The ideal of the Rule of Law depends upon the formulation and application of rules that are general, neutral, certain, and clear (Tamanaha 2004, 97). Where the legal rules and their application fail to meet these standards, or where excessive official discretion weakens the generality, certainty or clarity of the law, it undermines citizens perception that they are governed by the rule of law rather than the rule of man. Although the rule of law typically refers to a system of formal norms, informal norms within society and specific institutions may affect the extent to which citizens and officials perceive and comply with particular formal rules. For this reason, it is critical to evaluate the manner in which law is mediated and shaped by individual behavioral norms within specific institutional settings. In the context of the judiciary, the informal norm of stare decisis operates to strengthen the rule of law by ensuring that judges treat like cases alike, by providing for predictability and stability in judicial decision making, and by constraining judicial discretion. Although stare decisis has been the subject of abundant descriptive and normative scholarship, we know every little about how and why the norm of stare decisis develops in courts, what factors perpetuate or diminish the norm, and how the norm influences the character and productivity of judicial institutions. The state supreme courts offer a perfect laboratory to study stare decisis, since substantial variation exists among the state courts institutional characteristics. This study explores adherence to stare decisis by examining judicial decisions to overrule precedent in the fifty state supreme courts. Overruling decisions constitute the most extreme deviation from the norm of stare decisis. As a consequence, developing systematic information about overruling behavior will substantially enhance our understanding of judges willingness to adhere to the norm. We shall develop and test hypotheses concerning the development of the norm of stare decisis based upon a simple model of judicial behavior derived from the game of prisoners dilemma. This model focuses on the trade-off judges must make between embodying their own policy preferences in the law and following their predecessors decisions. From this theoretical base, we will derive hypotheses concerning the relationship between overruling behavior and a host of institutional variables such as court size, caseload, docket control, term length, selection method, and membership change. To evaluate these hypotheses, we propose to gather data on overruling decisions in the fifty state supreme courts over the period of their histories. This will produce a comprehensive database of all overruled and overruling cases decided by each state Supreme Court. Using these data, and employing a variety of statistical methods including event count modeling, we will evaluate the connection between these state supreme courts institutional characteristics and their propensity to overrule their own precedent (and consequently destabilize the rule of law). We hope to determine that state supreme court justices are more inclined to overrule their brethrens (or former brethrens) opinions in (1) systems in which the justices are elected and (2) systems with shorter term lengths. If we are successful, the results will be consistent with the game theoretic model presented in our proposal, which rests upon the notion that judges are rational actors who will pursue their own policy objectives unless they are otherwise constrained by their institutional environment. Where judges work in an environment that promotes inter-judge cooperation and the opportunity for retribution in the event a judge defects, it fosters the stability of precedent and hence the rule of law.