Productivity,mechanization and skills: A test of the Hirschman hypothesis for Latin American industry

This paper attempts to test the paradoxical Hirschman hypothesis relating to the suitability of certain capital-intensive industries for less industrialized countries based on the ‘process-centred’ or ‘machine-paced’ nature of those industries which makes their productivity less dependent on the quality of local labour. Following extensive discussion of the methodology (e.g. problems of selection and specification of the variables explaining productivity differentials, and problems in specifying the appropriate variables to be explained), the author presents and analyses results for seven Latin American countries. These countries were chosen because of their more highly developed industrial structure, i.e. higher industrial value added per capita and a more diversified product mix, in order to provide a more severe test of the hypothesis. Although the results are not definitive, they seem to indicate that the industries which did better in relative productivity terms were those based on abundant local resources or of a traditional character on the whole; rather a reaffirmation of comparative advantage with little paradox.