The organizational structure of the bureaucracy is a key determinant of policy outcomes. Bureaucratic agencies exhibit wide variation in their organizational capacity, which allows politicians to strategically shape policy implementation. This article examines what bureaucratic structure implies for the ability of voters to hold politicians electorally accountable. It explicitly models differences in organizational capacity across bureaucratic agencies and considers a problem where a politician must decide not only which policy to choose but which agency, or combination of agencies, will implement it. The choice of implementation feeds back into the choice of policy and this, in turn, affects how voters perceive the performance of the incumbent. This creates a chain of interdependence from agency structure to policy choice and political accountability. The formal model shows that the variation in organizational capacity serves the interests of voters by improving electoral control of politicians.