By Joanna Woronkowicz and Doug Noonan
NORC’s new National Survey of Artists puts the number at 18.3% of U.S. adults. That is dramatically higher than standard federal counts. The real question is not which number is “right.” It is what, exactly, we are trying to count—and why.
If 18 percent of U.S. adults are artists, then the United States is home to a much larger artistic workforce than our standard statistics have ever captured, and we are using the word artist to mean something quite different from what federal labor statistics mean by it.
That is the puzzle raised by NORC’s new National Survey of Artists, which estimates that 18.3 percent of U.S. adults qualify as artists. Set beside the usual federal benchmark—roughly around 1 percent of adults, depending on the measure—that number is startlingly high (for example, that is roughly the share of U.S. adults who work in management, business, legal, or office-support jobs).
The gap is too large to shrug off as a technicality. It forces a more basic question: what is the count for?
This is not just a methodological debate. Definitions shape policy. If we are trying to understand artistic labor markets, design income supports, target grant programs, or think seriously about working conditions, the right definition of “artist” matters a great deal. And the biggest possible number is not automatically the most useful one.
This is not just a larger version of the usual artist count
The federal artist count is mostly an occupation count. It asks, in effect: what is your job? If your current primary job falls into one of a limited set of artist occupations, you count.
NORC’s survey is doing something broader. Its screener does not begin with occupation. It begins with practice. It asks whether respondents engaged in artistic activity over the last year, in what settings, for what reasons, and in what capacity. People can come into view as hobbyists, teachers, current professionals, former professionals, or culture-bearers. They can qualify through a mix of practice, public orientation, recognition, and sustained engagement—not just through a current arts job title.
That means the 18.3 percent figure is not simply the federal artist count plus a few side-gig creatives. It is built on a different idea of what artistic work is.
And to be fair, there is a good reason to want that broader lens. Federal occupation data are not great at capturing portfolio work, intermittent work, community-rooted artistic practice, teaching, self-employment, or artistic labor that does not show up neatly in a single current job. If your concern is that official statistics miss too much of how artistic work actually happens, NORC is answering a real need.
But solving one problem can create another. A broader practice-based frame may be better for understanding artistic activity. That does not mean it is automatically the right frame for labor policy.
The gap is not mainly about people between gigs
A natural first guess is that the difference between NORC and federal counts must mostly be about intermittency: artists between projects, artists whose arts work comes and goes, artists whose occupational identity is hard to catch in a survey snapshot.
That is part of the story. But it does not look like the main story.
Our quick pass through the public-use data suggests something more interesting. Roughly half of the gap between NORC’s estimate and a federal-style count seems to live outside a current arts-job frame altogether: people with artistic practice and sometimes art income, but no arts-coded job in the observed job slots; people with non-arts jobs who still maintain serious artistic practice; and people doing unpaid, public-facing, or community-rooted artistic work that labor-force statistics were never designed to capture.
The other half of the gap seems to live inside NORC’s current arts-job universe. Even when we try to limit our attention to just the part of the NORC sample that would fit in a federal occupation frame, the apparent overlap still comes out several times larger than the official benchmark. In other words: even after stripping away some of the obviously broader parts of the NORC definition, the numbers still do not line up cleanly—the NORC sample flags 5X the number of workers in arts jobs!
That matters because it suggests this is not mainly a matter of federal data missing a few artists at the margins. It suggests the two systems are not simply counting the same population in slightly different ways. They are capturing different kinds of artistic labor.
Who shows up in the broader count?
Importantly, the “extra” artists in NORC are not all casual dabblers.
Some are people with current arts jobs that fall outside a narrow federal-style artist definition: arts educators, administrators, managers, craft artists, and workers in other arts or design-adjacent roles. Some are people whose work is real and often monetized, but does not register as an arts occupation. Some are intermittent or off-cycle artists whose work does not happen to align with a labor-force snapshot. And some are unpaid or public-facing community artists—the kinds of practitioners federal occupation counts were never built to see.
That is a heterogeneous population. Which is exactly the point.
“Artist” here is not one thing. It includes livelihoods, practices, identities, public roles, and forms of cultural contribution that overlap without being identical. NORC’s broader frame surfaces that complexity. But it also makes the resulting headline number easier to use carelessly.
If a count combines both a self-employed working musician piecing together income from gigs and teaching, and someone whose main artistic activity is unpaid community singing, that may be perfectly appropriate for one research purpose. It may be much less appropriate for another.
So which number should we use?
That depends on the question.
If the question is, “How many Americans make or perform art in some meaningful way?” then 18 percent may not be especially large at all. Other national participation surveys suggest that artistic creation and performance are much more widespread than occupation counts imply. In that context, a broad practice-based definition makes sense.
If the question is, “How many people depend on artistic labor for a meaningful share of their livelihood?” then 18 percent starts to look too expansive.
If the question is, “Who should be included in artist workforce policy?” the answer almost certainly should not be “everyone who qualifies under the broadest possible definition.” Policy tools are not infinitely elastic. Portable benefits, unemployment systems, tax treatment, labor protections, grant targeting, and income supports all require sharper distinctions than the generic category artist can provide on its own.
That is why the most important issue here is not whether NORC’s number is too big. It is whether we are being clear about the purpose of the count.
The arts field has a weakness for very big numbers
This is a familiar pattern. The arts sector has long had a habit of treating the largest available number as the most politically useful one.
We see it in economic impact studies. We see it in advocacy language. We see it whenever a claim about the arts becomes more persuasive, supposedly, by becoming bigger.
But bigger is not always better. A giant number can be rhetorically satisfying while still being analytically muddy.
That is the risk here. The point of counting artists is not simply to prove that there are a lot of them. The point is to understand who is being counted, what kind of work they are doing, and what problem the count is supposed to help solve.
If we want to understand labor precarity among working artists, we need a labor-focused definition. If we want to understand cultural participation, community creativity, or cultural transmission, we need a different one. If we want to understand the full ecosystem of artistic practice, we may need several overlapping counts rather than one master figure. There is a book that argues this very same thing.
Count, but count for a reason
So: how many artists are there in the United States?
There is no single answer. There is an occupation count. There is a practice-based count. There is a participation count. Each tells us something real. None should be mistaken for the others.
NORC’s 18.3 percent estimate is useful precisely because it shows how much artistic practice spills beyond the narrow federal occupation lens. But that does not mean 18.3 percent is the number we should automatically carry into every policy conversation. A count built to illuminate broad artistic practice is not necessarily the count we want for artist labor policy.
The better move is not to hunt for one definitive number. It is to get more explicit about the purpose of the count.
Count artists, absolutely. But be clear whether you are counting occupations, livelihoods, practices, identities, communities, or participation. Otherwise, we are not really measuring the same thing at all. We are just arguing over whose big number wins.