Labor market regulations and the duration of employment in Brazil

The objective of the paper is to estimate the impact of an increase of dismissal penalties, implemented by the 1988 Brazilian constitution, on the duration of employment spells. After a simple comparison of estimates of this variable in periods pre and post 1988 we made use of two alternative methods to estimate the impact we are looking for. The first method use control groups in order to separate the impact of underlying changes in the macroeconomic environment from the impact of the constitutional change. Accordingly, we proceed by breaking down the overall population into two groups, the so-called the treatment and control groups and then apply what is known as differences-in-differences methodology. The second method would essentially consist of regressing monthly or yearly estimates for the duration of employment spells on an indicator for the constitutional change, controlling for a set of macroeconomic indicators. The estimated coefficient on this indicator would then be an estimate of the impact of the constitutional change on the hazard rate. This procedure also allow us to test the validity of all control groups on which we base part of our empirical analysis.