Bank of America internal data shows the lower-income share of new business applications rising from roughly 16% in 2020 to 22% in 2026 and the higher-income share falling from 53% to 45%.
Census BFS, on the same rolling twelve-month basis, shows total applications up 15.3% year on year, while high-propensity applications are up only 2.5%, and applications with planned wages are down 15.5%.
What does this imply?
1) Minimum efficient scale has genuinely fallen. Payment infrastructure, marketplace distribution and now AI tools let one person run what previously required a payroll. Firms that would once have needed staff now run on none. On this reading the falling planned-wages share is a productivity result.
2) This is necessity self-employment. People register businesses when hiring is weak, and entry-level hiring has been weak. On this reading a rising lower-income share of applications is a labour market symptom, and the falling planned-wages share reflects firms with no capacity to hire rather than no need to.
The data cannot separate them yet. The survival rates past two years, and whether these firms later transition to payroll will provide a clearer picture. If minimum efficient scale has fallen, non-employer firms should survive at normal rates and generate normal revenue per firm. If this is necessity entrepreneurship, they should fail at elevated rates and never cross into employer status.
The bottom line is that business formation is rising and its composition is getting weaker. A record number of Americans are registering businesses, and a falling number of them plan to pay anyone a wage. Read as lower barriers to entry, that is a productivity story. Read as the absence of alternatives, it is a labor market warning.
By Amir Kh.
Leave a Reply