Dionissi Aliprantis
Job Market Candidate
Research:
My areas of interest are Applied Econometrics, Economic Development, Education, Labor, and Program Evaluation. I am currently working on projects related to racial disparities in education and labor market outcomes, early childhood interventions and related methodological issues, and the provision of primary education in Haiti.
Job Market Paper:
Human Capital in the Inner City
Download: pdfAbstract: There is a large divide in the education, labor market, and personal security outcomes of black and white young males in the United States. Previous empirical literature in economics explores the sources of these disparities while abstracting from non-market considerations. A smaller and mainly theoretical literature in economics has been influenced by work in sociology to study how non-pecuniary rewards affect these outcomes. This paper builds on both literatures to develop and estimate a dynamic model of black young males' joint decisions about schooling, labor force participation, and personal security. The formulation of the model is inspired by Elijah Anderson's ethnographic research regarding the incentives black young males face to ensure their personal security in environments where it is not provided by state institutions. I operationalize Anderson's notion of the "code of the street" by defining the set of skills and knowledge useful for providing personal security to be a distinct type of human capital, street capital. In the model agents decide whether to attend school, work, and engage in street behaviors, and accumulate both regular human capital and street capital through these decisions. The model also includes a probability of incarceration that depends on street behaviors. The model is estimated using longitudinal data from the NLSY97, which includes unusually rich information on participation in street behaviors. Using the estimated model, I quantify the influence of the "code of the street" on black males' schooling and labor market choices, and I examine potential policies to influence such choices. The estimated model is used to simulate a world in which children do not face incentives to engage in street behavior, which may be interpreted as allowing children to grow up in safe neighborhoods. In this world about 20% more black young men after the age of 20 choose to work, about 7% more graduate from high school, and there is also a decrease in incarceration rates. An additional counterfactual experiment is performed in which agents are given the choice at age 16, without prior knowledge, to either keep their current stocks of street capital or to set them to zero. In this scenario about 7% more black males choose either to work or to attend school, and an additional 12% choose to graduate from high school. Finally, counterfactual experiments are performed to test the effects of wage and education subsidies. Such interventions are found to have important impacts on their targeted outcomes, but little effect on street behavior or incarceration rates. The large effects from the code of the street indicate that interpersonal violence is an empirically important factor influencing the education and labor market outcomes of black young men.
Additional Research Papers:
Redshirting, Compulsory Schooling Laws, and Educational Attainment (Revise and Resubmit at Journal of Educational and Behavioral Statistics)
Download: pdfAbstract: A wide literature uses date of birth as an instrument for educational attainment. Considering the assumptions of independence and monotonicity within the context of a latent index model, we find that this identification strategy produces biased estimates if parents choose to redshirt, or to delay their children's initial enrollment in kindergarten, based on treatment effect heterogeneity. An inspection of the ECLS-K data set indicates not only that such redshirting is common, but also that heterogeneity in the treatment effect of educational attainment is almost certainly a factor in parents' redshirting decisions. These findings have important implications for a broad assortment of parameter estimates in the literature.
When Should Children Start School?
Download: pdfAbstract: Recent attention given to early childhood education highlights the importance of deciding the age at which children enter kindergarten. Although parents, schools, and state governments have chosen during recent decades to increase the average age at which children enter kindergarten in the United States, little is known about the consequences of these choices. By selecting a quasi-random sample of children from the ECLS-K data set, this paper is able to separately identify coefficients related to both absolute and relative entrance age effects. The identification strategy employed avoids the problems redshirting creates when using date of birth as an instrument for educational attainment. For children in the sample, estimated absolute age coefficients provide no evidence that increasing the average age at which children begin kindergarten increases achievement test scores. These results, while far from conclusive, suggest that increasing kindergarten entrance age need not be a top priority of early education policy.