Thursday, August 17, 2006

Pricing of iTunes and iPods: Vive la France!

I’ve wanted to write a post on iTunes/iPod pricing for some time now. That the French actually passed a law aimed at forcing Apple to allow songs purchased from iTunes to play on other portable music devices makes the issue all the more interesting. The more general issue however is simply the economics of pricing in this kind of “hardware/software” paradigm. Warning: this gets a bit intense on the economic theory front.

The basic point that bothers the French, and could in fact bring up antitrust issues in the US as well, is that songs purchased from iTunes are not playable on devices other than iPods. While not exactly fitting the typical story, this practice is a form of tying or bundling, where the purchase of one product is conditioned on the purchase of another, or two products are only sold together. In this case, we really have the ability to use one product (songs)conditional on the purchase of another (the players). The competitive concern with tying/bundling, and with the iTunes situation, is that of extension of monopoly power in one market to another market. Apple has market power in the online music industry, the claim would go, and by not allowing its songs to be played on anything other than iPods, it is restricting competition in the portable music device industry.

Before I get to some more interesting economics of this practice, I have to note that Apple would seem to have a very strong “efficiency” defense – at least on the basis of what I understand about the technology. The efficiency defense is that Apple must restrict its music only to its own devices, for in order to let the music be playable on other devices, Apple would have to release the code for its digital rights management (DRM) scheme. This is the technology that, for instance, keeps us from transferring a song to an iPod and then downloading it to a computer, from whence we could make even more copies.

Truly one of the genius aspects of iTunes is that it has made legal downloading of music competitive with piracy. If forcing Apple to make iTunes music playable on other devices made piracy even easier, then one would hope that regulatory authorities would use a “rule of reason” approach and conclude that the restriction, while possibly causing competitive harm in the device market, creates outweighing benefits in the online music industry overall.

(There should also, of course, be an assessment of whether Apple does have market power, in the antitrust sense, in any significant market. It is not clear at all that they do, if you consider the broad market for music.)

I also should point out right away that there is a sort of “straw man” argument against Apple restricting iTunes only to iPod. Why not make iTunes songs playable on any device – this can only increase the demand for songs from iTunes and increase Apple’s profit?

To answer this without resorting to the DRM point, let me get into the more interesting economics of the practice. To do this, I want to begin with the essential pricing analysis of songs from iTunes and the devices, i.e., the iPods. As we will see, this pricing theory relies upon the fact that consumers will be different in their demands/values of music, especially in terms of price sensitivity, and it will also rely on the inability of Apple to price discriminate on the devices.

To start, let’s suppose that, contrary to the last statement, Apple could identify different types of consumers and charge them exactly what an iPod is worth to the individual consumer. Of course, an iPod is not valuable by itself; its value derives from its ability to play music. And while the value of an iPod to an individual consumer will depend upon many things, it will most certainly depend on how much music will be played on the iPod, and that in turn depends on the price of music from ITunes. Thus enters the key dependency between iTunes and iPods pricing.

With the ability to differentiate consumers in this kind of situation, Apple would want to price its songs from iTunes at the marginal cost of the songs (essentially, the royalty that Apple must pay, which I believe to be about 67 cents per song). This induces consumers who have bought iPods to consume the “right/efficient” amount of music – so long as the value to a consumer of a song exceeds the marginal cost, it should be purchased and consumed.

Then comes the grand whammy, which is that all the net value for consumers created from their ability to buy songs at marginal cost is taken away in the device pricing. If Consumer A values the combination of an iPod and the right to buy unlimited songs at marginal cost at $300, then Apple would charge that consumer $300 for the iPod. Another consumer who values the package at $400 would pay $400, and so on.

This pricing model works wonderfully – under the rather unrealistic assumption that we can differentiate perfectly across consumers. What happens if we can only set one price for the device itself? (Note that I am not dealing with the “versioning” aspect of iPods, which is that you can in fact differentiate across consumers a bit by offering different versions of the same basic device, at vastly different prices. How much more does a 4 GB Nano cost, by the way?)

So, let’s take the easy case and suppose there are two kinds of consumers (obviously I am being restrictive here, but if you want more realism, you are going to force me to use math rather than words). Type I consumers don’t really want a lot of songs; they just want the device and want to listen to their favorite music. These consumers will be insensitive to the song price, for again, they really want, and just want, their favorite music. Type II consumers like a lot of songs, and they are price sensitive: if you lower the song price, they buy a lot of music, and if you raise the song price, they cut back a lot.

So suppose we still have the songs priced at marginal cost. What do we price the iPod at, if we can set only one price to the entire market? The key point is that the value of the iPod will be different for the two types of consumers, and I am going to suppose that it is lower for Type I consumers.

So Apple would have two choices: price iPods at the lower value, driven by Type I consumers, or at the higher value of the Type II consumers. So long as there are enough Type I’s out there, the optimal choice will be to set the iPod price at the lower Type I value.

And note how Type II’s feel about this: they are getting a bargain, for they buy the iPod at the value of Type I consumers.

That is what makes Apple want to revisit the song pricing. Remember that as of now we have the songs priced at marginal cost, which means that Apple makes no profit at all from the songs. Now this might be what people believe to be true about iTunes, but I don’t think it is right. Apple might not make a huge amount from the songs itself, but it clearly does not have them priced at marginal cost (67 cents).

In order to capture some of the “consumer surplus” from the Type II consumers, Apple will in fact want to raise the price of songs above marginal cost, thereby making a profit contribution from each song sold. Yes, this reduces the value that every consumer places on the iPods, but because Type I’s are not price sensitive, it does not reduce their value by very much. This means that Apple does not have to reduce the price of iPods by very much, and it ends up getting a lot more revenue and profit contribution from the Type II consumers who buy a lot of music.

OK, that is all fine. Basically we end up pricing songs somewhat above marginal cost (99 cent price versus 67 cent marginal cost) and we still take in a lot of profit from the devices as well. It is not the classic razor/razor blade paradigm, where we give away the razor and instead take in all our revenue from razor blade sales, but that is because the economics of the music industry is different (I think the main difference is in price sensitivity. I will use one blade every couple shaves almost no matter what the price is, but I will vary my music purchases a lot on the basis of song prices.)

Now we get to the punch line on the tying aspects of iTunes/iPods. There are two angles to be discussed. First, note that since we are saying that Apple prices its songs so as to actually make some money off of them, it would not want its device owners to be able to buy songs elsewhere. That would defeat the bundled/tied pricing scheme. And if you don’t think the profit from songs is large enough to make this argument interesting, let’s wait until we start getting videos in a big way. Also, if Apple sells 1 billion songs per year, a contribution of 33 cents per song is $330 million dollars profit contribution – not exactly chopped liver.

But the more interesting tying aspect to the case, and what really ticks off the French, is that iTunes songs cannot be played on other portable devices. What about our pricing scheme makes Apple want to put that restriction in place – especially when allowing other devices would only increase demand for songs?

It is a subtle argument that I want to make, but let me try. The price per song, in the pricing scheme as I have described it, will be above marginal cost of a song, but it would be below the price per song that Apple (or anyone else) would set if it were only selling songs. Apple holds the price of songs below what I might call a “market power only in songs” level because by pricing songs lower, it increases the value of an iPod, and by pricing iPods higher, it extracts value through that channel.

So with the tying of iTunes and iPods together, the price of songs is lower than otherwise would be. But think of the advantage that this gives other manufacturers of portable music devices! The essential complementary product to, say, Sony’s portable music device, the songs, is being priced lower than otherwise would be because Apple is trying to capture revenue through its iPods sales. Well, just as this increases the demand for iPods, it also increases the demand for other portable music devices. So essentially, Apple is subsidizing the demand for competing manufacturer’s products. This does not make sense in even a static sense. And in a dynamic sense, it is only going to contribute to the entry and growth of other portable device manufacturers, and all of that competition is going to destroy the basic pricing model for the package of iTunes and iPods.

To keep this from happening, and capture at least some of the value of its innovation, Apple will want to restrict its iTunes product to use only on iPods.

Is that anticompetitive? Does it restrict entry of other device competitors? Well, I think it depends on your vantage point. Yes, it restricts entry relative to what we would get with the same song pricing and no restriction. But my argument is that if you do away with the bundled iTunes/iPod scheme, you would see even higher prices of songs and probably even less entry.

Interesting economics, if you ask me. The French might not agree – but that is their problem.

6 comments:

Anonymous said...

Bob:

Can you analyze why Apple does do either or both of the following:

1) Create a club for buying itunes. Membership in the club is required to buy itunes, and is purchased either by buying an Ipod or by paying some fee.

2) Sell 2 versions of each song, one playable only on an Ipod, one playable on any mp3 player. The former could remain priced at $0.99, the other costs more, say $1.99.

Ignore the DRM issues that (2) might create. Either or both of these would allow them to keep their current profits and make money off those who choose to buy another mp3 player, no? If they do (1) alone, imagine that they sell tunes that can play on any mp3 player.

Finally, doesn't Apple's current approach resemble Oi's Mickey Mouse Monopoly?

Paul W

Anonymous said...

Of course I meant "doesn't do".

Robert G. Hansen said...

Just a quick reply for anonymous: Yes, my ideas are building off of Walter Oi's famous article on Disneyland pricing.

I will skip the club idea for now ans address the second issue. Why doesn't Apple sell the same song at two different prices. One might be a legal issue, as that is pretty blatant price discrimination. It might actually get them into a Robinson Patman violation as it could be construed as harming competition.

But I think the bigger issue is what I bring up later in my post. Apple is worried about entry into the portable music device industry; that is one reason why it does not want to allow iTunes songs to play on anything. I grant that a price of 1.99 might not help entry too much, but it does make the library available to new entrants.

I suspect the artists/publishing houses would really not like the two tier pricing scheme either.

Robert G. Hansen said...

Just a quick reply for anonymous: Yes, my ideas are building off of Walter Oi's famous article on Disneyland pricing.

I will skip the club idea for now ans address the second issue. Why doesn't Apple sell the same song at two different prices. One might be a legal issue, as that is pretty blatant price discrimination. It might actually get them into a Robinson Patman violation as it could be construed as harming competition.

But I think the bigger issue is what I bring up later in my post. Apple is worried about entry into the portable music device industry; that is one reason why it does not want to allow iTunes songs to play on anything. I grant that a price of 1.99 might not help entry too much, but it does make the library available to new entrants.

I suspect the artists/publishing houses would really not like the two tier pricing scheme either.

Aquabhi said...

Am I intrepreting it correctly, towards the end of the article you argue that Apple's monopoly is benefitting the consumers at this stage? By letting the songs play on other platforms, Apple will be forced to raise price to produce at Profit maximising price; since it will no longer have the option of extracting the remaining consumer surplus through iPOD sales. But isn't it that the consumer is paying anyways, either through iPod or song prices

pathrecords said...

Thank you for this article, pretty helpful information.