What’s the Future of the Music Industry? A Freakonomics Quorum
By Stephen J. Dubner
... the music industry of today looks almost nothing like the music industry of 20 years ago. There are a ton of reasons, most of them having to do with digital technology. If you are a young journalist starting out today, you may still aspire to get a big publisher to give you an advance and widely publish your book; but if you are a young musician starting out today, do you want to get a big record advance or do you want to sell the music yourself, like these folks do, and like Jane Siberry does? If you are a record label, what do you do about illegal downloads, and do you keep putting out “albums” that nobody buys or do you instead try to release only individual songs, as many people seem to prefer?
It strikes me as ironic that a new technology (digital music) may have accidentally forced record labels to abandon the status quo (releasing albums) and return to the past (selling singles). I sometimes think that the biggest mistake the record industry ever made was abandoning the pop single in the first place. Customers were forced to buy albums to get the one or two songs they loved; how many albums can you say that you truly love, or love even 50% of the songs — 10? 20? But now the people have spoken: they want one song at a time, digitally please, maybe even free (yikes: big can of worms, which is addressed ably below).
So what really happened to the music industry, and what will it look like in five or ten years?
That’s the question we put to five smart people in our latest Freakonomics Quorum. I found their answers to be incredibly interesting, full of real information and clear-eyed thinking. (If you haven’t already done so, you should also read Lynn Hirschberg’s really good recent profile of Rick Rubin in the Times Magazine.) Huge thanks to all our participants.
Let me begin by discussing the current state of the U.S. record industry. As has been widely reported, sales are down. According to Nielsen SoundScan, album sales fell 18 percent between 2000 and 2006, after accounting for paid digital downloads from online stores like iTunes. While these numbers are not good, other industries have experienced similar downturns. For example, new car sales are down 22 percent for U.S. automakers.
It is important to remember that sales downturns are not atypical in the music business, and that investors remain interested in selling records. The current situation closely mirrors the post-disco bust in the early 1980s. Specifically, real revenues fell by the same percentage during the years 1979 to 1985 and 1999 to 2006. The record industry also continues to generate profits and attract interest from investors. For example, a private equity firm just last month completed a ₤3 billion takeover of EMI, and an investment group purchased the Warner Music Group in 2004 for $2.6 billion.
Investors seek out high returns, and these large investments suggest that many believe that they can make money in the record business. It also implies that the industry is still profitable. While profit data can be hard to come by, we get a small window from Warner, the only publicly traded standalone record company in the U.S., which enjoyed operating margins of 7 percent and 10 percent from its recorded music segments in 2005 and 2006.
Putting profitability aside for now, what is the explanation for the sales reduction that has occurred? The most obvious culprit is illicit file-sharing on networks such as Napster, *****, *******, and BitTorrent. While linking the two seems tantalizing — file sharing rose to prominence at roughly the same time that record sales started to fall — there is surprisingly little evidence to support the claim that file sharing has significantly hurt record sales. I co-authored a paper with Felix Oberholzer-Gee of the Harvard Business School in which we studied this link using download data from file-sharing networks. If file sharing hurts record sales, then albums that are more heavily downloaded should experience lower sales than comparable albums that are less downloaded. But, after controlling for the role of popularity, we found that downloads had little effect on album sales.
There are several other factors that might explain recent sales trends. First, recall the industry’s similar problems in the early 1980s. Then, as now, sales were down as consumers stopped purchasing albums from a previously popular genre (in the ’80s it was disco; now it’s teen-pop). So one explanation is that the industry has failed to find genres that capture the interests of consumers.
Second, much of the reduction in sales is the direct result of industry cost-cutting. The major record labels have cut large numbers of staff and severed ties with many artists. Such moves are not necessarily bad business choices, but they suggest that less attention should be given to revenues and more to profits.
Third, recorded music has had trouble competing against other products that vie for consumers’ entertainment spending. Consider home video products like the DVD. It does not seem implausible that a good chunk of the $11 billion rise in spending on home video products since 1999 represents foregone CD sales. (Music industry revenues only fell $2 billion over this period.) Entertainment spending was also likely channeled into cell phones and video games, both of which experienced large sales growth and have been particularly popular with the key teen demographic.
A fourth and final factor to consider is the rise of paid digital downloads made popular by iTunes. While this model is often described as a competitor of illicit downloading, there is little evidence that file-sharing users also use iTunes (plus genres like classical music, which are largely ignored on file-sharing networks, are very popular on iTunes). More problematic is the likelihood that music consumers who used to purchase whole albums now download only one or two songs, so rather than getting $15 for an album sale, the industry gets two downloads at $2. While there is no direct evidence that cannibalization is occurring, the growing size of paid downloads makes this factor an important one to consider.
As for the future, I am dubious about making forecasts. Much will depend on the choices the major labels make on key issues (will they run experiments to determine the optimal pricing of digital downloads?) and the arrival of still-unforeseen technologies (which could allow labels to more cheaply distribute music, or lead to new forms of piracy). At the same time, I reject the argument that recorded music is close to death, simply because the financial incentives to create music have never been particularly high. In 2005, less than one in five albums were released on a major label, and even among those releases, fewer than one in fifteen went gold (the usual measure of record success). With such daunting odds, recording an album may have seemed like a pointless task. But in that year, nearly 44,000 albums were released — enough to provide almost three consecutive years of listening. Regardless of what happens to companies that produce and distribute music, I am sure that recorded music will continue to be made.
By Stephen J. Dubner
... the music industry of today looks almost nothing like the music industry of 20 years ago. There are a ton of reasons, most of them having to do with digital technology. If you are a young journalist starting out today, you may still aspire to get a big publisher to give you an advance and widely publish your book; but if you are a young musician starting out today, do you want to get a big record advance or do you want to sell the music yourself, like these folks do, and like Jane Siberry does? If you are a record label, what do you do about illegal downloads, and do you keep putting out “albums” that nobody buys or do you instead try to release only individual songs, as many people seem to prefer?
It strikes me as ironic that a new technology (digital music) may have accidentally forced record labels to abandon the status quo (releasing albums) and return to the past (selling singles). I sometimes think that the biggest mistake the record industry ever made was abandoning the pop single in the first place. Customers were forced to buy albums to get the one or two songs they loved; how many albums can you say that you truly love, or love even 50% of the songs — 10? 20? But now the people have spoken: they want one song at a time, digitally please, maybe even free (yikes: big can of worms, which is addressed ably below).
So what really happened to the music industry, and what will it look like in five or ten years?
That’s the question we put to five smart people in our latest Freakonomics Quorum. I found their answers to be incredibly interesting, full of real information and clear-eyed thinking. (If you haven’t already done so, you should also read Lynn Hirschberg’s really good recent profile of Rick Rubin in the Times Magazine.) Huge thanks to all our participants.
Let me begin by discussing the current state of the U.S. record industry. As has been widely reported, sales are down. According to Nielsen SoundScan, album sales fell 18 percent between 2000 and 2006, after accounting for paid digital downloads from online stores like iTunes. While these numbers are not good, other industries have experienced similar downturns. For example, new car sales are down 22 percent for U.S. automakers.
It is important to remember that sales downturns are not atypical in the music business, and that investors remain interested in selling records. The current situation closely mirrors the post-disco bust in the early 1980s. Specifically, real revenues fell by the same percentage during the years 1979 to 1985 and 1999 to 2006. The record industry also continues to generate profits and attract interest from investors. For example, a private equity firm just last month completed a ₤3 billion takeover of EMI, and an investment group purchased the Warner Music Group in 2004 for $2.6 billion.
Investors seek out high returns, and these large investments suggest that many believe that they can make money in the record business. It also implies that the industry is still profitable. While profit data can be hard to come by, we get a small window from Warner, the only publicly traded standalone record company in the U.S., which enjoyed operating margins of 7 percent and 10 percent from its recorded music segments in 2005 and 2006.
Putting profitability aside for now, what is the explanation for the sales reduction that has occurred? The most obvious culprit is illicit file-sharing on networks such as Napster, *****, *******, and BitTorrent. While linking the two seems tantalizing — file sharing rose to prominence at roughly the same time that record sales started to fall — there is surprisingly little evidence to support the claim that file sharing has significantly hurt record sales. I co-authored a paper with Felix Oberholzer-Gee of the Harvard Business School in which we studied this link using download data from file-sharing networks. If file sharing hurts record sales, then albums that are more heavily downloaded should experience lower sales than comparable albums that are less downloaded. But, after controlling for the role of popularity, we found that downloads had little effect on album sales.
There are several other factors that might explain recent sales trends. First, recall the industry’s similar problems in the early 1980s. Then, as now, sales were down as consumers stopped purchasing albums from a previously popular genre (in the ’80s it was disco; now it’s teen-pop). So one explanation is that the industry has failed to find genres that capture the interests of consumers.
Second, much of the reduction in sales is the direct result of industry cost-cutting. The major record labels have cut large numbers of staff and severed ties with many artists. Such moves are not necessarily bad business choices, but they suggest that less attention should be given to revenues and more to profits.
Third, recorded music has had trouble competing against other products that vie for consumers’ entertainment spending. Consider home video products like the DVD. It does not seem implausible that a good chunk of the $11 billion rise in spending on home video products since 1999 represents foregone CD sales. (Music industry revenues only fell $2 billion over this period.) Entertainment spending was also likely channeled into cell phones and video games, both of which experienced large sales growth and have been particularly popular with the key teen demographic.
A fourth and final factor to consider is the rise of paid digital downloads made popular by iTunes. While this model is often described as a competitor of illicit downloading, there is little evidence that file-sharing users also use iTunes (plus genres like classical music, which are largely ignored on file-sharing networks, are very popular on iTunes). More problematic is the likelihood that music consumers who used to purchase whole albums now download only one or two songs, so rather than getting $15 for an album sale, the industry gets two downloads at $2. While there is no direct evidence that cannibalization is occurring, the growing size of paid downloads makes this factor an important one to consider.
As for the future, I am dubious about making forecasts. Much will depend on the choices the major labels make on key issues (will they run experiments to determine the optimal pricing of digital downloads?) and the arrival of still-unforeseen technologies (which could allow labels to more cheaply distribute music, or lead to new forms of piracy). At the same time, I reject the argument that recorded music is close to death, simply because the financial incentives to create music have never been particularly high. In 2005, less than one in five albums were released on a major label, and even among those releases, fewer than one in fifteen went gold (the usual measure of record success). With such daunting odds, recording an album may have seemed like a pointless task. But in that year, nearly 44,000 albums were released — enough to provide almost three consecutive years of listening. Regardless of what happens to companies that produce and distribute music, I am sure that recorded music will continue to be made.
