A dozen states have joined California and many countries in passing legislation to ban the sale of conventional cars and push everyone into electric vehicles (EVs), many within the decade. Similarly, in a feat of regulatory legerdemain, the U.S. Environmental Protection Agency has proposed emissions rules that would effectively require automakers to sell mostly EVs. And of course, the ill-named Inflation Reduction Act, a.k.a. the Green New Deal, gushes subsidies across the EV ecosystem.

The rush to subsidize and mandate EVs is animated by a fatal conceit: the assumption that they will radically reduce CO2 emissions. That assumption is embedded orthodoxy not just among green pundits and administrators of the regulatory state but also among EV critics, who take issue with a forced transition mainly on grounds of lost freedoms, costs, and market distortions.

But the truth is, because of the nature of uncertainties in global industrial ecosystems, no one really knows how much widespread adoption of EVs could reduce emissions, or whether they might even increase them. (And no, this has nothing to do with the truth / joke that Teslas are coal-fired when fueled at night in many places.) While grid realities will indeed matter more than most realize, the relevant and surprising emissions wildcard comes from the gargantuan, energy-hungry processes needed to make EV batteries. This is one of those technical issues that tends to attract slogans, simplifications, and illusions of accuracy; a better understanding requires some patience.

EV emissions realities start with physics. To match the energy stored in one pound of oil requires 15 pounds of lithium battery, which in turn entails digging up about 7,000 pounds of rock and dirt to get the minerals needed—lithium, graphite, copper, nickel, aluminum, zinc, neodymium, manganese, and so on. Thus, fabricating a typical, single half-ton EV battery requires mining and processing about 250 tons of materials. (These figures hold roughly true for all lithium chemistries.) For the carbon-counters tracking such things, the global mining and minerals sector uses 40 percent of all industrial energy—dominated by oil, coal, and natural gas—and that’s before we take into consideration the massive expansion that would be required to supply all the battery factories planned for widespread EV adoption.

The inherent uncertainties about calculating real-world EV emissions arise from myriad “known unknowns” about mining and refining activities. Those all happen elsewhere, upstream, before assembly at a battery or EV factory—that is, before the first mile driven on a grid-supplied kilowatt-hour. Of course, a conventional car also has upstream emissions, though these derive mainly from steel and iron, which account for 85 percent of its weight. For conventional cars, those upstream emissions are a minor factor; burning gasoline dominates the CO2 footprint. But the need for far more materials, and different types, dominates an EV’s total footprint. Production of those metals, such as copper, nickel, and aluminum, uses on average three to ten times more energy per pound than does steel production. All the other EV minerals are similarly energy-intense.

The International Energy Agency (IEA) flagged these realities in a 2021 report. While that report focused on the inadequacy of the supplies of “energy minerals” (something that has since, finally, become widely known), the researchers noted that upstream CO2 emissions from fabricating an EV can “vary considerably across companies and regions.” Indeed. Changing the source of copper or nickel, for example, can lead to doubling or more than tripling those metals’ emissions intensities, depending on a facility’s age, process types, and locations. Building an EV requires several hundred more pounds of copper than building an internal-combustion car. Assumptions about aluminum matter too, because EVs also typically require several hundred pounds more of that material, and two-thirds of global aluminum production comes from coal-fired grids in China, Russia, and India. (The U.S. produces just 2 percent.) In general, refineries in China, which account for 50 percent to 80 percent of global “energy minerals” supply, have emissions 1.5 times greater than those in the European Union or U.S.

A review of dozens of studies of upstream emissions revealed that the bottom-line estimates of EV lifecycle emissions varied by fivefold. It gets worse. That same review found that, across those studies, the median size of the battery assumed for the analyses was 30 kilowatt-hours. But the overwhelming majority of U.S. EVs bought last year sported batteries two to three times bigger. Tripling battery size triples the upstream emissions.

None of these variabilities appears in government forecasts for “zero emissions” cars. In fact, the range of upstream emissions is so wide that it renders meaningless any use of an average number to calculate an EV’s overall carbon footprint. But that’s what analysts do, whether at the IEA or EPA.

Some automakers—notably Volkswagen and Volvo—have published their own studies that take into account both upstream emissions and grid realities. Those analyses found that an EV powered on Europe’s grid creates more CO2 emissions than a conventional car until at least 50,000 miles of gasoline-free driving. After 120,000 miles, the studies estimated that total cumulative emissions reductions finally reach about 15 percent and 25 percent, respectively. Hardly “zero.” Those savings shrink for cars using batteries significantly bigger than small one in the Volkswagen and medium-sized one in the Volvo. And the calculated CO2 reductions collapse, and even evaporate entirely, if one factors in the higher ranges of known values for upstream emissions in mining and processing, rather than the low, average values chosen in those studies.

Even the Wall Street Journal was seduced by the appearances of precision and the consequences of “hidden” assumptions that emerge from the complexities of upstream emissions. In a featured 2021 investigation directed at unearthing the truth of EV emissions, the Journal concluded that, even accounting for both grids and mining, “the data show that switching from gas to electric vehicles will make a huge impact.” That confidence was based on a study that the Journal commissioned from a University of Toronto team. Notably, however, that study chose an extremely low—arguably wildly unrealistic—value for upstream emissions from battery fabrication. (The authors identified the assumptions in their primary research.) In the above-noted review of 50 such studies, the median value for upstream emissions was threefold higher, and the maximum value sixfold higher than the Toronto team used. Making the Journal’s lifecycle emissions calculation using that median value completely neuters the “huge impact.”

The Journal investigation was in good company, since the IEA also asserted that emissions “along the mineral supply chain do not negate the clear climate advantages of clean energy technologies,” a claim parroted by governments and proponents and enshrined in mandates and bans. But as the IEA’s own data shows, that assertion is anything but clear. It’s difficult, perhaps impossible, to identify where specific minerals are sourced and thus their actual associated emissions.

Worse still for carbon counters, we know that upstream emissions are rising. The IEA acknowledges that the impending demand surge for energy minerals comes as global ore grades continue to “fall across a range of commodities,” due to geological realities. Poorer ore grades mean that more of the Earth must be dug up and processed to yield the same pound of metal. So, per the IEA, “lower-grade ores require more energy . . . greenhouse gas emissions and waste volumes.” Already, for example, over the past decade in Chile, the world’s top supplier of copper, mining-energy use has grown ten times more than the tonnage produced.

EV enthusiasts respond that the technology will get better. Of course, better, lighter batteries are not just possible but inevitable—eventually. So, too, is better mining technology that will lead to lower costs and a lighter footprint. But progress is slow in big industrial domains. Rushing to subsidize and mandate yesterday’s technologies won’t make that future happen sooner; indeed, doing so usually stifles innovation. Perhaps someday the U.S. will re-shore a minerals industry that would be both cleaner and more transparent. Don’t hold your breath.

Meantime, hundreds of billions of dollars designated for wildly premature all-EV mandates will likely become stranded capital because the quantities of minerals needed won’t be available soon enough. Along the way, those stranded billions will do little or nothing to cut CO2 emissions. In the end, the rush to EVs could even increase global vehicle-related emissions.

Photo by Justin Sullivan/Getty Images


City Journal is a publication of the Manhattan Institute for Policy Research (MI), a leading free-market think tank. Are you interested in supporting the magazine? As a 501(c)(3) nonprofit, donations in support of MI and City Journal are fully tax-deductible as provided by law (EIN #13-2912529).

Further Reading

Up Next