Why do some investors delegate financial decisions to supposed experts? We report a laboratory experiment designed to disentangle four possible motives. Almost 600 investors drawn from the Prolific subject pool choose whether or not to delegate a real-stakes choice among lotteries to a previous investor (an ``expert'') after seeing information on the performance of several available experts. We find that a surprisingly large fraction of investors delegate even trivial choice tasks, suggesting a major role for the blame shifting motive. A larger fraction of investors delegate our more complex tasks, suggesting that decision costs play a role for some investors. Some investors who delegate choose a low quality expert with high earnings, suggesting a role for chasing past performance. We find no evidence for a fourth possible motive, that delegation makes risk more acceptable.