The minimum wage is equivalent to a non-means-tested low-wage subsidy, financed by a on low-wage employers. This paper, building on insights from both labor economics and tax-transfer policy, assesses the main reasons for adopting such a subsidy, and shows that the financing method largely defeats the presumed goals. The assessment includes a review of Card and Krueger's recent work on the minimum wage, which is shown to be empirically weak and theoretically implausible despite various complexities in labor markets. Consideration of a more appropriate low-wage subsidy requires looking comprehensively at U.S. tax-transfer policy. Due to phaseouts of various social welfare benefits, the United States has long imposed confiscatory marginal rates on many poor households, sometimes exceeding 100 percent. Phaseouts are generally a flawed way to target benefits, since they result in anomalously high bubble rates during the phaseout range. A better approach would seek a plausible overall rate structure (defined in Mirrleasean optimal terms) based on at least conceptual tax-transfer integration, as in the income tax that was widely discussed in the United States in the 1960s and 1970s. Such a system would likely involve a demogrant that was gradually phased out and then converted into a positive net liability as income increased. Marginal rates on people's wages would be low or even negative (under certain assumptions about positive externalities from increased workforce participation by the poor) at low income levels, and they likely would not bounce repeatedly up and down.