John McCain has introduced a bill to “encourage the wholesale and retail unbundling of programming by distributors and programmers.” Would a la carte pricing result in lower prices and greater consumer welfare or would it raise prices and result in less investment in television media? Time to take a look at the economics of bundling. In this video from our MRUniversity course on media economics I review the theory of bundling and then apply it to cable TV.


Even assuming McCain forces providers to offer a la carte channels he cannot set the price, can he?

What if their rate-card says $50 for the 100-channel bundle or $45 for each. Is there a bill clause that prevents this? Or is this just yet another toothless law?

I had a similar thought. Like eBooks or music, cable companies could artificially "flatten" prices between the popular and unpopular channels as a way to discourage undercutting and to protect high value channels. Perhaps some sort of "common carrier" framework could avoid this, but I think you'd need an auction to make something like that work. If a cable had 200 slots, it should just sell them, and then those slot-owning channels would in turn set their own price with the consumer.

That common carrier and channel controlled price structure is pretty much what I have with Rokus, and a data only connection, I think.

Does McCain's bill have anything in it that prevents providers from declaring "$50 for the bundle or $40 each"?

If not, it is somewhat toothless.

Do you think that anyone would buy either if that was the choice?

For cable we pay one price for a bundle of product. For mobile phone service we pay a bundle of prices for one product. Call your cell phone company and ask them to make a single change to your plan. It's easier to predict the weather than the total cost of cell phone service. This bill is dinosaurs regulating dinosaurs. Let's talk about something relevant for today. What are the economics of distraction?

All true as long as the consumer does not value all but one products of the bundle at 0. If I only ant CNN the bundel forces me to pay extra.

Some posts from a few years ago by Matt Rognlie on the subject:

Several thoughts on this. Cable channels are themselves bundles. Regardless of the granularity of the bundling, you buy something you don't want. However, all of this is irrelevant to my children - they bypass the channels completely. Many of the channels are available in some way online. If you are willing to wait a while, almost everything is available for free. Or, for a small price, sites like Hulu or Amazon make programming available (the Spotify model).

What about physics?

McCain wants ever more Federal rules & dictates (..what a surprise).

Core problem was the grant of exclusive monopoly franchises to cable-TV companies, by local governments years ago.
That sorry legacy continues to harm most consumers.

Less than 10% of American consumers have any real choice in local cable providers. That lack of competition prompts price gouging.

Local governments greatly interfered with the development and distribution of cable-TV ... ultimately harming their citizens & taxpayers.
The "brilliant" minds in Congress will screw things up even worse with more Federal interference.

Years ago Houston decided that it would allow an unlimited number of cable operators to enter the market. What happened is that the operators started in different parts of the city, but chose to stop expanding when they reached the boundary of another operator. Why? Cable is a "natural monopoly." With high fixed costs for stringing cable, and low marginal costs for serving more viewers, the larger firm in any area can spread its fixed costs over more viewers, and undercut any smaller competitor.

Oh, and by the way, it's been 30 years since the Feds told municipalities that they couldn't grant exclusive monopoly franchises for cable.

Yes, the naive view is fixated on paying for channels you never watch, when this is exactly why bundling works: everyone wants the disparate elements of the bundle to different degrees, and marginal costs are close to zero.

Why is McCain pushing this now? Cable viewership is declining in the US. Americans, particularly younger cohorts, are increasingly receiving their media through the internet (Netflix, Hulu, Amazon Instant, Apple Store, etc).

Besides being misguided, it is also 10 years behind the times.

Nice video.

When I think of cable companies and bundling I think not so much of a package of cable channels, but rather cableTV+internet+VoIP.

I really don't need VoIP, but the price of any two of the three almost equals the three together. Plus if I didn't have VoIP I'd need to increase my mobile plan expenditures by perhaps more. So for my household, they've found a nice sweet spot.

Whatever the typical benefits of bundling in general, there's nothing to say cable TV offers the best style of bundling for this kind of programming. We've reaching a point where bundle prices are driving consumers out of the cable market, especially given the alternatives. I'm tired of paying over $100 a month for a service that's only occasionally used at all much less one for which I never watch 95% of the channels included. That's the real advantage of Spotify, btw, they bundle, but based on a much better understanding of what I want to consume.

Since I can buy (or otherwise acquire) the content I want more cheaply and when I want it rather than on someone else's schedule, time to cut the cord and do business the competition that's moved to the next generation business model.

Let's get rid of all government mandated bundling before we start government mandated unbundling. Start with health insurance.

Why so impatient? If you've lived through the past 25 years, you should know by now that disruption of the cable industry is coming. It might take 5 or 10 more years, but it's coming. This might sound a tad optimistic (as a sit typing this on my tablet, on the patio) but I remember when you could rent any color telephone you wanted, as long as you wanted black.

Maybe a short coming of "the great stagnation" thesis is our time scales are off kilter. I think it takes a good 25 to 50 years for significant "revolution" to be fully effective.

What is the reasoning behind adding up the individual MWTPs?

I may have a MWTP of $20 for Word and $30 for Excel, but for both $50 seems too much. Especially if I "know" that adding Word to Excel is a zero-cost activity.

I could very well claim my MWTP for the bundle to be no higher than my highest valuation, given the zero marginal cost condition.

If you do this for the example presented in the video, the profits from bundling equal 180 (=2*90), *lower* than for the sell high strategy.

What gives?

If you "know" that adding Word to Excel is a zero-cost activity, then you also "know" by the same logic that initially adding Excel to your shopping cart is also a zero-cost activity... so following your logic you would have to claim that your MWTP is zero.

If you assume that consumers act rationally (yes, it's a stretch, but still a better starting point than anything else), then your willingness to purchase is driven by the value you get from it. If having Word makes your life $20 better, and having Excel makes your life $30 better, then having Word+Excel makes your life $50 better, and you make your decision to purchase accordingly.

Thanks, but I still don't understand the assumption of adding MWTP.

Producing excel or word individually is not a zero-cost activity, but Microsoft simply putting them together is. So why should I value the bundle more than each individual item?

And taking this to the Cable TV - if I have zero value from channel A, and positive value from channel B and some other person has the opposite preferences - then overall profits are unaffected by the bundle. Overall profits from the bundle reduce, if the actual values differ, accepting the addition of MWTP as a valid procedure.

I am not sure, but I guess my point is reasoning from preferences is pretty arbitrary.

Scratch that, I ignored consumer surplus - if you factor that in, then the profits are no lower than under the sell high strategy.

Still my total MWTP from a bundle of two channels one of which I don't care about could be lower than my individual MWTP - let's say I hate sports and under the bundle my MWTP reduces than under an un-bundled regime.

In fact, under such a scenario, where the MWTP for the bundle may be lower than the individual MWTP, the results become a little weird. But this is a very primitive analysis; nevertheless, it points out how useful a 2 by 2 table can be! ha ha!

I love this topic, I gotta say. It always pulls at my heartstrings. Cable company price discrimination and bundling were the topic of my Senior thesis. Readers of this blog wouldn't be surprised to find that the current legislation explicitly permits price discrimination if a cable provider has less than 90% if the market share. That's supposed to mean that the market is competitive (really, it's on the books). It also turns out that if you complain enough to the cable company, you can consistently pay a sum far below the introductory rates. The zero marginal cost of providing cable make this a beautiful and mathematically convenient topic for monopolistic welfare analysis.

Great video, Alex. Too few people understand the economics of bundling, especially for low MC goods.

People still pay for cable when they have Internet access?

Very good post/video.

However, I have a few concerns. Mainly, why is it that we don't see more bundling, and secondly, why do we see the bundling we do.

Especially today, we see very very many zero MC goods being produced (recorded music, e-books, movies, tv shows, video games, and software). So why is it that we don't see more bundling for these? You would expect to see a lot of video game bundles for instance, when they are in reality very rare.

Secondly, why is it that the bundles we do see (except for cable TV) seem to have incredibly low diversity within them. Go to the "boxed set" area of any video store. Almost every bundle will be multiple movies either of the same genre, or featuring the same actor/director, often both. According to bundling theory, we'd see box sets of an Ingmar Bergman movie, Robocop, Dumb and Dumber, Disney's Pocahontas, and Oklahoma. But we never do.

Similarly, we should see video game bundles spanning multiple genres, (an RTS, an RPG, a nFPS, a sports game, an MMO (apologies if you have no idea what any of those things mean, but you should get the idea)), but again, the bundles we do see are either things like a "Warcraft Battle Chest," which has games that are straight sequels to each other (and have almost identical fan bases), the Steam "Every game from the same Studio for $200 sale," (again, as many studios focus on a single type of game, there is small amounts of diversity between games), or the five dollar cds with 50 subpar games at the value rack at Wal-Mart. Again, the example you gave, of Microsoft Office, again has products with very similar costumer bases, (people who need excel are the same people, in general, who would need powerpoint). According to bundling theory we should see bundles consisting of Excel, AutoCAD, Turbotax, etc., but we don't.

Finally, (apologies of I'm beating a dead horse here), the old Time Life collections were inevitably of one genre (100 country classics! The sounds of the 80's). Shouldn't, if diversity in bundles is a good thing, we have seen mixed Time Life offerings (Jethro Tull, Hank Williams and Duran Duran in the same set). Yet again, those series were always of the same genre/time period.

Again, a very good lecture on the theory of bundling, but there is a very large dog here which isn't barking.

Excellent question! I am not sure I have a full answer but bear in mind two things. First, bundling requires that both goods be monopolized by the same producer. If one of the goods can be competitively supplied or if a close substitute can be competitively supplied bundling isn't going to work.

Second, and I think this does answer many of your cases, note what the video says about never wanting to sell someone something when they value it at less than cost. In other words, if you are bundling something that costs you $20 to produce and the consumer values it at $10 then even if you can sell it to them in a bundle you don't want to because if you could take it out of the bundle you could price the bundle $10 less, the consumer would still buy, and your costs would fall by $20 raising profits.

Now bundling says put diverse goods together but the never sell someone something when they value it at less than marginal cost says doesn't put diverse goods together and when MC is significant the latter effect can easily win out.

Now here is what I think is the clincher. You say we don't see bundles combining Ingmar Bergman movies, Robocop, Dumb and Dumber, Disney’s Pocahontas, and Oklahoma and that is true for boxed sets when MC>0 but it's not true when MC is zero. In fact there are lots of bundles like this! Namely, Netflix, Amazon Prime, HBO etc. so I think the theory looks pretty good.

The explanation of bundling in this video, similar to others that I've encountered, strikes me as correct but missing something important. It ought to start off by saying that if we have a market which is not entirely free because people have been prohibited from reselling products that they buy then we have the following consequences of bundling ... And somewhere in there it should say that in a completely free market if Microsoft priced Office at $120 then they would sell only one copy to either Amanda or Yvonne who would then sell the least desirable part of the bundle to the other person.

Correct. Price discrimination, of which bundling is an example, require that resale is expensive or ineffective. We cover price discrimination at greater length in our textbook, Modern Principles.

Isn't preventing people from reselling products socially non-optimal? Why does our legal system agree to enforce these restrictions?

The problem with the presentation is the assumption that value is only created by suppliers and 1005 capture of consumer surplus is a good market setting.

This is incorrect.

Bandwidth is now cheap enough that I expect, for cable television, the marginal cost of delivering to the average customer all of the channels he doesn't watch, is actually negative due to the additional transactional costs associated with a la carte delivery. And for the customer, there are non-trivial transactional costs in figuring out which cable channels you really do watch often enough to want to subscribe.

That being the case, it is difficult for me to see how an efficient market would result in a la carte pricing would result in anything other than the average customer paying more and getting less. Essentially, the cable bundle is priced at the fair market value for the programs the average customer watches, less a discount for not whining about "paying for channels I don't watch" and making the cable company selectively switch them off. Ask them to switch off those channels from your feed, and you will pay extra.

That said, a la carte pricing will be extremely popular right up until people see the price tag on account of, A: the average customer doesn't really understand economics, and B: the average customer really, sincerely believes that he or she watches about half as much television as the average customers. Once people do realize they are paying more and getting less, the will predictably believe the cable companies must be greedily collecting windfall profits and demand that Something Must Be Done About This. I suspect McCain will oblige them.

Much of the internal structure of the cable industry is dictated by an insane set of long-term contracts going back a couple of decades or more. As of a few years ago, the most widely used billing system would accommodate only a relatively small number of billable packages -- on the order of 100. When you start multiplying out the combinations of service level, plus a handful of foreign-language packages, then reserve space for special marketing offers... well, those 100 slots get used up very fast. In most cases, you can't change billing systems because either the contract doesn't allow for it, or the billing system has to interface with the system that talks to each of several million addressable set-top boxes so the box knows what programming it's allowed to present, and the contract for the box-control system doesn't allow the interface to be revealed to any other software companies. The cable industry, and particularly the digital cable industry, was dominated by a handful of individuals who made out like bandits due to these contracts.

At some point -- behind the scenes, never in public -- the cable companies will inform the bill sponsor(s) that the bill will have to set aside a variety of private contracts in order to allow the cable companies to implement a la carte pricing. And then the bill quietly dies.

In Singapore competition increased in the cable business a few years ago. Net effect: viewers are paying a lot more with most of the gains going to content sellers. Prices in Singapore for EPL, soccer world cup etc. are now among the highest in the world.

yes, but presumably everyone that's not a football fan is getting cheaper TV? or are all of the prices up?

Great video, wonderful content. Also pleased to see Ali Y's work getting attention.

Yes, a great video. If McCain wants to tackle the problems of Cable, what he has to do is to separate the physical network an the content, like we get on internet TV. When the monopoly goes away, consumers face many options for many niches, and we really let the market choose. A pity that politics can't let that happen.

Comments for this post are closed