Local measures of intergenerational mobility of income, cognitive and non-cognitive skills

Sunday, October 11, 2015: 9:00 AM
Christa Marr, Ph.D , Economics, History, and Political Science, Fitchburg State University, Fitchburg, MA
Andros Kourtellos, Ph.D. , University of Cyprus, Nicosia, Cyprus
Chih Ming Tan, Ph.D. , University of North Dakota, Grand Fork, ND
Using National Longitudinal Survey of Youth (NLSY) data we investigate whether the observed patterns of intergenerational persistence in cognitive and non-cognitive   abilities are consistent with the predictions of the genetic hypothesis.   In doing so we employ the varying coefficient model to estimate nonparametric (local) measures of intergenerational mobility of income, cognitive, and non-cognitive skills as smooth functions of log parent permanent income. 

The main contributions of this paper are to consider how cognitive and non-cognitive skills influence our understanding of intergenerational mobility and to account for parameter heterogeneity and nonlinearities  in the measurement of intergenerational mobility along the parental income distribution.  We take advantage of the maternal linkage between two cohorts in the NLSY to measure intergenerational mobility of income as well as the intergenerational  transmission  of cognitive  and  non-cognitive  skills from mothers to their children.

 Similar to past literature, we find a benchmark  intergenerational income elasticity measurement of 0.41 for sons and 0.33 for daughters. However, our findings show that intergenerational mobility exhibits nonlinear patterns.    Individuals with different parental income are characterized by different degrees of intergenerational mobility.  Intergenerational income elasticities are nonlinear  and  follow an inverted  U-shape over the parental  income distribution.   Children from middle-income  parents  experience  the highest  intergenerational  income persistence.

Moreover, we find evidence that suggests that the genetic component in the overall intergenerational transmission mechanism is much stronger than the epigenetic for both sons and daughters. For both sons and daughters, we find that the persistence in the transmission of non-cognitive skills (i.e. Rosenberg Self-Esteem and Pearlin Mastery scores) is much smaller than the persistence in the transmission of income and cognitive skills (i.e. Peabody Math and Picture Vocabulary scores).  This suggests  a potentially  important role for a genetic  component  in the overall intergenerational transmission mechanism.