Technical efficiency in the Mexican manufacturing sector: a stochastic frontier approach

Abstract

This paper analyzes the technical efficiency in the Mexican manufacturing sector in which determinants and changes of the efficiency since NAFTA (North American Free Trade Agreement) are studied using the Panel Data Stochastic Frontier Analysis in its time-invariant and time-variant versions, comparing each other. It was used the Annual Industry Survey (AIS), which panel data information allow us to model the efficiency performance of firms in the period 1994-2001. Our main findings show that Mexican manufacturing firms worked, in average, at almost 23% of its potential product (compared with the best firm performance) and that there was a slight lost of capacity along time (1994-2001). Additionally, could be detected structural change, understood it as the change of firms' ranking observed in the model and the coefficient of the production function. Moreover, could be detected those firms that were consistent, winners or losers in the process of openness. Finally, under the assumption of Cournot duopoly competition, there are studied determinants for R&D investment.

Un supuesto esencial del cual se parte en microeconomía es que las empresas son homogéneas. Esto equivale al supuesto de la competencia perfecta y que todas las empresas operan bajo el mismo nivel de eficiencia. Sin embargo, muchos estudios demuestran que este supuesto no se aplica en todos los casos de la realidad (Caves, 1989), y México no es la excepción. Con la apertura comercial, las empresas mexicanas entraron en una nueva fase competitiva. Este trabajo pretende medir el impacto de la apertura comercial en la eficiencia técnica de las empresas, con base en la metodología de los modelos de Frontera Estocástica con Panel de Datos, de 1994 a 1991. Se encuentra que algunas empresas son consistentes después de la apertura; sin embrago otras pierden eficiencia.