We propose a new three-dimensional approach to party newness and an interval index of party congruence/novelty. Building on that, we also propose a split-vote-by-congruence (SBC) approach to electoral volatility that employs the index. Four elections from different countries that exhibited different forms of party change are used to illustrate the approach. The congruence/novelty index corresponds to our qualitative case knowledge and the SBC approach leads to meaningful volatility scores. Ad hoc coding of parties as new/old or singular successors/predecessors can seriously over- or underestimate volatility that explains divergence among volatility scores in literature. By eliminating the need for dichotomous and often controversial coding decisions, the SBC approach allows for a substantially more reliable calculation of volatility.