We are continuing the second half of our long-running series on what various file-sharing studies really have to say. This study examines the effect file-sharing has had on live performances or live concerts for the artists and finds that profits have increased dramatically.
The study of the day bears the title, “Supply Responses to Digital Distribution: Recorded Music and Live Performances”. It was posted online in 2005 and was written by Julie Holland Mortimer of Harvard University and Alan Sorensen of Stanford University.
Towards the beginning of the study, one part states the following:
However, the debates about copyright protection for information goods have tended to overlook (or at least underemphasize) the simple fact that these goods typically have many different uses and means of consumption. For example, recorded music can be downloaded easily from the internet (through legitimate means or not), but the experience of attending a concert cannot be downloaded.
This is technically true even though videos of live performances can be downloaded. Personally, though, my gripe of live performances is that the music tends to be too loud when the technicians cranks the music up to something like 120 decibels. I can appreciate how great it is sometimes to just crank up the volume of the music, but it’s always like they took what is sufficiently loud and increased the volume by about 20%. That’s usually why I always pay attention to positions of the speakers and looked for parts of the mosh pit or parts of the stadium seats least likely to be the loudest and hang out there. Also, on a personal note, when your moshing in the mosh pit, pick people up who fall over to prevent them from getting trampled. I know I’ve probably personally saved a few souls picking people up off the ground in the middle of the mosh pit.
So, on with how the data was collected:
We have collected a detailed dataset covering sales of both recorded music and live performances for 2,135 artists. The data span 10 years (from 1993 to 2002) and include all popular music concerts performed in North America during this period, as well as weekly CD sales from 100 cities, for each artist. The detail provided in the data is very rich: for each concert (ranging from small jazz clubs to stadium tours of international rock stars), we observe revenues, ticket quantity, high and low ticket prices, the identities of all performing bands (the headline act as well as any supporting acts), and the place and time of the concert. The data on CD sales provide the band and album name, and the quantity of each album sold, by week, in 100 Designated Market Areas in the U.S. (similar to an MSA). The merged dataset contains all album sales and concert activity for every band in each of the 100 markets in the U.S. in every week over ten years.
So, this is a very wide array of music – though still mainstream type music. So, next, the study touches on the revenue an artist makes on music albums:
Royalty rates range between 10-18% of retail, with the typical rate being 12%; however, artists earn somewhat less than this due to various deductions that are usually built in to the contract. A reasonable estimate is that the artist earns around $1.00 for every CD she sells.
This, in my view, highlights one of the biggest problems in the music business. Why does the artist always hardly get any money from album sales? If you want to discuss why artists earn so little when they sell millions of albums, the contract and the percentage is a great place to start. I would think that it’s fair to have the percentage more towards 50%, not barely get mere pennies per album. Let’s put this kind of math into another perspective:
An album could cost, say, $10. The artist in question could find that, through their major record deal, sell 10,000 albums. That’s $100,000 in revenue gross from sales for the year. For the band, that earns $10,000. Since this is a small band, this will be divided three ways which would amount to $3,333.33. That’s not tying in any other costs like the down payment for the loan that the band makes with the label which has been known to throw the band into millions in debt. Now, compare this with going independent. Let’s say the album earns the band $10 after the cost of manufacturing. How many albums must the band sell to earn the same amount as with the record contract? The answer is simply 1,000 albums. One tenth of the number of sales. If you’re touring all over, that’s not that huge of a proposition. There’s no big loans for the band to pay off the record label either. The only factor to take into account after this is simply taxes. No file-sharing to take in to account here. It’s strictly dealing with the label.
The study later on discusses revenue from live performances and says that the data for their sample period between 1993 and 200 “show concert revenues and average ticket prices in each year, which also increase sharply from an average ticket price of $18.87 in 1993 to an average ticket price of $35.77 in 2002 (reported in December 1997 dollars). Note that the number of concerts and concert revenues both increased quite sharply in 2001 and 2002, at which time file-sharing had become widespread. The increases in price were also most dramatic in 1999-2002.”
Even further down, we have the comment on what happened to CD sales. The study says, “CD sales rise quickly until 1998. In 1999, sales of CDs drop modestly, and in the years after 1999, CD sales drop quickly, back to 1996 levels.”
The study also wanted to be clear in that “we wish to convey the following three significant trends in this industry. First, concert prices and quantities are increasing over time. Second, CD revenues and quantities are decreasing in the second half of the sample. Finally, the aggregate importance of concert revenues is increasing over time.”
The study went on to discuss how live performances boost CD sales. Later on, the study goes back on the topic on if there was a relationship between file-sharing and the drop of CD sales. What was said was this:
Although little concrete evidence has been brought to bear on the question of which artists or genres have been most affected by file-sharing, the conventional wisdom seems to be that file-sharing was more active for popular music genres (like rock and rap) than it was for jazz or classical music. If in fact this was the case, we should expect the spillovers to have declined more sharply for rock and rap artists than for jazz artists. The table suggests the opposite: the change in the spillover effects of concerts appears to have been most pronounced in the Jazz/Latin category. However, we are reluctant to draw strong inferences from genre comparisons, because our grouping of artists into genres is very coarse, and (more importantly) we suspect the accepted wisdom about file-sharing—i.e., that it was most prevalent for pop music—is at best an oversimplification.
I think it is interesting that in genre’s that weren’t as well pirated experienced an even greater decrease in music sales. However, this only remains to be an interesting note.
Another point further down in the study is certainly interesting as well:
Assuming concert demand is an increasing function of listenership, digital sharing of recorded performances should increase the overall demand for live performances. With sharing, each purchased CD potentially results in many listeners. This could result from direct sharing, as when an individual purchases a CD and then burns copies for several friends using a CD-writer, or from broad distribution via peer-to-peer networks. In principle, even one purchased copy of a CD could translate into tens of thousands of actual listeners. In the absence of such digital redistribution technologies, by contrast, the number of consumers listening to an artist’s album would roughly equal the number of albums purchased.
I would argue this point alone should be an excellent argument for strong provisions of fair use. With a CD encoded with DRM (Digital Rights Management), the album also restricts activities that would increase potential sales and general knowledge of the bands existence legally speaking. So, when a country introduces anti-circumvention laws, those laws are, in fact, harming artists abilities to promote themselves. The moment DRM is put on an album, that artist is effectively saying goodbye to potentially hundreds of thousands of new fans or criminalizing the activities that would give artists potentially hundreds of thousands of new fans.
What I found a little cryptic, though, was the conclusion which starts by saying this:
While changes in distribution technology appear to have eroded the profitability of selling recorded albums, our preliminary findings suggest that these changes may have simultaneously boosted demand for live performances.
In decrypting this, we were able to find that this study is not saying file-sharing is responsible for any decline in music sales – at least, as far as this line is concerned. The conclusion further says:
For artists, the decline in revenues from recorded music after 1998 is striking, but appears to have been more than offset by a concomitant increase in concert revenues. Total industry revenues, on the other hand, have not fully recovered, despite the increasing contribution of concert revenue to the total.
artists usually contract with independent promoters to produce their concerts, with little (if any) of the concert revenues reverting to the artist’s record label. Not surprisingly, the loudest complaints about the effects of internet file-sharing have come from record labels and their parent distributors. Since concerts capture returns to investments at least partially made by record labels, it seems likely a new equilibrium will emerge in which those labels play a larger role in concert promotion and claim a larger share of concert profits.
For the music industry, some of the most interesting unanswered questions concern the differential impact of internet file-sharing across artists. It is quite likely that file-sharing is a boon to some artists and a bane to others, but to date there is little empirical evidence indicating which types of artists gain vs. lose.
I one takeaway message one can get from this study is that nothing in here makes a correlation between a drop in music sales and the arrival of file-sharing. At best, you could say that there is a coincidence that they both emerged at similar times, but to point to the two and say that there is a cause and effect is simply misreading what is being said here in this study. As I read this study, I kept asking, “OK, is there a relationship between file-sharing’s rise and a drop in CD sales?” and ended up not having that question answered. If you were hoping to find a connection between a fall in music CD sales and file-sharing, you’ll never really get that in this study.
Meanwhile, there is this rise in live performances in this time period. In reading this paper, there was a few possible explanations for this including the fact that the artist makes more money from touring than in album sales. I don’t think that is unreasonable. If an artist makes $5,000 on album sales in a year and $250,000 in live concerts, where do you think the artist is going to focus their efforts on? As for whether file-sharing has caused the increase in ticket sales or not, that was actually not entirely clear. All we know is that in this time period, public performance revenue has been going up. So, I would say that this is clear evidence that there is money to be made in music.