Picture a musician who teaches piano lessons during the day, picks up shifts at a coffee shop on weekends, and squeezes in time to write her own songs whenever she can. The teaching pays better than the songwriting, and the coffee shop pays more reliably than either. So which work gets cut when money is tight, and which work would she rush back to if a check arrived every month with no strings attached?
That question sits at the center of a study by Douglas Noonan and Joanna Woronkowicz, both of the O’Neill School of Public and Environmental Affairs at Indiana University. In a working paper, they examine what happened when 2,400 artists across New York State received $1,000 per month for 18 months, no work requirements attached, and asked a simple question: when financial pressure eases, do artists pull back from work altogether, or do they shift toward the creative work they care about most?
A decades-old theory about why artists work
Economists have a standard story about wages and work. Pay people more, and the effect on how much they work is hard to predict. Some put in extra hours to capture the higher pay, while others decide they can now afford more free time. The relationship is, in the textbook phrase, ambiguous.
But artists may not fit the standard story. In 1994, the economist David Throsby proposed what he called a “work-preference model.” His idea was that artists get something from creative work beyond a paycheck, a kind of satisfaction that economists call “psychic income.” Because of this, artists often take lower-paying or unstable creative work and then top up their income with non-arts jobs to cover the basics.
Throsby’s model makes a specific prediction. As long as artists need to cover subsistence costs, they will spend more time in higher-paying non-arts work than they would otherwise prefer. Relax that pressure by handing them income they did not have to earn, and the theory says they will redirect their hours toward creative work, even if it pays less or nothing at all.
The authors point out that most research testing this idea has relied on surveys asking artists about job satisfaction or how they divide their time. That kind of evidence is suggestive, but it does not show how artists actually behave when their financial situation changes. A guaranteed income program offered a chance to watch that shift happen.
The program and the people in it
The intervention was Creatives Rebuild New York’s Guaranteed Income for Artists program, funded with $125 million from the Mellon, Ford, and Stavros Niarchos foundations. It gave 2,400 artists unconditional monthly payments of $1,000 over 18 months. The program deliberately reached vulnerable and often overlooked groups, using a weighted lottery that gave extra entries to applicants facing what the program described as “multi-point oppression,” including Black, Indigenous, and other artists of color, LGBTQIA+ individuals, deaf and disabled people, caregivers, and those who had been involved with the criminal legal system.
More than 22,000 people applied. Eligibility required proof of artistic practice, New York residency, and financial need measured against a cost-of-living benchmark called the Self-Sufficiency Standard. The applicant pool was economically precarious: about 92 percent reported having no financial safety net, and many received public assistance.
The lottery design gave the researchers a built-in comparison. Because selection was random, the artists who got the money and the applicants who did not were similar at the outset, which lets the authors attribute later differences to the payments themselves. They surveyed both groups using the same questionnaire, collecting data from 1,315 enrolled artists and 4,384 control-group applicants. To handle the weighted lottery and differences in who responded to the survey, they applied statistical weights designed to balance the two groups on background characteristics like age, race, location, and artistic discipline.
What the data showed
The survey asked artists to report their time and earnings separately for three kinds of work: artistic or cultural practice (like writing music), arts-related work that uses creative skills without being creative itself (like teaching music), and non-arts work (like retail).
Artists who received the payments spent more time on creative work. On average, they put in 3.9 more hours per week on arts work than the comparison group. At the same time, their non-arts work fell by about 2.4 hours per week. They were also more likely to be doing any arts work at all, and less likely to be doing non-arts work. In short, the money appeared to let artists trade hours in outside jobs for hours in the studio.
One point stands out: the recipients did not simply work less. The increase in creative hours slightly outweighed the drop in non-arts hours, so total work engagement held up or rose. This contrasts with some other guaranteed income studies, which the authors note tend to find modest reductions in overall work hours.
The earnings results tell the other half of the story. Income from non-arts work fell by roughly $3,680 a year for recipients, a statistically significant drop. Earnings from arts and arts-related work did not change in a way the analysis could distinguish from zero. The estimate for total earnings pointed downward by about $11,591 per year, though this figure carried a wide margin of uncertainty.
The authors highlight that the total earnings decline lands close to the $12,000 the program paid out annually. They interpret this as artists holding their net income roughly steady while reshuffling their labor, working as much or more but earning less from the market because they shifted toward creative work that pays less. That pattern, they argue, is what Throsby’s model predicts.
The findings held up across a series of additional checks. The authors re-ran their analysis with extra control variables, with methods that account for relationships between the different outcomes, and with statistical corrections for testing many hypotheses at once. The basic pattern stayed stable throughout.
Important limits on what this means
The authors are careful about how far these results travel. The program used a broad definition of “artist” that included community arts practitioners and traditional culture bearers, not only painters and musicians with commercial careers. And by design it reached economically vulnerable applicants. The authors caution against assuming higher-earning, more commercially established artists would respond the same way.
Timing matters too. The program ran in the immediate aftermath of the COVID-19 pandemic, a period of unusual disruption for arts work. Some of what the data captured may reflect artists rebuilding routines and recovering rather than making long-term choices about how to allocate their work. The payments were also explicitly temporary, which the authors suggest may have shaped how recipients responded, treating the money as a short-term cushion rather than a permanent floor.
There is also a measurement caveat. Recipients might have overstated their creative engagement to reflect well on the program, which could inflate the estimated effects. And the authors stress that more artistic hours should not automatically be read as greater productivity. For some participants dealing with health problems, caregiving, or immigration precarity, the income may have mainly restored their basic capacity to keep working at all. Related research on the same program found improvements in physical and mental health among participants.
The researchers frame their contribution modestly. The hours and earnings shifts they document are behavioral responses consistent with a long-standing theory, not a verdict on whether guaranteed income is a good use of public or philanthropic money. Judging that, they write, would require valuing outcomes beyond work hours and comparing the program against other ways of spending the same dollars. What the experiment does offer is a direct look at a question that had mostly been answered with surveys: give artists a financial floor, and many of them appear to spend the freedom making more art.
