The dissertation views the varied start dates of airport departure taxes as a series of natural experiments, in each of which plane tickets are subject to different levels of taxes. It shows that airlines are able to raise fares of nonstop tickets by more than the taxes applied while leaving fares of connecting tickets little changed. These results provide empirical support for two propositions in economic theory: First, suppliers may "over-shift" the tax burden onto consumers under imperfect competition. Second, when two products are taxed, the tax-inclusive price is higher for the product that has a less elastic demand (i.e., nonstop tickets). The dissertation argues that there are efficiency and equity gains to be made by revising the current tax schedule. One potential adjustment is to allow taxes to vary by demand and other direct costs; the revenues can then be used to invest in projects that lower long-run costs and to subsidize airports and airlines that serve smaller communities.