Association for Public Policy Analysis & Management

Analyzing cohort data on all U.S. employers, we find that a small fraction of firms account for most employment both at firm birth and at age seven. Relative size is strongly persistent over time, and little of the size variation is accounted for by industry or amount of finance. We link universal and survey data on 55,800 founders of 37,100 start-ups and define “high growth” as the largest 5% of employers at age zero and seven. Other things equal, female-founded firms are 34% less likely to be high-growth, but a similar gap for African-Americans disappears by age seven. At entry, founder age is positively associated with high growth, but the profile flattens over time and even turns negative. The education profile is initially concave, with graduate degree recipients no more likely to found high growth firms than high school graduates, but the former catch up to those with bachelor’s degrees by firm age seven, while the latter do not. Most other relationships of high growth with founder characteristics are highly persistent over time. Prior business ownership is strongly positively associated, and veteran experience negatively associated, with high growth. A larger founding team raises the probability of high growth, while, controlling for team size, diversity (by gender, age, race/ethnicity, or nativity) either lowers the probability or has little effect. Controlling for start-up capital raises the high-growth probability of firms founded by women, minorities, immigrants, veterans, sole proprietors, and novice, younger, and less educated entrepreneurs.