The paper offers a critical discursive and pragmatic analysis of a corpus of hateful Facebook and Twitters status updates of politicians, political activists and voters in the 2016 pre-and-post election period, in Macedonia. Aiming to determine how power is exerted on social media, the paper focuses on identifying the stance social media users take when posting messages with political content. The analysis first attempted to unveil what speech acts the hateful posts are predominantly composed of, what roles the authors of (...) the posts normally assume, who the hateful political discourse in the given socio-political context is directed to, as well as what are some of the predominant linguistic strategies underlying the analysed hateful comments. The results show that, by using mostly assertive and expressive speech acts, social media users assume mainly the roles of analysts and judges and only subsequently the one of activists, they mostly address politicians directly and they use a lot of negative lexis, rhetorical figures and boosters as interpersonal metadiscourse markers to express their negative stance and exert power and dominance. (shrink)