Review of Industrial Organization
12: 389–398, 1997.
1997 Kluwer Academic Publishers. Printed in the Netherlands.
Do Firms Diversify Because Managers Shirk? A
Reinterpretation of the Principal-Agent Model of
DAVID C. ROSE
Associate Professor, Department of Economics, University of Missouri-St. Louis, U.S.A.
Abstract. In the principal-agent model of ﬁrm structure proposed by Aron (1988), diversiﬁcation
is optimal because it reduces managerial risk exposure. Diversiﬁcation has this effect because it
improves the monitoring function of output-contingent compensation which exists to alleviate the
managerial moral hazard problem facing the ﬁrm’s owners. But what if managers never shirk or
can be easily observed by ﬁrm owners so the moral hazard problem vanishes? Does the optimality
of diversiﬁcation vanish as well? This paper develops a self-selection variant of Aron’s model to
show that the answer to this question is no. A by-product of this exercise is that it suggests how one
might craft a test of whether the principal-agent model of diversiﬁcation is driven by the problem of
managerial moral hazard or managerial self-selection.
Key words: Diversiﬁcation, principal-agent, managerial compensation.
Debra Aron’s (1988) principal-agent model of ﬁrm structure provided industrial
organization economists with a remarkably general explanation for the puzzle of
ﬁrm diversiﬁcation. According to her model, diversiﬁcation reduces managerial
risk exposure by making observations of ﬁrm output a more reliable indicator of
managerial effort. Diversiﬁcation does this by increasing the number of sample
points over which the ﬁrm infers managerial effort, effectively “averaging-out”
idiosyncratic ﬂuctuations in output that have nothing to do with the manager’s
effort level. Since risk averse managers must be compensated for risk bearing, this
beneﬁts the ﬁrm because it reduces the cost of using output-contingent compensa-
tion without attenuating the incentive aligning properties of such compensation.
Like all principal-agent models that involve managerial behavior, Aron’s mod-
el assumes that the manager’s utility function is declining in effort and that the
manager’s effort level is unobserved by ﬁrm owners. Together these two assump-
tions make managerial shirking a potential moral hazard problem for ﬁrm owners,
For an example of a model in which ﬁrm diversiﬁcation is a consequence of managerial risk
exposure but diversiﬁcation is optimal even in the absence of output-contingent compensation, see
Rose (1992). This model also shows that even risk neutral managers have an incentive to reduce their
risk exposure by diversifying their ﬁrms.