Updated dire

Hey, Greg, haven't you heard? Down 30 is the new flat.

Now that is interesting. It's so easy to see different statistical snapshots when you work with a completely different set of numbers.

Factor in what is surely the increase of more community based and non-professional groups and who knows--there might actually be an absolute increase in classical music audiences if, albeit, the audiences are more distributed over an [supposed] greater number of ensembles/events. Who knows?

Greg: I think there are two major flaws in your implication that a modest 1.3% reduction in the number of people attending classical music performances between 1982 and 2008 does not represent a major box office problem. I do not have macro classical music data (and could not find it in the NEA reports), but I believe that in the last 25+ years:

1. the total number performances has increased dramatically and

2. the average frequency of attendance by classical music audience members at performances has decreased substantially

As a result, I believe that average attendance has probably decreased substantially for many (though obviously not all) classical music groups. Unfortunately, absent the complete data, it is difficult to make much use of the overall audience size statistics.

Anything Joe says has to be taken seriously. And certainly in the orchestra world, starting more or less in the mid-'90s and continuing till at least five years ago (when the available data stops), there have been more performances with fewer people going to them. It's hard to know what that means. Overall attendance is down, that's clear. But why more concerts? And what does that do to average attendance per concert? Seems like that's down, but what does the decline mean? Fewer people going to main subscription concerts? Or a more varied smorgasbord of events, with more performances given in smaller venues, so the attendance would naturally be fairly small?

Joe is certainly right to think that a 1.3% decline in attendance to top of the line concerts -- orchestra subscription concerts, for instance -- wouldn't be trivial. Could affect the bottom line, and could affect donations, especially if the decline continued, as the NEA numbers suggest that it might. Fewer ticketbuyers means -- in the normal way that orchestras recruit donors, looking for entry-level donors among their subscribers -- fewer donors in the future.

But the NEA figures don't tell us anything about particular kinds of classical performances. We have to get that data -- which can be hard to come by -- separately. What the NEA figures do show is a troubling decline in the proportion of people in almost any age group who go to classical concerts, indicating that classical tickets should be harder and harder to sell, in large numbers, as time goes on. The surprise, given how long this decline has been going on, is that the classical music world isn't fully plunged into a visible attendance crisis, and that, I thought, could be explained at least in part by the increase in population, which offsets the decline in the proportion of people who buy tickets.

But that doesn't mean that the situation isn't serious, or that individual segments of the classical music world -- including some highly visible institutions -- might not already be suffering.

Is it possible to make any meaningful conclusions based on this data? My guess is not.

If 'classical audiences' (whatever that means) actually are getting smaller in the US, does this mean the death of classical music? Again, probably not.

Is anecdotal evidence relevant? Consider: if you are Lincoln Center subscriber, you might notice that the audiences are getting smaller (theoretical example). What does this tell you? It tells you that Lincoln center audiences are getting smaller. You might give other examples of audiences getting smaller. Other people might give examples where they are getting bigger. What does this tell us? That anecdotal evidence is next to useless and cannot be extrapolated to other situations.

Let's leave the music to the musicians, and the stats to the statisticians!

Hi Greg,

I covered this issue quite thoroughly in my PhD thesis.

There are some problems with looking at the data the way you have. You haven't taken into account the change in the age make-up of the population over the period in question, for example. The NEA's own data shows that attendance is much higher in certain age ranges, notably the 55+ range. I don't have the latest data to hand but I suspect the proportion of the US population in that age range might have increased since 1982.

Secondly you haven't taken into account the change in socio-demographic make-up of the population in the period. The NEA 2002 data I analysed for my PhD for example shows that there was a much higher classical music attendance record in what we in the UK call demographic class "AB" at 61% attending at least once in 12 months (this class made up, in 2002, 34% of the US population) compared to the lowest class "DE" at 17%, which made up 36% of the US population. The problem with this data supplied by the NEA is that they don't tell you how often each class attended. "Once in 12 months" is really not very often, and I would be very interested to learn the average attendance within each age band. At home in London I know of people in every age group who go more than once a week, it's not something confined to retired folks!

I actually concluded, in the chapter concerned with this issue, that there really wasn't an audience crisis, not least because the decline (if any) in concert attendance was more than made up for by a number of other factors including the "attendance" at live streaming events (for example The Met got 170,000 to watch La Bohéme and 80,000 to watch Peter Grimes). Bayreuth (admittedly not in the US!) even got 10,000 people to pay €49 for a 6-hour Meistersinger riddled with technical glitches (if reports are to be believed, I didn't see it myself).

Anyway, there are of course as Disraeli said, lies, damned lies, and statistics! But they are very interesting to study.

The Bayreuth stream did have technical problems. My wife watched it. The Bayreuth tech support staff were helpful, but another live Internet stream happening about the same time -- from Verbier -- was problem-free.

At least in the US, live streams, however successful, don't yet offset any decline in ticket sales. First, they don't bring in comparable revenue. Major classical music institutions in the US get (roughly speaking) from 30% to 45% of their income from ticket sales to live performances. If that money starts to shrink, it's not easily replaced elsewhere. Live streams don't bring in comparable income. And they may also bring considerable expenses, above and beyond the cost of putting on a live performance. That's certainly true of the Met's live performance streams to movie theaters, which lost quite a bit of money at first. Eventually they may bring in net revenue, but that's only starting to happen.

Second, classical music institutions in the US depend quite a lot on donations from individual supporters, and from corporations and foundations. These (especially the individual and corporate donations) are strongly tied to live performance. The individual donors are recruited from subscribers. Large individual donors and corporate sponsors are wooed by bringing them to performances, and having them meet the performers, especially conductors and soloists. Can't do that with a live stream. And if the house is increasingly less full, that's not impressive to donors, to put it mildly.

As for the changing age breakdown of the population, I've done a little work with that. Using NEA figures, I found that -- when attendance data is adjusted for changes in the population age breakdown -- it becomes clear that a single cohort has become the largest segment of the US classical music audience. That is, the largest age segment in the audience in 1992 was the group 35 to 44 years old, and in 2002 the largest age segment was 45 to 54. This isn't a good sign. It would seem to show that people born from 1948 to 1957 are to some extent the mainstay of the live classical audience, and that people born in later years aren't taking their place.

I don't have updated figures from 2008, but the statistical tables published with the latest NEA study show something comparable. In 1982, the age group with highest percentage of classical music attendance was people 35 to 44. In 1992, the group with the highest attendance rate was 45 to 54, in 2002 55 to 64, and in 2008, 65 to 74.

That suggests that a single cohort has, during these years, been most apt to attend classical performances, in this case people born from 1938 to 1947. We'd then want to ask why one cohort is most likely, on a percentage basis, to go to classical performances, while another one makes up the largest part of the audience. That's surely answered by figures showing the relative proportion of each cohort in the total population. And what I'm reporting here is very, very rough, because my percentage-of-the-audience figures are adjusted for changes in the age makeup of the population, while the percentage-attending figures aren't. Still, the picture emerges of two cohorts being dominant factors in the classical audience, and not being replaced.

Sorry if all this is confusing. That's the nature of statistical research -- you have to nitpick the details.

One last point. People I know in the classical music business, the few people who think about this data, are struck by something, anecdotally. Given the growth of babyboomers as a percentage of the population, and the dominance of baby boomers in the classical music audience, the classical music audience should be growing more than it is. So even if numbers should be up, they're not up as much as population figures suggest that they should be, which then might (again) suggest a diminishing of interest in classical music as the years go by.