日本語原文： ＣＯＲ(候補者超過達成比率)の分布（参院選2025年福岡のデータ）
- Relationship between Party Vote Share and the COR
In municipalities where the number of proportional representation (PR) votes for a party is extremely small, the COR tends to be overstated. In such cases, it is necessary to carefully distinguish whether a high COR reflects the candidate’s genuine mobilization capacity or merely the thinness of the party’s baseline support. Accordingly, the COR should not be interpreted in isolation; its meaning becomes analytically determinate only when evaluated in conjunction with the spatial relationship between parties and candidates in correspondence analysis (CA). Nevertheless, taking the numerical values as they are, the following interpretations can be made.
4.1 Operational Rules for Analysis
In correspondence analysis (CA) plots, a short distance between a candidate and their affiliated party indicates structural similarity in vote distribution—that is, the candidate receives votes in the same locations where the party performs well. If, under this condition, only the COR of a particular municipality is high, it is highly likely that this reflects statistical fluctuation caused by an exceptionally small party vote base in that municipality, rather than the candidate’s individual strength.
By contrast, when a candidate is plotted far from their affiliated party in the CA space, this suggests that the candidate is drawing votes from a constituency distinct from the party’s original support base. If such a candidate records a high COR, this provides strong evidence that the high vote-gathering efficiency is attributable to the candidate’s personal mobilization capacity rather than reliance on the party label. In this sense, CA distance functions as a contextual condition that qualifies the substantive interpretation of the COR.
4.2 Interpretation of the Reverse Case
First, when the distance between a candidate and their party is large in correspondence analysis, this implies that the candidate is campaigning outside the party’s traditional support base. A low COR in such cases can be explained as a structural decline in vote-gathering efficiency, reflecting the misalignment between the candidate’s electoral strategy and the party’s established territorial support.
Conversely, if the distance to the party is small—meaning that the support bases overlap and the distribution structures are similar—yet the COR is extremely low, this suggests either a failure to fully absorb the party’s baseline votes within its own stronghold or an inability of the candidate’s personal appeal to amplify those baseline votes. Here, the explanation shifts from structural conditions to factors related to candidate-level performance or campaign effectiveness.
Furthermore, from a purely numerical perspective, when a party’s baseline support in a given region is exceptionally strong (i.e., the denominator is very large), the COR may be calculated as relatively low even if the candidate’s absolute vote count is substantial. This phenomenon should be understood as a mathematical property of the ratio itself rather than as an evaluative judgment about candidate performance.
