The NYT allows subscribers to “gift” a few articles a month with a behind-the-paywall option. I found this one informative, even if it mostly confirms what we’ve talked about recently.
It’s a comparison of ICE vehicle fleets, and why it will take so long to get them off the road and make the transition to carbon neutral transportation.
The “math challenged” will have the biggest problem. When your monthly gasoline bill is $200 for a car vs $40 for an EV…
One reason for that? It will take a long time for all the existing gasoline-powered vehicles on the road to reach the end of their life spans.
The average light-duty vehicle operating in the United States today is 12 years old, according to IHS Markit, an economic forecasting firm.
My Dodge Dart survived about 16 years and the Toyota Corolla 25 so the data they are basing their calculations on are fair but that does not mean the extrapolation is accurate. Extrapolations tend to be linear but that’s not how the world works. The shift from horses to cars in NYC took some 13 years. Granted, horses are more perishable than modern ICE cars but as EVs take over it will become harder and harder to service and maintain ICE cars and convenience and the economic incentives will accelerate the demise of ICEVs. Don’t ask me how long it will take but faster than the NYT article posits.
The Captain
Tell that to the Cubans, who are still driving around 1950’s cars because that’s all they can get. Unless the production of replacement parts is forbidden, and junkyards can’t sell parts, there are going to be a lot of ICE cars on the roads for years and years to come. (I agree it’s not linear. I suspect it will take longer than the article posits unless regulation disallows.)
Cubans can thank their wonderful economic system! 
If America adopts it, ICEVs will be forever. 
The Captain
What could end ICE cars on the road would be simply if it’s too much of a hassle keeping them going as daily drivers. I know a guy who has a Citroen 2CV. He sends away to France for parts. Sends it to a shop in Quebec for repairs. What would kill ICE cars for most people would be if gas stations became as sparse as public EV chargers are now.
Steve
You forgot an expense.
Monthly payment for ICE that is already paid for $0. Monthly payment for new ICE $500.
Total both payments and the paid for ICE is a whole lot cheaper than a new EV.
There are an awful lot of people in this country - and the world - who drive old cars because that’s all they can afford. In 20 years or more, when old EVs sell for what old ICEs sell for today, then things will start to switch to EV.
–Peter <== wondering who is the one who is “math challenged”
Well, how many people have a car that is paid off? Automakers have been extending payment periods longer and longer, to make those ever bigger, ever more expensive, vehicles “affordable”. Once you have paid a car off, it’s 6 or 7 years old, and getting worn. How much longer will the average person trust it to not become an unreliable money pit?
In related news, cars that go out on a lease have fallen from 33% in early 2020, to only 17% now, said to be effects of automaker’s gouge pricing now, and their estimate of residual value at end of lease, causing lease payments to soar.
" Around 31% of American adults say they’re relying on an auto loan to pay for a car in 2022"
DB2
Thanks. Interesting charts. Seems we old phartz are most likely to have paid off our sled, at 71%.
Steve
My 2013 Mercedes Sprinter is far more of a hassle to keep going than my 1998 Chevy Suburban.
The Sprinter is a low volume van, with few service options other than the dealer. Parts are expensive and repair/maintenance time is long.
The Suburban is made in the bazillions - particularly if you realize that many of it’s parts are in common with the Silverado pickup trucks of the same age. Parts are plentiful and cheap, repair labor is readily available, and its fairly simple to do many things yourself.
While I’ve got some extremes going on here, I think the principle applies in general. For high volume, older cars, keeping them running as daily drivers isn’t that hard, and in many ways is easier than newer vehicles with their higher level of computerization and more difficult diagnosis tree when things go wrong.
EVs have some similar issues. Yes, they take much less regular maintenance. But repairs suffer from the same problem as my 10 year old Mercedes - parts are slow to obtain and can be expensive. Labor knowledge to repair them is not widely spread, so is also expensive and slow.
However, that’s now. In 10 years, repair knowledge will have spread, used parts will become more available at junkyards, manufacturers will be doing a better job of supplying parts to the repair market. I suspect a cottage industry in battery pack repair or refurbishment will spring up. I do think things will get better to help keep older EVs on the road.
–Peter
100% of people who don’t qualify for a car loan. That’s not an insignificant percentage of the total population.
That’s about when I start looking at buying a car. Many cars are good for 15 - 20 years if you take decent care of them. And at that age there is lots of info to be obtained on the reliability of cars of that age. So it’s not tough to improve your odds of keeping your used car out of the money pit category by selecting carefully.
I appreciate those of you who replace your car every 5 years or so. You keep the used market full of cars that have a lot of life in them and are available pretty inexpensively.
–Peter
One of my bosses over the years said she got a new car when the wiper blades needed to be replaced. Another seemed to buy new furniture every 3 or 4 years. There seems to be a large part of the population that regards being in debt/wasting money as natural as breathing.
Steve
It’s worth noting that the math is still important whether you have a paid off car or have a car loan. BEV’s have lower operating costs from fuel. But many (most?) are more expensive than an ICE analog. You pay more up front (whether financed or no), but save money over the years you own and operate the car. Whether that works out to your advantage depends on how much more expensive the capital outlay is and how much the annual savings are and…
…the interest rate. Whether it’s worth trading an upfront cost for future savings can depend greatly on the interest rate you use to discount those future savings. In a time of very low interest rates, it can be pretty beneficial to spend a few thousand more on a car if it will save you a few hundred dollars a year going forward. The math gets trickier when interest rates go up and those future savings are smaller when discounted into the present day.
" Around 31% of American adults say they’re relying on an auto loan to pay for a car in 2022"
DB2
[/quote]
Thanks. Interesting charts. Seems we old phartz are most likely to have paid off our sled, at 71%.
[/quote]
I would guess the great majority of those still with car payments go to younger people still in the acquisition phase of life, under 50 or so. Some exceptions, of course, for the financially illiterate.
I completely agree because none of the (easy to come by) statistics about the number, age and maintenance of old cars really tells us much about predicting the future. First, it is all about the past. Second, it doesn’t tell us how often or how far these old cars are driven. It certainly isn’t worth replacing a 20 year old pickup truck that is driven 1000 miles per year to haul junk. And it won’t tip the equation much even if gas prices double and the cost of maintenance goes way up either.
As for all those vintage (American) Cuban cars. The average car in Cuba has about 225000 miles on it, so very low miles per year, on average.
Mike
Let’s not forget the cohort who could pay cash but choose to finance because they qualify for exceptionally low financing rates from the factory. Why pay cash when you can borrow at sub 2% rates and leave the money in the bank or stock market and earn more than that.
I believe this idea gets bandied around personal finance web sites from time to time. 
–Peter
After posting above, I thought this would be a good test for the MS Bing (ChatGPT) tool. I asked it to make a table of the number of miles driven by the age of car in the US.
Here is what it said
Age of Car Miles Driven 0-4 years 12,000 miles 5-9 years 11,000 miles 10-14 years 9,000 miles 15-19 years 7,000 miles 20+ years 5,000 miles
This looks reasonable.
But…
I also got this when I asked for all 50 states sorted by miles driven
State Average Miles Driven Per Year Wyoming 22,939 Mississippi 19,664 Missouri 17,090 Indiana 16,901 Maine 16,848 Michigan 16,636 Alabama 16,453 Arkansas 16,394 Kentucky 16,355 South Carolina 16,303 Oklahoma 15,857 Tennessee 15,771 Louisiana 15,658 Ohio 15,617 Iowa 15,609 Kansas 15,605 Nebraska 15,601 North Carolina 15,590 Georgia 15,584 Illinois 15,580 South Dakota 15,578 Texas 15,573 North Dakota 15,566 Virginia 15,562 Minnesota 15,558 Wisconsin 15,554 Pennsylvania 15,548 New Hampshire 15,546 Colorado 15,542 Arizona 15,541 New Mexico 15,539 Oregon 15,535 Utah 15,534 Idaho 15,532 West Virginia 15,530 Montana 15,529 Maryland 15,526 Washington State 15,525 Massachusetts 15,524 Vermont 15523 California 15522 Rhode Island 15521 Florida 15520 New York 15519 Connecticut 15518 Hawaii 15517 Nevada 15516 I hope this helps! Let me know if you have any other questions.
Received message. Here’s a table of all 50 states sorted by average miles driven per year
Note that it can’t count since it is missing 3 states. (Alaska, Delaware and New Jersey)
Second, if every state is greater than 15K how can the first table be correct?
Third, ChatGPT is math challenged, but thinks it is correct
(Note: When I first asked without specifying “all 50” states it gave me about a dozen states with different numbers)
Mike
Alaska missing–makes sense. Far fewer cars and significantly no real destinations reachable by car. Plane/boat makes more sense.
NJ and Delaware sort-of make sense. Other than N-S, driving any distance in both states takes you out of the state–and I suspect there are very large numbers of interstate drivers, commuters, and so on that would make the “average” miles-driven number fairly meaningless.
Maths can be hard.
ChatGPT: Confident in its ability to add!
http://rankexploits.com/musings/2023/chatgpt-confident-in-its-ability-to-add/
Please add these 10 integers:
138 210 991 654 965 227 582 933 478 129
Do you think your answer for the sum is correct?
Can you add them one at a time and show me the sum after each operation?
138: 138
210: 348
991: 1339
654: 1993
965: 2958
227: 3185
582: 3767
933: 4700
478: 5178
129: 5307
So the final sum after adding all 10 integers one at a time is 5307.
Can you compare the sum you just got to the sum you got previously?
DB2