The intellectual merit of this proposal emerges from its multimethod, data-based approach to investigating a puzzle of great theoretical and policy significance: When do gavels defy guns? When do judges invalidate actions of government? In doing so, it can make a major contribution to theories of democracy and authoritarianism and to policy-relevant knowledge. That judges are or should be able to defy rulers is one of the foundational assumptions of efforts to promote democracy and the rule of law. Yet rulers can refuse to implement judicial rulings, cut the court's jurisdiction, reduce the budget allotted to the court, and impeach the magistrates to bend the gavel to their will. These threats are particularly compelling in new or re-established democracies where judges do not enjoy the institutional safeguards or the popular support that protect their counterparts in more developed polities. Our understanding of when and how judges do defy rulers is underdeveloped. This proposal develop a theoretical framework that uses contextual, attitudinal, and legal variables to account for when and why judges try to limit the exercise of governmental power, applying the framework to the Supreme Courts of the Philippines and South Africa. Statistical analyses of the decision making of these courts during periods of authoritarianism and transitions to democracy will be supplemented by field research and interviews with Philippine and South African judges to construct a culturally grounded and parsimonious model of judicial control. The broader impact of this study is evident in the contribution it can make to promoting a better intellectual understanding of the role of courts and judges in democratization and establishing the rule of law the potential contribution it can make to efforts by policy makers to promote more democratic and lawful governments. Completing the research proposed will also build the infrastructure of scholars who come from and understand the struggles of democratizing nations recovering from authoritarianism.