In this work we explain the size distribution of business firms using a stochastic growth process that reproduces the main stylized facts documented in empirical studies. The steady state solution of this process is a three-parameter Dagum distribution, which possibly combines strong unimodality with a Paretian upper tail. Thanks to its flexibility, the proposed distribution is able to fit the whole range of firm size data, in contrast with traditional models that typically focus on large businesses only. An empirical application to Italian firms illustrates the practical merits of the Dagum distribution. Our findings go beyond goodness-of-fit per se, and shed light on possible connections between stochastic elements that influence firm growth and the meaning of parameters that appear in the steady state distribution of firm size. These results are ultimately relevant for studies into industrial organization and for policy interventions aimed at promoting sustainable growth and monitoring industrial concentration phenomena.

在这项工作中，我们使用随机增长过程解释了商业公司的规模分布，该过程再现了经验研究中记录的主要程式化事实。此过程的稳态解决方案是三参数Dagum分布，它可能将强大的单峰性与Paretian上尾相结合。与传统的模型通常只关注大型企业相比，由于其灵活性，建议的分布能够适应公司规模数据的整个范围。对意大利公司的经验应用说明了Dagum分布的实际优点。我们的发现超出了拟合优度本身，并揭示了影响企业成长的随机因素与企业规模稳态分布中出现的参数含义之间的可能联系。