20 January 2012

Record Company Economics

Morning Edition, in a segment borrowed from sister program Planet Money, explored the revenues, expenses and profit margins that went into Katy Perry's 2011 alblum, the best selling title last year.

About $4.5 million of expenses probably went into making the record and in house marketing costs - a good share of this would have gone into advances to Katy Perry and big name guest artists and to producers who would mix the alblum (the best of the best producers charge about $100,000 a track for the twelve tracks on an alblum).

The record company earned about $8 million from the U.S. sales of about 8 million CDs and 25 million digital downloads. The balance of the gross revenue from those sales (which would be about $25,000,000 of gross sales for digital downloads and something like $75,000,000-$100,000,000 of gross sales for CDs once discounting and special rates are figured in and before the cost of actually manufacturing the CDs and their packaging themselves) would have gone folks in the CD distribution chain, Apple's percentage of digital download sales and something on the order of 20 cents per track of artist royalty.

So, the record company makes a roughly $3.5 million profit from these sources.

The record company also makes profits from CD sales and digital downloads in future years, foreign sales, licensing to radio stations and Internet radio stations, licensing for specific uses of the songs in movies, commercials, and merchandise related to the song (e.g. alblum T-shirts).

There are transaction costs that go into the deals that provide additional revenue, like fees paid to licensing agreement lawyers, fees and expenses paid to executives who negotiate these deals, creative involvement to make artistic judgments on licensing for specific uses and merchandise, accounting and collecting payments for licensees and record stores, etc., as well as overhead for the record company as a whole (somebody has to pay for the record company's downtown Manhattan rent, H.R. department payroll, janitors, etc.).

This is pretty much as good as it gets for a record company. Every single other artist is going to produce fewer gross sales. The "cost of goods sold" marketing and manufacturing and royalty costs after the initial creative production and in house marketing effort costs are pretty much a fixed percentage of gross sales that aren't going to very much without regard to the artist in question.

It is possible to make a commercial grade alblum for much, much less money than was spent on this bankable star. Katy Perry's advance was on the order of $1 million to $2 million, but B list musicians or musicians with less of a recor of commercial success would happily do it for 10% of the advance or less, and anybody else would be happy to get paid 1% of this amount or less as an advance. Similarly, there are plenty of perfectly decent producers who would charge less than $100,000 a track. One could record a record while paying everyone involves a living wage for their time and a premium that reflects long term investments by the artists in developing their talents and unique styles for $35,000, and even if this has in house marketing costs as much as half of the $250,000 that went into promoting a sure thing bankable big star like Katy Perry, it is probably possible to put together and promote a credible alblum from a debut band for something on the order of $150,000, if it generally uses a similar business model for that kind of artist.

Also, notably, a lot of the costs of making an alblum, like advances to creative talent, recording studio time, producers, and so on, scale very well. Pretty much, the production costs for a given grade of artists and producers are a certain amount per song, so you can make a single for perhaps 10% of the cost of making an entire alblum. And, the widespread popularity of digital downloads have improved the economics of trying to sell singles v. trying to sell alblums although the costs of marketing a single to radio stations v. marketing an alblum are similar. One can imagine producing a professional quality single for an unknown band with competent but not famous producers and running a credible but not very fancy marketing campaign to radio stations for a single for perhaps $50,000 total.

This analysis of the economics of a big time star suggests that a sales figure necessary to make the effort worthwhile for a very bones alblum at a record coming is going to be something on the order of 5% of Katy's Perry's sales, i.e. about 400,000 CDs and about 1 million digital downloads. With a single produced at bare bone costs, you might get buy with 800,000 digital downloads and a 200,000 CD singles in the U.S. as a target that would cover the $50,000 of up front cost and make the effort worth the record company's while. Throw in radio and Internet radio licensing fees, future year sales, foreign sales, a few minor league commercials, and a few T-shirts and maybe you can reduce the first year sales a bare bones single needs to have been worth the effort to make down to 150,000 CD singles and 500,000 digital downloads in the first year in the U.S.

As the radio segment noted, the problem in the record industy isn't that it can't make a profit from A list stars who have top hits. The problem is that the industry is struggling to have a high enough percentage of alblums and singles that are profitable, some of which have to be big hits to make the venture as whole profitable.

Record company profits are extremely front loaded. The percentage of all of the revenues from a song that come in the first year is huge, and almost all of the revenue from all but a tiny percentage of songs will come in the first decade.

We care about an example like this one out of curiousity about how the world works, and because it is pertinent to the policy implications of intellectual property laws, particular intellectual property law rules regarding the duration of a copyright, and intellectual property law provisions regarding the proportionality of money damages and criminal penalties for copyright violations.

Copyrights are easily ten times as long as the need to be for the duration of a copyright to have any material impact on the economic viability of the record industry. And, it would also be possible to dramatically increase the size of the public domain at trivial cost to the industry, by requiring that copyrights be renewed after a much shorter time period. As Bill Patry recently explained:

In 2002, Jason Schultz did a study of books still in print that were first published during the period of 1927–1946. He found that of the 187,280 books published during that period, only 2.3% were still available in 2002. Thus, the 1998 term extension kept under copyright 97.7 percent of books that were no longer in print, but which could nevertheless not be used. . . . Mandatory formalities, such as affixing a copyright notice and filing a statement of a continued interest in the work can help here. There was a requirement of filing a timely renewal application with the U.S Copyright Office from 1790 to 1992 (the latter date for only some works). Failure to renew meant that copyright owners only got one term of protection, originally that was 14 years, and later it was expanded to 28 years.

The failure to renew was an empirical, market signal about the value copyright owners themselves placed on copyright. The renewal rates also showed a consistent difference in renewal rates for classes of works. The lowest renewal rates (0.4 percent) were for technical drawings, lectures, sermons, and other oral works. The highest renewal rate was for motion pictures (74 percent). Music was 48 percent and books only 7 percent. Our current one-size– fits-all approach ignores this significant data about how copyright owners have themselves valued copyright.

The amount of money that an illegal download deprives artists and record companies of is also highly pertinent to the issue of the appropriate amount of statutory damages for copyright violations. A song on iTunes can be downloaded in most cases for a dollar, of which a substantial proportion of that cost is avoided in the case of an illegal download because no IT expertise or server infrastructure or bandwidth at the iTunes store is utilized in an illegal download. The economic harm for a single illegal song download is less than a dollar and roughly comparable to the theft of a candy bar. It may be illegal, but it is petty theft at best. Few states would authorize a criminal fine or more than $1,000 for a petty theft, and many would typically impose a criminal fine closer to $100 to $500 on a typical person who shoplifts a candy bar.

The U.S. Supreme Court has held that the constitutional boundary on punitive damages is generally on the order of three times the compensatory damages in a case, although higher ratios may be permissible when the compensatory damage amount is nominal. Copyright violation statutory damages imposed without regard to the economic value of the work appropriated involving illegal downloads are immense relative to the economic harm involved, for an offense that is quintessentially non-violent, never puts the victim at any apprehension of physical harm, doesn't breach the peace, and in general, causes are harm that is almost entirely economic harm to large publicly held companies and very wealthy individual artists, as opposed to non-economic emotional distress or economic harm to widows and orphans that deprives economically struggling people of the necessities of life. From a harm perspective, the harm associated with illegal downloading is quite comparable, for example, to the harm that credit card companies suffer from people who fail to pay their credit card companies as agreed. It is illegal to do so and unjustly enriches the person who has done so, but copyright enforcement in illegal download cases is a purely economic harm that is a cost of doing business and makes some kind of percentage dent in the profitability of the company selling it.

Keep in mind that in the typical illegal download case, the copyright owner is already entitled to reasonable attorneys' fees, costs of collection and that there is the option of awarding compensatory damages. Would it be so horrible to deny the availability of statutory damages that are not calculated with reference to the fair market price of the illegally downloaded work in cases where it is possible to establish from a fair market with many recent transactions in an identical work, in much the same way that contract provisions that establish liquidated damages in lieu of actual compensatory damage awards are disallowed as against public policy in cases where the amount of damages is not difficult to determine after the fact with reasonable certainty?

Sure, if someone has been involved in 100,000 illegal downloads of songs, a hefty non-economic damages award, perhaps $300,000, might be appropriate. But, in the case of someone who illegal downloads 100 songs, it is hard to see why non-economic damages of much more than $300 or perhaps $1,000 make any sense. Intellectual property laws should not encourage record companies to make mountains out of mole hills.

No comments: