After a solid decade of seemingly unstoppable inevitability, The EV Transition (capitalized, so you know we are in the presence of a construct) is having a few issues. No, it’s not that EV sales are falling, or that the global shift toward electrification has lost all momentum, or anything so dire. But here in the United States at least, there’s a correction brewing that should give us pause to reconsider a few basic principles.

I don’t want to get too bogged down in the details of the problem, because it’s not actually that complicated: there are too many premium EVs chasing too few buyers, and all the classic signs of oversupply are building up. Higher inventories and incentives, lower transaction prices, and even more red ink in a segment where anything else was already a rare exception. If this were happening in trucks and SUVs (as it has in the past) the media would have no problem calling it what it is, although even then at least trucks and SUVs have profit margins to give.

More worrying than the prospect of a correction in the US market premium EV segment, which would hardly be an existential threat to the established auto industry, is the reaction from EV supporters. It’s not surprising that a sector defined by Tesla would be riddled with magical thinking, but the denial around the current situation is still shocking. Way too many smart people are trying to pull an Elon Musk here, hoping that high confidence and glib narratives will push the cherished EV Transition narrative past this growing speedbump.

And hell, maybe it will. None of us know for sure. For myself, I certainly do appreciate that The EV Transition narrative is a load bearing construct for anyone hoping that America can make progress on climate goals. At a time when our political unity and policymaking chops seem to be at all-time lows, the alternatives to an organic, consumer-driven replacement of gas cars with battery electrics all feel like fantasy.

Be that as it may, I am firmly committed to the principle that you should learn something from every crisis, and I’ve been thinking a lot about The EV Transition lately. Unfortunately that line of thinking has led me toward (ugh) economics, a field in which I am far from well-practiced these days, particularly questions about how the EV market will develop. One issue in particular seems sufficiently provocative to hack together my crude understanding of it into a blog post, with the understanding that it’s still very much just a theory.

The theory starts with the problem the US EV market currently faces: the oversupply of premium market EVs. Actually that’s only half the problem, the other half of which is the undersupply of affordable EVs. People blame the failure of every affordable EV from the original Nissan Leaf on down on execution, but the reality is that the demand curve for EVs in this country makes them look more like Veblen goods than more typical products.

Of course “more like” is doing a lot of work here, and the issue isn’t actually price at all, as the ongoing red ink-fueled growth proves. In fact, the issue is that the US market for EVs has a capability floor, a minimum viable product line, below which an EV simply won’t sell. The simplest metric for this capability is miles of range, although the charging network is a fuzzier factor as well, and it appears to be somewhere around 250 miles. This minimum level of capability demands a large battery pack, which increases cost, which is why everybody is competing in the premium segment and there is no real competition in low cost EVs.

The problem of the affordable EV is where Transitionheads fall back on a surprising level of faith in technological innovation and efficiencies of scale, which tends to kill off actual discussion. But even if we can drive down the cost of batteries to the point where a 250 mile minimum viable EV is genuinely affordable on a sustainable positive margin basis, there will be weird factors in that market that nobody seems to be even considering. That’s because the entire EV Transition narrative is built on a false equivalency between gas and battery electric cars that no amount of wishing will magic away.

If you strip the two technologies down to first principles, the differences between the two start to become more obvious. The difference between a cheap gas car and an expensive gas car boils down to two factors: creature comforts and power. Things like the size of the gas tank and the refueling experience are utterly commodified, being no different if your car costs $30k or $100k. In stark contrast, performance is more or less commodified in EVs but the size of the battery and the quality of the charging network are central to the difference between a cheap EV and an expensive one.

This distinction is incredibly important given that American consumers have been conditioned to associate cars with freedom and the open road to the tune of billions of dollars every year in advertising for more than a century. The depth of this association, and its resonance with the central themes of our national mythology (individualism, the frontier, etc), should not be underestimated. The idea that we can shift a market for products where freedom is commodified to one where you have to pay more for more freedom without seeing changes in consumer behavior is absurd.

A quick aside on terminology here: the French call a vehicle’s range its “autonomy,” which I think is a useful way to think about this without all the vagueness and baggage of “freedom.” Mobility is the ability to move, but autonomy is the ability to move where you want when you want. Autonomy in this sense is effortless for gas cars of all prices, but it defines the demand curve for EVs. It is literally what you are buying more of as you spend more money on an EV.

In survey after survey, consumers have no problem identifying the combination of range, charging and cost as the fundamental challenges to EV adoption. The problem is that there is no way to solve the range and charging problems without adding to cost. More importantly, the more range you give customers the more they demand. The iron relationship between these factors proves that EVs simply aren’t drop-in replacements for gas cars.

These constraints aren’t obvious in the wealthy early adopter market, where social and other factors blur the line between want and need, but as you work down into the affordable mass market the demand for the autonomy that gas provides becomes more real. Tesla owners aren’t driving three hours to work a service job in a neighborhood they can’t afford to live in, and the people who do have to live like that can’t afford to be sentimental about the value their car provides. The more someone relies on one car to survive, the harder it will be to overcome their “autonomy anxiety.”

Until we reach the point where the marginal cost of batteries is zero, the EV demand curve will always be defined by the autonomy that every gas car gives you for free. EVs more than make up for this with their own “superpowers,” like starting every morning with a full battery, so the point here isn’t that EVs are bad and gas is good. The point is that we’ve been pretending that this transition can happen without a fundamental shift in consumer attitudes and behaviors, and the sooner we accept that we’ve been fooling ourselves about this the better.

There’s a similar problem with driving automation technology: one ride in a Waymo in downtown San Francisco proves the technology is real, but because we don’t have self-driving cars people think it isn’t. The technology is only a failure in the context of a single use case: the car, the ultimate universal mobility easy button. Both driving automation and battery electric drive will require new markets defined by new consumer attitudes in order to thrive, and will look like failures for as long as we compare them to cars.

At a high level, both of these technologies are pushing toward an “unbundling of the car,” something that my friend Horace Dediu has been talking about for a while. No single new technology can replace the totalizing dominance of the gas car, which leaves only a future where we must begin to think about using the right vehicle for each trip. That will take a shift in consumer attitudes which won’t be easy, but at least we will be able to finally stop trying to shove new technological pegs into a car-shaped hole with increasingly frustrating results.