The Voter Behaviour: an Analysis of Presidential Opinion Polls

Abstract

This study sought to investigate the efficacy of opinion polls in influencing swing voters in presidential contest in both Dagoretti South and Kibra constituencies in Nairobi County. Anchored on David Morley‟s Audience Reception Theory, the study illustrates how the social positioning of a voter would influence their interpretation of opinion polls forecast from a dominant, negotiated, or oppositional position. This study used explanatory research design and the target population comprised 225,515 registered voters in Kibra and Dagoretti South constituencies. The 278 voters were selected through simple random and convenience sampling methods. The study achieved a response rate of 91 percent equivalent to 254 registered voters. Statistical Package for Social Sciences software was used to key data that enabled the generation of frequency distribution tables and bar charts for interpretation and analysis. The key informants were purposively sampled from a political party, academic scholar, opinion polling firm, political analyst and the Independent Electoral and Boundaries Commission personnel. The study established that the kind of questions asked by opinion polling firms, a voter‟s political party of affiliation and their choice of a presidential candidate are some of the major factors that informed their perception of public opinion polls. Based on these factors, the voters agreed that opinion polls are a tool for testing public opinion but hold the view that they are not correct if they do not interview equal number of voters across the political divide. The study revealed that opinion polls influenced voting behaviour in presidential elections with the influence mainly at perception level and presidential level. Depending on a voter‟s inclination to a given presidential candidate or political party, they will perceive any opinion poll that rate their candidate as leading to be true and polls that rate them trailing their close competitor as false. The study recommends that for public opinion polls to be perceived empirically, their autonomy and impartiality is only guaranteed by them not undertaking polling sponsored in any way by special interest groups like political parties since analysts and voters in general have come to associate them with either favouring or anti particular to a given political establishment. Secondly, the accuracy of the pre-election survey is highly hinged on opinion polling‟s carried out closer to the elections day, as voters are more decided then. Lastly, the National Cohesion and Integration Commission, and the Independent Electoral and Boundaries Commission should make deliberate and definite move to abate the prevalent influence of kinship cohesions and set-ups that manifest in the choice of a presidential candidate and political party formation in the country