The example of 2 people, one willing to pay $10 for ESPN and $3 for The History Channel and the other willing to pay $3 for ESPN and $10 for The History Channel disregards the likely majority who would pay $10 for the one and $0 for the other.
The problem with bundles is that they force you to pay for things that have zero value to you. And given that the incremental cost of providing the product approaches zero, this is strongly in the seller's interest.
However, the idea that a company can force you to pay for things with zero value is bunk. If you value ESPN at $10 and History channel at $0, then they should charge you $10 regardless of bundling, because that is the profit maximizing price.
(Note I'm not claiming the bundling prevents purchase, just that there is an economic opportunity in non-bundled alternatives, such as Windows boxes without crapware, or solving the problem of a clogged experience due to poor bundling such as Netflix's focus on surfacing desirable content.)
Imagine being an English only speaker and getting a Spanish channel - what would you pay for it?
Fallacy 2: There is no relationship between the act of bundling and the happiness of consumers.
Fallacy 3: The additional money saved by consumers if the goods were not bundled would not be allocated to higher happiness inducing products.
And so on.
There's an argument to be made regarding the paradox of choice and the inconvenience of micro-payment systems, but I don't find this one very convincing.
People don't want to pay for something that has no value to them but they will accept features that have no value if they feel they are included in a total price of what they are buying (except of course when they can compare products and make a decision not to buy a product because a competitive one is cheaper etc.)
Of course even if the existing market bundles, you may have little choice. Your minimum viable product pretty much by definition cannot be a viable bundle.
If all my favorite shows were available on ITunes, I'd kiss cable goodbye in an instant.
All the author is saying is that there is a set of assumptions that when met, bundling benefits both consumers and sellers. Specifically, this set of assumptions includes 0 transaction costs. Thus, the decrease in transaction costs does not mean the end of bundling.
Here's another example: 100 channels for $1 each. The average customer chooses 30. Bundling all 100 for $40 would benefit everyone.
Bundling all 100 channels for $40 doesn't benefit everyone. There are a ton of assumptions embedded in that analysis that may, or may not be true. You assume that every channel yields positive marginal utility, for example.
Personally, I would pay for maybe three cable channels. But each additional channel would lower my utility because of the knowledge that I was contributing to its existence (I sleep better at night knowing that I am not contributing to the development of shows about obese people with 14 children, etc.). Thus, I do not have cable at all.
You also assume that each channel costs the same to provide. Under an a-la-carte pricing scheme the prices would be all over the map because cable providers couldn't afford to offer certain channels for $1. So "the average customer chooses 30" would become a bit more complicated.
There are also probably psychological issues in play. For instance, why aren't HBO and Showtime part of "basic" cable? Surely they could make more money through bundling! Starz was included (bundled) with Hulu (or was it Netflix?) for awhile, but it was removed because the network feared losing its image as a "premium" brand. I suspect HBO feels the same way.
Beyond this, a television channel is a bundle in itself. You get a set of shows (or even episodes) and movies that you can't alter. So even if channels were un-bundled, the bundling would just move from cable providers to networks themselves, or to other intermediaries such as Hulu.
