The Incorporation Premium: Descriptive Evidence on Business Structure and Self-Employment Earnings

Abstract

This paper provides descriptive evidence on the earnings gap between incorporated and unincorporated self-employed workers in the United States. Using 2022 American Community Survey data on 135,952 self-employed workers aged 25–65, I document a raw income gap of \$41,350 annually. Adjusting for demographics, education, and hours worked using regression and propensity score methods, the conditional gap is \$23,843 (95% CI: \$22,416–\$25,270). Sensitivity analysis using the Cinelli-Hazlett framework yields a robustness value of 0.11, meaning an unmeasured confounder would need to explain 11% of residual variance in both treatment and outcome to fully eliminate the conditional association. I emphasize that these are descriptive correlations, not causal estimates: unobserved factors like business scale, growth orientation, and entrepreneurial ability plausibly explain much of the remaining premium. The contribution is methodological—demonstrating how sensitivity analysis can transparently quantify the conditions under which observational estimates would fail—and descriptive, establishing a new stylized fact about heterogeneity within self-employment.