Testing Gibrat's law for small, young and innovating firms

This article analyses whether small, young, and innovating firms have experienced a greater employment growth than other Spanish firms over the period 1990-2000. The study draws upon a sample of 1272 manufacturing firms in which only 967 of the firms survived for the entire ten year period. The analyses test Gibrat's law, both by least squares and by utilizing the procedure proposed by Heckman in which a probit survival equation is first estimated to correct for sample selection bias. Two estimators correcting for selection bias are utilized: one that incorporates the inverse Mill's ratio and the other that employs maximum likelihood methods. All the results reject Gibrat's law and support the proposition that small firms have grown larger. Additionally, the results show that old firms grow less than young ones, and innovating activity - both process and product - is a strong positive factor in the firm's survival and its employment growth.