This paper shows that employment in cohorts of US firms is strongly influenced by aggregate conditions at the time of their entry. Employment fluctuations of startups are procyclical, they persist into later years, and cohort-level employment variations are largely driven by differences in firm size, rather than the number of firms. An estimated general equilibrium firm dynamics model reveals that aggregate conditions at birth, rather than post-entry choices, drive the majority of cohort-level employment variation by affecting the share of startups with high growth potential. In the aggregate, changes in startup conditions result in large, slow-moving fluctuations in employment.

About one-half of all startups fail within five years, and those that survive grow at vastly different speeds. Using Census microdata, we estimate that most of these differences are determined by ex ante heterogeneity rather than persistent ex post shocks. Embedding such heterogeneity in a firm dynamics model shows that the presence of ex ante heterogeneity (i) is a key determinant of the firm size distribution and firm dynamics, (ii) can strongly affect the macroeconomic effects of firm-level frictions, and (iii) helps understand the recently documented decline in business dynamism by showing a disappearance of high-growth startups (“gazelles”) since the mid-1980s.