[seadan83]

Ah, I think i just realized how to fix the analogy!

The issue is with how many books are checked out a time. If the line is absurdly long, at some point you will make fewer trips to the library to avoid paying the cost of waiting in line. You would check out more books so you would go less frequently. You would be trading storage space at home in exchange for time (not having to wait in line). If the line were infinitely fast, then why not go to the library exactly after you have finished one book to then go get one new book.

If an automated checkout then exists, the line time would be less, making it less expensive to go to the library, which means a person would be willing to increase their trip frequency to the library. Suddenly, you have a line full of people all checking out exactly one book, and returning the next day to do the same thing again (rather than checking out one weeks worth of books, and coming back a week later instead of the next day).

[thereisnospork]

I think that's a fair extension of the analogy, from which I would make two observations:

First is that with frictionless checkout the library's efficiency is maximized (books are only checked out when being read and people's time isn't spent unproductively).

Second is that there is a limit to the demand of the library. A book will be finished before checking out a new one and a person can only read so many books a day. No matter how entertaining there is a fairly hard cap based on a persons need to sleep and reading rate. So a library would only ever need enough automated checkout lanes to match the populations awake time and reading rate before book demand is fully sated.

[seadan83]

I really appreciate the dialog! I find your observations interesting.

Though, I do think you might be overemphasizing the number of checkout lanes.

To torture the analogy, let's consider the variables at play:

- how often a person goes to the library

- how many books they check out at a time

- how long they wait to checkout (we can potentially include travel time with this number, and this number is a function of the number of checkout lines)

- how many books can a person use at a given time (can a person read 3 books at a time?)

- how fast a person consumes books

- what is the max number of books that can be transported

- what is the max number of books that can be stored

The variables of "how many times do you go to the library" and "how many books do you get each time", and the cost to do so - "the time to check out", I think are the 3 really interesting variables to demonstrate induced & latent demand. The others are a factor, but we can hold them constant in order to demonstrate the relationship of the other three.

Diving in now - latent demand are people who would rather not read at all rather than spend more time than X waiting to check out. If the time to check out decreases, some people will start making the trip - this is latent demand.

As time to check-out decreases further, some people will start making multiple trips instead of just one - this is induced demand. The people making multiple trips are still checking out the same number of books overall - illustrating there is demand for high frequency and not just an absolute number of books. If all you wanted was 'X' many books, why take more than one trip to get them? The frequency provides flexibility and relieves other costs (carry capacity cost & storage cost).

For example, an expedition to antarctica would be very willing to pay a high cost to have a high carry capacity in order to transport a lot of food (they'll buy a literal boat to carry it) and another high cost to store it (storage space is not free). The expedition is willing to pay these costs because the cost to get more food is so high.

One can also ponder, why not get a lifetime supply of books from a library (ignoring late fees)? To one extent, fitting that many books in your arms and then getting them to your home, and then storing them are costs. Why do that when you could make another trip a week later?

Though, let's say for some reason you knew this was your last ever trip to a library. In this case, you would be highly incentivized to invest in higher carry capacity (eg: rent a truck) and to also invest in storage for the books. (Or, the person would be very incentivized to find alternatives). The really interesting part of induced demand IMO is that typically for existing road resources, we are pretty well into the state where many people would prefer to make more trips than they do today. So when we think about the inverse, as travel costs go down, the incentives to carry more and store more go down - which results in more trips being made. As more trips are made by more people, congestion increases, which creates a balancing effect and a steady state of traffic congestion. Naturally, other factors can break the steady state, the fact there is this counter-balancing force is the (IMO) interesting part of induced demand.