Does silencing a group’s peaceful expression of radical ideology prevent further radicalization and political violence in democracies? How might a democracy’s use of repressive measures against non-violent political dissenters affect the likelihood of subsequent political violence? How do repressed activists groups adapt in a democratic environment? This project investigates how repressive measures are associated with the subsequent likelihood of radicalization and political violence in a democratic context. I argue that when democratic states use repressive measures to constrain the behavior of non-violent activist organizations, the activists resort to informal channels for reorganization and recruitment. These informal channels (such as anonymous web forums, chat rooms and personal friendship networks) are both decentralized and in use by those affiliated with other proscribed groups. When repressed activists adapt by using such informal channels in democracies, their personal networks increasingly overlap with those of individuals affiliated with other proscribed groups, including violent organizations. Thus as social solidarity seekers who are unable to organize openly, once-peaceful activists are more likely to experience radicalization and support or even carry out acts of political violence when they are repressed in democracies. I argue that this phenomenon occurs both domestically and internationally, because the fluidity of informal channels leveraged by repressed activists transcends borders. As a result, proscription of one group domestically can also lead to an unintended consequence of radicalized activist proliferation abroad. To test these arguments, I take advantage of a natural experiment and measure variation in repression of the non-violent Islamist group Hizb ut-Tahrir (HBT) across four democratic countries from 2002-2010. Known Islamist organizations are coded on an ordinal scale that differentiates those that are peaceful from those that are violent. Cutting-edge automated text analysis methods are then used to facilitate the extraction of network data from open source materials and archived Islamist web forum posts. After approximating the HBT network, dynamic network analysis is used to measure changes in the organizational affiliations of HBT members over time. Membership affiliations with violent Islamist organizations are expected to increase as a function of repression within democracies. This paper thus combines comparative case study research with cutting edge analytical techniques in order to demonstrate that democratic repression is not only associated with the further radicalization and potential violent behavior of political dissenters, but is also associated with the proliferation of radicalized political dissenters abroad.