An Algorithmic Measurement of Technical Progress

In this paper we propose a measure of technological progress which is based on the information embedded in standard input-output tables. Well known duality properties enables one to establish a connection between the quantities necessary as inputs and the associated output and some auxiliary prices (like the wage-profit curves). We claim that properly tailored wage-profit frontiers may provide a basis for the measurement of technological progress. But the computation of these wage-profit frontiers is not trivial. A brute force algorithm for the computation of the wage-profit frontiers has high combinatorial complexity that would make its precise computation intractable. But thanks to an efficient algorithm that we have been able to devise we can now compute it. We consider this to be an important and original contribution. Here we present and apply this algorithm. Due to this improvement we can now use these wage-profit frontiers as benchmarks against which to measure technological progress: two new indices have been defined. These new tools have have been applied to the OECD input-output data 1970-2005 and the reslts are presented here.