Efficience technique et incitations salariales. Analyse empirique sur un panel incomplet des industries textiles en Tunisie

This study combines the stochastic production frontiers approach with the efficiency wage approach to identify two technical inefficiency components. The first component is observable and varies over time. It measures the technical inefficiency resulting from an inadequate wage incentive. The second component remains stable over time and concerns unobservable company-specific technical inefficiency. This paper sets out to estimate the time-stable technical efficiency component whilst controlling the company-specific effects using instrumental variable techniques applied to incomplete panels. Labour and capital production factors are measured by some of their qualitative characteristics, i.e. technological progress incorporated into equipment and manpower by skills level. The study then empirically checks the wage-productivity relationship before estimating the technical inefficiency attributable to a lack of effort. The empirical analysis considers a non-cylinder panel of 619 Tunisian textile companies studied from 1983 to 1994. The estimation results point to a decline in autonomous technological progress from 1983 to 1990. Moreover, productivity gains are generated by technological progress incorporated into equipment and manpower skills. The unobservable efficiency is approximately 61% on average. The inefficiency ascribable to a lack of effort is approximately 4% on average.