Determinants of spoiled and rejected ballots in Ghana's presidential elections: Explaining the impairment on voter participation

Résumé

This paper examines the determinants and electoral effects of rejected ballots in Ghana's presidential elections from 1992 to 2024, introducing the concept of voter impairment to theorise how rejected ballots undermine democratic participation. Integrating protest voting theory and ballot complexity theory. It analysed gazetted electoral data from the Electoral Commission of Ghana across nine election cycles (N = 9), combined with illiteracy and poverty data from the Ghana Statistical Service. Correlation analysis reveals that the number of presidential candidates on the ballot exhibits the strongest positive association with rejected ballots (r = 0.778, p = 0.014), followed by registered voters (r = 0.658, p = 0.054), while illiteracy (r = 0.533, p = 0.139) and turnout ratio (r = 0.563, p = 0.115) show moderate associations. Notably, poverty incidence demonstrates a negative correlation with rejected ballots (r = -0.631, p = 0.089), challenging assumptions about socioeconomic disadvantage as a primary driver. Trend analysis shows rejected ballots have consistently exceeded the combined valid votes of all minor candidates in elections since 2000, including the 2024 election, where 239,109 rejected ballots (2.1% of votes cast) surpassed the total of ten minor candidates and contributed to the 2000 and 2008 presidential run-offs. The paper concludes that rejected ballots reflect both protest voting (intentional dissatisfaction) and ballot complexity (accidental error), with the number of candidates emerging as the dominant predictor. The paper advances voter impairment as an analytical framework distinguishing inadvertent errors, deliberate protest, and administrative failures. It recommends reducing ballot complexity through access requirements, targeting voter education for low-literacy populations, and further study using district-level data to test causal relationships.