Review of Industrial Organization 17: 249–265, 2000.
© 2000 Kluwer Academic Publishers. Printed in the Netherlands.
Firms’ Growth, Size and Age: A Nonparametric
JOSÉ C. FARIÑAS and LOURDES MORENO
Universidad Complutense and PIE-FEP, Plaza Marqués de Salamanca, 8, 28006 Madrid, Spain
Abstract. This paper offers empirical evidence of ﬁrm failure rates as well as the mean of the
distribution of realized growth rates, distinguishing between the sample of non-failing ﬁrms and the
sample of all ﬁrms, failing and non-failing. Attention is directed at identifying a set of characteristics,
in particular the size and age of ﬁrms, systematically related to the patterns of ﬁrm growth and exit,
using a panel of Spanish manufacturing ﬁrms. The two main contributions of the paper are the use
of nonparametric techniques and the analysis of issues ignored in other studies like the regression-
to-the-mean bias and the measurement of learning effects. We ﬁnd evidence that failure rates and
the mean growth rate of successful ﬁrms decline with size and age. When failing ﬁrms are integ-
rated, there are no signiﬁcant differences in the mean growth rate across the age and size of ﬁrms.
Regression-to-the-mean does not prove to be a substantial factor behind the negative relationship
between size and growth of surviving ﬁrms.
Key words: Age, exit, ﬁrm growth, learning, size.
This paper examines the patterns of ﬁrm exit and growth, paying attention to tradi-
tionally major issues in theoretical and empirical research on ﬁrm growth. Do small
ﬁrms grow faster than large ﬁrms? Do young ﬁrms grow faster than old ﬁrms? For
several decades the conventional wisdom has been that expected growth rates are
independent of ﬁrm size, a property known as Gibrat’s law. Theoretical works that
followed this path took the law as an assumption or as a desirable implication; in
particular Simon and Bonini (1958) assume that Gibrat’s law holds for ﬁrms above
the minimum efﬁcient size level.
More recent theoretical literature on industry evolution has emphasized the im-
portance of learning as far as ﬁrm growth and changes in market structure are
concerned. Different models emphasize different factors determining ﬁrm dynam-
ics: Jovanovic (1982), the learning process about innate efﬁciencies; Ericson and
The authors are particularly grateful to Miguel A. Delgado, Jordi Jaumandreu, Mark J. Roberts
and two anonymous referees for helpful comments and suggestions. An earlier version of this paper
was presented at the 24th Annual EARIE Conference in Leuven and at the Jornadas de Econom
Industrial in Madrid. This research has been partially funded by the project CICYT SEC497-1368.