PRODUCT-MIX AND TECHNOLOGICAL CHANGE WITHIN THE LEONTIEF INVERSE

ABSTRACT. Intertemporal differences in input-output coefficients can be attributed to technological change and to changes in the mix of products composing the aggregate sectors of input-output models. In this work, we develop the theoretical foundation necessary to separate these changes for both the structural and Leontief inverse matrices. Using this foundation, we examine the relative empirical importance of technological and product-mix change. The product-mix effect is then combined with RAS estimates of the technological effect to form updated estimates of the inverse. Results show that the accuracy of updated input-output coefficients can be improved in this manner.