I present three essays on income taxation in Pakistan. The first essay investigates how taxes influence agents’ earnings, compliance and business organization choices. Using a tax reform introduced in Pakistan in 2010, which raised tax rates on partnership earnings as compared to sole proprietorship income, as a natural policy experiment, I (i) identify a full range of behavioral responses to the tax rate changes (ii) study the determinants of tax compliance (iii) investigate if VAT causes firms to be more tax compliant. Relying on administrative tax records that comprise the universe of income tax returns filed in 2006-11 and a rich set of firm characteristics, I find that the reform induced substantial extensive and intensive margin responses including reduction in earnings, income shifting, movement into informality, and spillover effects on VAT base. I also find that the firms that have greater fraction of their tax withheld at source, are registered for VAT, or withhold taxes of other agents are more tax compliant. This highlights the importance of the notion that information trails on arm-length business transactions facilitate enforcement. Comparing short-term responses of partnership firms – which arguably identify tax evasion – on both sides of the VAT exemption threshold, I find that the evasion changes discontinuously at the cutoff suggesting that the VAT causes firms to be more compliant. In the second essay, I along with my co-authors, analyze the design of tax systems under imperfect enforcement. A common policy in developing countries is to impose minimum tax schemes whereby firms are taxed either on profits or on turnover (with a much lower tax rate on turnover), depending on which tax liability is larger. This is a production inefficient tax policy, but has been motivated by the idea that turnover taxes are harder to evade. Such schemes give rise to kink points in firms’ choice sets as the tax rate and tax base jump discontinuously at a profit rate threshold. Analyzing responses to one such scheme in Pakistan, we find large bunching of corporate firms around the minimum tax kink. We show that the combined tax rate and tax base change at the kink provides small real incentives for bunching, making the policy ideal for eliciting evasion. Based on the methodology that we develop, we estimate that turnover taxes reduce evasion by up to 60-70% of corporate income. In the third essay, I along with Henrik Kleven, develop a framework for non-parametrically identifying optimization frictions and structural elasticities using notches – discontinuities in the choice sets of agents – introduced for example by tax and transfer policies. We apply our framework to tax notches in personal income tax schedule of Pakistan to estimate structural elasticities of taxable income.