Department of Labor Logo United States Department of Labor
Dot gov

The .gov means it's official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you're on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Article
December 2022

Jobs, jobs, jobs: what’s an analyst to do?

Analysts and economists often face the task of using employment metrics to characterize industries of interest. Some key challenges can be understanding where to find employment metrics, the differences in various employment metrics, and when each metric should be used. This article analyzes a variety of publicly available employment data for the United States and compares these data. A detailed description of the intricacies of each data source is provided, which covers factors such as regionality, industry breakout, periodicity, and the types of jobs included. This article provides several case study examples, using the oil and gas extraction, coal mining, and chemical manufacturing sectors to portray challenges data users may face when developing employment estimates that suit their needs. Data users should be aware of a variety of data sources to understand alternative analysis options when data limitations are present and to determine which data source best meets their needs. Instances may occur in which information from one dataset may be used to help impute missing values.

Several labor market statistics may be used to characterize an industry, including wages, benefits, occupational categories, productivity measures, output, employment, geographical characteristics, and numbers of full-time workers versus part-time workers. This article focuses on the metric number of employees, examining a variety of publicly available employment data series for the United States and comparing these data. A detailed description of the intricacies of each data source is provided, and several case study examples are included that use employment data from the oil and gas extraction, coal mining, and chemical manufacturing sectors. In addition, the following key questions are addressed:

·       What publicly available data sources provide U.S. employment estimates at the industry level?

·       What are some examples of challenges and issues data users may face when developing employment estimates?

·       How do these data sources differ in terms of regional representation, level of industry disaggregation, types of employees that are included (e.g., proprietors’ employment), periodicity (e.g., monthly, quarterly, annually), and the length of the time series (e.g., 1980-2020)?

·       When the data series of interest are not publicly disclosed, what are some alternative options to develop employment estimates?

Comparison of publicly available data sources

This section compares several publicly available employment datasets. Table 1 summarizes the data sources examined in this article.

Table 1. Summary of public employment data sources examined
PublicationSourceMost disaggregated seriesPeriodicityBeginning of time seriesMost granular level of detail

QCEW

BLS6-digit NAICSMonthly, quarterly, and annually2001County

SAEMP25

BEA3-digit NAICSAnnually1998[1]State

SAEMP27

BEA3-digit NAICSAnnually1998[1]State

CAEMP25

BEA2-digit NAICSAnnually2001[1]County

Annual coal report

EIAAnnually1989[2]State

Mine injury and worktime quarterly statistics

MSHAQuarterly and annually1993National

[1] Year is in terms of NAICS. Standard Industrial Classification codes are available in earlier years.

[2] Employment estimates are available for 1984, but a gap exists in data until 1989.

Note: BEA = U.S. Bureau of Economic Analysis, BLS = U.S. Bureau of Labor Statistics, CAEMP25 = total full-time and part-time employment by industry by county, EIA = Energy Information Administration, MSHA = Mine Safety and Health Administration, NAICS = North American Industry Classification System, SAEMP25 = total full-time and part-time employment by industry by state, SAEMP27 = full-time and part-time wage and salary employment by industry by state, and QCEW = Quarterly Census of Employment and Wages.

Source: “Employment by state” (Washington, DC: U.S. Bureau of Economic Analysis), https://www.bea.gov/data/employment/employment-by-state; “Employment by county, metro, and other areas” (Washington, DC: U.S. Bureau of Economic Analysis), https://www.bea.gov/data/employment/employment-county-metro-and-other-areas; Quarterly Census of Employment and Wages” (Washington, DC: U.S. Bureau of Labor Statistics, updated quarterly), https://www.bls.gov/cew/; “Annual coal report,” “Table 18: Average number of employees by state and mine type, 2021 and 2020” (Washington, DC: U.S. Energy Information Administration, October 2021), https://www.eia.gov/coal/annual/pdf/table18.pdf; and “Mine injury and worktime, quarterly” (U.S. Mine Safety and Health Administration, January–December 2020), https://arlweb.msha.gov/Stats/Part50/WQ/2020/MIWQ-2020.pdf.

The U.S. Bureau of Labor Statistics (BLS) Quarterly Census of Employment and Wages (QCEW) employment data include estimates by region and industry, with regional breakout, including county, metropolitan statistical area (MSA), state, and national levels.1 Data are available monthly, quarterly, and annually. QCEW data are available to the six-digit North American Industry Classification System (NAICS) level. (Note: Data are not disclosed in instances in which they do not meet BLS or state agency disclosure standards.)

QCEW counts are derived from administrative reports submitted by employers to state unemployment insurance agencies. Workers such as proprietors and the self-employed are not covered by unemployment insurance. They are therefore not included in these administrative reports or the resulting QCEW counts.2 For instance, if users wish to examine employment data for oil and gas extraction (NAICS 211), coal mining (NAICS 2121), and chemical manufacturing (NAICS 325) at the state level, then they would use the BLS QCEW series as follows:

·       Oil and gas extraction (NAICS 211) data are available monthly from January 2001 to June 2021 and annually from 2001 to 2020.3

·       Coal mining (NAICS 2121) employment data are available monthly and annually from 2001 to 2020.

·       Chemical manufacturing (NAICS 325) data are available monthly from January 2001 to June 2021 and annually from 2001 to 2020.

This article analyzes three series of annual employment by industry by either state or county.4 (Note: The U.S. Bureau of Economic Analysis [BEA] also provides a complete listing of BEA regional data availability.5)

·       Total full-time and part-time employment by industry by state (SAEMP25)

·       Full-time and part-time wage and salary employment by industry by state (SAEMP27)

·       Total full-time and part-time employment by industry by county (CAEMP25)

SAEMP25 includes both wage and salary employment and proprietors’ employment by industry, whereas SAEMP27 only includes wage and salary employment by industry. Proprietors’ employment is the number of sole proprietorships and the number of general partners.6 The proprietors’ employment category includes categories such as independent contractors.7 It is also important to point out that BEA estimates of proprietors’ employment are counts of how many proprietors are active during any portion of the year.8 Wage and salary employment includes all jobs for which wages and salaries are paid and includes both part-time and full-time jobs.9

For many industries, the estimates between SAEMP25 and SAEMP27 are similar. In certain industries, however, this similarity is not the case (e.g., oil and gas extraction [NAICS 211]), as table 2 demonstrates. It displays estimates for both series by industry for the United States in 2020. The far-right columns of table 2 show (by NAICS code) the difference between the two series in employment estimates and the ratio of employment estimates. Both of these BEA series tend to have data available at the two- and three-digit NAICS code levels but not more disaggregated in most instances. For instance, if users need to examine employment data for oil and gas extraction (NAICS 211), coal mining (NAICS 2121), and chemical manufacturing (NAICS 325) at the state level, then they would use either the SAEMP25 or SAEMP27 series as follows:

·       Oil and gas extraction (NAICS 211) and chemical manufacturing (NAICS 325) data are available annually from 1998 to 2020.

·       Coal mining (NAICS 2121) data are not available. The most disaggregated industry available in this instance is mining, except oil and gas (NAICS 212), which is available annually from 1998 to 2020.

Table 2. Employment estimate comparison of U.S. Bureau of Economic Analysis datasets SAEMP25 and SAEMP27, 2020

The major geographical breakouts of the CAEMP25 series include the United States, states, counties, MSA, micropolitan statistical area, combined statistical area, metropolitan division, metropolitan and nonmetropolitan portions, and BEA regions. One potential limitation of CAEMP25 is that disaggregated industry breakout is not available at the same level of industry detail as BLS QCEW, BEA SAEMP25, or BEA SAEMP27. For instance, if one examines CAEMP25 employment data at the state level for oil and gas extraction (NAICS 211), coal mining (NAICS 2121), and chemical manufacturing (NAICS 325), the following issues arise:

·       Oil and gas extraction (NAICS 211) and coal mining (NAICS 2121) data are not available. The most disaggregated industry available in this instance is mining, quarrying, and oil and gas extraction (NAICS 21), which is available annually from 2001 to 2020.

·       Chemical manufacturing (NAICS 325) data are not available. The most disaggregated industry available in this instance is manufacturing (NAICS 31 to 33), which is available annually from 2001 to 2020.

Case studies

This section provides four case studies that illustrate examples of issues and potential solutions analysts may face when developing employment estimates that suit their needs.

Case study 1: oil and gas extraction employment by state (Alabama)

Case study 1 provides an estimate of 2020 Alabama oil and gas extraction employment on the basis of BLS QCEW, BEA SAEMP25, and BEA SAEMP27. (See chart 1.) The largest employment estimate occurs in the SAEMP25 series, likely because of its inclusion of proprietors’ employment, which tends to be a large share of overall employment in the oil and gas extraction sector (NAICS 211), as shown in table 2. Analysts must either determine which series best represent the research needs or provide a range of employment estimates.

Case study 2: coal mining employment for the United States, including proprietors’ employment

Case study 2 provides two examples of how to estimate U.S. 2020 coal mining employment (including proprietors’ employment). The BLS QCEW estimate for NAICS 2121 U.S. employment for 2020 is 40,109. However, BEA SAEMP25 and SAEMP27 only disaggregate as far as NAICS 212. One option to develop an employment estimate that includes proprietors’ employment is to multiply the BLS employment estimate in NAICS 2121 for the year 2020 by the ratio of BEA SAEMP25 mining, except oil and gas (NAICS 212), employment to the BLS employment estimate for NAICS 212 employment. The BEA SAEMP25 U.S. employment estimate for NAICS 212 in 2020 is 218,200, whereas the BLS QCEW estimate is 176,114, implying a ratio of 1.23897. Multiplying this ratio by the BLS QCEW employment estimate of 40,109 yields an estimate of NAICS 2121 employment (including proprietors’ employment) of 49,694. (See chart 2.)

Another option is to seek an alternative data source. The U.S. Department of Labor’s Mine Safety and Health Administration (MSHA) publishes the Mine Injury and Worktime Report,10 which provides estimates for both coal mining operator employment and contractor employment at the national level. The current article assumes contractor employment falls into the proprietors’ employment category. The MSHA report shows operator coal mining employment of 43,749.11 (Note: This value includes employment in mines, preparation plants, independent shops/yards, and offices.) The report also indicates that U.S. mining contractor employment was 19,886 in 2020.12 Combining data from the MSHA report results in an estimate of coal mining employment (inclusive of proprietors’ employment) of 63,635, which is on the same order of magnitude as the estimate calculated in the previous paragraph (49,694) but is 28 percent higher. These results of case study 2 are summarized in chart 2.

Case study 3: coal mining employment by state (Pennsylvania)

Case study 3 demonstrates two options to estimate 2020 Pennsylvania coal mining (NAICS 2121) employment and discusses data issues that arise. As in case study 2, BEA SAEMP25 and SAEMP27 are only disaggregated as far as mining, except oil and gas (NAICS 212). Conversely, BLS QCEW coal mining (NAICS 2121) data are available monthly and annually during most years from 2001 to 2020 (not disclosed in either all or part of 2016 or 2020 because they do not meet BLS or state agency disclosure standards). The nondisclosure of data can challenge the data user.

In this instance, one option is to examine an alternative data source, such as the Energy Information Administration (EIA) annual coal report, which provides estimates of coal mine employment by state.13 The EIA estimate for 2020 Pennsylvania coal mining employment is 4,818. To provide a soundness check, Pennsylvania coal mining employment data can be compared for 2019, in which the EIA coal mining employment estimate was 5,432, whereas the BLS QCEW estimate for NAICS 2121 was 5,087. This match is relatively close because the EIA employment estimate is 6.8 percent higher than BLS QCEW estimate. In the absence of an alternative data source (EIA in this example), another possibility is to take the ratio of 2019 employment in NAICS 2121 to a more aggregated industry (NAICS 212, assuming it was published) and use that to impute the missing 2020 value. For instance, the 2019 BLS QCEW employment estimate for Pennsylvania employment in NAICS 212 (mining, except oil and gas) was 9,876 and the estimate for NAICS 2121 (coal mining) was 5,087. Dividing NAICS 2121 employment by NAICS 212 employment (0.515087) and then multiplying the ratio by 2020 NAICS 212 employment (8,934) results in an estimate of 4,602 employees.

The analysis requires further thought if an employment estimate that includes proprietors’ employment is desired. As previously discussed, BEA SAEMP25 and SAEMP27 are only disaggregated as far as NAICS 212. A second option to obtain an estimate for NAICS 2121 inclusive of proprietors’ employment is to take the following steps:

  1. Determine the ratio of BEA SAEMP25 mining, except oil and gas (NAICS 212), employment to the BLS employment estimate for NAICS 212. BEA SAEMP25 Pennsylvania employment estimate for NAICS 212 in 2020 is 11,933, while the BLS QCEW estimate of NAICS 212 employment is 8,394, implying a ratio of 1.4216.
  2. Multiply the ratio (1.4216) by employment estimates obtained in the previous paragraph (4,818 and 4,602) for 2020 Pennsylvania employment in NAICS 2121. This calculation results in imputed estimates of 6,849 and 6,542, respectively, which include proprietors’ employment.

The various estimates of case study 3 are summarized in chart 3.

Case study 4: chemical manufacturing employment by county (Washington, Pennsylvania)

Case study 4 provides estimates of 2020 Washington County, Pennsylvania, chemical manufacturing (NAICS 325) employment that either include or exclude proprietors’ employment. The BLS QCEW employment estimate is 581 in 2020. BEA SAEMP25 and SAEMP25 series are not available at the county level.

If analysts want to also include proprietors’ employment, an option is to use the ratio of the SAEMP25 Pennsylvania NAICS 325 employment estimate (43,730) calculated from the BLS QCEW Pennsylvania NAICS 325 employment estimate (42,233), which is 1.03545. This result would imply an employment estimate (including proprietors’ employment) of 602. The difference in estimates is much smaller than that in case study 1, as the ratio of proprietors’ employment to total employment is smaller in the case of NAICS 325 than NAICS 211. (See table 2.) The various estimates in case study 4 are displayed in chart 4.

Conclusion

This article summarizes publicly available data sources that will allow economists, analysts, and other data users to better understand options available to provide employment estimates of various regions and localities at the industry level, as well as understand differences in types of employees included (e.g., proprietors). Several publicly available employment data sources for the United States are examined and compared. A detailed description of the intricacies of each dataset is provided, including a description of factors such as regionality, industry breakout, and types of jobs included. This article provides several case study examples, using the oil and gas extraction, chemical manufacturing, and coal mining sectors to portray potential solutions to challenges analysts may encounter when developing employment estimates. Data users need to be aware of a variety of data sources to understand alternative analysis options when data limitations are present and to determine which data source contains the information that best meets their needs. This article provides an example of using values from one dataset to impute missing values for another dataset.

Future research may consider exploring additional employment data sources and could expand to include an examination of nonpublic data sources (e.g., IMPLAN14) and their potential uses by analysts. In the future, this article’s research could be broadened to add examples that users of publicly available employment data (e.g., BLS, BEA) could encounter, such as instances in which employment data are not readily available for a particular year. One option to impute employment is to use the ratio of production per employee if production is known in the year that employment is missing. For example, if coal mining employment is known only in 2019, but coal production is known for 2019 and 2020, then a 2020 employment estimate could be derived based on the 2019 production to employment ratio.

Future research could also examine how various pieces of literature use employment data. An example is economic input-output analysis, which estimates the direct, indirect, and induced economic impacts of the interindustry relationships among the various sectors within an economy.

Suggested citation:

Gavin C. Pickenpaugh and Justin M. Adder, "Jobs, jobs, jobs: what’s an analyst to do?," Monthly Labor Review, U.S. Bureau of Labor Statistics, December 2022, https://doi.org/10.21916/mlr.2022.31

Notes


1“Quarterly Census of Employment and Wages” (Washington, DC: U.S. Bureau of Labor Statistics, updated quarterly), https://www.bls.gov/cew/.

2 For more information, see U.S. Bureau of Labor Statistics, “Quarterly Census of Employment and Wages: QCEW overview,” https://www.bls.gov/cew/overview.htm.

3 Third quarter 2021 Quarterly Census of Employment and Wages (QCEW) data were released February 23, 2022, and 4th quarter 2021 QCEW data were released May 25, 2022. For more information on QCEW release dates, see U.S. Bureau of Labor Statistics, “Quarterly Census of Employment and Wages: schedule of news releases and full data availability of county employment and wages,” https://www.bls.gov/cew/release-calendar.htm.

4 “Employment by state” (Washington, DC: U.S. Bureau of Economic Analysis), https://www.bea.gov/data/employment/employment-by-state; and “Employment by county, metro, and other areas” (Washington, DC: U.S. Bureau of Economic Analysis), https://www.bea.gov/data/employment/employment-county-metro-and-other-areas.

5 “Regional data table availability” (Washington, DC: U.S. Bureau of Economic Analysis), https://apps.bea.gov/regional/docs/DataAvailability.cfm.

6 “State personal income and employment: concepts, data sources, and statistical methods” (Washington, DC: U.S. Bureau of Economic Analysis, September 2022), https://www.bea.gov/system/files/methodologies/SPI-Methodology.pdf.

7 Robert Habans, “Is California’s gig economy growing? Exploring trends in independent Contracting” (Los Angeles, CA: UCLA Institute for Research on Labor and Employment, June 2016), https://irle.ucla.edu/wp-content/uploads/2016/03/Is-Californias-Gig-Economy-Growing-Exploring-Trends-in-Independent-Contracting.pdf; and Ted Egan, “The gig economy in San Francisco: prevalence, growth and implications” (San Francisco, CA: Office of the Controller, Office of Economic Analysis, July 5, 2016), https://sfcontroller.org/sites/default/files/Gig%20Economy.final_.pdf.

8 “Local area personal income methodology” (Washington, DC: U.S. Bureau of Economic Analysis, November 2021), https://www.bea.gov/system/files/methodologies/LAPI-Methodology.pdf.

9 “Regional economic accounts: regional definitions” (Washington, DC: U.S. Bureau of Economic Analysis), https://apps.bea.gov/regional/definitions/.

10 “Mine injury and worktime, quarterly” (Department of Labor, U.S. Mine Safety and Health Administration, January–December 2020), https://arlweb.msha.gov/Stats/Part50/WQ/2020/MIWQ-2020.pdf.

11 Ibid, p. 4 (in table 1).

12 Ibid, p. 19 (in table 5). This employment number includes the same data categories as those in endnote 11, table 1.

13 “Annual coal report,” “Table 18: Average number of employees by state and mine type, 2021 and 2020” (Washington, DC: U.S. Energy Information Administration, October 2021), https://www.eia.gov/coal/annual/pdf/table18.pdf.

14 For more information, see IMPLAN at https://implan.com.

article image
About the Author

Gavin C. Pickenpaugh
gavin.pickenpaugh@netl.doe.gov

Gavin C. Pickenpaugh is an economist at the National Energy Technology Laboratory, U.S. Department of Energy, Pittsburgh, PA.

Justin M. Adder
justin.adder@netl.doe.gov

Justin M. Adder is an economist at the National Energy Technology Laboratory, U.S. Department of Energy, Pittsburgh, PA.

close or Esc Key