An Optimally Classified Supreme Court: Teaching Optimal Classification as an Introduction to Models of Spatial Voting
Published:
This work provides an overview of optimal classification, a rank-ordered scaling procedure that can be used for voters in any parliamentary voting environment. Originally developed by Keith Poole as a computational approach to scaling members of the US Congress (see Spatial Models of Parliamentary Voting 2005), I translate the approach to scaling justices of the Supreme Court. At its core, optimal classification in a single dimension uses a cutting point procedure and a legislative procedure to optimally rank order voters (though it can theoretically be applied to more than one dimension). The benefits of the approach are found largely in its ease of implementation. Unlike statistical approaches (e.g., NOMINATE or Martin-Quinn Scores), the procedure can be completed with pen and paper if the number of votes and voters is small enough. However, using a computational approach (as opposed to statistical) means that the rank ordering will not produce cardinal distances - merely an optimal rank ordering. I ultimately address how the procedure can serve as a good tool for introducing students to models of spatial voting.