• 2018-07
  • 2018-10
  • 2018-11
  • 2019-04
  • 2019-05
  • 2019-06
  • 2019-07
  • 2019-08
  • 2019-09
  • 2019-10
  • 2019-11
  • 2019-12
  • 2020-01
  • As shown in Chart most studies concerning international


    As shown in Chart 1, most studies concerning international and domestic Deferoxamine cost draining have been carried out by assessing the brain-drain rate in the aggregate, mainly by including locational features in the origin and destination places. Brazilian studies follow the international literature both on domestic and international spheres.
    Empirical model As said before, skilled labor mobility could be verified by identifying individuals having higher learning education in t, whose federation unit in t was different from that in t+1, i.e., skilled out-migrants. Therefore, two sets of comparison comprised in this phenomenon should be taken into account when estimating mobility determinants, as follows: the decision to migrate or remain in a given place and the skilled individual\'s decision to migrate as compared to the unskilled individual\'s choice. A highly significant stylized fact in the literature refers to that migrants do not comprise a random sample of individuals (Borjas, 1999), as the decision to migrate makes them distinct from non-migrants. Self-selection becomes even more evident when studying skilled labor mobility as these workers had already shown significant personal characteristics in a positive selection related to skillfulness. As panel data were available, estimating labor force decision to migrate could be carried out by means of a fixed-effect logit model, an efficient mode of treating migrant selection bias. It is reasonable to assume that other non-observed personal characteristics might also have influenced the skilled worker\'s decision to migrate, such as individual preferences and abilities and education quality as well. Estimates could thus be inconsistent and biased in case non-observed variables were not included in the regression. Therefore, in a first moment, the dependent binary variable ascribed value 1 to out-migrants and 0 to the remaining individuals. Additional variables – concerning features of individuals and their origin and destination states – were then aggregated to the decision to migrate as described in Eq. (1).where i represents the individual, t the calendar years, y represents the binary variable indicating out-migration, α the fixed-effect vector, X is the vector explaining variables of individual characteristics, Z represents the vector of variables related to employment, R is the vector of locational variables, relates to the error term, T represents the dummies for the calendar years and, β, λ, τ and γ are the parameters. In a second moment, the skilled individual\'s decision to migrate was estimated. Therefore, only individuals who have out-migrated at any time were selected for the sample so as to make it possible to compare the skilled worker\'s decision to migrate with an individual\'s decision to migrate, no matter their qualification. As for the new estimation, which was also based on Eq. (1), the dependent variable showed value 1 to skilled out-migrants and 0 to the remaining out-migrants.