It's demand, all the way down
Cool concept! Might work even better explaining real estate. I'll have to think it through.
When prices go up, isn't the supply (for sale) supposed to increase? But that doesn't seem to work in real estate except maybe over the long run, years.
In the short run, following the double-demand idea, it could be that higher prices increase the demand from current owners. Owners like to hold stuff more when it's increasing in value more, so supply falls instead of increases when prices are increasing fast.
Then combine that with incredibly inelastic stock of houses (versus supply of houses for sale).
Wait, wait, I got it!!
It all depends on how we define the demand for used cars.
If the demand for used cars is ONLY the demand to BUY a used car, then the supply of used cars will decrease.
But if the demand for used cars ALSO includes the shadow demand by people who already own their used car that they could have sold, then the supply of used cars will NOT decrease. In this case, the demand for used cars will increase greatly because it includes two sources of demand - used car buyers and also shadow demand by people who are refraining from selling the used car they were considered selling.
When the price of a new car increases, then current owners of used cars are willing to pay more to hold onto their used cars. The willingness to pay to hold onto the car is an increase in the reservation price, which is a decrease in supply. So we could model the demand for used cars increasing due to two separate reasons. In that case, the supply won't shift. But if we model the demand for used cars as exclusively arising from people who don't own a car, then the shadow demand - the willingness by current owners to pay more to hold onto their car - must be represented as an increase in the reservation price, i.e. a decrease in supply.
In the end, price and quantity of used cars will change by the same amount. The difference is merely how we model it. We can either have a very large increase in demand but no change in supply, or else we have a small change in demand plus a change in supply.
So this actually highlights the fact that supply is just the inverse of demand. If we define demand to include only demand by people who are buying a car they don't own, then supply has to shift as well. But if we define demand to include shadow demand by people who already own their used car and can potentially sell it - or not! - then this demand will increase by more, and this increase in demand will itself capture the decrease in supply.
So it's not double-counting. Rather, it's a question of how to define demand.
Author has changed underlying definitions of demand and incorrectly. Demand presumes you do not have and wish to have. Supply means you already possess it. Therefore, you can’t ‘demand’ something you already possess: if so, pay yourself $100,000 for your own car. Also, a supply curve at the start is near zero…
This may be a little stretched and risks turning economics, like philosophy, into an elegant game played with words.
In Brazil, it is a well established fact that used car prices rose far more than new car's. Makovi above is right! The supply of used cars is on balance stable or declining given that stock is in fact fixed prices rise.
I think you were right the first time: the supply of used cars might shift as well.
The supply of used cars is NOT a fixed vertical line. That's the physical stock, but the physical stock is not economically relevant. Rather, the supply of used cars is the schedule of people who are willing to sell their cars, which is less than the total physical stock in existence.
And as you yourself point out, the supply curve is just an upside down demand curve. When the supply of new cars increases, this increases the price. And this increases the demand for a substitute, viz. used cars.
But since the supply of used cars is itself an upside down demand curve for used cars, an increase in the demand for used cars implies a decrease in supply. At any given price of a used car, people will be less willing to sell it than before, because they know that they will have to go buy a replacement car (new or used) that is more expensive than it used to be.
But it's hard to say how much the supply of used cars will shift. It depends on relative elasticities and expectations.
For example, suppose the price of new cars increases by $5,000. It's possible that the price of used cars - after all demand and supply shifts are completed - will also increase by precisely $5,000, so that sellers of used cars will be completely compensated. The seller makes a $5,000 gain upon sale that is immediately dissipated by a $5,000 loss upon purchase.
Alternatively, if the demand for a new car is relatively inelastic, then the seller of a used car might be willing to take a loss on the sale of their used car, because they want a new car so badly. In that case, the supply of used cars will decrease less, and the price of used cars - after all shifts are completed - will increase by less than the price of new cars.
And the opposite is true if the demand for used cars is relatively inelastic. In that case, the price of used cars might increase by more than the price of new cars.
And it probably also depends on people's expectations for what future prices are going to be. Someone who has multiple cars might be willing to sell one today with the expectation of buying another one in the future, or they might be willing to buy a car today with the expectation of selling their current car in the future. (I am doing that myself. I just bought a new car, while the sale of my used car is still pending, while I wait for the title to arrive in the mail.) Depending on the expectations of the future, the supply of used cars may shift more or less.
Ordinarily, when the price of one good (good A) increases, then only the demand for its substitutes (good B) increases, but the supply of its substitutes does not change. I think this is because the suppliers of good B are not trying to turn around and buy good A. They just want to earn their opportunity cost. So let's say the price of hot dogs increases because of a decrease in the supply of hot dogs due to a problem manufacturing hot dog making machines. (Crucially, the decrease in supply of hot dogs is due to a decrease in the supply of a specific factor of production that is not used in the substitutes for hot dogs.) Therefore, the demand for hamburgers increases. But the supply of hamburgers will NOT decrease. It is not as if hamburger makers are contemplating exiting the hamburger industry and entering the hot dog industry. Nor will they sell their hamburgers in order to buy hot dogs, the way someone sells their used car and buys a new car. It is particularly implausible to imagine that hamburger makers are going to exit the hamburger industry and enter the hot dog industry because by assumption, the problem in the hot dog industry is a problem with a specific factor of production that has technologically reduced the supply. So the hamburger makers will not change their supply. They are relatively fixed in their industry - at least in the short run - so they just want to earn their opportunity cost, which may simply be their marginal/variable costs of production. So in this scenario - when the price of hot dogs increases due to a decrease in the supply of a hot dog-specific factor of production - only the demand for hamburgers shifts, but the supply does not.
And even if the hamburger makers COULD enter the hot dog industry, the effect would probably be to restore the supply of hot dogs closer to its original level, not to decrease the supply of hamburgers. Maybe it could be more complicated if we consider complement factors of production. E.g., the hamburger maker has a fixed workforce, so if they buy hot dog making machines, they have to reallocate some labor from the hamburger machines over to the hot dog machines. That would mean that increasing the supply of hot dogs requires decreasing the supply of hamburgers. But then we have to consider whether the hamburger makers could just hire workers away from the hot dog makers. The hamburger makers could increase their output of hot dogs without decreasing their supply of hamburgers by hiring newly-unemployed workers from the hot dog makers. So whether the supply of hamburgers will shift becomes really, really complicated.
But in the case of used cars, I think it's more clear-cut. When the price of a new car increases, it's not merely that more people want to buy a used car. It's also that fewer people want to sell their used car because they know that they will have to go buy a new car themselves. That's very different from the case of hot dogs versus hamburgers, where it isn't so obvious that an increase in the price of hot dogs makes hamburger makers raise their reservation price out of greater reluctance to sell hamburgers than before.