Premium
SIZE‐CHANGES AMONG BUSINESS FIRMS
Author(s) -
Engwall Lars
Publication year - 1970
Publication title -
metroeconomica
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.256
H-Index - 29
eISSN - 1467-999X
pISSN - 0026-1386
DOI - 10.1111/j.1467-999x.1970.tb00149.x
Subject(s) - ranking (information retrieval) , gini coefficient , econometrics , mathematics , index (typography) , section (typography) , distribution (mathematics) , statistics , estimation , economics , inequality , mathematical economics , computer science , mathematical analysis , management , machine learning , world wide web , economic inequality , operating system
S ummary This paper deals with transition‐matrices as a tool to describe the changes within size‐distributions of firms. In a theoretical section the estimation technique is discussed and further uses of the matrices are described. The latter means derivation of steady‐state distributions and measures for mobility. The theoretical part is followed by another section, where empirical results are given. The approach has been applied to data concerning the largest firms with respect to sales within five areas, viz.:1) The United States 2) Countries outside the United States 3) Europe 4) Scandinavia 5) Sweden.In this section it can be seen that the probability of staying within a certain class is the largest followed in size by the probability of a transition to an adjacent class. Other probabilities are mostly estimated as zero. During the investigated period 1956–1965 movements towards a steady‐state can be seen. There still exist, however, some differences between actual and theoretical distributions. The average time spent in the different classes varies about 10 years. An index for mobility shows values, which mean mobilities, which are 10–20% of the mobilities of a perfect mobile industry. The indices follow fairly well a ranking based on Gini's coefficient of concentration. It is most necessary to obtain further knowledge concerning the process. Thus one extension of this study would be a study using the approach of Monte‐Carlo simulation.