The effect of the Ebola outbreak on political trust, attitudes, and voting behaviour

Lack of trust between communities, government, and health workers could be one of the key reasons why the Ebola outbreak in West Africa went out of control. It is only when the death toll reached unprecedentedly high levels that communities started to (partially) trust health workers and government educational campaigns. The importance of trust and strong relationships between communities and the government has been revealed to have been crucial in fighting the Ebola epidemic, as well as other emergencies in other countries.

This research project exploits the intensity and dynamics of the Ebola Virus Disease (EVD) in Liberia to shed light on how this outbreak influenced individual trust in the government, local authorities, health workers, international organisations, and religious leaders, as well as individual perceptions about these agents’ responsibilities and performance. Liberia is now considered an “Ebola-free” country. Nevertheless, understanding the causes and consequences of changes in citizens' attitudes and trust toward the government is of crucial relevance for the present and future political stability of the affected countries.

Given that senatorial elections were last held in Liberia in December 2014, and that presidential elections are scheduled for 2017, we can study how EVD affected turnout and voting over time in a rather clean setup. We will do so by relying on a novel and unique dataset that combines publicly available electoral results, own data collection, and EVD case data provided by the Liberia Ministry of Health and Social Welfare. We aim to causally estimate the impact of the epidemic on political outcomes using a combination of empirical strategies.

To overcome some of the implementation challenges, we will adopt an innovative sample selection and data collection strategy. About 2,000 respondents will be selected via a "random dialing of numbers" approach, implemented by our partner organisation VotoMobile. The sample is selected through a short Interactive Voice Response (IVR) survey. Once the list of mobile phone numbers is generated, enumerators from the local NGO Parley will perform interviews through live calls. To increase response rates, we will keep the survey concise, and we will provide airtime incentives. To reduce priming effects, the order of questions will be randomised.

This project primarily contributes to a growing body of research investigating the impact of large unexpected shocks, such as natural disasters, on political outcomes. The Ebola outbreak provides a unique setting to explore novel research questions, mainly because of its rapid diffusion and the “fear” that accompanied the epidemic. We also relate to the literature that studies retrospective voting, in an attempt to understand how citizens evaluate and act on their perceptions of government performance. Finally, a related strand of literature is characterised by the use of IVR and voice-based phone surveys to collect data in developing countries.

This study aims to inform government, politicians, and policy-makers about the political consequences of the EVD epidemic and the relevant role played by trust among citizens and other main stakeholders.