By converting industry-level data to jobs-level data, we discovered some interesting trends both before and during the COVID economy.
For this exploration, we used industry-level monthly employment data to estimate
the growth and decline of specific jobs over the past 19 months.
In the graphic below, each line shows the rise and fall of employment levels. Jobs paying in the top 25% are shown in green, jobs paying in the bottom 25% are shown in orange, with jobs in the middle shown in yellow.
This view of jobs weathering COVID
gave us some big take-aways:
Some of the top and bottom 10% in employment change were continuing 2019 trends
You can see these by selecting the Top growth 2019 (highest 10% change) and Top decline 2019 (worst 10% decline) buttons below. We wondered why library workers have been weathering COVID with few employment losses, and this chart tells us that these jobs experienced tremendous growth prior to the pandemic. On the other hand, some jobs, including a number of energy-related jobs, fell steadily throughout 2019, and more dramatically during COVID.
Education offers protection
If you select the High Ed button below, only the jobs in which two-thirds of workers
have earned a bachelor’s degree will be highlighted, and you’ll observe two things. First, high education jobs typically offer higher salaries, weeding out the lower-salary careers. Second, the high education jobs have had modest losses from COVID in comparison to the significant overall losses.
If you select the Low Ed button, we show you jobs for which no more than a third of the workforce has earned a bachelor’s degree. In many of these jobs, workers were already earning in the bottom 25% of Americans, and on top of that these workers continue to suffer the most significant COVID losses.
As a final note, we observe that the jobs that were steadily declining through 2019 and accelerating
during COVID represent some of the few high-paying low-education jobs available. These are represented by the green lines that appear in the “Low Ed” chart that have been steadily falling for the entire time span shown.
Employment levels over time
Use the filters to highlight jobs by their COVID performance
Unfortunately, monthly employment data is typically available by
industry rather than by career, so we had to get a bit creative to estimate the
change in employment.
We began with our aggregations from the American Community Survey. With
a few exceptions, we’re able to classify each worker surveyed by occupation and
by a 3-digit NAICS industry identification. In some cases, the 2-digit classification is all that is
available, and in a few cases a hybrid designation is assigned.
For each ACS job designation, we calculated the percentage of workers within
each industry classification as described above. We also calculated a distribution
for the education attained and the 10th, 25th, 50th, 75th, and 90th salary percentiles for each
ACS job designation.
In our final step, we combined the BLS employment data with the ACS survey data. For
each job designation, we used the percentage within each industry classification
as weights to average the percentage of employment change calculated using the BLS
data. For the few industry designations that did not match the 2- and 3-digit NAICS
data from the BLS, we simply omitted those workers from our calculation.