| Since Hall (1978), there has been growing interest in the consumer's decision on durables and nondurables consumption under the life-cycle hypothesis (LCH). However, almost all of the empirical tests of the LCH model have rejected this hypothesis.;The empirical results show that the assumption of additive separability across durables and nondurables is rejected against nonseparability. The hypothesis that the imputed services of durables stock are more sensitive to changes in prices than are nondurables is not confirmed by the estimated joint-decision model, in which a constant depreciation rate and proportionality are assumed. On the other hand, the utilization model supports this hypothesis. Also, the orthogonality conditions of the LCH, examined by including lagged incomes, are confirmed in the utilization model.;Since any statistical LCH model involves a joint hypothesis, the unfavorable empirical evidence against the LCH may be attributable to components other than the LCH itself. Hence, we examine whether or not the relaxation of several critical assumptions in earlier work is economically meaningful. Also, the interesting hypothesis that services out of durables stock are more sensitive to changes in prices than are nondurables is examined. We relax the assumptions of additive separability across durables and nondurables in the consumer's utility function, a constant depreciation rate of durables, and proportionality between durables stock and services rendered by the durable stock in our utilization model; in so doing, we derive a reduced-form relation for the varying depreciation rate with which to construct the stock of durables. |
