Features
Do US Arts Fellowships Favour New York and California? A Method for One Reporter
A desk method for testing geographic skew in US arts fellowships with recipient lists, census population and a small tally, plus the claims such data cannot support.
What to take away
- In MacArthur Fellowship classes from 2015 to 2025, 90 fellows worked in arts fields and were based in the 50 states or DC at the time of award. 37 were in New York and 14 in California, together 56.7 percent. Those two states hold 18.0 percent of the 2020 US resident population.
- The skew is mostly a New York story. New York had 41.1 percent of those fellows against 6.1 percent of the population. California had 15.6 percent against 11.9 percent.
- A different program tells a different story. On the NEA's National Heritage Fellowships list, New York and California together make up 18.1 percent of the profiles, close to their population share.
- Neither result shows bias. The MacArthur program takes no applications, so there is no applicant pool to compare, and one reporter cannot see the nominators.
- Say what a tally shows (where recipients were based) and stop there. Cause needs applicant data and interviews.
The question, and what one reporter can answer
"Coastal bias" is two claims. The first is descriptive: recipients cluster in a few places. The second is causal: the selection process favours those places. A reporter with no travel budget can test the first with public lists. The second needs data the funders mostly do not publish.
Start with programs that publish recipients with a location. Check three things for each: whether location is listed, how the program defines it, and whether it publishes anything about applicants. Programs that list only names, or that block automated access, cannot be tallied without extra work. Record that as a finding about transparency rather than skipping it.
For a UK comparison of the same exercise, see how regional funding in England has been mapped. For the funders' own money, the reading arts finances guide explains filings and award lists.
Worked tally 1: MacArthur Fellows
The MacArthur Foundation's fellows search defines the location field. "Location at Award" is the location of the fellow's primary place of employment, or permanent home address if unaffiliated, at the time the award was announced. The page also offers a download of fellows data. The figures below come from the search index behind that page, for class years 2015 to 2025 (accessed 25 October 2026).
How the tally was built:
- Take all 257 fellows in classes 2015 to 2025.
- Keep fellows whose primary area of focus is one ofFiction and Nonfiction Writing, Poetry, Music Performance and Composition, Theatrical Arts, Choreography and Dance, 2-D Visual Art, 3-D Visual Art, Photography, Moving Image, Performance Art, or Comics and Graphic Narratives. That is a desk judgment. It leaves out architecture, art history, curation and journalism.
- Drop two fellows listed outside the United States and one in Puerto Rico, which sits outside the 50-states-and-DC population base used below. That leaves 90.
- Count by state, and divide by the state's share of the 2020 resident population.
| State | Arts fellows | Share of 90 | Share of US population |
|---|---|---|---|
| New York | 37 | 41.1% | 6.1% |
| California | 14 | 15.6% | 11.9% |
| Illinois | 6 | 6.7% | 3.9% |
| Texas | 2 | 2.2% | 8.8% |
| Ohio | 1 | 1.1% | 3.6% |
| Indiana | 0 | 0.0% | 2.0% |
| Michigan | 0 | 0.0% | 3.0% |
Population shares use the Census Bureau's 2020 apportionment table: resident population of 331,449,281 for the 50 states and 689,545 for DC (332,138,826 together, the denominator used here), with New York at 20,201,249, California at 39,538,223, Illinois at 12,812,508, Texas at 29,145,505, Ohio at 11,799,448, Indiana at 6,785,528 and Michigan at 10,077,331. One year's census count is used for an 11-year window, which is rough.
Within New York, 30 of the 37 were in New York City boroughs (New York, Brooklyn, Queens). The other seven were elsewhere in the state, including three at Annandale-on-Hudson. Fellows were based in 25 states plus DC. Counts this small move a lot with one or two awards, so give the numbers and not a ranking.
Worked tally 2: NEA National Heritage Fellowships
The NEA's National Heritage Fellows list can be filtered by state. On 25 October 2026 the unfiltered list linked 502 profiles across 34 pages. Filtered by state, it linked 38 for New York, 53 for California, 6 for Ohio and 4 for Indiana. As shares of 502, that is 7.6 percent, 10.6 percent, 1.2 percent and 0.8 percent. New York and California together are 18.1 percent, close to their 18.0 percent of population.
This program honours master folk and traditional artists and has run annually since 1982, according to the NEA, so it is not a like-for-like comparison. The list includes group entries. The page does not define what the state filter means, so check a sample of profiles before you print a state count. Ohio and Indiana sit below their population shares here too, which is worth a question to the NEA, not a conclusion.
What the data cannot support
The MacArthur FAQ says the program does not accept applications or unsolicited nominations. It limits consideration to people nominated by invited external nominators, who serve confidentially, and says it receives about 2,000 nominations a year. It also says typically 20 to 30 fellows are chosen each year. So:
- You cannot compute a success rate by place, because applicant and nominee geography is not published.
- You cannot say panels favoured a place. Nominator networks are private.
- "Location at award" is where someone works or lives. A professor at a New York college may have trained and made their work elsewhere.
- The program covers all fields, and the arts subset is your own classification.
- Eleven classes are a small sample. A year-by-year chart would show noise.
Do not write that a fellowship is "biased", that moving a program would fix the gap, or that a state is "underserved" on this evidence alone. Write what you counted, the years, the definition of location and the limits.
A one-reporter workflow
- Choose two or three programs from their own sites. Save the page or export, with the access date.
- Write down each program's location definition. Quote it.
- Pick your arts categories before you look at results, and say what you excluded.
- Tally by state and by city. Use the Census Bureau's population table for shares, and say which year.
- Ask each funder two questions by emailhow many applications or nominations came from each state, and how do they define location. A refusal is reportable.
- Show New York City and the rest of the state separately. Otherwise "state" hides the real concentration.
- Pitch a follow-up only if readers ask for it. Editorial suggestion: stop at a desk story until a local arts group or editor asks you to widen it, since the interviews that explain the pattern need money and time.
For context on how arts coverage rules govern claims like these, see the arts criticism and reporting hub. For the English system of funders, see UK arts councils explained.
