Skip to main content
The FRED® Blog

Moonlighting in the spotlight

Trends for multiple jobholders

Today we’ll try to better understand moonlighting—that is, holding multiple jobs. The Bureau of Labor Statistics records the number of multiple jobholders, and FRED has the data all the way back to 1994. What can we learn from the graph?

Most multiple jobholders hold a full-time plus a part-time job (blue line in the graph), and this group now makes up about 3% of the working population in the U.S. The percentage of workers with this particular work arrangement has declined since at least 1994, when it was over 3.5%.

Those in the next-largest group hold two part-time jobs (red line). The percentage of workers with this arrangement is significantly lower than the first group—a little less than 1.5% of all employees—and has been quite stable over time.

Finally there’s a small group of workers with two full-time jobs (green line), which accounts for about 0.25% of workers. The percentage for this group has also been quite stable since 1994.

We can also see that recessions don’t seem to have a significant impact on these groups of workers with multiple jobs.

How this graph was created: Search for and select the monthly series “Multiple Jobholders, Primary Job Full Time, Secondary Job Part Time.” From the “Edit Graph” panel, use the “Edit Line 1” tab’s “Customize data” section to search for and add an additional series: “All Employees: Total Nonfarm Payrolls” (not seasonally adjusted option). Then type “a/b*100” into the formula box and click “Apply.” Repeat this process for lines two and three, with “Multiple Jobholders, Primary and Secondary Jobs Both Part Time” for line two and “Multiple Jobholders, Primary and Secondary Jobs Both Full time” for line three. All series should be not seasonally adjusted. Use the “Format” tab to select alternative colors for the lines.

Suggested by Makenzie Peake and Guillaume Vandenbroucke.

View on FRED, series used in this post: LNU02026625, LNU02026628, LNU02026631, PAYNSA

Something changed in Black unemployment

Unemployment data reveal several differences for race and gender

This post is a bit long, with a puzzling observation and, even after five FRED graphs, no definitive explanation. But sometimes the journey must be the destination…

The (first) graph above shows unemployment rates by race and gender since the start of the Great Recession. It’s clear men’s rates overall are higher than women’s, possibly due to factors such as women’s less-harmonious attachment to the labor market and different gender composition across industries and occupations. Also, Whites overall enjoy a lower unemployment rate than Blacks, which is at least partly due to the differences in the industries and occupations Blacks and Whites tend to work in.

The movements in the unemployment rates also differ, and this is the puzzle we focus on here. Look closely and you’ll see that unemployment rates didn’t start to decline until late 2011 and early 2012 for Black men and mid-2013 for Black women. The decline for Whites occurs much earlier: early 2010 for men and gradually from 2010 to 2012 for women. Has this difference always existed? We’ll need to look back in time to investigate…

…and fortunately we can use FRED’s time slider at the bottom of the graph to do that. The (second) graph above shows the previous recession in 2001; again, we see a longer lag for Blacks than for Whites for unemployment rates to decline.

Stepping back a bit farther… The (third) graph above shows the 1990-91 recession; it looks like all unemployment rates peaked simultaneously and declined over the same time frame.

The (fourth) graph above shows the 1981-82 recession, where the unemployment rates again peaked simultaneously and declined over the same time frame. We can skip the 1980 recession, as it’s so close to this one.

The (fifth) graph above shows the 1974-75 recession, where the unemployment rates again peaked simultaneously and declined over the same time frame, except for Black women, whose unemployment rates don’t seem to have recovered at all.

Let’s summarize: For the 1974-75, 1981-82, and 1990-91 recessions, Black and White unemployment rates essentially peaked and declined over the same time frame. For the 2001 recession and Great Recession, Black rates took longer to decline than White rates.

What changed?

Even if we can’t provide an answer here, we can suggest where you might do some additional research on the topic. Look to FRASER, FRED’s sibling site, for a deeper examination of historical demographics related to employment: The statistical publications “Employment and Earnings” (1954-2007) and “Women in the Labor Force: A Databook” (2004-2010) are good examples. The latter focuses mainly on differences between the sexes, but also provides statistical tables that relate to race, including one on multiple jobholders.

How these graphs were created: From the employment situation release table, select the series you want according to race, gender, and age and click “Add to Graph.” For all FRED graphs, you have three ways to select the dates you want to display: (1) the date picker above the graph, (2) the time slider below the graph, or (3) selecting the range to highlight within the graph (click and drag).

Suggested by Christian Zimmermann.

View on FRED, series used in this post: LNS14000028, LNS14000029, LNS14000031, LNS14000032

Juggling jobs

Some jobs pay less because they’re part-time. Some jobs just pay less. And these jobs may not provide enough income for workers to make ends meet, bring down debt, or pay for family health expenses. Whatever the reason, some workers have multiple jobs. Let’s consult FRED to see how common this is.

The Current Population Survey from the Bureau of Labor Statistics offers data on the fraction of workers (among all workers) who hold multiple jobs. It’s not a large number, but it’s not negligible either: Today, it’s about 5%. Its slow decline over time suggests that the need for multiple jobs, often some sort of financial distress, is becoming less frequent. By looking at the graph, we see that recessions (shown by the gray bars) seem to have no significant impact on this measure.

How this graph was created: Search for “multiple jobholders,” select the series, and expand the sample period to the maximum range.

Suggested by Christian Zimmermann.

View on FRED, series used in this post: LNS12026620


Subscribe to the FRED newsletter


Follow us

Back to Top