A large part of open data concerns statistics, such as demographic, economic and social data (henceforth referred to as Open Statistical Data, OSD). In this paper we start by introducing open data fragmentation as a major obstacle for OSD reuse. We proceed by outlining data cube as a logical model for structuring OSD. We then introduce Open Statistics as a new area aiming to systematically study OSD. Open Statistics reuse and extends methods from diverse fields like Open Data, Statistics, Data Warehouses and the Semantic Web. In this paper, we focus on benefits and challenges of Open Statistics. The results suggest that Open Statistics provide benefits not present in any of these fields alone. We conclude that in certain cases OSD can realise the potential of open data.