Abstract
Despite increasing public attention and media coverage of fraud, no solution to this problem has been designed specifically for the not-for-profit sector. This study focuses on charities in England and Wales and examines variables derived from a content analysis of 42 fraud and 42 no-fraud charities' annual reports and financial statements. We use logistic regression to explain and predict fraud in the charity sector. We examine whether a range of governance-type variables are significantly related to the likelihood of fraud in charities. We find that smaller boards, which imply a small cosy environment for governance, are associated with a greater likelihood of fraud, suggesting that larger boards enhance monitoring in the not-for-profit sector. Also, a low or zero-level of grant funding is associated with a greater likelihood of fraud, implying that monitoring by long-term donors is also important in curbing fraud.