This paper studies four classic fiscal-policy experiments within a qua
ntitatively restricted neoclassical model. Our main findings are as fo
llows: (i) permanent changes in government purchases can lead to short
-run and long-run output multipliers that exceed 1; (ii) permanent cha
nges in government purchases induce larger effects than temporary chan
ges; (iii) the financing decision is quantitatively more important tha
n the resource cost of changes in government purchases, and (iv) publi
c investment has dramatic effects on private output and investment. Th
ese findings stem from important dynamic interactions of capital and l
abor absent in earlier equilibrium analyses of fiscal policy.