Abstract The intrinsic problem in empirically analyzing Supreme Court justices' decision making is that cases before the Court are not necessarily independent of the justices. When a justice has taken part in deciding a precedent of a current case, her present decision should be affected by her past decision. This effect, the most common feature under the common‐law system, would impose a difficulty in doing empirical research about judicial decision making. Thus, without controlling for this path‐dependent effect, any test cannot help but be incomplete. Focusing on the votes of justices categorized by ideological direction, in this paper I develop a model that explicitly considers individual justices' voting in the precedents. Using about 600 relations of Supreme Court cases between a later decision and a precedent, I quantify the effects of institutional and individual stare decisis and, furthermore, decompose various factors affecting individual justice's decision making.