In this letter we argue that negative partisanship and affective polarization are intertwined yet distinct concepts: the first refers to identity, and the second refers to attitudes. Ignoring this distinction poses two problems at the empirical level – one of validity and one of analysis – because negative partisanship is a likely determinant for affective polarization which is often excluded from analyses, thus leading to an overestimation of the effect of positive party identification. However, current operationalization does not allow separate measurements. This depends partly on the under-conceptualization and the consequent imprecise measurement of negative partisanship. We discuss appropriate measures to use in future analyses.