According to the Wall Street Journal "compact-disc sales for the first three months of this year plunged 20% from a year earlier." Now I'm not sure why anyone would care: presumably sales of LP's, 8-track and cassette tapes has been zero for some time, and nobody seems especially out of sorts over it. Will the WSJ have an article about the sad decline of the buggy whip market next?

Buried deep within the article is the real news "Overall, sales of all music -- digital and physical -- are down 10% this year. And even including sales of ringtones, subscription services and other 'ancillary' goods, sales are still down 9%, according to one estimate; some recording executives have privately questioned that figure, which was included in a recent report by Pali Research." While perhaps not so dramatic as 20%, 9% - which seems to be the real number - is still a substantial fall.

If you want to understand why the fall in sales, I don't think you have to look much further than this To summarize:

I must admit to being completely flummoxed. There I sat, a loyal music fan who has shelled out actual money to a business that is supposed to be having financial problems, and the best they can do is tell me to wander the streets of Seattle looking for different internet providers who might allow me to download the music that I have already paid for, music that I have spent the better part of three house trying to listen to, and which is still unusable?

Apparently the music industry takes the maxim "no one ever got poor underestimating the intelligence of the marketplace" a little too seriously. Nobody ever got rich by offending their best customers either.

This is disingenuous. The point of the article is not to fetishize the CD over digital distribution (nor the buggy over cars). The point is that this data point, if accurate, says a considerable amount about the changing market and the future of music. As you no doubt gathered, part of the implication here is that illegal downloads are to blame. The real meat of the article, which you conveniently gloss over is actually the second sentence:

The sharp slide in sales of CDs, which still account for more than 85% of music sold, has far eclipsed the growth in sales of digital downloads, which were supposed to have been the industry's salvation.

Perhaps you are correct that the problem is the record companies' mishandling of online distribution. I think a pretty elementary logic would suggest some of the blame is the availability of FREE PERFECT SUBSTITUTES.

You may think this is a salutary development. Fine. But it seems unlikely that piracy isn't involved here. Similarly, whatever the industry's failings in developing perfect DRM (and, European interventionism aside, I'd have to say that iTunes and Zune both offer a pretty simple, pretty effective and pretty useful platform to buy music), it's hardly the evil monster it's often (in places like this blog) made out to be.

I mean really--the implication on the blog you linked to and, I imagine, right here, as well, that our rights--our rights!--are being infringed because we used to be able to do anything we wanted with music on lps, but now we might be restricted in what we can do with digital downloads is the true fetishization of the past over the present.

Had the article led with a 9% decline in sales and then pointed out that this was primarily due to the large drop in CD sales not being fully replaced by the increase in legal paid downloads I'd agree with you - but it was written the opposite way, with the information about the overall decline in sales buried far down in the article.

I'm against copyright. I'm not in favor of government regulation to limit DRM. I think companies are within their rights to put DRM on their music if they like. I just think they are being stupid. The evidence that piracy reduces sales is pretty weak to say the least. The evidence that DRM reduces sales is pretty persuasive. If the music companies want to cut their own throats, that is their right. But their crodile tears won't get much sympathy from me.

Professor Manne [a.k.a. Geoff] -

Here is the problem I have with all of the current platforms that charge for distributing music digitally: The rules of supply and demand don't apply. For all practical purposes, there is an unlimited amount of music available. If song A is popular and gets 5,000 requests for downloads, it essentially costs the same amount to distribute it as the niche song that only gets 4 requests for downloads. In that instance, how do you determine the pricing mechanism? No rational mechanism exists since there is no inherent economic scarcity in this instance. Many platforms have a flat rate per song. But why? How is this price determined when supply and demand doesn't apply? Because of copyright restrictions, the economic model does not allow for other companies to enter the marketplace and offer the same goods at a more competitive price. Doesn't your background in antitrust litigation give you the least amount of pause over any of this?

You write: "I mean really--the implication on the blog you linked to and, I imagine, right here, as well, that our rights--our rights!--are being infringed because we used to be able to do anything we wanted with music on lps, but now we might be restricted in what we can do with digital downloads is the true fetishization of the past over the present." If rights have been restricted today when compared to previous years, then I see nothing wrong with yearning for a time when such rights will be returned to their proper ballance. If you wish to characterize it as a "fetish", so be it. I make no apologies for my belief that as distribution technology becomes cheaper, fair use rights should correspondigly expand - not retract. So yes, a growing number of blogs (like this one) will continue to criticize the music industry and make the case for vibrant fair use in order to counter what is currently being taught in most of our law schools today.

You say:

No rational mechanism exists

I wouldn't say that.

See: www.wikipedia.org/Auction

"Digital art auction: In this indefinitely long auction, designed for unreleased works that are trivially reproducible at zero cost (recordings, software, drug formulas), bidders openly submit their maximum bids (which may be adjusted or withdrawn at any time). The seller may review the bids and close with a price of their choosing at any time the successful bidders that pay this price are those whose bid meets or exceeds it, and these are the only bidders who receive a copy of the item."

Justin:

To agree with Crosbie, I want to draw attention to a link I posted to Amie Street a while ago. It is music distribution platform where the price of songs is determined by the amount of songs sold. So the first downloads are free and then the price slowly increments up. Currently, the site mostly features independent bands but the Bare Naked Ladies recently signed up up with them so hopefully that will provide them with some added legitimacy and encourage other, larger, bands to use their services.

Crosbie and Sean -

The Digitial Art Auction that you bring up is interesting - but I don't buy it as a rational alternative to free market principles. The Auction can certainly help determine demand, but that it only one part of the equation. In order to rationally determine a price, there must be a finite amount on the SUPPLY end. The Art Auction doesn't overcome this. For all practical purposes, there is still an unlimited supply of any given song that can be distributed for costs that are negligible. Supply and demand has two inseperable components. When you are able to overcome one of them (in this case, supply), then I don't see how you cannot rationally fix a price to it.

An 'alternative to free market principles'??

The Digital Art Auction is actually an alternative to non-free market principles. It is designed to operate without the monopolistic privilege of copyright. So, if you're a proponent of the free market, you're a copyright abolitionist and a supporter of the DAA.

A 'finite supply'? I suspect you're talking about copies. The DAA does what it says on the tin - it sells art, not copies - copies are provided free of charge as a means of delivering the IP within the art to the successful bidders.

The fact remains: a new, unreleased song by your favourite artist is in short supply. It doesn't matter how many CD burners exist in the universe, you do not have that song - nor does any other of the artist's many fans.

The DAA lets fans collectively procure the publication of the song. Each fan independently proposes how much they'd be willing to pay as an equal share if pooling their funds with others.

You have an artist on one side with art and a desire for money.
You have an audience on the other side with money and a desire for art.
Supply and demand. Demand and supply. A market. A bargain. An equitable exchange.

Art for money, money for art.

Beware of becoming fixated on the abundance of copies.
Copies are not valuable. Art is valuable.
A copy is simply a delivery mechanism.

So yes, you're right. It is not rational to fix a price to a copy.
You fix a price to a share in the cost of the art.

I agree with Crosbie that there is always a "finite amount on the SUPPLY end." That said, it may well be that copies flood the market so quickly that the price drops very quickly to effectively zero. (But I don't understand why this contradicts the laws of supply and demand: we don't pay for the air we breathe, the language we speak - the prices of some goods happens to be zero.) Certainly a song can be sold in advance of release, and that can generate revenue. But in a mass market there is a real free-rider problem of people who won't contribute to getting the song released, and just wait for the release to get free or nearly free copies. My impression is that in practice live performances and tie in sales are more likely to be big revenue producers, and that sales of recordings may well become loss leaders. Amazingly enough the architect of the DMCA Bruce Lehman seems to agree.

Let's divide an artist's audience up:

A) people so interested in the publication of an artwork that they will pay for it
B) people interested in the publication of an artwork, but not so much that they'd pay for it
C) people who know about it, but aren't interested
D) people who don't know about it, but if they did, might be A,B or C.

Despite the protestations of the RIAA, the B group of the audience is not a source of revenue.

The artist really has to focus on group A. Moreover, once published, the As will help find more As hidden in group D (promotion).

Now then we have this dreadfully pejorative misapplication of the term 'free rider' to art. There is no consumption of art! Mankind's knowledge and culture is there to be ridden freely upon - with all our blessing!

It is only the most invidious of the corporate monopolist's corruptions that persuades us that art is commoditised entertainment to be consumed - never to be retained, shared or built upon.

So, when an interested audience commissions the publication of a work of art (even though it is out of their own interest), it is no burden to them or the artist for the work of art to be added to mankind's cultural commonwealth for all eternity - to be ridden freely upon by this and countless future generations.

If anything a larger audience both helps share the expense and increase the revenue for the benefit of artist and audience alike.

There is no 'free-rider' problem.

For example, look at free software. This is commissioned or developed by those interested in it. It is freely distributed and through viral promotion builds an ever larger audience of people interested in commissioning its enhancement.
People who use free software without paying for it are free riders - but NOT in a pejorative sense. They are simply in group B. They may of course one day move to group A, or even group C.

I meant free-rider in a fairly narrow sense. An artist will release his work for say $1000. There are 100 of us each willing to pay $100 to see the work released. But each of us refuses to donate, hoping that someone else will. Or ten of us donate $100 each, and the other ninety free-ride off of the ten who paid.

This example is defective in that if we all know how much the art is worth to each other, it isn't so hard to come up with a scheme to make sure the artist is paid. For example: the artist says: you each pay $10, take it or leave it. So we each know that if we don't donate our $10 the thing won't be released. However, in practice there is a lot less certainty about how much art/music and so forth is worth to other people. This has been studied pretty carefully by economists - the key result is that if we fix a particular, possibly small, degree of uncertainty, and consider increasingly large populations, the amount that any mechanism can extract becomes a diminishing small fraction of the total willingness to pay. I don't insist on calling this a "free-rider problem" but I think that is a reasonably descriptive term to describe what is going on.

I'm unaware of any historical examples of significant revenue raising through take-it-or-leave it type auctions. The one case I know of where this was tried was in the case of Daguerre who invented photography: in this case the mechanism seems not to have raised enough money to persuade him to reveal the secret. On the other hand: the internet can bring together people of like taste in a way that was impractical in the past, so it may be that a take-it-or-leave it type auction will prove practical now and in the future.

I'm going to take a wild guess and say that for the Rolling Stones or some other wildly popular band, the amount of revenue they could extract by selling their songs in advance through some kind of take-it-or leave it auction is a lot less than they could earn through live performances at football stadiums. The latter we know is pretty lucrative.

Justin,

I don't quite believe the argument that supply, at least in the instance I pointed out, has been overcome. I have owned several websites and the standard operating procedure with the various hosting companies has been that they provide X amount of bandwidth per month and each addition gigabyte of bandwidth costs Y.

We should also assume that the site, in addition to its fixed costs, incurs additional costs for each band it lists. These costs come in the form of the fee paid to the web designer who has to design the addition page, create clips of the songs (which use bandwidth when people preview them), advertise the band, etc. Finally, assuming that there are people out there who are going to download music simply because it is free, we see that additional bandwidth, hence additional cost, is added. Finally, we should also probably assume that after a certain point the bands themselves start receiving royalties.

Understanding the costs, we can logically understand why price discrimination and marginally increasing cost occurs. Bands, when agreeing to use the services, are implicitly agreeing to support the other artists and the parts of site which are money losers, in the event that their songs turn out to be popular.

Yes David, it's all pretty untested.

Moreover, your two statements "This has been studied pretty carefully by economists" and "I'm unaware of any historical examples of significant revenue raising through take-it-or-leave it type auctions" are slightly at odds with each other, i.e. 'careful' is not necessarily 'sufficient'.

Given no history of the fully price-flexible nature of digital art auctions, I'd be wary of drawing conclusions concerning their potential from 'careful' study by economists. I daresay manned flight had been carefully studied prior to its achievement - contrary to much skepticism concerning its feasibility or potential.

The very same mechanism that facilitates extremely fluid diffusion of IP among an artist's audience is the very same mechanism that can facilitate collective bargaining on the part of their audience, i.e. collective patronage.

Moreover, the audience can effectively haggle with itself and the artist through continuous feedback.

Anyway, I insist, there is no inequity, no unfairness in 'free-riding'. Each participant in the collective bargain enabled by the DAA says how much they'd be willing to pay. If the price is lower, great. If the price is higher - no deal.

1) The artist will only sell if they're happy with the total revenue.
2) Each member of the audience will only do a deal if they're happy with the price.
All those who share in the purchase, pay the same price.

That's what I call a 100% equitable and compulsion free bargain.

There is no prisoners' dilemma except in the minds of those who still believe copies hold value.

In your example...

Is the release of the art worth $100 to you, personally? If yes, then bid $100. You never know, it may well go for much less, e.g. 100 bidders of $10 or more.

If you're one of those few strange people who can't bear the thought of others enjoying your beneficience who hadn't paid a dime for the privilege (despite no burden upon you) then perhaps you've got a hang-up that precludes your involvement in collective patronage? Remember, this applies today. You can go into a shop and pay $20 for a CD whereas others copy it from their friends for nothing.

The thing is, how much the art is worth to you is truly independent of whether anyone else gets it for nothing or less than they value it. All you are doing is expressing a willingness to participate in a collective procurement of a work of art you fancy - with the proviso that you can only participate if the cost of a share is $100 or less. Many others may well have lower valuations, e.g. $10.

The two statements aren't really at odds: one is a theoretical statement about what happens with a lot of people. Problem is how many is a lot depends on such things as how much uncertainty there is. So how much revenue is raised in a specific instance is an empirical question, which as I indicated and for reasons you point out, isn't really known. You say "Anyway, I insist, there is no inequity, no unfairness in 'free-riding'." I agree with this: I'm an economist, to me free-riding is a technical term not a statement about values. As a theoretical matter there can be free-riding in this setting; as an empirical matter there almost surely is going to be some free-riding. To an economist the issue is whether there is going to be so much free-riding in specific instances that art that has a lot of social value doesn't get produced. Most of the empirical evidence suggests that there isn't a big problem because sources of revenue by selling others things such as live performances generate enough revenue to cover the costs of most of the things produced by copyright. Revenue by selling things in advance might be important, but we don't have much evidence about it.

You say "the issue is whether there is going to be so much free-riding in specific instances that art that has a lot of social value doesn't get produced".

I still think 'free-riding' is generally received as a pejorative term, inescapably laden with negative connotations. You may be able to make it a value free term, but I suspect most don't - unless pressed to think really hard with a lot of nudging in the right direction.

Anyway, your issue can be reconstructed:

"The issue is whether there are going to be sufficiently interested sections of audiences (and/or broadcasters on their behalf) willing to participate in a collective funding process that very expensive public works can be produced".

The free-rider bogie-men don't need to be mentioned - they're entirely irrelevant. Whether works are commissioned is entirely independent of free-riders. What matters is whether there is enough financial interest.

And yup, you're absolutely right, whether there's actually any revenue to obtained from selling things in advance is a matter of evidence - which we haven't had much of yet.

I'm working on it...

I've also been thinking of another type of auction: the ransom model. http://www.danielsolis.com/meatbot/ransom.html

The artist sets a fixed price, the ransom, and if the market will bear it, the artists releases the art to the public.

Even if it is certain or uncertain that the work will eventually be released to the public, people interested still have incentives to contribute, since that will result in a faster release. People that wants things tend to want it ASAP.

What is your economist view on this model?

The innovative feature here seems to be to give the money to a third party (the charity) if the ransom isn't paid by the deadline. That seems awfully clever - when I talked about the theory earlier in this thread, I'm not sure the relevant theory considered payments to third parties - I need to think this through.

As a practical matter, if you want this to become widespread, would it be better to allow the contributor to designate a charity of their choice? What seems like it could drive the scheme is the idea that I'm going to give some money to the charity anyway, so why don't I do it by contributing to the book; if that doesn't work out, then I've donated to the charity; if it does then I get the book and I can still donate to the charity.

Yup, all these revenue models will be supported by The Contingency Market.

It will be easy for anyone to create a site that uses this (free) web service where either the site can choose the charity or each contributor can.

As to which approach turns out to be more popular, well that's a matter of empiricism. :)