School Attendance, Child Labor and Local Labor Market Fluctuations in Urban Brazil

In School or at Work? Evidence from a Crisis

“…The significantly higher probability of working with respect to a higher urban area level wage (or employment) confirms that for Argentina, the labor market is contending for youths' time. While the effect of transitory household income is not found 40 Average predicted probabilities for the 13-18 year-olds are shown in Table 5 to be significantly different during the crisis years (same result as Duryea and Kuenning, 2003) the substitution effect as measured by the urban area wage is significantly smaller in magnitude (the interaction is positive in columns 2 and 3). This suggests that opportunity costs of the labor market provide an important countervailing effect during "normal" economic downturns (1995) but that this effect is attenuated during a crisis when families expectations change in response to a deep shock.…”

“…Most of the earlier studies have focused either on the binary choice between school and work (de Janvry et al, 2006) ; while some consider the four-fold choice between school only, work only, school and work, or neither (Duryea and Arends Kuenning 2003). Although these studies have provided policy makers with crucial information about the factors influencing schoollabor decisions of school-age youth, there is more to be learned by further disaggregation.…”

“…Some studies include the child's wage as a proxy, but because not many children work, estimates of children's wages are subject to serious selectivity biases. In Brazil, the effects of local labor market opportunities on schooling have been identified by the variation across states in wage movements (Duryea and Arends Kuenning 2003). However, in Argentina unemployment has been one of the most important results of crises, even though "discouraged worker" effects were also a byproduct of the mid-1990s crises.…”

Industrial Structure and Child Labor Evidence from the Brazilian Population Census

“…Some studies analyze temporary changes in aggregate local labor demand or local economic conditions. There is evidence for both Brazil (Duryea and Arends-Kuenning 2003;Neri and Thomas 2000;Parikh and Sadoulet 2005;Manacorda and Rosati 2007) and other countries (see for example Guarcello et al 2005 for Ethiopia) that on average child labor tends to increase when local labor demand is stronger. In a similar vein, Kruger (2007) shows that increases in the value of coffee production induce a rise in child labor in coffee producing regions of Brazil, especially among children of less educated 2 parents, implying that a temporary increase in labor demand tends to increase child labor, despite a potentially counteracting income effect.…”

“…Differently from existing work on Brazil that focuses on short term or cyclical changes in child labor (Duryea and Arends-Kuenning 2003;Duryea et al 2007;Krueger 2007, Manacorda andRosati 2007), our paper investigates long term trends in child labor using data from the 1980, 1991 and 2000 population censuses. There are two major advantages associated to using micro data from the Census compared to the more widely used data from the Brazilian Pesquisa Nacional por Amostra de Domicílios (PNAD).…”