Learn More about the Methods Applied to Develop the Resources Included in the Code Set

We applied the Federal Reserve Bank of Philadelphia’s list of NAICS industries and essential industry designations¹ corresponding to v3.0 of the CISA advisory list as a starting point for development of an updated list using CICs. We identified essential industry designations corresponding to v4.0 of the CISA advisory list for all six-digit NAICS codes and cross-walked NAICS codes and essential industry designations to CICs. We also identified estimated numbers of U.S. workers in all six-digit NAICS and Census industries. Specifically, we completed the following steps:

1. Broke each four-digit or five-digit NAICS code from the Philadelphia Federal Reserve list into all six-digit 2017 NAICS component industries² (see Table 1: Essential industry designations by six-digit NAICS codes). The Philadelphia Federal Reserve list includes a mixture of four-, five-, and six-digit NAICS codes, where the level of specificity was determined as the level at which all component industries shared an essential industry designation (i.e. all industries within that code were considered essential or non-essential under v3.0 of the CISA advisory list).¹ We applied the same essential industry designation as the Philadelphia Federal Reserve designation for each higher-level industry to all component industries
2. Identified changes to the CISA advisory list from v3.0³ to v4.0⁴ and updates to the Philadelphia Federal Reserve essential industry designations
   1. Updated the essential industry designations for industries that were added to or removed from the list between versions
   2. Noted industries for which the essential industry designation for a six-digit NAICS code did not correspond to the Philadelphia Federal Reserve designation for the higher-level NAICS code, or for which our coders and CISA reviewers applied a different designation than the Philadelphia Federal Reserve coders. All essential industry designations were reviewed by three occupational health scientists and CISA representatives to minimize coder error
3. Identified the essential critical infrastructure sector and sector-based major bullet number (solid black bullets) in v4.0 of the CISA advisory list⁴ under which each industry was considered essential, or determined that the industry did not correspond to any essential worker roles as defined by CISA
   1. Many NAICS codes may be appropriately classified as essential under more than one CISA essential critical infrastructure sector or worker role; each is identified under the sector and major bullet that most closely describe it
   2. Some six-digit NAICS codes include both essential and non-essential sub-industries; we designated NAICS codes as essential if they contained any sub-industries that corresponded to worker roles in v4.0 of the CISA advisory list because overinclusion at this stage captured only a minimal portion of the non-essential workforce and no standardized coding system distinguishes between sub-industries beyond the six-digit NAICS level. As an example, we designated “Other Scientific and Technical Consulting Services” as an essential industry because it includes biological consulting services and chemical consulting services that may be critical to biomedical research and vaccination and medication manufacturing and distribution, among other essential worker roles. However, it also includes motion picture consulting services, which do not fulfill an essential worker role as defined by v4.0 of the CISA advisory list. These cases account for many of the discrepancies between our designations and the Philadelphia Federal Reserve designations
   3. Several worker roles in the CISA advisory list are designated essential through their contributions to other essential industries. For example, bullet #5 in the Critical Manufacturing sector (p20) reads “Workers manufacturing or providing parts and equipment that enable the maintenance and continued operation of essential businesses and facilities.” We deferred to the Philadelphia Federal Reserve essential industry designations for manufacturing industries which may be applicable under these roles and could not be clearly identified under a separate worker role in v4.0 of the CISA advisory list. These designations were derived from the Bureau of Economic Analysis’s 2012 input-output tables⁵ identifying buyers and sellers of intermediate products and its 1997 capital flow table⁶ tracking product sales
4. Obtained estimated population sizes for each six-digit NAICS industry from the BLS Quarterly Census of Employment and Wages (QCEW) 2019 annual averages⁷
   1. The QCEW is a quarterly census of employers conducted using data reported to state unemployment insurance programs⁸
   2. QCEW data are estimated to include >95% of all US workers. Data include full and part-time workers in permanent and temporary positions. Workers who earned no wages during the entire applicable pay period are not included and workers employed by multiple eligible employers may be included multiple times⁸
   3. Populations not covered by the QCEW include proprietors, self-employed owners of unincorporated, non-employer businesses, unpaid family members, certain farm and domestic workers not covered by state unemployment insurance programs, railroad workers covered by the railroad unemployment insurance system, members of the armed forces, and select categories of government employees⁸
   4. The Bureau of Economic Analysis estimated that there were approximately 9.5 million persons in the US who primarily worked as self-employed owners of unincorporated, non-employer businesses in 2019.⁹ These workers are predominantly present in the Professional and business services (2.0 million), Construction (1.7 million), Educational services, health care, and social assistance (1.0 million), and Other services, except government (1.0 million) sectors.⁹ Many of these workers likely work in essential critical infrastructure industries and some may be considered frontline essential workers, including many who participate in electronically mediated employment, also termed “gig workers”.¹⁰ Other statistics on non-employer businesses are available in the US Census Bureau NES-D Tables¹¹
   5. Population estimates are not available for the NAICS industries “Dual-Purpose Cattle Ranching and Farming” and “Offices of Notaries”
   6. QCEW data may also be applied to generate population estimates by six-digit NAICS industry codes at the metropolitan statistical area (MSA), county, and state levels. Some data suppression is present at small geo-scales to protect worker privacy
5. Collapsed six-digit NAICS industries into the highest level at which all component industries shared a single essential industry designation to identify condensed NAICS codes
6. Cross-walked condensed 2017 NAICS codes to 2012 CICs¹² (Table 1: Essential industry designations by CIC and condensed NAICS)
   1. CICs are broader than NAICS and many contain multiple NAICS codes,¹³ some of which may not share the same essential industry designation. We identified Census industries containing essential and non-essential NAICS codes as mixed essential/non-essential to avoid overinclusion of broad portions of the non-essential workforce. For example, “Miscellaneous Manufacturing, N.E.C.” (CIC=3980) includes “Burial Casket Manufacturing” (NAICS code=339995) and “Musical Instrument Manufacturing” (NAICS code=339992), among others, and was designated as mixed essential/non-essential
7. Identified the estimated numbers of U.S. workers aged ≥16 years in 2018 in each Census industry using the National Institute of Occupational Safety and Health (NIOSH) Employed Labor Force (ELF) Query System,¹⁴ which applies data from the Bureau of Labor Statistics (BLS) Current Population Survey (CPS) public use microdata sample
   1. The CPS is a monthly household survey of the civilian, non-institutionalized population conducted to measure employment. The NIOSH ELF query system applies a subset of CPS data including only employed persons¹⁵
   2. Industry and occupation are reported for workers’ primary jobs and estimates are derived using composited final weights. Annual worker estimates are calculated by dividing the designated weights for all records in the specified year by 12, then summing. Estimates below 1,000 should be considered unstable and population estimates are reported in thousands¹⁵
   3. Estimated number of workers in the Census industry “Not specified type of mining” and all U.S. armed forces industries are not available in this data source
   4. Some CICs correspond to multiple condensed NAICS codes and are included multiple times in the Essential industry designations by CIC and condensed NAICS Do not sum this column directly to identify estimated population sizes
   5. The NIOSH ELF Query System may also be applied to generate population estimates by Census industry for U.S. regions, states, and territories

Essential critical infrastructure workers (ECIWs) may experience differential risk of exposure to COVID-19 by both industry and occupation. Census codes are well-suited to simultaneous evaluation of risk by industry and occupation because Census occupation codes are predicated on Census industry codes, and workers in the same occupation in separate industries may thus have different occupation codes.¹³ For example, laborers in construction industries are coded under a different COC than laborers in agricultural industries. We identified Census occupations within each Census industry as a foundation for public health researchers and practitioners evaluating COVID-19-related factors by both industry and occupation. This resource may also be applied to identify populations of interest by occupation within specific industries that may experience greatest COVID-19 exposure risk or be more likely to be overlooked by targeted interventions. Specifically, we completed the following steps:

1. Applied the NIOSH ELF Query System¹⁴ to cross-tabulate 2012 Census industries and 2010 Census occupations for US workers aged ≥16 years in 2018 (Table 1: CIC-COC pairs)
   1. This approach revealed some uncommon pairs; some industry and occupation pairs may be very rare, misreported, or miscoded in the BLS CPS public use microdata sample used to fulfill ELF queries
   2. 52 of 535 non-military 2010 Census occupations are not reported in ELF industry and occupation cross-tabulations due to small population sizes. Consider reviewing these occupations for potential inclusion as appropriate if evaluating a specific industry of interest
   3. ELF estimates below 1,000 are considered unstable. Although we retained all industry and occupation pairs to capture the largest possible breadth of occupations within each industry, others may choose to apply a threshold to the estimated number of workers in an industry-occupation pair in the US or refine the list of occupations within a specific industry of interest
2. ELF data applied for industry and occupation cross-tabulation are subject to the same limitations reported above under estimation of Census industry populations

1. Atalay E, Fujita S, Mahadevan S, et al. Reopening the economy: What are the risks, and what have states done?pdf iconexternal icon Philadelphia, PA: Federal Reserve Bank of Philadelphia; 2020.
2. Office of Management and Budget. North American Industry Classification System: United States, 2017pdf iconexternal icon. Washington, D.C.: Executive Office of the President, OMB; 2017.
3. Cybersecurity and Infrastructure Security Agency. Advisory memorandum on identification of essential critical infrastructure workers during COVID-19 responsepdf iconexternal icon. Washington, D.C.: US Department of Homeland Security, CISA; 2020.
4. Cybersecurity and Infrastructure Security Agency. Advisory memorandum on ensuring essential critical infrastructure workers ability to work during the COVID-19 response.pdf iconexternal icon Washington, D.C.: US Department of Homeland Security, CISA; 2020.
5. US Bureau of Economic Analysis. Input-Output Accounts Dataexternal icon. Suitland, MD: US Department of Commerce, BEA; 2020.
6. US Bureau of Economic Analysis. Capital Flows in the U.S. Economy, 1997external icon. Suitland, MD: US Department of Commerce, BEA; 2020.
7. US Bureau of Labor Statistics. Quarterly Census of Employment and Wages: QCEW Data Filesexternal icon. Washington, D.C.: US Department of Labor, BLS; 2020.
8. US Bureau of Labor Statistics. Quarterly Census of Employment and Wages: Employment and Wages, Annual Averages 2019external icon. Washington, D.C.: US Department of Labor, BLS; 2020.
9. US Bureau of Economic Analysis. National Income and Product Accounts: Table 6.7D. Self-Employed Persons by Industry. Suitland, MD: US Department of Commerce, BEA; 2020.
10. US Bureau of Labor Statistics. Labor Force Statistics from the Current Population Survey: Electronically Mediated Employment.external icon Washington, D.C.: US Department of Labor, BLS; 2020.
11. US Census Bureau. NES-D Tables.external icon Suitland, MD: US Department of Commerce, US Census Bureau; 2020.
12. US Census Bureau. Industry and Occupation Code Lists & Crosswalksexternal icon. Suitland, MD: US Department of Commerce, US Census Bureau; 2020.
13. US Census Bureau. American Community Survey: Recent changes in the Census industry and occupation classification systems. Technical Paper 78, March 2020. Suitland, MD: US Department of Commerce, US Census Bureau; 2020.
14. US Centers for Disease Control and Prevention. The Employed Labor Force (ELF) query system. Atlanta, GA: US Department of Health and Human Services, CDC; 2019.
15. US Centers for Disease Control and Prevention. The Employed Labor Force (ELF) query system: Technical Info. Atlanta, GA: US Department of Health and Human Services, CDC; 2019.