We test whether outside experts have information not available to insiders by using the voting record of the Bank of England’s Monetary Policy Committee. Members with more private information should vote more often against conventional wisdom, which we measure as the average belief of market economists about future interest rates. We find evidence that external members indeed have information not available to internals, but also use a quasi-natural experiment to show they may exaggerate their expertise to obtain reappointment. This implies that an optimal committee, even outside monetary policy, should potentially include outsiders, but needs to manage career concerns.