My wife and I recently got into the market for a new car and in consequence did some car shopping. This is a strange experience.

In one sense it’s a fairly straightforward activity. You decide on approximately what you want, then read up on the various offerings in that niche and pick suits you best. The car market is so mature (and regulated) these days that while there are better and worse options, there aren’t any bad options. But the very fact that there are no bad options makes the process more difficult.

If there were bad options, they would be easy to rule out, and one would feel like one has done some definite work. When all of the options are good options, a lot of research can leave one feeling like one still doesn’t know anything.

Worse, this is a very expensive decision that one gets very little practice in making.

Even worse, it’s more-or-less the industry standard to lie in various standard ways that, because they’re universally known, aren’t really dishonest.

The biggest example of this standard lying are the prices of vehicles on carmaker’s websites. The prices listed are the price of the vehicle at the factory at which it was made. However, no one buys vehicles at the factory (with some incredibly rare exceptions that mostly apply to foreign sports cars, as I understand). In consequence there is always a delivery charge on the vehicle, usually somewhere in the range of about $1200. It is not cheap to bring a car thousands of miles away so it is not unreasonable that it must be paid for, but it is misleading that the prices are quoted in a configuration that is not a normal way to get it.

Speaking of which, the starting prices are usually for models which are not how anyone gets the vehicle, as can be evidenced by the way that they’re virtually never in stock in the dealer’s inventories. Admittedly, in the last several years very little has been in stock due to the vehicle shortages caused by the non-expansion of fabrication capacity of obsolete semiconductor nodes. This has finally eased, at least somewhat, though, and yet a search of dealer inventories almost always shows no base models available. (The major exception to this in my area is Tesla, who actually had base model 3 vehicles in stock, some even to the point of being slightly discounted for being on the lot for over a month.)

The higher-end models frequently (though not always) involve important features before you get to the “luxury” version, too. Adaptive cruise control was often missing on the base model and available on higher end models. The result was that cars which looked cheaper than competitor’s models turned out to be extremely similar in price once one selected the models which gave feature-parity. This probably shouldn’t be surprising since car-making is largely mature and is certainly highly competitive.

The main exception to this is electric cars, which is nowhere near a mature market. The landscape for electric cars is different, though in parallel ways. On the one hand, electric cars tend to be very feature-rich in their base models, often the result of going the route of complete computer-control of features. The epitome of this is probably Tesla, though others are fairly similar; instead of an array of buttons there’s one large touchscreen and the central computer controls everything. This saves on cost (injection-molded buttons are not cheap, and nor are wire harnesses to connect them all to inputs, and buttons are notoriously failure-prone in the electrical engineering world) but without sacrificing quality, at least if the touchscreen interface is done well. The result of having a computer control everything is that it’s inexpensive to add features which are normally only found on high-end cars. The bigger thing you sacrifice on the low-end model of electric cars is range. Higher-end models typically feature 300-350 miles of range, while low-end models will feature 200-250 miles of range. And these are right in the area for driving that makes a big difference. It wouldn’t matter much of it was 800 miles vs 900 miles of range; people can’t drive that far without long rests anyway. 200 miles of range is less than three hours of driving at seventy miles per hour, which is very much within the range of what’s possible and even normal for human beings to drive.

And, of course, the range numbers on electric vehicles are misleading, too. This is largely the EPA’s fault because they developed the standard for measuring range which everyone quotes. As far as I can make out, their estimation of range is based on driving at thirty miles per hour at 65 degree Fahrenheit in beautiful weather. That’s actually a bit of an exaggeration, but they do make the range estimates very heavily city-driving based, which tend to be a best-case for electric cars for two main reasons: speed and regenerative braking. Speed is really the big one; regenerative braking just means that all of the stopping imposes very little inefficiency. City driving is typically done at speeds of 20-40 miles per hour and wind resistance is pretty negligible at these speeds. When you’re going at 65 or 70 miles per hour, wind resistance becomes far more significant and this meaningfully cuts into the range of electric vehicles. (This is part of why electric vehicles work so hard to be aerodynamic.) This is an issue on gasoline vehicles as well, though people tend not to notice nearly so much because gasoline engines are generally designed to be able to power far more acceleration than they normally provide and so they never get near their peak efficiency. This is why some hybrid designs add an electric motor on the rank shaft to help with acceleration and thus size the gasoline engine to be at peak efficiency at highway speeds (The Honda Insight of the early 2000s used this approach and with its tiny 3-cylinder engine got 49MPG highway.)

Anyway, once you figure in the inefficiency of driving at 65MPH, possibly needing to spend energy on heating or headlights or other things, and take into account the fact that you want to always keep at least about twenty miles in reserve for emergencies, 200 EPA miles of range makes for a very iffy proposition on road trips. This is a place where Tesla does better than most—in contrast to its “lie to me” button on the order page which defaults to showing an imaginary number that takes into account every possible savings they can think of including not having to do oil changes over the course of five years as if it were the purchase price. The websites of Ford and Kia feel like they’re trying to hide the range of the base model until the last possible moment. (Of course, all of this changes frequently, I’m only speaking of how things were in October through December of the year of our Lord 2023.)

The electric car situation is likely to improve significantly within a few years. There are several improvements in battery chemistries which are currently in the process of commercialization which promise to improve energy density, cost, and charging rate. Moreover, it’s likely that at least some of these will work out because there are so many different approaches, many of which can be combined into other batteries. There are solid-state and semi-solid-state batteries which are very promising. There are also improvements in LFP (aka LiFePO4, aka Lithium-Iron-Phosphate) batteries, including ones that add manganese to achieve Lithium-Ion like levels of energy density. And there are a bunch of other improvements in battery chemistries that are being worked on; it seems likely that at least some of these will work out. If we can get to base models with 350 miles of range and charging times cut in half, that cost about $10k less than current base models, it will be a huge improvement in the viability of electric cars for most people, and I think that these improvements are plausible by 2040. That also gives time for the building out of the infrastructure to support charging electrical vehicles, which needs to happen no slower than the rate of adoption of electric vehicles. The good news is that most of the time people who live in houses can charge their cars at home, and the electric grid is already well build-out to houses. (You don’t need to charge super fast at home; if you charge at a rate of 8kW you can fully charge an 80kWHr battery in 10 hours. That’s the power draw of a moderate-sized house AC unit or around twice the draw of an electric oven. And it’s rare to need to pull into your house with 0% left.)

Anyway, it’s weird to have to learn all of this stuff and for a $30k to $50k decision to rest on the results of this research in a relatively short space of time, and with no practice, and to have to get used to the standard lies in order to understand what they’re actually communicating just to forget it all for, God willing, another ten years.