Donor countries have a poor track record of meeting their foreign aid commitments. Yet the discrepancy between aid commitments and actual delivery of aid has received little scholarly attention. Traditionally, studies of aid allocation have assumed that donors provide an amount of aid that they deem satisfactory. In this thesis, I explore the possibility that the current lower-than-promised quantities of aid may be unsatisfactory for donor countries, i.e., that they would prefer an alternative state of the world with higher aid levels. Specifically, I consider how strategic interaction can lead donors to give less aid than they would ideally prefer. Using game theory, I find two main reasons for why a donor may give less aid than it would ideally prefer: 1) a donor may only want to provide aid on the condition of certain behavior by the recipient. If the donor believes or experiences that the recipient does not conform to this behavior, the donor may restrict aid even though it would ideally prefer to give more; 2) donors may see aid as a contribution to a public good, and be tempted to free ride on the efforts of each other. If the voluntary production of the good is Pareto suboptimal and effective enforcement mechanisms are absent, the resulting quantity of aid is unsatisfactory for the donor countries.