Refuting Liebowitz: Part 3

And now the thrilling third part of my Refuting Liebowitz trilogy!

In Part 2, I wrote about some points that Liebowitz made that I at least found to be thought provoking. As I move ahead in his paper, though, I find that Liebowitz is really going off the deep end.

Liebowitz attempts to use British record sales to bolster his argument that listening to U.S. radio stations hurts U.S. record sales. He really makes some incredible assumptions. On page 32, I don't understand why he combines U.S. and U.K. record sales. Why not show them individually?

Then on page 33, he assumes that an average album spawns 10 singles! 10! Anybody with any knowledge of the music industry knows that a successful album yields 2 or 3 singles on average. Not every song on an album is released as a single.

Page 38, brings Liebowitz back to his faulty premise:

An increase of one hour per day in listening to prerecorded music would
more than double the daily amount of time the average person spent listening
to prerecorded music. It is hard to believe that such a doubling would not
dramatically increase overall sound-recording sales. And this is just for
automobile usage of radio.
Looked at in this light, therefore, it is easy to imagine that radio broadcast
might decrease the purchase of sound-recordings.

Again, time spent listening to recorded music does not necessarily mean increased sales. Most music fans have albums that they listen to repeatedly. So, all of that time listening to CDs generates zero dollars for the recording industry.

Liebowitz makes some strange assumptions about payola, such as this one on page 40:

Radio stations want to maximize their profits, which requires balancing the
audience size, which is maximized by playing records that listeners most
prefer, against any revenues that might be generated by ‘selling’ airplay to
record labels a la payola.

He's assuming that all payola goes to a station's bottom line. That's not always the case. A lot of the time it goes right into the pocket of the person he who has control over the music. In today's business, it's the music director or program director. A station that doesn't already play Mariah Carey isn't going to start just because of payola. But a station that already plays Mariah Carey will play her regardless of whether payola is involved, because a skilled PD or MD knows that their audience wants to hear her newest stuff. Payola would only influence the number of times that the PD or MD plays her new song. The station is not going to suffer, since other labels/artists aren't going to be excluded. The listener might just hear other artists a little less often. And, by the way, those other artists/labels might be getting increased play across town thanks to payola at a station reaching a similar audience.

Finally, Liebowitz reaches a conclusion on page 41:

I have examined two episodes in which the impact of radio should be
relatively easy to observe. The evidence from this empirical examination
indicates that, contrary to common beliefs, radio broadcast does not enhance
the market for sound recordings.

The conclusion isn't very strong. He claims that there's no evidence supporting the argument that radio airplay increases music sales. Yet, his two examples claiming that radio listening is a substitution for recording listening are weak. His first example deals with US music sales in the 1920s! The 1920s?! He mentions the impact of the Depression, but dismisses it. There's also numerous cultural factors that could have affected sales. Artist rights certainly improved. Blacks became less victimized by the industry which led to an altogether different landscape. And while he mentions the popularity of non-music shows, again he dismisses those. Radio in its infancy can't be compared to today's radio. Radio probably did substitute for music sales back then since it was such a new and exciting medium. It was an experience that couldn't be replicated on a recording. I don't think that the same can be said about today's radio. (Which is really something that today's radio programmers need to take into account. Radio needs to find what can't be replicated by iTunes and the like. But that's for another time.)

And maybe I just don't understand his section on British radio, but that's an even weaker argument with his nonsensical charts that use US sales data. As he says, there wasn't even reliable UK sales data until 1973. But, sales were pretty much flat in the UK after the rise of commercial radio. Then radio didn't hurt sales, right? Why does the absence of a rise in sales mean substitution.

And that's it. Two arguments that he thinks supports his theory. This is far from the definitive word on the subject. And I think he even says as much.

I agree with the part of his conclusion about rethinking the relationship between radio and the music business. Any idea should be revisited from time to time. A new study would be interesting. But we need to know exactly what the listener is listening to and would increased listening to recorded music actually translate into more sales. I think it would depend on the individual. We'd probably end up with a wide range of responses. Someone who listens to a talk show for three hours and then pops in an old Pat Boone album probably wouldn't buy anything new whether they were listening to the radio or not. There are also listeners who actually listen for what's in between the records: contests, humor, traffic, weather, music info. They might not listen to a recording as long as they would listen to the radio considering the absence of those between the records elements. And let's not discount they fact that a great many listeners do listen to radio to find out what's new. How do they know what they want to buy without hearing it on the radio first?