The apparent failure of economists thus far to establish a positive em
pirical link between interest rates and saving does not, by itself, di
scredit the hypothesis of a direct structural relationship between the
two, ceteris paribus. This structural relationship may be shifting ab
out in response to changes in exogenous variables such as tastes and t
echnology in a way that is consistent with any type of reduced-form co
rrelation between interest rates, and saving in the data. This point i
s demonstrated within a simple model of optimal saving, interest rates
and economic growth. The different implications of endogenous versus
exogenous growth are explored in this context.