ESSAYS ON HIGHER EDUCATION, PARENTAL ALTRUISM, AND GRANGER CAUSALITY

Abstract

This dissertation consists of three distinct, publishable ‘papers’ included as separate chapters. Although all of the three chapters can be read and understood independently, the first two chapters are on related themes. The essence of the first two chapters are, therefore, to analyze the individuals’ higher education expenditures over their life-cycle. It considers the standard life-cycle human capital accumulation theory as a benchmark model and tests the rationale and validity of it by applying empirical methodologies and calibration. The third topic employs the Granger causality test to find out what causes the stock price volatility to become more sensitive, whether it is for interest rate or exchange rate risk.

More specifically, the first chapter uses Consumer Expenditure Survey (CEX) data to estimate the effects of an individual's age on higher education expenditures. Using a synthetic panel created from the CEX data and controlling for fixed effects and different demographic level characteristics, the regression results are then used to construct the life-cycle profile for higher education expenditures based on age. The estimated coefficients of age and its polynomials are found to be statistically significant. Special emphasis is placed on finding the turning points by using the polynomial regression coefficients that indicate a change in the age pattern over the life-cycle. The turning point analysis indicates a “hump” shaped nature for education expenditures, counter to what theory suggests.

The second chapter considers the life-cycle higher education expenditures profile for individuals as a baseline model and investigates the possible impact of parental altruism and borrowing constraints on the human capital investment. The analysis is conducted by employing a standard overlapping-generations model in which “parental altruism” is emphasized over borrowing constrains for children. This quantitatively calibrated model predicts that the intertemporal substitution effect dominates the income effect for the individual of middle age groups, causing allocation of their income towards the higher education expenditures of their young-aged child. The latter effect generates the hump shaped human capital expenditures pattern for individual parents who decided to enroll for college education at a later period of their life.

The third and final chapter explores the Granger causality test and the bi-variate as well as the multivariate co-integration to determine the interactions between interest rates, exchange rates, and the composite stock volatilities of different traded contracts under the Chicago Board Options and Exchange. Using the daily sector data for all observed variables from the St. Louis Fed over the period of 2007-2017 and introducing the method of simple vector autoregression, this study examines various aspects of the correlation where the current and the previous values of volatility indices, interest rates, and exchange rates have shown significant Granger causality effects to the return behavior of those volatility indices, interest rates, and exchange rates. The estimated result indicates that, under the absence of any long-run relationships, interest rates have more unidirectional and bi-directional causal effects with the stock market volatility indices than in comparison with the exchange rates, although both of them are identified as significant determinants of stock price volatility.