The Coverage Mechanism: Why Home-Care Work Appears as a Surge
The aggregate BDM–BDS gap shows that the two frames count different establishment landscapes, but it does not say whether the difference comes from missing employers or from different ways of organizing common firms. Education and Health Services provides a concentrated coverage case. The focus here is NAICS 624120, Services for the Elderly and Persons with Disabilities, where the linked analysis finds many establishments associated with firms in the QCEW-derived frame but not in the Census-derived frame. The case helps readers identify a coverage boundary without treating one industry as a complete explanation of the national gap.
The central lesson is that an establishment count can depend on how a worker’s employer reports employment, wages, and organizational status. In home-care and related services, the relevant reporting relationship may involve a private household, a state unemployment-insurance system, and a Census employer frame built around a different federal tax form. The observed “surge” is therefore not just a story about entrepreneurs opening storefronts. It is also a story about which employment relationships become statistical establishments.
Why this industry changes the diagnosis
Education and Health Services stands out in the paper’s comparison of the BDM and BDS establishment totals. The difference is concentrated in a detailed industry rather than being evenly distributed across every kind of business. That concentration matters diagnostically. If the national gap were simply caused by a universal tendency of one frame to split every firm into more locations, the same pattern would be expected broadly across industries. A sharply concentrated case points first toward a particular coverage boundary or reporting practice.
The detailed industry is Services for the Elderly and Persons with Disabilities, classified as NAICS 624120. The paper describes activities including day care, non-medical home care, homemaker services, social activities, group support, and companionship. These services can be delivered outside a conventional residential facility and may involve a worker serving a household directly. The economic activity is real, but its statistical representation depends on how the employer and work relationship enter the administrative systems.
The linked comparison for 2016 is striking. The Employer Characteristics File, which carries establishment-level QCEW information inside the Census Bureau’s LEHD infrastructure, contains about 539 thousand establishments associated with firms present in the ECF but absent from the LBD. The reverse category contains about 6 thousand establishments. The resulting net difference is about 532 thousand establishments in this detailed industry.
The categories matter as much as the totals. These are not simply 532 thousand ordinary firms that one agency forgot to count. The comparison says that many establishments are attached to firms visible in the QCEW-derived ECF but not visible in the LBD. That is evidence of a coverage difference between the frames.