Review of Industrial Organization 13: 271–293, 1998.
© 1998 Kluwer Academic Publishers. Printed in the Netherlands.
An Applied Econometrician’s View of Large
P. A. GEROSKI
London Business School, Regent’s Park, London NW1 4SA, U.K.
Abstract. This paper contains a brief survey of recent empirical work on the performance of large
companies. It tries to pull together the literature in the form of six stylized facts, illustrating them with
data drawn from a single sample. The paper concludes by highlighting the issues which are thrown up
for future work. These are: accounting for persistent heterogeneities between ﬁrms, accounting for
the apparently erratic performance of many ﬁrms and, ﬁnally, moving away from hypothesis testing
driven empirical agendas.
Paul Geroski has just ﬁnished a distinguished year as President of the Indus-
trial Organization Society, in line with the contributions by earlier distinguished
Presidents. Professor Geroski here offers a ﬁrst: the Presidential Address by the
outgoing head of the Society. The Society is the parent of the Review of Indus-
trial Organization, and members of the Society receive the Review. The Review
is pleased to make this new tradition possible. We hope that it will develop and
continue. Ancient and modern experience teaches that presidential addresses can
vary widely in style and subject. Professor Geroski has here chosen to present a
substantial survey of recent research on a speciﬁc and important topic: the perfor-
mance or large companies. He develops a series of important facts, and he draws
interesting and important lessons for future research. He has chosen not to use the
occasion as a platform for giving his thoughts about the industrial organization
ﬁeld as a whole, or about industrial policies, or any of the other interesting and
contentious matters that the readers of the Review probably enjoy. Perhaps future
annual Presidential addresses will explore more of the forms, styles and issues –
staid or spectacular – which this space can host. For now, this Address offers an
impressive and signiﬁcant debut.
Many of the ideas and arguments developed in this paper ﬁrst surfaced during work with some
of my recent co-authors, particularly Chris Walters, John Van Reenen, Steve Machin, Paul Gregg and
Giovanni Urga. I am also obliged to Steve Martin, Martin Conyon, Geoff Shepherd, Jose Mata, Steve
Davies and Karl Aiginger, for helpful comments on an earlier draft. The usual disclaimer applies.