This dissertation consists of three essays at the intersection of behavioral economics and public finance. Two essays focus on the topic of income taxes, in particular, income-tax salience and income-tax perceptions, and how these differ for federal vs. state income taxes. These perceptions have potential implications for optimal policy design and labor-supply decisions. The third essay uses a novel approach to examine potential nonresponse bias in large government surveys that are a frequent source of economics data. In Chapter 1, I examine income-tax salience. Income taxes distort labor market outcomes, sometimes as a primary goal (as in the earned income tax credit) and sometimes as a secondary consequence (as a by-product of raising revenue). However, for an individual’s decisions to be influenced by a tax, he needs to be incorporating it in his decision-making. If the source of the tax--e.g., the state or federal government--affects the consumer’s awareness of the tax in his decisionmaking, then different income taxes might have different labor-market effects. This chapter provides experimental survey evidence that state taxes are less salient than federal taxes. Reduced-form estimates show a large difference between state and federal taxes. Direct estimation of salience parameters in a preferred subsample indicates that federal income taxes are close to fully salient, while state income taxes are substantially and significantly less salient. The chapter also explores evidence for individual-level explanations of salience and tax knowledge. In Chapter 2, I examine income-tax perceptions. Income taxes are complicatedand potentially confusing to consumers. Individuals may rely on heuristics or make systematic mistakes in their perceptions of their taxes. This chapter uses experimental survey data to add to a literature seeking to understand federal income-tax perceptions and adds novel evidence on state income-tax perceptions. I find broad agreement in perceptions between state and federal income taxes. Generally, individuals perceive too much progressivity in both federal and state tax schedules. Estimating a model of federal tax perceptions yields results consistent with a more complicated mental model than suggested in previous literature. In Chapter 3, we examine nonresponse bias in large government surveys. Economists are often interested in measuring outcomes such as unemployment, labor force participation, obesity, and household expenditures. In this chapter we study the sources of the relevant official statistics|the Current Population Survey (CPS), the Behavioral Risk Factor Surveillance System (BRFSS), and the Consumer Expenditure Survey (CEX)|and find that measurements of these outcomes depend on whether we look at easy vs. difficult-to-reach respondents. These results empirically substantiate the theoretical warning against making population-wide estimates from surveys with low response rates.