Voting and elections are extremely important for democratic societies. If democracy is to be effective, it is essential to assess and mitigate the threats of fraud, manipulation, and coercion. However, formal analysis of voting infrastructures usually misses the fact that voting serves humans, is executed mostly by humans, and its outcome has serious impact on human welfare. Thus, we need to formally model and analyse not only the technological side of elections (i.e., protocols, architectures, and implementations), but also the human and social context in which it is embedded. The impact of the social factor has become especially evident in the view of some recent controversial elections, including the latest US presidential election. In particular, the Cambridge Analytics scandal showed to what extent a combination of technology and skilful social engineering is already being used to manipulate public decisions.In this project, we propose to use techniques from game theory, multi-agent systems, and theory of socio-technical systems to redefine and analyse various requirements in public decision-making procedures, such as voting and elections. In particular, we will propose and investigate variants of confidentiality, coercion-resistance, and voter-verifiability that take into account the social environment in which the election procedures and the voting infrastructure are embedded. The goal is threefold. First, we want to improve our understanding of the mutual influence between technological and social components in a democratic society. Secondly, we aim at development of formal methods for identification and analysis of threats in such systems. Thirdly, we will propose how to generalize existing, technology-centred solutions for secure and verifiable voting to face the challenges on the socio-technical level.