Produtividade agregada brasileira (1970-2000): declínio robusto e fraca recuperação

This study explores the productivity performance of the Brazilian economy between 1970 and 1998. We assess how much of the TFP downfall can be explained by some departures from the standard procedure. We incorporate to the standard measure utilization of capacity, changes in the workweek of capital, services of capital from electricity consumption, relative prices distortions, human capital, and investment in specific technology. We conclude that the downfall in productivity is quite robust to those specifications. The only case that presents a marked difference from the standard TFP measure occurs when relative prices of capital are corrected. The implications of this finding are a topic for future research.

Os efeitos da redução da jornada de trabalho de 48 para 44 horas semanais sobre o mercado de trabalho no Brasil

The reduction of standard weekly working hours from the curreilt leveI of 44 hours to 40 hours has recently been proposed by the main central unions as a way to create jobs and reduce unemployment in Brazil. The idea, known as work-sharing, is that the reduction in average hours per worker allows the same tasks to be performed bymore employees. However, the notion ofwork-sharing has been challenged by the theoretical and the empirical literature. Theory says that, in general, a reduction in the duration ofthe workweek tends to decrease employment. Work-sharing is even less likely in the case of no wage restraint, when monthly wages are not proportionally reduced. Recent cautious econometric evidence also conflict with the notion of work -sharing (Hunt, 1999, and Crépon and Kramarz, 2000). The objective ofthis paper is to study the effects ofthe workweek length reduction from 48 to 44 hours, prescribed by the 1988 Constitutional change, on the Brazilian labor market. We find that: i) the reduction of standard working hours was followed by a significant decrease in actual working hours: 60,7% ofthe employees that worked between 45 and 48 weekly hours in 1988 (the affected group) and that remained employed in 1989 switched to a 40-44 hours workweek; ii) belonging to the affected group in 1988 had no statistical effect on the probability ofbecoming unemployed, exiting the labor market, or switching to an informal job in 1989; iii) the reduction in working hours implied a 8,8% increase in hourly real wages with respect to those that remained employed at 40-44 hours a week. In sum, the reduction of standard working hours in 1988 reduced actual working hours, did not affect the probability ofa typical worker to lose hisjob in 1989, and implied a relative increase in hourly real wages.