UC Berkeley

My dissertation, "Essays in Public Economics," is comprised of three chapters. The first one, titled "Altruism, Reciprocity, and Equity: A Unified Motive for Intergenerational Transfers" is to address the following question: Why do parents divide bequests equally while transferring inter vivos gifts unequally? Across times and places, why have there mainly been only two extreme choices of distribution of bequests: either to give them to just one child (unigeniture) or to divide them equally (equigeniture)? How can a motive for intergenerational transfers explain both "equal division puzzle" (the former) and polarized inheritance patterns (the latter)? This chapter presents a behavioral model that coherently rationalizes these empirical realities. Namely, as head of a family, a parent altruistically cares about children but also wants them to spend effort for family. However, effort is costly and individual level of each child is unverifiable to a third party adjudicator. Given this incomplete information, there rise only two stable equilibria: either equigeniture or unigeniture. When the productivity of effort rises, the evolution of inheritance pattern from unigeniture to equigeniture occurs. So equigeniture is eventually adopted due to a rise in the productivity throughout industrialization. Furthermore, if the parent wants to counterbalance inequality among children who exert equal effort, the greater amount of inter vivos gift is transferred to a child with lower relative income compared to his siblings, while bequests remain equally divided. This model is consistent with the aforementioned empirical realities but also lends itself to further empirical tests. First of all, with a data set of pre-industrial agrarian societies, we find that a rise in the productivity of effort causes equigeniture to be chosen over unigeniture, which is consistent with the model. Second of all, through an empirical analysis on a micro-level data on inter vivos transfers in contemporary families, we find supporting evidence as follows: (i) income inequality among children increases the probability that their parent gives any inter vivos gift; and (ii) the amount of the gift is negatively associated with relative income of each child compared to his siblings.

The second chapter "Retirement and Exposure of Pension to Financial Market Fluctuations" studies how exposure of pension wealth to stock market fluctuations affects retirement behavior both theoretically and empirically. Characteristics of optimal plan for retirement are elaborated with reflecting that liquidizing pension wealth is more tied to retirement decisions than non-pension wealth as well as embodying time-sensitive restrictions on availability of pension benefits. Theoretical analysis finds that exposure of pension to financial market fluctuations does not always entail perfectly symmetric response of retirement. Exposure of pension to a positive shock actually brings responses of retirement only if the magnitude of the positive shock is large enough to compensate for foregone labor earnings and demand for resources necessary for post-retirement consumptions. In particular, whereas exposure of pension to a small negative shock leads to a decrease in retirement, exposure of pension to a positive shock with the same magnitude might not yield an increase in retirement. Next, empirical analysis is conducted with Health and Retirement Study, micro-level biennial panel data of senior workers in U.S., to examine actual retirement responses over the recent business cycle. Little evidence is found on a discernible increase in retirement rate owing to exposure of pension to the 2004 and 2006 positive shocks. However, the 2002 and 2008 negative shocks prove to lead to a decrease in retirement rate. In the view of theoretical findings, this is not self-contradictory but still can be consistent with a positive wealth effect on retirement; rather, it points to a case where these positive shocks are not sharp and large enough to bring substantive earlier retirement.

The third chapter, titled "Optimal Income Taxation and Optimal Revenue Mobilization," analyzes characteristics of nonlinear optimal income taxation and optimal revenue mobilization when the tax enforcement of a government is not costless (and thus not presumed to be perfect). The government cannot observe and verify an individual's innate ability although that ability turns out to cause inequality amongst them. This prevents the government from avoiding efficiency loss in the taxation, since each taxpayer can take advantage of private information over their own ability by reducing working hours to pretend to be less able than he truly is. Optimal income tax schedule is designed to minimize the efficiency loss from deterring such behavior to maximize social welfare. Moreover, the desired expenditure of the government is set for enhancing minimum living standard of society. In executing the tax schedule to finance this, however, tax evasion occurs due to imperfect enforcement. Although the government can verify the true amount of taxpayer's earnings, unlike their ability, it is costly to increase the enforcement rate. The optimal rate equalizes a gain of net increase in the tax revenue with a loss of decreased utility of risk-averse taxpayers from an increment in the rate. Notably, this chapter shows that aggregate loss of tax revenue can theoretically justify non-zero tax rate on top earners.