Donald Trump has consistently performed better politically than his negative polling indicators suggested he would. Although there is a tendency to think of Trump support as reflecting ideological conservatism, we argue that part of his support during the election came from a non-ideological source: The preponderant salience of norms restricting communication (Political Correctness – or PC – norms). This perspective suggests that these norms, while successfully reducing the amount of negative communication in the short term, may produce more support for negative communication in the long term. In this framework, support for Donald Trump was in part the result of over-exposure to PC norms. Consistent with this, on a sample of largely politically moderate Americans taken during the General Election in the Fall of 2016, we show that temporarily priming PC norms significantly increased support for Donald Trump (but not Hillary Clinton). We further show that chronic emotional reactance towards restrictive communication norms positively predicted support for Trump (but not Clinton), and that this effect remains significant even when controlling for political ideology. In total, this work provides evidence that norms that are designed to increase the overall amount of positive communication can actually backfire by increasing support for a politician who uses extremely negative language that explicitly violates the norm.