This paper revisits the ambiguous relationship between tourism and economic growth, providing a comprehensive study of destinations across the globe which takes into account the key dynamics that influence tourism and economic performance. We focus on 113 countries over the period 1995-2014, clustered, for the first time, around six criteria that reflect their economic, political and tourism dimensions. A Panel Vector Autoregressive model is employed which, in contrast to previous studies, allows the data to reveal any tourism-economy interdependencies across these clusters, without imposing a priori the direction of causality. Overall, the economic-driven tourism growth hypothesis seems to prevail in countries which are developing, non-democratic, highly bureaucratic and have low tourism specialization. Conversely, bidirectional relationships are established for economies which are stronger, democratic and with higher levels of government effectiveness. Thus, depending on the economic, political and tourism status of a destination, different policy implications apply.