This essay is based on the opening remarks delivered at a recent SOHO Forum Debate on electric vehicles.
If we could imagine a time machine bringing to New York City, an American citizen from the 19th century, odds are the one thing that would seem the most amazing about our time would be the proliferation of the personal automobile. Big buildings, big cities, roads, nighttime illumination would all be imaginable, even if different looking and greater in scale. But the one thing radically different about modern daily life is the convenience and freedoms that come from a car.
Yes, that 19th century citizen would probably be puzzled by people staring at glowing rectangles in their hands. In fact, the personal computer and the personal car are co-equal in their transformative impacts. MIT historian Leo Marx put it well when he wrote that: “To speak . . . of the ‘impact’ of ... the automobile upon society makes little more sense . . .than to speak of the impact of the bone structure on the human body.”
The centrality of the car in the social and economic structure of society is evidenced by how citizens have voted with their pocketbook. A car is the single most expensive product that 98% of consumers ever purchase. Over 90% of American households own or have access to a car. Average household spending on personal mobility is the second biggest expense after mortgage or rent.
And there’s nothing to the trope that the rising generations will abandon automobiles. A recent MIT analysis found Millennials exhibit no difference in “preferences for vehicle ownership” and in fact drive more miles per year than Boomers. As for Gen Zs, the share of cars bought by that cohort has increased five-fold in the past five years.
Finally, to finish framing why cars matter, let’s consider what used to be called telecommuting, Zooming and remote working, especially following the epic exodus caused by the destructive Lockdowns of 2020. Surveys show the lockdowns accelerated a trend that was already underway, one of a huge swath of Americans moving to suburbs or rural areas. It’s a trend that invariably leads to a greater need for cars and distances driven.
Now come politicians in a dozen states—and the EPA via a creative exercise of regulatory authority—with plans to to ban the right to purchase a car with an internal combustion engine, the kind of car that 98% of Americans own, and the kind of car that 98% of average-incomed Americans still buy. The goal of the bans is not, we’re told, to deny any citizen the ability to own or afford a useful car. Instead, as everyone knows, it is in service of the goal to cut carbon dioxide emissions by mandating the use of so-called zero-emission electric vehicles. EVs for all. The process of “transitioning” to EVs for everyone, everywhere, we’re also told will be painless because EVs are inevitably taking over the entire automobile market because they are—it is said—simpler, better, and easier to use, and “cleaner.”
And now the Orwellian-named Inflation Reduction Act (IRA) promises a gusher of money to induce that transition. Let’s stipulate the obvious – one doesn’t need subsidies and mandates to convince people and businesses to buy products that are inherently radically better and cheaper.
But, assuming that the inflationary legislation isn’t reversed by a future Congress, the IRA’s push for an energy transition will deploy $2 to $3 trillion, (when it’s fully ‘costed’ to include unfunded mandates and in-perpetuity subsidies), half of which will be for EVs and related infrastructures. You can buy a lot of obeisance with that kind of money. With so much money combined with political mandates and PR momentum, we should not be surprised to find no auto maker dare avoid genuflecting to the grand vision of an all-EV future. But money can’t buy a change in the laws of physics and underlying engineering realities.
The ostensible inevitability, the enthusiasm, the subsidies, and the mandates for EVs are anchored in three claims:
- That EVs will radically reduce global CO2 emissions.
- That EVs are cheaper and easier to fuel because, well, you can “just plug them in,” and
- That EVs will soon be cheaper than conventional cars because they are inherently simpler.
All three of these claims are simply wrong.
Start with the core claim that “they’re simpler.” Yes, conventional cars have complex thermo-mechanical systems. Engines and automatic transmissions are made from hundreds of components, although mated with a very simple fuel system, a tank holding a liquid with a one-moving-part pump. EVs, inversely, have a simple motor made from a few parts. However, the EV fuel tank is a complex electrochemical system made from hundreds, sometimes thousands of parts including a cooling system, sensors, safety systems, and a boatload of power electronics. EVs aren’t simpler, they’re just differently complex.
The illusion of EV simplicity has relevance to the strike underway by the United Autoworkers Union. EVs do not entail less labor to build, but instead shift where the labor takes place. The data show that, overall, while about 80 people are employed per 1,000 conventional cars produced, Tesla, the world’s biggest EV maker—for now—employs about 90 people per 1,000 cars produced per year.
Seem strange? Consider just the labor to make the two different drive trains. Again, take Tesla and specifically its trend-setting Nevada “gigafactory” where public data shows about 8 people are employed per 1,000 EV drivetrains produced – that’s electric motor plus battery. The combined employment at conventional engine and transmission factories is just 4 people per 1,000 drivetrains. That’s the inverse of the EV labor argument. And there’s more to the labor story, realities that have implications for emissions and costs.
Look upstream at the primary materials sent to the factories to fabricate the vehicles. Steel and iron make up about 85% of the weight of a conventional car. That upstream supply chain requires less than one person per 1,000 vehicles produced. Meanwhile, most of the weight of an EV is found in more exotic, so-called energy minerals, from copper and aluminum to, obviously, lithium, and also nickel, cobalt, manganese, and rare earths like neodymium. That upstream supply chain employs roughly 30 people for every 1,000 EVs. Of course, all that labor is elsewhere since the mines and refineries are not in America.
But before turning to cost and emissions implications of the upstream realities, we need to address the claim that EVs are simpler to fuel.
It’s obvious that the imagined all-EV future requires on-road fast-charging. First, the total labor to deliver the same energy to EV fueling stations is greater than it is for the gasoline infrastructure… something that will necessarily, ultimately impact costs. But setting that aside (and that’s a lot to set aside), the lie of the simpler-to-fuel is in the nature of electrical engineering for fast-charging batteries. The so-called superchargers offer, instead of overnight fueling, 80% charge in 30 to 40 minutes. This is only fast if it’s not compared to the 3 to 4 minutes it takes to fill up a gasoline tank. Long refueling times will translate into long lines at EV fueling stations as well as the need for five to 10 times more charging ports than fuel pumps.
That won’t be convenient, simple, or cheap. Each supercharger costs two to three times more than a gasoline pump. And, because superchargers necessarily operate at 100 times the power level of an overnight home-charger, that translates into staggering requirements for grid infrastructure upgrades. Today roadside fuel stations have the electric demand of a 7-Eleven; but convert those to EV fueling station and every one of them will have the electric demand of a steel mill – and highways will need thousands of them.
Enthusiasts are either unaware of or profoundly naïve about the time and cost challenges of all that. The naivety extends in particular regarding the materials demands for the quantities of copper needed for all the wires and transformers that will be required to replace cheap steel pipes and tanks. And the metal demands of the electric infrastructure will necessarily be piled on top of an unprecedented increase in demand for metals and minerals to fabricate the EVs.
While copper is the long pole in the tent, it is only one of the mineral challenges. The realities of costs and emissions for EVs is dominated by a simple fact: a typical EV battery weighs about 1,000 pounds to replace the fuel, and the tank weighing together under 100 pounds. That half-ton battery is made from a wide range of minerals including copper, nickel, aluminum, graphite, cobalt, manganese, and of course, lithium. And to get the materials to fabricate that half-ton battery requires digging up and processing some 250 tons of the earth somewhere on the planet. Those numbers, it’s important understand, are roughly the same no matter what the specific battery chemical formulation is, whether it’s lithium nickel manganese, or the popularly cited lithium iron phosphate.
And yes, that EV tonnage should be compared to the combined tonnage of metals and the weight of the oil used to produce and fuel a conventional car. Even if you compare those numbers, over a 10-year lifespan of both kinds of cars the EV still entails a ten-fold greater extraction and handling of materials from the earth, and far, far more acreage of land disturbed and, unfortunately, often polluted.
The astronomical quantity of materials needed for EVs has led proponents to claim that there are, after all, enough minerals on the planet and there’s nothing to worry about. And anyway, they say, we can recycle to reduce the monumental materials requirements. But recycling will be irrelevant for a long time since manufacturers claim EV batteries will last a decade. That means there won’t be anything significant available to begin recycling until the early 2030s, long after the world has had to face up to the biggest expansion of global mining in history.
As for the underlying geological resources to supply the suite of energy minerals: Of course there are enough of all those on planet Earth, and even in America. That’s irrelevant. What’s relevant is that the data show that, overall, the mines operating and planned can’t supply even a small fraction of the 400% to 7000% increase in demand for minerals that will be needed within a decade to meet the ban-the-engine goals. What’s relevant is that the IEA has told us we’ll need hundreds of new mega-mines, and that it takes 10 to 16 years to find, plan and open a new mine. You can, as they say, do the math on that.
The need to supply astonishing quantities of battery materials is also where we find the core problem with the claims for big EV emissions and cost savings.
Since over 70% of the price of an EV battery comes from the costs of just buying the basic materials needed – that means the future price of EVs is dominated by the future costs of those basic materials and is dependent on guesses about the future of foreign mining and minerals industries, not the labor and automation prowess at domestic assembly plants. Over the past half-dozen years, the often-cited long-run, rapid decline in battery costs has slowed dramatically. And now prices have increased some 20% since 2021. So far, that’s with EVs still under 10% of total vehicle production. We’re still in early days of minerals demands.
And it’s with the acquisition of key materials where we find flaws with the core claims for emissions. The energy used and thus emissions from producing a pound of copper, nickel, and aluminum, for example, is two to three times greater than for steel. The numbers are higher for the other minerals. Importantly, as researchers at the Argonne National Labs have pointed out, relevant data for all the battery materials “remain meager to nonexistent, forcing researchers to resort to engineering calculations or approximations.”
That means every emissions claim is a rough estimate or an outright guess based on averages, approximations, assumptions, or aspirations. There are huge variables and uncertainties in the emissions from energy-intensive mining and the processing of minerals used to make EV batteries. The simple fact is that no one knows how much CO2 emissions will decline as materials production rises to build more EVs. And all of the key variables point to higher, not lower, emissions in the future.
The energy used to obtain a pound of metal depends on the size, nature, and location of the mine. For copper, that number can vary at least two-fold, and for nickel by three-fold. Getting accurate information is complicated by the fact that 80%–90% of relevant minerals are mined and refined outside the U.S. and E.U. and will be for a long time regardless of subsidies. And, since China refines 50%–90% of the world’s suite of energy minerals for EVs, it’s relevant that its grid is two-thirds coal-fired—and will be for a long time.
There is a dishonesty at the center of all the facile claims about big EV emissions reductions. In fact, nearly all studies making emissions claims are worse than guesses, the estimates frequently cherry-pick low numbers for what’s really happening upstream. A meta-study of 50 different technical studies found the estimates of emissions varied by over 300%. And, worse, that analysis exposed the fact that most emissions claims were based on assuming use of a small 30 kWh battery. That’s one-third the size of batteries actually used in most EVs. Triple the battery size and you triple the upstream emissions – and you triple the demand and thus price-pressure for the minerals.
Upstream minerals emissions not only offset the savings from not burning gasoline but, as the demand for battery minerals explodes, the net emissions savings shrink and could even vanish. Reasonable, even likely scenarios will lead to EVs causing a net increase in global emissions. Geologists have long documented that ore grades have been and will continue declining. That’s because global ore grades are declining – for the non-cognoscenti, that means for each new ton of mineral there’s a steady and unavoidable increase in the quantity of rock dug up and processed. A decrease of just 0.4% in copper ore grade will require seven times more energy to access the copper.
Unlike cars with internal combustion engines, it’s impossible to measure an EV’s CO2 emissions. And, unlike cars where those emissions are the same wherever or whenever the car is made or used, EV emissions vary wildly depending on how it’s made and where it’s used. While, self-evidently, there are no emissions while driving an EV, emissions occur elsewhere—not only upstream before the first mile is ever driven, but also when the vehicle is parked to refuel. The latter, emissions from the grid, is also far more complex than simplistic forecasts assume. Real-world emissions from charging depend on precisely where and exactly when it’s done. The refuel emissions can vary from near zero on a sunny day in some states and, in other states and times, to the same or more than would have come from burning gasoline.
None of this means that the lithium battery doesn’t deserve a pivotal place in history. Its invention was consequential for many reasons, nearly all of which have little to do with EVs. Nor do the inconvenient complexities of mining and grids mean there won’t be many more EVs in the future. Today’s fleet of nearly 20 million EVs globally—notably half of which are in coal-burning China—will doubtless balloon to hundreds of millions of EVs on global roads in the coming couple of decades. But even such dramatic growth would mean that EVs would by then account for barely 15% of all consumer vehicles, and far tinier share of industrial and commercial vehicles.
By way of analogy, the future of EVs in land transportation ecosystems will end up echoing, in market share terms and for similar operational reasons, the role of helicopters in aviation. Helicopters offer very different and, in many applications, far more useful even essential features compared to conventional fixed-wing aircraft. That’s why there’s a very significant $60 billion a year global market for helicopters. Even so, helicopters are only 15% of the overall global aircraft market. While helicopters, like EVs, are useful for a large number of applications, one would no more expect all aviation to use helicopters than for all drivers to use EVs.
The realities of physics and engineering mean that politicians pushing for an all-EV future run a high risk. Quite aside from the eventual discovery that EVs will disappoint with only a tiny impact on global CO2 emissions, the bigger impacts will come as consumers find vehicle ownership costs and inconveniences both escalating. That will lead to unhappy voters motivated by the key underlying reality: A car is the single most expensive and critical product used by the overwhelming majority of citizens.
Mark P. Mills is a Senior Fellow at the Manhattan Institute, author of the 2023 report, Electric Vehicles for Everyone? The Impossible Dream, and author of the 2021 book (Encounter Books), The Cloud Revolution: How the Convergence of New Technologies Will Unleash the Next Economic Boom and A Roaring 2020s. Mark is also part of RealClearEnergy‘s BrainTrust.