Legal Commentary

In honor of the release of the latest Dave Matthews Band album, and my upcoming concert visit, I thought I would blog a bit about an area of interest that I don't often get to write about.  As the title implies, this topic combines three areas of interest: DMB, Bootleg Trading (which prefer to call concert trading, as I don't really view permissive recording of concerts to be a bootleg), and network economics.  The punchline is a critique of the popular argument that record companies should want free distribution of music because it will increase sales.  But first, some background.

Dave Matthews Band, as most people know, makes music.  Despite releasing new albums more slowly than most groups (about every 3 years, and 4 since the last one), the group is extremely "sticky" among its fans.  It is routinely one of the top grossing concert acts in the world, and its fans continue to go to shows.  I've been to at least 25 shows in 5 states (I've lost count), and the only reason the number isn't 50 is that I got married and had kids.  I've been a member of the fan club for 9 years, yet when seats in Pittsburgh were doled out by seniority, I'm still at the back of my section, meaning that there are many fan club members sitting in front of me with 10+ years in the club. 

But here's the thing, they don't have a "lot" of record sales, at least not as compared to many other pop acts.  Few songs hit the charts, and the albums rarely crack the top 5, let alone the top 10, and if they do it's not for very long.

The band's popularily among its fans is largely attributed to its loose concert taping policy.  From the beginning, anyone with a tape recorder (and now more sophisticated devices) could record any show, and share it (for free) among friends.  Most are high quality these days - every show will have someone in the first 10 rows setting up three microphones (left, right, center) 15 or 20 feet in the air.  Trading is fun, and because the band changes the setlist nightly and improvises significantly, no two concerts are ever the same.  For example, I have some 200 concerts on CD or hard disk, even though I'll never get to listen to them all. Trading used to be a quid pro quo - I'll send you one show if you send me another. 

Over time, trading has lost some of its steam, as cheap storage and high bandwidth allows new recordings to be released within a couple days via BitTorrent. Thus, there's no need for actual trading - it's all about downloading now.

If what I'm describing sounds familiar, perhaps you heard about it the first time when it was called the Grateful Dead, or maybe the second, when it was called Phish.  GD, Phish, and DMB are all examples of bands with devoted fans, high-activity trading, but relatively weak record sales.

This leads me to the core academic point in this story - the network benefits associated with trading.  The typical economic argument used to argue that record companies should favor music sharing is that the more people that have a song or two, the more likely people are to buy the album. The effect is two-fold:

1.  Sampling: the more people that listen to a song, the more people that will buy the album

2. Status: everyone is listening, I'd better get the CD and listen, too

I don't want to get into whether this is true or not in general, though empirical studies seem to imply that it is not. Instead, I want to comment on a more peculiar aspect of sharing - what happens when there are two markets?

As far as I know, DMB keeps almost all of the money it gets from concerts, but a much smaller portion of the money it makes from record albums (maybe more now than in the past, but the label is still in there). Based on this fact, it stands to reason that if full-sharing is encouraged by the band, then it will make more money, even if that sharing substitutes for record sales. Thus, where you have a band known for great concerts, the dual markets benefit the band, but not the record label. This appears to be the network effect working, but only in one market.

Thus, the band has an incentive to encourage sharing, and record sales may suffer despite popularity.  This implies that the network benefit model fails, but only partially. At least my 3 band sample says so. It may also be the reason why many bands favor sharing even though their labels do not.

This may be why artists that are not known for good concerts, or that do not get to keep most of the concert revenue do not want any kind of bootlegging or other music sharing - where there is a single market, the substitutive effect outweighs the network benefits in general (unless you believe the studies that say otherwise).

These are off-the-cuff thoughts - I haven't studied this much, and probably never will. So, if anyone has some thoughts on the topic that can guide my meanderings, please post!