Short cuts to dynamic factor demand modelling

By means of so-called virtual or shadow prices, short-run factor demands, short-run marginal costs, etc. can be derived from any long-run cost function. The traditional approach (short-run/restricted/conditional/variable cost functions) is criticized, and it is also shown that technological change, scale effects, etc. can be added to any cost function by means of disembodied factor-augmenting efficiency indexes, easing the interpretations of the effects, but without loss of flexibility. It is shown that the trend- and scale-parameters of the (long-run) translog cost function can be directly translated into trend- and scale-parameters of such efficiency indexes. The techniques are illustrated on the well-known Berndt–Wood data set, using a (Diewert) long-run generalized Leontief (GL) cost function, and assuming that capital and labour are quasi-fixed.