Why should we care what candidates tweet during a presidential campaign?(JMP)
This paper aims to inform about the new challenges that our democracies are facing. In the twenty-first century, the danger to democracies does not resemble old generals and tanks taking the streets; democracies now are eroding from within. Once in power, politicians elected through competitive and fair elections start to dismantle institutions and undermine society’s democratic foundations. This scenario is common but not exclusive to Latin America. Countries such as Hungary, Poland, Turkey, Russia, and India are experiencing democratic backsliding led by elected leaders. These leaders do not share a political ideology or a government style, but they have a common value low commitment to democracy (LCD). Which manifests in disdain towards democratic institutions and limitations imposed by them. Elected leaders with LCD take advantage of their political strengths or exceptional external circumstances, such as crises, to erode democracy generally or make attempts to stay (illegally) in power. This paper aims to show a new approach to identify candidates with LCD based on politicians’ behavior on Twitter and using Supervised Machine Learning Models. Moreover, I demonstrate the effect of having candidates with LCD on the likelihood of experience democratic backsliding regardless of contextual or institutional settings.