Embodiment, productivity, and the age distribution of capital

An important theme in modern research on productivity has been that technological progress may be embodied in capital in the sense that traditional measures of TFP growth reflect unmeasured improvements in the quality of capital inputs as well as pure disembodied technological progress. It is commonly believed that an implication of this embodiment hypothesis is that there should be a negative relationship between measured TFP and the age of the measured capital stock. This paper presents empirical evidence which suggests that an increase in the age of the capital stock is actually associated with higher TFP. This surprising result may be due to the presence of a mis-measurement normally overlooked in this literature: With mis-measured improvements in capital quality, the usual depreciation rates used to construct empirical capital stocks are incorrect for growth accounting. This effect dominates the usual average age effect.