Mexico and Vietnam, although China’s weak labor market in 2023–2024 may have led to some recent softening of wages. Within the APAC region, Beijing industrial rents exceed everywhere except the major cities in Australia, Taiwan, and New Zealand; Shanghai rents top Singapore.
U.S. industrial electricity prices are low relative to other advanced economies. In 2023, U.S. industrial electricity prices were the lowest of 24 International Energy Agency (IEA) member countries, according to a UK government study; 16 had prices at least twice that of the United States. U.S. industrial electricity prices are, however, higher on average than in China. As of December 2022, the IEA puts U.S. electricity prices at $85 per megawatt hour, versus $65 in China.
China was the obvious place to relocate manufacturing production if a company was contemplating doing so, before the first Trump administration imposed tariffs, starting in 2018. That was the start of a series of changes that made China a less attractive destination. And this trend may continue.
President-elect Trump’s call to impose across-the-board tariffs on all imports from China, of 60 percent or more, will have a chilling effect on foreign manufacturing investment in China for export to the United States. The risk of universal, escalating tariffs will be especially potent for U.S. corporations when considered in combination with the hawkish policy on China that could make doing business in the United States while having production in China a more uncomfortable position. On November 25, 2024, Trump wrote that he would impose a 10 percent tariff on all goods from China, and raise the tariff rate on Chinese imports already subject to tariffs by 10 percentage points. The memory of the shifting outlook for tariffs on Chinese imports during his first administration will leave companies hesitant about producing in China for export to the United States, even if an initial policy is set.
Beyond the threat of new U.S. tariffs, the combination of U.S.-China tensions, existing tariffs, reshoring and friendshoring, and recent supply chain challenges reduce the likelihood of moving new production from the United States to China, especially for export to the United States. Changes over the last decade include the following:
The same factors, especially tariffs and higher costs in China, have reduced the odds of U.S. plants losing business to China-based competitors.
Additionally, Mexico was the second place a U.S. company might have historically considered relocating a manufacturing plant, but it too has diminished its attractiveness. Mexico could be a big beneficiary from friendshoring and efforts to shorten supply chains. But the administration of former president Andrés Manuel Lopez Obrador was far from business friendly and saw crime rise. The new president, Claudia Sheinbaum, has yet to demonstrate the intention to make Mexico a more investment-friendly destination. Moreover, president-elect Trump has threatened to impose tariffs on imports from Mexico—a threat he made, then pulled back from in 2019, during his previous term in office. On November 25, 2024, he wrote he would impose a 25 percent tariff on imports from Mexico immediately after taking office until the drug trade and illegal immigration end.
Today, many large companies, especially large public companies, have set rigorous targets to reduce emissions. Those targets force them to consider emissions as well as the cost of procuring energy. Many corporates have made commitments to the Science-Based Targets initiative to reduce emissions by a fixed date, usually 2030 or 2035. The initiative helps companies define targets in line with limiting their greenhouse gas emissions proportionally to limit global warming to 1.5 degrees Celsius, then monitors compliance.
Companies that set these targets create internal and external pressure to shift toward renewable energy and manage the costs of doing so. If fossil fuels and carbon capture or removal achieve those reductions less expensively, companies will also look at those solutions.
Over 800 large U.S. companies have set or agreed to set science-based targets (SBTs), including 325 manufacturing companies. Nearly half of those 800 have set targets.
Looking at my own former industry of automotive as an example, 11 big global automakers have plants in the United States. Of these, seven have set SBTs: Ford and General Motors (GM) as well as BMW, Mercedes, Toyota, and Volkswagen. Tesla has committed to setting targets but has not yet set them. Honda, Hyundai/KIA, Nissan, and Stellantis have not yet committed.
When I was chief economist of GM, the company translated those into specific goals, like the share of its energy use from renewables, and then started negotiating power purchase agreements. GM has reached agreements to source 100 percent of the energy to power U.S. sites with renewable electricity by 2025 and is working to achieve that globally by 2035.
Large companies that set targets likely accept that they may need to pay higher prices for renewables, but also recognize that those prices are falling. To the extent that their energy costs rise, setting SBTs means the company needs to find the best price that meets its climate goals and find cost offsets elsewhere. Admittedly, some U.S. companies could withdraw their SBTs in response to the incoming Trump administration’s shift of focus away from climate.
Manufacturing was the clear energy-intensive industry of the past. Today and in the future, AI and bitcoin mining may be the most energy-intensive industries in terms of both share of energy demand and energy’s share in costs. While the world is very early in our understanding of how generative AI will affect our economy and society, it has powerful potential to be beneficial (as well as to be harmful). In contrast, bitcoin mining involves a transfer of value to successful miners, but no evident social benefits. Meanwhile, if electricity-hungry, deep-pocketed tech companies become more deeply involved in energy, they may yield new solutions for clean energy and efficient transmission investment.
Carbon taxes are current policy in the European Union and elsewhere and will affect the relative importance of raw energy prices versus the carbon intensity of that energy. For exports to the European Union and other markets that follow its example of introducing carbon border adjustment mechanisms, clean energy will become an advantage in terms of final product competitiveness. By the same token, it may be a headwind to China, whose grid is far more coal and carbon intensive than the U.S. grid. China, however, is simultaneously investing in clean energy (especially wind and solar) and coal generation plants, with the intention that the mix of generation will steadily shift toward renewables. How quickly China achieves that shift will affect China’s manufacturing competitiveness relative to the United States.
Elaine Buckberg, PhD, is a senior fellow of the Salata Institute for Climate and Sustainability at Harvard University and is on the Geoeconomic Council of Advisers at the Center for Strategic and International Studies in Washington, D.C.