Over the past month or so, we’ve been treated to a series of reports saying that classical music, and opera in particular, is doing quite well. As I’ve commented on my blog before, Chicago’s Lyric Opera has had a record-breaking year. Opera Theatre of St. Louis had a record-breaking year. Houston Grand Opera has also had a fantastic year. All kinds of opera companies are having great years. And across the pond, several English opera houses have openly mocked the Met’s Peter Gelb for his dire warnings about the death of opera, saying they too are having great years.
And yet there are still those who continue to argue that opera is dying. Dying!
(For those who have not seen it, Charles Rosen created this handy chart for the New Yorker explaining the various predictions of death for classical music over the centuries.)
The worst part of this nay-saying is that it is coming from people who purport to be opera supporters. And yet they continue to spout warnings and threats like a modern day Savonarola.
Yes, I get it. As a classical performer, as an arts administrator, and as a board member of a large arts organization, I am quite aware of all the potential threats that opera and classical music are facing, thank you very much. But I cannot, and will not agree with those who seem to think the art form has one foot in the grave. Not when there are so many encouraging trends taking place.
One of the more recent articles foretelling the death of opera is Greg Sandow’s The Peter Gelb Furor.
A bit of background. Peter Gelb, the General Manager of the Met, has made a whole series of comments recently that opera is a dying art form with shrinking audiences. Because of these difficulties, he has demanded steep cuts on the musicians, singers, and workers who make the operas happen… although he has conveniently excluded himself from the list of people that must endure painful economizing. (For my commentary on his remarks to the press, visit here and here.)
Sandow rushes to Gelb’s defense, arguing that opera really is teetering on The Abyss. His article sets out to show that the good news coming from the opera world is misrepresenting the true state of affairs, creating an illusion that threatens to lull people into a false sense of security.
With respect, I think Sandow is wrong, and I’d like to explain why.
* * *
“Yes, some companies are doing well. That would be true even if the entire field was doing badly. It’s a basic principle of statistics. And of common sense. If you look at a large enough sample of opera companies — or shoe manufacturers, or dairy farmers, or any group you care to name — a few will be doing much better than the norm, and a few will be doing much worse.”
Here’s where I start to have concerns. Yes, Sandow is correct that there will always be statistical outliers, regardless of whether or not the overall trend is positive or negative.
But that’s the point. There will always be outliers. So his framework can easily be inverted… this is essentially a half-full, half-empty philosophical debate that very much depends on the observer’s frame of reference.
But what worries me is that he isn’t approaching the subject with a neutral mindset—his hypothesis is already well-established in his mind. Looking at his blog entries as a whole, it seems clear that he adamantly believes classical music and opera are dying. So is he a neutral observer trying to accumulate data to determine a trend, or a fierce partisan trying to gather anecdotes to persuade an audience of his view?
At what point do you determine all those various “outliers” are, in fact, the trend? The last month has brought good news from a whole variety of opera companies—are they all outliers? Is he dismissing everything that doesn’t align with his view as an outlier?
“Back in September, I asked my invaluable assistant, Caroline Firman, to contact classical music service organizations in the US, and ask if they had information on ticket sales in the area of classical music they deal with. Kate Place, research director of Opera America, supplied a chart showing attendance from 1988 to 2012 at all the larger US opera companies, taken together.”
And again, I hear warning bells. To be sure, Sandow is absolutely correct to use broader data to test his hypothesis. But I have questions about the data he gathered.
For one, he told his assistant to gather data from specific sources—did his guiding philosophy unduly influence his choices? Did the fact that Caroline Firman worked for him influence her choices? There’s no methodology listed, no listing of the questions asked or data requested, no list of sources contacted or even how many were contacted, no mention of the return rate, no indication of who responded, or what years the data covered. Plus, I don’t know what mechanisms were put into place to minimize confirmation bias or agreement bias.
And what, exactly, is meant by “larger opera companies?” And what allowances are made for different repertoire, different forms of presentation, different geographies, and so forth? How many opera companies fit into this definition—and are represented in the results? In short, how did he determine who is in the data pool?
And, obviously, the Met is in a category of its own… so how does it fit into the sample?
Another point. I’m intrigued by the dates used for the analysis. How were they determined? Would a different timespan support Sandow’s hypothesis? It’s also curious that Sandow is using data he collected in September to discuss a problem that has popped up now. This is hardly a fatal flaw, and the timing be fortuitous. But it is much more problematic that he is presenting a huge statistical view that ends in 2012—two years ago. That is a significant gap, particularly if the narrative is that opera is failing right now. The exclusion of two current years, when the survey had no problem finding data from as far back as 1988, suggests that something might be hiding in those two critical years… something that the researcher doesn’t want to reveal.
Before I even see the data that’s presented, I have serious concerns about it.
And then he presents the data, and things really get murky.
Again getting to my first point, I’m intrigued that Sandow sees the trend as proof of decline. Really? Looking at this chart I can see that opera attendance has risen and fallen, but it has done so within a fairly narrow band, centered around 1,000,000 attendees. Peaks rose to slightly over 1,100,000, but fell back again to around 1,000,000. Moreover, between 2003 and 2008 the needle barely moved at all, holding steady at the baseline of 1,000,000. Based on Sandow’s set-up, I expected a much sharper decline over a much longer timeframe.
The decline that is apparent in this graph comes primarily in one single year: 2009. But… well, everything went haywire in 2009 as a result of the Great Recession, the largest economic downturn since the 1930s. Of course you’d expect to see attendance falling. But what I find really remarkable is that dip wasn’t nearly as bad as I would have expected. The decline actually halted in 2010 and remained steady ever since. That to me represents astonishing durability for the art form—even in the midst of an enormous economic crisis, ticket numbers dipped, but did not collapse.
And from my earlier point, I have to wonder if the slow—but real—economic recovery helped return attendance to its former benchmark….
“Maybe the numbers picked up since 2012, but what would that mean? A two-year increase wouldn’t reverse the trend.”
Sandow must have read my mind. But this isn’t a satisfactory answer. As noted, I see the overall trend of this graph as one of stability. If the attendance figures returned to their former level, or at least moved in that direction, it would essentially prove my point, and completely undercut Sandow’s.
“But how did they do the year before? Really badly. In their 2012-13 season, ticket sales were only 83% of capacity, down from 88% the season before. That’s a 9.4% drop.”
Here is where my most serious grievances begin. On the one hand there is an important question to ask—what is the cause of this decline? Leaving the specific case of the Lyric Opera aside… when there is a sharp decline in an opera company’s ticket sales, is that the result of large-scale, impersonal forces? Or is it the result of a very localized case of bad management and/or bad luck?
This is a particularly important point with the Met. Sandow seems to think that the decline in ticket sales is due to the larger “death of opera,” while many of the musicians and workers are saying the decline is specifically the result of Gelb’s mismanagement.
But there is a far worse problem here. As get gets into his analysis, Sandow begins to blur three separate criteria for measuring success. They are:
- Ticket revenue. The total amount of money that is brought in by selling tickets, over the entire year. This number is directly tied to an organization’s budget. It assumes over the course of the year some shows will do better than others, but sets a final benchmark for where the organization hopes to land at the end of the year. If the company doesn’t hit its ticket revenue goal, there are serious, direct budgetary consequences.
- Number of tickets sold. This simply reflects the number of tickets an organization sells over the course of the year. This number is usually tied to marketing goals, but it doesn’t have to tie directly to the organizational budget. In part this is because different tickets cost different amounts. And, it doesn’t necessarily indicate who actually attends the performance—there is always a number of people who buy tickets but ultimately don’t use them, as a result of inclement weather, last minute emergencies, or such. Or, a company could buy a block of tickets to give to its employees, only to find the employees aren’t interested or available. Again, this is a useful metric, but hard to tie directly back to the budget.
- Percent capacity. Percent capacity refers to how many people are actually in the theater as a percentage of the total seating capacity. This is the number that really makes a difference to the performers, who love to perform to a full house. Conversely, few things are more depressing for performers than working to a house that’s only half full, or worse. But just because the house is full, or at near-capacity, that doesn’t mean the show is doing well. If a show is struggling, administrators may choose to “paper the house” or give away tickets to improve the optics of the situation. Or, it could offer a wave of last minute discounts to lure people in for a fraction of the listed ticket price. There are other issues that can eat into inventory, such as stage extensions or closed-off sections that actually reduce the number of seats available. This can be a useful metric, but as a result of the qualifications I mentioned it is usually the least important metric, and has the least impact on a company’s budget.
But again, the key to remember is that all of these things are different, and don’t directly relate to each other… or necessarily to the overall budget. As a result, your final analysis will look very different based on which criteria you choose.
Sadly, Sandow seems to base his argument for the death of opera on paid capacity—a particularly problematic choice. Again, capacity in and of itself doesn’t tell the whole story. Were those seats occupied because the opera company papered the house? Or were the ticket prices relatively cheaper in the 1990s, leading to more tickets being sold but far less overall revenue? What were the total number of seats sold, and what was the average price per ticket? Was the price per ticket appropriate for the performance, leading to acceptable amounts of revenue? After all, it is possible, particularly when an expensive superstar performer is headlining a production, that a company could sell every ticket in the house and still lose money.
And, have the sizes of the auditoriums stayed constant since the 1990s, an era when many arts organizations renovated their facilities? The Minnesota Orchestra, for example, just completed a refurbishment of its home, Orchestra Hall that reduced the total number of seats in the hall in order to make those that remained larger and more comfortable. Therefore, it will be impossible to compare paid capacity pre- and post-renovation.
The point is that while obviously opera houses want full houses, paid capacity alone doesn’t tell if an organization is financially healthy or not… or relevant to the community or not. A company could have full houses back in the 1990s, but have been overextended and increasingly marginalized. Was it making wise investments in building audiences or patting itself on its back assured that good times would last forever? Was it investing in artistic partnerships or going on autopilot? Was it working to understand its audience’s tastes or relying on conventional wisdom?
As we all know from the business world, sometimes a company’s glossy exterior is hiding the rot below, and sometimes a smaller, leaner company is having a greater impact.
“The Lyric might now announce what seem, in the present climate, like wonderful results, but in the past 20 years they’ve taken a big hit to their bottom line.”
But there are so many issues going on here. You can’t compare the 1990s to today—yes, it is a different climate, and yes that does make a difference. Financial models, resources and aspirations are wildly different today, particularly coming out of the Great Recession.
And more to the point, even if everything Sandow said about the Lyric Opera’s finances—and by extension, the Met’s—was true, it wouldn’t mean that opera was “dying.” Again, the fact that its paid capacity was larger in the 1990s than it is now tells us nothing about how relevant the organization is today. Nothing about how effectively it is acting on is mission. Nothing about its artistic activities… or their quality. Nothing about what it’s doing long-term to build audiences. Nothing about what it’s doing to create a new generation of musicians, composers, arts administrators or arts workers. Nothing about what it gives back to the city of Chicago.
These are important issues—it is, after all, a non-profit, right? So why is the only criteria being used to determine whether it’s dying or not its paid capacity?
* * *
At this point, let me back up and address Sandow’s larger question: What does all this mean for the Met? On the one hand, given its size, reach and national importance, I don’t know how well it can fit into generalized trends—it will always skew the sample.
This is important. The Met should always be above the curve. As the largest opera house in the country, with resources most other organizations can only dream of, it should be leading the way in nearly every category. It should attract the best talent, of course, but it should attract the most innovative administrators, who are ahead of the curve in finding ways to reach new audiences. It should be better at building revenue. Better at meeting customer demand.
The fact that it isn’t outperforming everyone else should raise some serious questions.
But let’s say that Sandow’s trend is correct and the Met fits the pattern the same as everyone else. That still doesn’t answer the real question at hand—what is causing the Met’s problems? As I’ve mentioned before, Gelb’s administrative and artistic choices have had a far greater impact the Met’s financial situation than the abstract notion that “opera is dying.” It’s clear that his shows have been disproportionately expensive and poorly attended, particularly in revival. They represent a poor return on investment. Using capacity, the same measurement that Sandow uses, his 16 revival shows sold less than 65% capacity, and six revival shows sold at less than 50% capacity.
Is that due to the decline of opera? Or simply bad choices?
And related to that, what will do more to right the ship: stopping these bad managerial decisions, or simply forcing union workers to take a pay cut?
* * *
So in the end, I am not convinced by Sandow’s presentation. In fact, I suspect that he has fallen into the same trap he warns us about… cherry picking facts and ignoring a larger context.
I see an art form that has had ups and downs, but has remained remarkably secure over time. Looking at Sandow’s own data, I am encouraged by opera’s resiliency; it has successfully adjusted to the Great Recession and based on many examples, has begun to thrive again.
To me, the glass is more than half-full.