Music Unit Sales Analysis

Note: 20110831:

US recorded music sales finally turn up in 2010!!


Pro-Piracy advocates often point to the increasing unit sales of the music industry as prima facie evidence that online music piracy is not hurting the music industry. Because this is such a bedrock argument of the "opposition," in addition to our general comments here, we felt it necessary to go into greater depth on this page to fully explode the fallacy of their "proof."

Album Equivalent Unit Sales Analysis

Let's first look at the data. We took the industry unit sales data as reported by the RIAA (see the graph at the top of the Unit Sales page here). We then converted everything into album equivalents (based upon there being 12 tracks in an album). So, for instance, we divided singles downloads each year by 12 to get to the equivalent albums digitally downloaded and then added that to the actually reported digital album downloads.

Based upon those calculations, the following graph reflects the volumes by year.



So, what are we to take away from this chart and the textual comments? The online piracy advocates argue unit music industry sales are up, and conclude that obviously online piracy is not affecting the music industry. This chart puts the lie to that. When you compare apples and apples (all unit sales converted to equivalent albums)—unit sales since the onset of online music piracy are clearly down.

Another thing that is abundantly clear from this chart: when the new CD format was introduced, unit sales went up dramatically. When the new digital download format (Napster) was introduced, music industry unit sales dropped like a rock. Of course, the latter makes abundant sense as the Napster activity was theft. Artists and their labels and the songwriters and anybody else that should have been in the revenue stream—got screwed.

Faced with our previous argument, the "opponents" then argue that the run up in CD sales up to the year 2000 were in large part "artificial" as they just reflect music consumers replacing their album and cassette collections with the "new-fangled" CD. In other words, the decline that started in 2001 just reflected the end of an artificially high sales level inflated by consumers replacing their LP collection with CDs.

The chart below is the exact same data, but we have converted the unit sales data to an index starting at 100 in the year 1973. This approach makes it easier to see the volumetric changes over time.


What this look tells us is that 2008 music industry unit sales are back to the level of 1978 sales. Now regardless of why (the supposed CD ramp up from the inception of CD sales in 1983 or whatever), an industry reverting to its sales level three decades earlier is evidence of something seriously endemically wrong. What's wrong is NOT some dissipation of LP/cassette sales momentum. Let's offer some proof of that.

The following chart shows the waves of unit sales by product type: vinyl, cassettes, CDs and finally digital downloads. This is also an indexed chart. However, to get apples and apples considering that we are looking at different subsets of data (cassettes, vinyl, CDs, and digital downloads), we need to "normalize" the data. We did this by setting the index to zero for each subset in the year that each subset's unit sales approximated an index value of 100. Those years are as follow:

  • Cassettes - 1973
  • Vinyl - 1987
  • CDs - 1987
  • Digital Downloads - 2006


So, what does this data tell us? It starts to call into question the fundamental assertion that the CD sales ramp up was vinyl and cassette replacement. We know that vinyl fell out of favor due to the advent of cassettes. Yet, cassette unit sales, prior to the advent of CDs barely reached the level of vinyl sales. If all of these sales were merely replacing vinyl records or some significant percentage were, then newly recorded products during this period must have been limited.

Why is that important? Because, if cassettes were replacing vinyl, then when CDs came along, any replacement sales they were accounting for was mostly for vinyl, not newly recorded cassette product. You really can't have it two ways. You can't argue that a significant portion of the CD unit sales increase was replacing cassette and vinyl if cassette was mostly replacing vinyl—that's double counting.

So, let's look at the unit sales levels for CDs. The index goes over 900 for both the years 1999 and 2000. The best vinyl years were 1977 and 1978 when their index slightly exceeded 300.

Is it reasonable to presume that CD sales were triple vinyl sales, but somehow all or a significant part (say a majority) of those CD sales were replacement sales?

The argument is just not logical. If it's not logical, then what accounts for the drop off in CD sales starting in 2001? Especially this is a salient question in 2001-2005 as legal digital downloads don't start at all until 2004 and don't reach the 100 mm unit mark (our threshold index level) until 2006.

Clearly, something else was going on and we maintain the "something else" was rampant illegal downloading.