I’ve blogged quite a bit about the issue of declining audiences for classical music here (see especially my post, What if there’s really no “decline” in Classical Music audiences?) but have only touched tangentially upon the purported “Aging Audience” issue. Most of what I’ve blogged about dealt with how much we as a culture tend to idolize youth and how that creates interesting economic consequences when we believe (erroneously, as I think) that this youthful demographic (the “Savior Demographic” as I have been facetiously been calling it) will, well, “Save the Day” for industries prone to the Cost Disease.
Matthew Guerrieri has taken some of the data as has been mentioned on numerous occasions at Greg Sandow’s blog and shown how the aging audience fits in perfectly well with the the rising age of first marriages (for both men and women) and average life expectancy rates. In other words, the trendlines match almost exactly. He also mentions the average first birth rate fits in pretty well with the average rise of the age of the audience. Note that these are all median ages, not arithmetic means. As Guerrieri states:
This intuitively jibes with the NEA’s “Audience Participation in the Arts” surveys, which show the median audience age going up for all surveyed forms of performance. In other words, the problem—if it even is a problem—would seem to be more a function of demographic evolution than a lack of cultural wherewithal on the part of classical music specifically.
In my comment on his blog, which I’d originally posted on a discussion I was having with Lisa Hirsch on her facebook page about how the rate at which the audience is aging almost perfectly matches the rate at which the US population is aging (again, median age)–here’s the relevant comment for reader’s convenience:
Oh, as an aside, the median age of classical music audience is proportionally the same relative to the median age of the US since the Baumol and Bowen Study (1963-1964)–going back to the earlier studies (LA Orchestra 1937; Grand Rapids 1937; Minneapolis Oymphony 1955) isn’t particularly helpful since those weren’t 1) random samples; 2) were self-reported; 3) were particularly small samples (n= approx 900; 1000; 1900 respectively) relative to the sample size of the later studies Baumol/Bowen, NEA SPPAs, MkKinsey studeis (n= 28,000; 18,000; 25,000 respectively)–and Greg should know better than to take the arithmetic mean of the median age of the two 1937 referenced studies (33 and 27) and refer to that as ‘median age’ (30) of that study as a whole–just doesn’t work that way. Of course the median age of audiences has risen faster than the population–otherwise the latter wouldn’t be consistently within a range of 74-76% of the former for all five time points (1963-64, 1982, 1992, 2002, 2008)–the average life-expectancy has maintained a consistent proportion to the median age since 1900 too (range between 46-49%) and that had to rise far faster relative to the median age of the population to remain so. I think the doom and gloom folks make far too much of the median age issue–the actual decline of audience issue is a bit trickier, but I think given how mass media has developed–and how entertainment industries have capitalized on them (or not as the case may be) has far more to do with the decline of population cohorts than anything else.
Since that conversation, I’ve managed to get copies of all the primary source materials (except the Minnesota piece) that Sandow has been using, as well as compiling a bibliography of a number of other sources that do discuss median age of audiences–most of which take place between the original 1940 book mentioned and 1976 (well before the first NEA SPPA in 1982) and will be taking a very critical look at each and every one of the sources and how they are being used to “prove” that the median age of the audience for classical music has been rising faster than the median age of the US population. In other words, as Guerrieri states:
Which circles back around to Botstein’s point—classical music has historically played to an adult audience, it’s just that the passage into adulthood—as indicated by first-marriage age—has been getting later and later, and the length of adulthood—as indicated by life expectancy—has been getting longer and longer.
There are so many inaccuracies as well as fallacious reasoning behind the usage and interpretation of this data that I think it might be useful for those of us interested to really understand what these numbers mean, and the limits of what the numbers can mean within a broader context of demographic data. Also, I’ll relate this to that other issue of the declining audience which, I believe, is a much more serious issue but one that may also be overstated though in ways (if you’ve read my first linked blogged post above) that are counter-intuitive since we don’t tend to think of populations as being heterogeneous with correspondingly heterogeneous preferences.
For more post in this series, visit the Aging of Orchestra Audiences page.