The earned income tax credit (EITC), which uses the federal income tax system to provide an earnings subsidy to low-income workers, has enjoyed support across the political spectrum as a ''pro-work, pro-family'' alternative to traditional welfare programs. Advocates of the EITC also claim that the EITC's tax-based administration is cheaper and less stimatizing than traditional welfare administration. In this Article, Professor Alstott argues that the case for the EITC has been oversimplified in two significant ways. First, the Article argues that both advocates and opponents of the EITC place undue emphasis on whether it discourages work and marriage among EITC recipients. Professor Alstott demonstrates that the conventional policy debate over the EITC's behavioral incentives is too narrowly framed: the debate relies on incomplete economic analysis and reflects important but unacknowledge normative tensions. Second, the Article shows that the EITC, as a tax-based income-transfer program, faces inherent institutional constraints not present in traditional welfare programs. Professor Alstott explains that, although economists and policy analysts have long advocated integration of the tax and transfer systems, they have overlooked the problems of inaccuracy, unresponsiveness, and noncompliance that are inherent in tax-based administration. The Article goes on to demonstrate that, although reforms might improve the EITC's performance along these dimensions, such improvements would require either compromising the benefits of tax-based administration or undertaking a major restructuring of basic institutions of the federal tax system. Professor Alstott concludes that, absent such changes, the EITC and similar tax-based transfers are likely to prove widely acceptable only if we radically revise our expectations about accuracy, responsiveness, and compliance.