Equalization transfers are a common and noteworthy feature of fiscal decentralization systems around the world, especially in developing countries. However, the way in which they are designed and implemented is not clearly rooted in mainstream public finance theory. In this paper we develop a formal framework to explain the rationale of the fiscal gap model (the difference between expenditure needs and fiscal capacity of a jurisdiction), arguably the most popular model used by applied economists for the design of equalization transfer programs. First we take into account the problem of accountability. If government authorities are selfinterested, transfers from upper levels are shown to reduce their responsiveness to taxpayers’ preferences. Next, we use a normative approach to derive the conventional fiscal gap formula. We argue that equalization transfers should only be used to finance limited types and amounts of public expenditures, called here standard public expenditures. In contrast, discretional public expenditures (not subject to equalization) should be financed exclusively with own revenues. Last, we describe the conditions that ensure the affordability of the system. We conclude that the fiscal gap model of equalization is sufficient to efficiently allocate the available funds across jurisdictions, setting the marginal cost of funds at the appropriate level. However, no information on the marginal cost of funds is required. Policymakers can simply rely on estimates of expenditure needs and fiscal capacity to design an optimal equalization transfer program.