Within the one-sector optimal growth framework with public services an
d labour-leisure choice, we find that the effects of temporary product
ive government spending on the real interest rate depend on the discre
pancy between actual and optimal government sizes as well as on the le
ngth of the action lag. This paper thus sheds light on the controversy
between the empirical findings and the theoretical predictions regard
ing the behaviour of the real interest rate within the neoclassical fr
amework.