The dissertation consists of three essays that study consumption choices and their policy implications. The first chapter tests the life-cycle permanent income hypothesis via an experimental approach. University students are asked to provide their consumption choices in hypothetical circumstances. Background information such as their numerical and financial literacy are collected in the experiment. The data suggest that the subjects’ consumption choices do not necessarily support the life-cycle permanent income hypothesis. The consumption levels chosen by some subjects vary greatly when they face the same total remaining lifetime resources, but different composition of these resources between net and human wealth. This consumption error varies across gender and is affected by the subjects’ numerical and financial ability. The results have important implications for policymakers. In the second chapter a censored quadratic model is used to examine whether soda taxes affect consumption. Households that do and do not consume soda are taken into account. The results show that in states that have a higher soda than food tax or that have just increased the soda tax rates, households consume significantly less soda. Households that have more children and are located in areas with a small population assign significantly larger weight to soda in their total utility. The price elasticity of soda consumption is high. The unconditional price elasticity is larger than the conditional price elasticity, which is measured using households that do purchase soda. Also, the price elasticity for low-income households is greater than for high-income households. The results presented may be important to policymakers setting soda taxes to prevent obesity. Following the second chapter, a modified three-good censored utility model is used to examine if beer and wine taxes affect alcohol consumption. This study provides empirical estimates of price elasticities on beer and wine, which suggests that beer taxes may not be a good avenue for policymakers to consider in counteracting the serious consequence of alcohol abuse. However, wine taxes may be effective in reducing alcohol abuse, especially for low-income households.