Incluye bibliografía

The theoretical analysis focuses on the distinction between temporary and permanent movements in government purchases. Under plausible conditions, the temporary case involves an output response that is positive, less than one-to-one with the change in government purchases, and larger than that generated by an equal-sized, but permanent, shift in purchases. The equilibrium real rate of return rises in the temporary case, but changes little in the permanent one. Defense purchases are divided empirically into "permanent" and "temporary" components by considering the role of (temporary) wars. No temporary shifts in nondefense purchases were isolated. Empirical results verify an expansionary output effect for temporary purchases that exceeds that of permanent purchases. The results for some other expectational hypotheses are found to be generally supportive of the theory