Economics literature has paid relatively little attention to the study of underemployment, which is defined as the situation of an individual who works less than 35 hours a week, but wants to work more hours. This can be explained by the scarcity of data on underemployment, and by the fact that labor statistics treat it as employment, and thus its numbers may not create so much expectation as those of unemployment. The study of underemployment is important in view of its relation to poverty, especially as it affects women, and because of the psychological damages that result to underemployed individuals, particularly young people. This paper analyzes the behavior of underemployment in a sample of five Latin American countries by means of the estimation of an unrestricted vector auto-regressive (VAR) model using 2000-2008 panel data from five Latin American countries. The variables included in the VAR are: the unemployment and underemployment rates, female and male participation rates, and the indexes of formal employment and the real wage. Impulse response functions are analyzed to make inferences about the behavior of the variables in response to shocks to one of them. In addition, the paper develops models to represent the existence of participation and poverty traps, which, respectively, result from the interaction between unemployment, economic growth, inequality and participation and from the interplay between underemployment, poverty and children’s school desertion. The main results show that there exists an inverse relationship between formal employment and underemployment; that underemployment and female participation have persistent responses to shocks to unemployment; female and male participation decrease in response to a shock to the real wage; and underemployment and unemployment have persistent response to shocks to female participation. In general terms, the results show different responses along gender lines. Another result was that underemployment is the initial means of adjustment to recessionary periods and as such it is an early warning indicator of rising unemployment. The results show the prevalence of the discouraged and additional worker phenomena in the sample countries. The results have shown that underemployment can lead to persistent poverty and that it acts as a break on economic growth, which indicate the necessity of eliminating labor market segmentation by gender and directing attention to the particular needs of women (as well as those of young people, the elderly, the disabled, and ethnic minorities) with respect to their labor market participation.
