During the 1990s, US income-transfer and tax policies shifted towards trying to encourage work among low-income families. Optimal tax theory, however, suggests that work subsidies are usually an inefficient way to raise the incomes of poor families unless the work effort of recipients has external benefits and/or taxpayer/voters prefer redistributing income to the working poor rather than the idle poor. This paper discusses the conditions under which work subsidies may be economically efficient and assesses empirical evidence suggesting that welfare reform and expansions of the EITC have increased work effort among low income families, but is inconclusive about whether the policy shift has enabled them to advance beyond entry-level jobs or benefited their children.