Weird (Data) Science: The Bigger Meaning of a New Study on Success in the Art World
This week, adding crucial context to a non-auction development…
BLINDED BY SCIENCE
Last weekend, the acclaimed academic journal Science published a study that used rigorous statistical analysis to argue that artists’ financial and critical success in the highbrow art world primarily owes to early approval from the industry’s most discriminating gatekeepers. But my concern is that much of the discussion I’ve heard about the research still considers it in too narrow a context to be helpful.
As covered by my colleague Sarah Cascone, the three-year study evaluated nearly half a million international artists’ careers through the lens of hundreds of thousands of gallery, museum, and auction data points between 1980 and 2016. The lead author was Albert-László Barabási, a Northeastern University data scientist specializing in network theory. The data came from author and art-market economist Magnus Resch’s namesake venture, Magnus.
To simplify as much as possible, Barabási and his co-authors daisy-chained together a series of institutional and market findings to create a cohesive argument about class and connections in contemporary art.
First, they mapped out a “coexhibition network” demonstrating which galleries and museums tended to show the same artists as other galleries and museums. Second, they assigned each gallery and museum a letter grade for its prestige level within the industry. And with that foundation in place, they established a strong correlation between institutional prestige and financial value by examining auction and gallery prices for the artists shared by the various entities in the network.
From there, Barabási and his collaborators analyzed how artists’ careers progressed—or didn’t—based on where in this ecosystem they managed to gain traction early on. The results statistically reinforce the popular notion of the legacy art world as a (mostly) closed social club that takes care of its own.
The study found that 39 percent of artists whose first five exhibitions embedded them in the upper echelon of the system—meaning they were at high-end galleries like Gagosian, Pace, and David Zwirner or sterling institutions like the Museum of Modern Art, Tate, and the Guggenheim—were still actively exhibiting a decade later. Of those survivors, almost 60 percent continued showing in “high prestige” venues, and almost none—just 0.2 percent—ended their career “on the fringes,” as Barabási put it to the Wall Street Journal.
In contrast, the data showed that artists who began as social outsiders had a hard slog ahead in infiltrating the structures of the legacy art world—but not an impossible one. Only about 14 percent of artists who started showing in a “low initial prestige” context were still showing at all 10 years later. Of those survivors, almost 73 percent eventually reached a “medium prestige” context, but only about 10 percent managed to ascend into the rare air of “high final prestige” venues by the close of their careers.
The chart below synthesizes a lot of this information into a single visual:
Now, in the past nine days or so, I’ve seen this study treated as everything from the unmasking of a grand international conspiracy to a truism on par with mathematically establishing a link between heavy marijuana use and enthusiasm for jam bands. I understand both extremes on the viewpoint spectrum. All that I’ll say in this regard is that, personally, I still think it’s valuable to find hard evidence verifying what most people already believe to be true. Without rigorously testing our assumptions, how can we know when they might be wrong?
Still, I’ve noticed that even many of the people who take polar-opposite positions on the study’s value tend to agree on at least one point: that the data proves the art world is an especially, if not singularly, rigged system.
In reality, though, I think the findings actually show another way in which the art world is more like the rest of the non-art world than many of those on the inside tend to believe.
Perhaps no economist has done more path-breaking work on social mobility in the modern United States than Raj Chetty, who heads Harvard’s Opportunities Insights initiative. Formerly known as the more easily understood Equality of Opportunity project, the initiative aims to quantify and, through public policy recommendations, reinvigorate what Chetty and his collaborators have dubbed “the fading American Dream.”
There are many sobering dimensions to this phenomenon. (If you’re a stat-head, a masochist, or both, click here to sort through 14 charts’ worth of them.) But out of the many, one of Chetty’s findings strikes me as particularly relevant to Barabási et al’s research on art-market mobility.
In a study published in 2014, Chetty and his collaborators set out to measure how likely Americans born into one socioeconomic class were to level up to a higher class across generations. (The technical term for this metric is relative intergenerational mobility, just in case you ever want to obliterate your odds of ever procreating with anyone in earshot.) For a large and diverse set of children born every year from 1971 through 1986, the researchers compared their subjects’ income at age 26 to their respective parents’ income when their subjects were born.
The findings revealed a classic “that’s good/that’s bad” dynamic. On one hand, Chetty’s results showed that this particular type of mobility stayed remarkably stable over the study’s entire scope. That’s good, right?
No, that’s bad—at least, in our current context. The problem isn’t just that American social mobility never substantially improved over the years. It is that any child born to parents in the bottom fifth of the population by income had only a small chance of making their way to the top fifth—a chance of about 10 percent, to be exact.
Why does a roughly 10 percent chance of progressing from the low end to the high end sound familiar? Because it matches the small proportion of surviving artists who managed to move from a “low prestige” context at the start of their careers to a “high prestige” context by the end in Barabási and company’s art-market mobility data.
I get that there are several major differences between these two studies, that correlation does not equal causation, and that I may be particularly prone to bias because of how staunchly I’ve already come to believe that the art economy is, in so many ways, often a reflection of the larger economy. In other words, this could be me doing something like the analytical equivalent of seeing Prince’s face in a potato chip because I was listening to Purple Rain on repeat on his death day.
At the same time, the US has been (and continues to be) the undisputed economic center of the art market by basically any measurement. Based on that link alone, it seems plausible to me that something stronger than random chance explains these eerily similar mobility ratings—one inside the art world, one outside it.
If I’m not just imagining phantoms in the comparison, then we should be a lot less scandalized by what Barabási’s findings mean for the arts than about what they mean for the greater power structures in which the arts exist. And the solutions we seriously consider should keep that broader context in mind, too.
That’s all for this week. No Gray Market next week because of Thanksgiving. So until we meet again, remember: Not everyone is trying to beat the odds.