We address the question posed in the title of this paper by investigating recent developments in the literature that estimates the NAIRU. A necessary condition for the existence of a NAIRU is dynamic homogeneity: the Phillips curve should be homogeneous of degree one in lagged and/or expected inflation. But contemporary approaches to estimating the NAIRU typically assume rather than test for dynamic homogeneity, thus assuming (rather than testing for) the existence of a NAIRU. We argue that these developments remove the NAIRU from the domain of testable hypotheses and transform the concept into an article of faith. This does not constitute scientific progress.