Both raw intuition and past experience suggest that the success of an employment guarantee scheme (EGS) in safeguarding the welfare of the poor depends both on the wage it promises, and the ease with which any worker can gain access. An EGS is thus at once a wage guarantee and a rationing device. We chart the positive and normative limits of such an EGS as an efficiency improving and poverty alleviating policy reform in a canonical labor market setting. At its core, an EGS provides an aggregate, not just EGS, employment target. Given the target, the EGS wage and access can be fine-tuned to deliver outcomes ranging from a contestable labor market to a simple universal unemployment benefit. The credibility of any such target, however, is shown to be triggered endogenously by a host of factors: the distributional concerns of the planner, private sector productivity, the prevalence of market power and the need for public works. Paradoxically, the outcome with a planner who cares only about efficiency can be less efficient than the outcome with a planner whose social welfare function also gives weight to poverty!