Overview

The matched-firm analysis shows that BDM and BDS can include the same parent organization yet assign it different establishment counts or single-unit/multi-unit status. For the change in the gap from 2004 to 2016, 53.3% was associated with firms present in only one source; the remaining contributions reflected organizational differences. In Professional and Business Services, common firms account for a large differential, but the 2016 tabulations do not prove more independent entry or a universally correct frame.

Learning outcomes

For the 2004–2016 change in the gap, the paper attributes 53.3% to firms present in only one source.
In 2016 Professional and Business Services, matched firms contributed a net differential of about 271 thousand establishments.
Different single-unit and multi-unit labels can reflect alternative organizational maps, not new independent companies.

Keywords

BDMBDSfirmsestablishmentslinkageProfessional and Business Services

The Organizational Mechanism: When One Firm Becomes Many Establishments

The home-care case showed a coverage difference: many establishments appeared in the QCEW-derived ECF among firms absent from the Census LBD. This chapter examines the other mechanism behind the BDM–BDS gap. It asks what happens when the same underlying firm appears in both sources but the two frames assign it different numbers of establishments or different single-unit and multi-unit status. The practical judgment is that more counted locations can reflect a different map of an existing organization, not a wave of independent competitive entry.

That distinction matters because a firm is not an establishment multiplied by a fixed number. A parent organization can operate several locations, and the statistical frame must decide which records belong together and how many establishments those records represent. If one frame sees branches and another sees fewer consolidated units, their establishment totals can diverge even when they agree that the parent organization exists.

The matched-firm decomposition: missing businesses are only part of the story

The linked analysis separates firms and establishments into categories that make the organizational mechanism visible. Some firms are single-unit in both sources. For those firms, agreement about the number of establishments does not contribute to the divergence. Other firms are multi-unit in both sources, but the number of establishments attached to them can differ. Still others are classified as single-unit in one source and multi-unit in the other. Finally, some firms appear only in one source.

These categories are not merely bookkeeping. They distinguish two questions that an aggregate total cannot answer. The first is whether the same firm appears in both business universes. The second is how many establishments each frame assigns to that firm. A firm can be common to both sources and still be a major contributor to the establishment gap if one frame records a larger network of locations.

For the change in the establishment-count difference from 2004 to 2016, the paper attributes 53.3% to firms appearing in only one source. That is the coverage component. But the remaining contributions are organizational: 23.4% comes from firms classified as single-unit in the LBD and multi-unit in the ECF; 21.2% comes from firms classified as multi-unit in both but associated with more establishments in the ECF; and 2.1% comes from the reverse pattern. The percentages are a decomposition for the studied period and scope, not a timeless formula for every establishment series.

The result changes how the national gap should be described. More than half of the measured change is associated with firms that appear in only one source, so coverage is important. But almost half is associated with firms that are present in both sources yet represented differently. The difference is therefore not just a list of employers missing from one register. It is also a disagreement about organizational structure.

The single-unit/multi-unit categories require a careful reading.

Published by DeepMox

FAQ

What does the matched-firm decomposition measure?

It measures how firms present in one or both sources contribute to the change in the establishment-count difference, separating exclusive coverage from alternative establishment structures.

Can the same firm be single-unit in one frame and multi-unit in another?

Yes. The frames can assign different establishment networks or statuses to a common firm without implying that its physical organization changed.

What did Professional and Business Services contribute in 2016?

Matched firms contributed a net differential of about 271 thousand establishments in the paper’s 2016 tabulation; the result is not a universal national estimate.

Previous10 minBDM vs BDS: NAICS 624120 Home-Care Coverage

Next11 minBDM vs BDS: How Measurement Shapes Policy Models

BDM vs BDS: One Firm, Different Establishment Maps

The Organizational Mechanism: When One Firm Becomes Many Establishments

The matched-firm decomposition: missing businesses are only part of the story

FAQ