As a first step towards understanding the evolution of small groups in firm
s, this paper develops and tests a simple decision-theoretic model of resea
rch-unit evolution in which, as managers resolve their uncertainty over tim
e, they shut down under-performing units and remove under-performing unit h
eads. The selection process generates several testable hypotheses about how
unit and head characteristics affect the assignment of heads to units and
resource allocation within the firm. Data on research units in Alms is used
to test the model and estimate the relative importance of the different ef
fects on resource allocation. (C) 2000 Elsevier Science B.V. All rights res
erved.