Substitution over Time: Another Look at Life-Cycle Labor Supply

Most studies of the intertemporal substitution of work use life-cycle data, and from those studies many have concluded that intertemporal labor substitution is unimportant for macroeconomics. This paper takes another look at life-cycle data and argues that a consideration of measurement errors, taxes, on-the-job training, older workers, hours reporting bias, and all of the "margins" composing aggregate labor supply over the life cycle suggests that substitution over time may be very important for macro fluctuations. The life-cycle data used include fairly standard male cross-section and panel-data samples as well as a sample of women experiencing the termination of AFDC benefits as their youngest child turns 18 years old.