Gibrat’s law and legacy for non-profit organisations: a non-parametric analysis

Abstract

Gibrat’s Law of proportional effect (i.e. growth is independent of initial size) has been tested for firms for several decades. In this paper I test Gibrat’s Law for charities in England and Wales through time. Using a data set based on the population of registered charities from 1998 to 2009, I am able to test the ‘ex ante’ hypothesis that Gibrat’s Law holds over the long run for a sample of charities as well as testing Gibrat’s Legacy (that Gibrat’s Law holds for large and mature organisations), the ‘ex post’ hypothesis. I use nonparametric local polynomial smoothing techniques which are more robust to the issues of autocorrelation, sample selection and truncation that make the conventional parametric approaches to testing Gibrat’s Law difficult in practice. Results suggest that the dynamic processes driving growth in the charitable sectors may differ from those driving the growth of firms. Unlike for-profit firms Gibrat’s Law is found to generally hold when controlling for selection both ‘ex ante’ and ‘ex post’. Results may be driven by the absence of a minimum efficient scale which charities must achieve to survive and the different funding profiles of charities.