Politicians, Parties, and the Persistence of Weak States: Lessons from the Philippines Gabriella R. Montinola ABSTRACT Under what conditions will politicians strengthen state capabilities through bureaucratic reform? This article presents a principal±agent model of state capacity that shows that unless competition to in¯uence policy is largely con®ned to a single dimension, politicians have no incentive or `political will' to adopt bureaucratic reform. The validity of this model is tested using the experience of the Philippines in the period 1946±72. It is found that politically relevant groups in the Philippines were competing to in¯uence policy over issues of social welfare, economic nationalism and control over public spending, and that groups' demands over these issues failed to align along a single dominant dimension. Consistent with expectations derived from the principal±agent model, there were numerous calls for and attempts to improve bureaucratic performance, but the calls went unheeded and the attempts failed miserably. The study highlights a lesson of potential use to countries currently undergoing democratization. It argues that state capacity is not a function of insulating the state from societal forces, but rather of clarifying lines of authority and accountability in the hierarchy of principal±agent relationships between society, politicians and bureaucrats. 1. INTRODUCTION Many Ð probably most Ð development theorists agree that a country needs a competent bureaucracy, often referred to as a strong state, in order to promote economic development. But how does one create and maintain a competent bureaucracy? And why do weak ones persist? Some scholars argue that bureaucratic capabilities are determined by the distribution of social control between the state and other societal organiza- tions (Kohli et al., 1994; Migdal, 1988). According to this `state-in-society' view, macro-historical forces, such as the spread of capitalism and colonial- ism from Western Europe, determined the distribution of social control within less developed countries at independence. This colonial legacy in turn determined whether states would have the capacity to `penetrate society, regulate social relationships, extract resources, and appropriate or use resources in determined ways' (Migdal, 1988: 4). Development and Change Vol. 30 (1999), 739±774. # Institute of Social Studies 1999. Published by Blackwell Publishers Ltd, 108 Cowley Rd, Oxford OX4 1JF, UK.