So, Is There a Startup Boom?

The question behind this series is not whether the United States has experienced meaningful business dynamism. The evidence clearly shows movement: establishments open, expand, contract, and close; the BDM and BDS publish different establishment totals; and the two frames describe sharply different distributions of small and large operating units. The harder question is what kind of dynamism those measures represent. A defensible answer must keep the counted unit, the administrative frame, and the parent-firm relationship together.

The conclusion is conditional. There may be a surge in establishments or in newly active local units under the BLS-derived frame. That observation is not the same as proof of a surge in independent firm formation. The sources support a richer statement: the apparent boom combines changes in observed operating units with differences in who enters the statistical frame and how common organizations are divided into establishments. The difference is economically important precisely because it changes the interpretation of scale, competition, and policy sensitivity.

What the evidence supports—and what it does not

Start with the BDM definition. An opening is a unit with positive employment in March of the current year after zero employment in March of the previous year. Gross job gains include jobs added at openings and expanding units; gross job losses include jobs lost at closings and contracting units. These are valuable measures of employment movement through operating units. They can show where new jobs appear and where existing employment shrinks or disappears.

But an opening is not a legal biography. It can belong to a newly formed firm, an established firm adding a branch, or an organization represented differently as a work site becomes visible in the frame. Establishment age begins when the local unit first reports positive employment. Firm age is classified using the oldest establishment in the firm. An old firm can therefore have a young establishment. The sources do not provide a universal conversion from openings to independent startups, and the definitions do not support one.

That distinction should not be used to dismiss establishment data. A new clinic, warehouse, restaurant location, or care-service site can create jobs and change local access to services even if its parent organization is old. Establishment dynamism is a real economic phenomenon. It simply answers a different question from independent entrepreneurship. The mistake is not counting establishments; it is silently treating establishment entry as firm entry.

The aggregate evidence shows why the distinction matters. The BDM and BDS series were close at roughly 6.5 million private-sector establishments in the early 2000s, then diverged. For March 2023, the paper reports approximately 9.2 million BDM/QCEW establishments and approximately 7.4 to 7.5 million BDS/LBD establishments. Its different sections present the rounded totals and gap slightly differently, so the responsible description is a difference on the order of 1.6 to 1.8 million, not a falsely exact single number.

The direction of the employment comparison is opposite. The appendix reports about 137.5 million employees in BDS and about 127.9 million in BDM for 2023, a difference of roughly 9.5 million. The exact figures are rounded and belong to a published-aggregate comparison, but the relationship is clear: the frame with fewer establishments can report more employment. Unit count and workforce count are not interchangeable measures of economic scale.

The size distribution makes that divergence concrete. Across the eight supersectors in the paper’s 2023 table, BDM reports 7.0902 million establishments with 1–19 employees, compared with 5.5803 million in BDS—a 23.8% surplus. For establishments with 500 or more employees, BDM reports 17.5 thousand while BDS reports 22.5 thousand, making the BDS count 25.1% higher.