Corruption is hidden action which distorts allocations of resources away from competitive outcomes. Hence the detection of such actions is both dicult yet important. In many economic contexts, agent actions are unobservable by principals and hence detection is dicult; sport offers a well-measured context in which individual actions are documented in great detail. In recent years the sport of cricket, which records a huge volume of statistics, has been beset by a number of corruption scandals surrounding the xing of matches. We use 18 one day international (ODI) matches that are known to be xed by one of the teams involved and anal yse a wide range of observed statistics from all ODI matches since 1971, in order to determine whether corruption manifests itself in recorded out comes. We nd that corruption does aect a number of observed outcomes in anticipated ways, suggesting that both the increased reporting of statistics, and the statistical analysis of them may be a useful tool in detecting corruption.
