Salary data presented in an interactive visualization by occupation, and by metropolitan area.
The Bureau of Labor Statistics, an agency of the United States Department of Labor, gathers data on employment and wages in a program titled Occupational Employment Statistics. BLS describes the program:
The OES program produces employment and wage estimates for over 800 occupations. These are estimates of the number of jobs in certain occupations, and estimates of the wages paid to them. These estimates are available for the nation as a whole, for individual States, and for metropolitan statistical areas (MSAs), metropolitan divisions, and nonmetropolitan areas; national occupational estimates for specific industries are also available. (1)Bureau of Labor Statistics. Occupational Employment Statistics. Available at https://www.bls.gov/oes/oes_ques.htm#overview.
OES data is gathered through a semi-annual mail survey of non-farm establishments. Data is released annually in May using three years of data to improve reliability. For example, “The May 2017 employment and wage estimates were calculated using data collected in the May 2017, November 2016, May 2016, November 2015, May 2015, and November 2014 semi-annual panels.” (2)Bureau of Labor Statistics. Occupational Employment Statistics. Available at https://www.bls.gov/oes/oes_ques.htm#overview.
BLS presents data in a hierarchy. (3)Bureau of Labor Statistics. May 2017 Occupation Profiles. Available at https://www.bls.gov/oes/current/oes_stru.htm. At the tops are groups, like “Production Occupations” in the nearby example.
There are then one or more subgroups like “Computer Numerically Controlled Tool Operators and Programmers.” Then, there are actual occupations, like “Computer Numerically Controlled Tool Programmers” and “Computer Numerically Controlled Tool Operators.”
The visualization I created uses the hierarchical nature of the data: Groups, Subgroups, and Occupations. In the visualization, when you hover the mouse of a column heading, a “+” or “-” may appear. Click on these to expand or contract the data. This is also known as “drill down.”
In “Table by Area,” this data is provided:
- Total Employees.
- Jobs per 1000. This is the number of jobs in the occupation per thousand jobs in the area.
- Location Quotient. BLS says: “As measured here, a location quotient shows the occupation’s share of an area’s employment relative to the national average. … For example, a location quotient of 2.0 indicates that an occupation accounts for twice the share of employment in the area than it does nationally, and a location quotient of 0.5 indicates the area’s share of employment in the occupation is half the national share.”
As an example, the location quotient for aerospace engineers in Wichita is 15.14. This is a relatively high value, which is not surprising given the concentration of that industry in Wichita.
- Mean salary. For each group of areas, these values are also provided: Rank, Difference from highest ranked, and percent difference from highest-ranked.
- Median salary. For each group of areas, these values are also provided: Rank, Difference from highest ranked, and percent difference from highest-ranked.
There are also tabs with charts showing graphs of mean salary, location quotient, and jobs per thousand.
This visualization holds data only for occupations with salary expressed as an annual value.
Comparing average salaries for groups of occupations in different cities has problems. One is the number of workers in occupations. Considering management occupations, there are few chief executive officers but many other managers. The weight of the number of workers needs to be considered.
Also, the magnitude of salaries is an issue. Chief executive officer salaries vary widely, by tens of thousands of dollars. The data tells us that in 2017, a CEO in Wichita earns $65,400 less than in Des Moines. That variation is greater than the average salary across all occupations.
Data is for the May 2020 release, the most current available. Data for each year is usually released in spring.
Click here to access the visualization.
For more visualizations, click here.