Steven Landsburg figures out why some hotels charge for internet access and others don't.
The goal, then, is to use amenities to equalize your customers' willingness to pay. Take a stripped-down example: You've got two customers, Jack and Jill. Jack will pay $100 for a room without net access; Jill will pay $120. That poses a pricing problem. But suppose you happen to know that Jack will also pay an extra $20 for net access, while Jill thinks computers are instruments of the devil. Now the pricing problem is solved: Make net access free. Both customers will pay exactly $120 for the room-plus-net package, and that's what you charge.
But he still hasn't figured out why popcorn is so expensive at the movies.
In every important respect, the popcorn pricing problem is identical to the Internet pricing problem. If Jack will pay $10 to see the movie and Jill will pay $12, but Jack really loves popcorn, then a greedy profit-maximizing theater owner will offer to shower Jack with free popcorn until he's willing to pay $12 to get in.
But if Jack and Jill are each willing to pay $12 to see the movie, that same greedy profit-maximizing owner will charge them both $12 and bleed Jack dry at the popcorn stand.
Sometimes the numbers should work out one way; sometimes the other. Yet in the real world, popcorn, unlike wireless Internet, is never free.
It's logically possible that by pure coincidence the numbers at every movie theater in the world all work out the same way, while the numbers at hotels work out one way half the time and the other way the other half. But "pure coincidence" theory is even less satisfying than the "differential greed" theory. There must be something I'm missing that makes popcorn essentially different from Internet access. I remain stumped.
UPDATE: Landsburg also mentions movie popcorn in this article.
Not even a consideration of marginal cost? Tsk, tsk.
The marginal cost of allowing you to use a local internet connection is essentially zero. It's mostly fixed costs. Willingness to pay is surely more important for pricing in this case.
As for popcorn, the marginal cost is surely small relative to the price. I mean, how much would they have to raise the ticket price to cover the cost of a bag per patron?
Clearly movie theaters are targeting consumers with inelastic demand. But as Landsburg writes in the article linked in the update above, "Why should popcorn eaters be less price-sensitive than the rest of us? I don't know the answer to that riddle, but there is a lot to be learned by exploring it."
So naturally there is no objection to offering free internet access. Would you like a bed with that room? Popcorn on the other hand has a small but finite cost and moreover competes with other food offerings on sale. Welcome to the buffetorama free movie.
BTW, the real expense of popcorn is probably cleaning up after it.
"BTW, the real expense of popcorn is probably cleaning up after it."
Very true.
And you raise a good point about bundling in general. I would argue that the bed is included because it's costly to move in and out of the room at the whim of whether people would choose to pay extra for it. (That and social convention, I suppose.) But note what type of things hotels do charge extra for... the mini-bar, for one. There is controlled access, the goods are consumed in a way that the bed or internet access is not "consumed", and the items have a significant marginal cost.
I would suggest that hotels would be more likely to charge extra for goods and services that have one or more of those characteristics.
Not to mention the fact that different patrons will have different elasticities of demand for the mini-bar! (But probably similar elasticities of demand for the bed!)