Art Works Blog

Taking Note: A WPA Legacy for Measuring Employment Among Artists

The poster depicted above was originally created by Katherine Milhous as part of the WPA Federal Art Project of Pennsylvania. The WPA (Work Projects Administration) was a New Deal agency in operation from 1935 to 1943.

Artists sponsored by the WPA's Federal Art Project include a Who's Who of 20th-century American art. In addition to Milhous, for example, WPA projects featured works by Sinclair Lewis, Mark Rothko, Bernice Abbott, and Orson Wells.

However, the WPA's link to the arts does not end with the Federal Art Project. Interestingly, the agency was also the origin of the Current Population Survey (CPS). Spearheaded by the WPA in 1940, the Sample Survey of Unemployment evolved into the CPS, which is now co-sponsored by the U.S. Census Bureau and the Bureau of Labor Statistics (BLS). The CPS serves as the primary source of labor statistics for the U.S. population, and it is used by the NEA to report employment conditions among artists.[i]

Take, for instance, the unemployment rate. Because the CPS is conducted monthly, it is well-suited to measuring conditions that are time-sensitive, particularly unemployment.

The BLS reports that the unemployment rate among artists was 7.1 percent in 2013. Although this marked an improvement over conditions in 2009 and 2010—when artist unemployment exceeded 9 percent—it was, nevertheless, well above the pre-recession unemployment rate of 3.6 percent of artists in 2006.

This pattern is also evident among several individual artist occupations. Architects, for example, were particularly hard hit by the housing bust of the Great Recession of 2007-2009. In 2009, the unemployment rate among architects shot up to 10.8 percent. It fell to 4.7 percent in 2013, but was still higher than the 0.8 percent unemployment rate reported for architects in 2006.

A similar pattern is shown among designers, whose unemployment rate in 2013 was 5.5 percent—down from 9.7 percent in 2009, but not yet recovered from the 2.5 percent rate reported for 2006.

Unemployment among musicians, alternatively, has remained at 8 to 9 percent since 2009, or three to four percentage points higher than the unemployment rate among musicians in 2006 (4.8 percent).

As a group, artists belong to the occupational category labeled "professionals." [ii] Yet the CPS reveals that artists experience unemployment rates similar to the higher rates reported for the U.S. civilian labor force. In 2013, the unemployment rate among all professionals was 3.6 percent; but it was 7.1 percent of artists and 6.6 percent of all U.S. civilian workers.


Among employed workers, the CPS shows that artists numbered 2.1 million in 2013, including 784,000 designers, 218,000 writers and authors, 202,000 musicians, and 158,000 producers and directors. It is important to note, however, that these figures refer to primary artists. (A primary job is defined as one at which the greatest number of hours were worked.)

Among all the federal government data sources on employment, the CPS stands alone in its ability to capture multiple jobholding. Tallies of the public CPS file show an additional 271,000 workers holding second jobs as artists in 2013.[iii]

Musicians account for the greatest number of moonlighting artists (84,000), while workers holding second jobs as announcers represent the greatest share of the total number of jobs in artist occupations (nearly 38 percent in 2013).

Bar graph

Because the population of workers holding second jobs as artists is small, it is difficult to estimate detailed information about them without incurring a large sampling error. However, a few key traits about secondary artists emerge.

For one, most workers holding second jobs as artists work in professional occupations in their primary jobs. In 2013, teachers (post-secondary, secondary, and "other" teachers such as those at music and dance schools) composed roughly 20 percent of all workers holding second jobs as artists; another 20 percent of secondary artists were also artists in their primary jobs.

And because moonlighting artists are holding down two jobs, they work more hours than workers with one, primary artist job. In 2013, secondary artist jobholders typically worked a 46-hour week in their primary (artist or non-artist) jobs and their second artist jobs, combined.

By contrast, workers holding one artist job in 2013 usually worked 37 hours per week, a slightly shorter workweek than the 39 hours usually filled by all workers with one job.

Also, secondary artist work stems from self-employment. In 2013, more than 60 percent of moonlighting artists were self-employed in their second artist job. That rate was considerably greater than the 35 percent of primary artists who were self-employed.

Demographically, secondary artists tend to mirror primary artists: men account for a larger share (55 percent); half are married; and the median age is 41.

Artists are well educated—in 2013, 65 percent of both primary and secondary artists held bachelor's degrees or higher levels of education. Among all U.S. workers, 32 percent held college degrees.

Drawing on CPS data, the NEA released on Monday our Arts Data Profile #3, "Keeping My Day Job: Identifying U.S. Workers Who Have Dual Careers as Artists." Thanks to the WPA, we have not only treasured American art, but also a gold-standard survey by which we can measure employment conditions among American artists.


[i] Artist occupations include: actors; announcers; architects; art directors, fine artists, and animators; dancers and choreographers; designers; musicians; other entertainers (e.g., jugglers and comedians); photographers; producers and directors; and writers and authors.

[ii] "Professionals" is a broad occupational category comprising positions that generally require college training. In addition to artists, the professional occupation category includes doctors, lawyers, teachers, librarians, and clergy.

[iii] Estimates reported by the BLS differ somewhat from those calculated using the CPS public file, which uses techniques (e.g., "top-coding" of age) to prevent disclosure of survey respondents.



Add new comment