Análisis de la duración del desempleo en Uruguay (1986 - 1999)

In recent years, the sharp increase in Uruguayan unemployment rates has been accompanied by the perception that the risk of unemployment is unevenly distributed so that some groups are more prone to undergo long periods of unemployment. This research aims to complement the information reflected in the unemployment rate with the one referred to its duration, for the period 1986-1999. First, it studies the so-called “incomplete duration” of unemployment, as reported by the National Statistics Institute. This indicator informs about the job-search time until the information is collected. However, it is important to have an effective indicator to describe the whole unemployment period. For that reason, an estimator to give account of the time the unemployed people should expect before returning to employment is calculated. It is based on the calculation of conditional probability of remaining unemployed proposed by Sider (1985), Baker (1992), Corak and Heisz (1996). The study shows that the average complete unemployment spell reached four months for the whole population, which is a long one, considering the dynamics of economic growth between 1986 and 1999. It also finds differences among demographic groups in terms of their entry to unemployment and their total unemployment spell. Finally, it is established that the increase in the unemployment spell in aggregate terms has not been led by an increase in the incidence of groups typically associated to longer durations (compositional effect), but to an increase in duration for all groups. In short, most of the variation in the total unemployment spell duration has been caused by variations in the aggregate unemployment rate rather than by changes in the individual characteristics of the unemployed people.