The components of elections: district heterogeneity, district-time effects, and volatility

The quality and stability of democracy are often considered products of particular traits of parties or party systems. Studies of the 'personal vote' or 'district effects', the 'nationalization of elections', and 'electoral volatility', however, have generally conflated the different concepts at both the theoretical and empirical levels, thus raising questions about the independent relationship of each trait to democratic functioning. In response, we decompose district-level electoral data for parties in 20 countries from Europe and the Americas to clarify the conceptual dimensions along which parties and party systems may be compared. We show that what we term the 'district-time effect', 'district heterogeneity', and 'electoral volatility' are theoretically and empirically distinct properties of political parties with separable impacts on democracy, and we draw on this analysis to offer a new classification of political parties.