Determinants of Gender Preferences at the Presidential, Parliamentary, Assembly Member, and Unit Committee Member Elections in Ghana

Abstract The study examines the determinants of gender preferences in representative democratic elections at the presidential, parliamentary, assembly member, and unit committee member levels in Ghana. The study uses the Ghana Living Standards Survey (Round 7) dataset with 13,833 observations. Multinomial Probit (MNP), Heckman Probit, and Binary Probit estimation techniques were employed in the analysis. The study finds that gender preferences of Ghanaian voters in elections are higher at the lower levels of elections and lower at the higher levels of elections. That is, gender preferences increase as one moves from presidential to parliamentary, assembly, and unit committee-level elections. At all levels of elections, voters prefer their own-gender candidates. Other factors found to be significantly associated with gender preferences in elections at various levels are age, religion, literacy status, rally attendance, and location (rural/urban). The study is unique to the extent that it examines gender preferences at all levels of elections in Ghana, which has not been done in previous studies. Policies that focus on reducing gender affinity and encouraging voting based on the abilities of candidates are essential. In general, policies aimed at reducing gender preferences in elections should focus on the correlates of gender preferences identified in this study.