This dissertation consists of two essays. The first essay studies the effect of different kinds of pension plans on the labor market decisions of the older workers. Due to the aging population, Social Security's projected annual cost is expected to increase to about 6.2 percent of the Gross Domestic Product by 2035, thus posing significant challenges to the U.S. policy makers. This has fueled an interest in research geared towards understanding the determinants of retirement. Past research has shown that pensions have a significant effect on retirement decisions. But the pension landscape in the U.S. has changed dramatically in the last few decades. From being once dominated by the traditional annuity-based Defined Benefit (DB) plans, the trend has now moved towards account-based Defined Contribution (DC) plans. This change has been accompanied by a reversal in the participation trend of older workers resulting in an increasing labor force participation of the elderly in the United States over the last thirty years. This essay investigates the link between the two by building a life cycle model of retirement and pension plan types. By conducting counterfactual experiment which changes all DB plans to DC plans, I hope to understand the role played by the differences in the nature of pension wealth accumulation under different pension plans in explaining the differences in retirement behavior observed across different pension plan holders. The second essay explores policy questions pertinent to the aging population in the healthcare field. Medicare Part D is a government program introduced in 2006 to offer outpatient drug benefits to Medicare beneficiaries. A lot of the brand-name drugs covered under Medicare Part D are also available in generic versions and it has been argued by policy makers that a higher level of utilization of these generic drugs would result in significant cost savings for the government. However, the cost savings of forcing consumers onto generics may lead to large welfare losses for consumers of non-generic alternatives if they highly value them. This issue is addressed in this essay where a structural model of drug demand that allows for heterogeneity in match quality between consumers and drugs and also allows for consumer learning about the stochastic match quality of the drug is estimated. The 2007-2008 administrative claims data for the 5\% Medicare Sample is used for demand estimation. By conducting counterfactual experiment which eliminates branded drugs for which the generic is available from the choice-sets of consumers, I hope to understand the effect of generic substitution on consumer welfare and the resulting cost savings for the government.