Political competition is expected to become less particularistic as prosperity rises and a middle class emerges. But particularistic linkages persist despite rising wealth in urban Ghana. Politicians are unable to commit to campaign promises with voters who want large-scale public policies, many of whom are in the middle class. This creates incentives to avoid mobilizing many of these voters and to ignore their preferences. As a result, voters who want major public policies rather than patronage differentially refrain from participation, allowing poorer voters to dominate the electorate and party organizations. This may only reinforce politicians’ disincentives to make policy appeals, and stall the emergence of more policy-based electoral competition even as the middle class grows.

By analysing the socially segmented party–voter linkages deployed by the Unión Demócrata Independiente (Independent Democratic Union, UDI), a Chilean conservative party, this article demonstrates the usefulness of combining Kitschelt's party–voter linkage framework with Gibson's conceptual approach to conservative party electoral coalition-making. In Latin America, parties take advantage of social fragmentation and the availability of non-state campaign financing to combine multiple linkage types and thus attract socially diverse constituencies. Although it is an opposition party, UDI's historical trajectory and organisation have enabled it to receive private funds from its traditional and party-identified core constituency (business and conservative sectors), whose programmatic preferences and interests it represents, and then use these resources in a ‘charismatic’ mobilisation approach and particularistic exchanges with a non-core constituency (low-income, non-traditional voters of the radical right), in a segmented, but nationally integrated, electoral strategy.