This PhD dissertation is a study of the individual level behaviour of labour supply, retirement, and consumption in different contexts. The first chapter studies the importance of intrafamily supports in elderly people's work and retirement choice. I build a dynamic programming model with extended families consisted of elderly parents and their adult-child households who do not live together, are imperfectly-altruistic toward each other and engage in a non-cooperative dynamic game. The two key innovations are allowing both parents- and child-household to provide transfers to each other and investigation of joint decisions making in older people's labour supply, savings, and intrafamily transfer choices. The structural parameters are estimated using the Korean Labour and Income Panel Study. I find that taking account of inter-vivos financial transfers can provide a better explanation of older people's life-cycle choices and reliance on government supports. The estimated model is used to evaluate the impact of two policies of social security expansion on elderly people's life-cycle choices. First, the expansion of guaranteed minimum income for the elderly results in crowding-out private transfers and unintended redistribution of resources rather than mitigating old-age poverty. Also, the welfare evaluation of policy can be biased if the strategic interaction between family members is not accounted for adequately. Second, the rise in state pension income amount just has a limited effect on older people's work incentives, and the vast amount of intrafamily resource sharing could be partly caused by high elderly poverty. The second chapter builds and estimates a dynamic model of older people's joint decisions of labour supply, savings, and social security benefit (SSB) application. One new feature of the model relative to the existing literature is that I allow for the selection into self-employment jointly with paid-employment and retirement. Agents in the model observe their own paid-sector productivity but are uncertain of their productivity in the self-employment sector. However, the learning process enables people to reduce initial uncertainty by observing the performance of their business. The parameters of the dynamic programming model are estimated using the U.S. Health and Retirement Survey. I find that allowing for the transition between paid- and self-employment delays the retirement of older workers. In the counterfactual simulation, I compare the effects of payroll tax-cut and self-employment subsidy program and find that these policies can contribute to strengthening security in retirement and have a large effect on the proportion of people who choose paid- and self-employment. However, they have limited effects on elderly people's retirement choice. The third chapter is co-authored with Kai Liu, Shawn Ni and Youn Seol. We estimate the wealth effect on consumption by exploiting the differential effect of housing price booms and busts across households with different holdings in housing wealth. We also extend the analysis of the wealth effect from the consumption growth to its inequality by allowing income shocks and the ability to smooth consumption against income shocks to vary over housing wealth and housing market-driven wealth shocks. Using household-level panel data on consumption, income, and wealth from the Korean Family Income and Expenditure Survey, we find that the house price change has a significant and large differential effect on consumption growth, and homeowners exhibit a stronger ability to insure consumption against income fluctuations. Also, the ability to insure consumption against income risks is imperfect, and the effect of transitory shocks on consumption is smoothed more than that of permanent shocks.