Newly established firms often try to secure their market position by buildi
ng up a base of loyal customers. While recessions may not destroy technolog
ical leadership, they may be harmful for such firm-customer relationships.
Without such customer bases, these firms find themselves more vulnerable to
attacks by competitors. We formulate this idea within an Aghion-Howitt-typ
e model of creative destruction and discuss its implications for growth. In
the context of this model, recessions might be good for growth since they
weaken the incumbent firm's position and, thereby, stimulate research by ou
tside firms. The model allows for the extreme case where the leading firm c
an be so entrenched that growth ceases, unless a recession shakes up its cu
stomer base. We find a one-to-one relationship between the average growth r
ate and the cyclical variability, a U-shaped relationship between the avera
ge speed of building up good customer relationships and the average growth
rate, and a positive relationship between the arrival rate of recessions an
d average growth. It is finally shown that an appropriate stochastic tax pr
ogram can implement the social planner's solution. In some cases, general-e
quilibrium effects may generate interesting results, conflicting with intui
tion from a partial-equilibrium approach: we show that, in some cases, a so
cial planner might want to subsidize research in order to discourage it.