The authors examine the $35 billion bank overdraft market. Specifically the authors are concerned with whether overdraft fees are sufficiently salient for most people. They issued 21 surveys about overdraft fees between 20042008 to over 7,000 participants.They find that when people take the surveys, they are less likely to incur a fee in the following month. When they take several surveys, the likelihood of paying overdraft fees is significantly reduced for up to 2 years. This reduction comes not from increasing balances but from making fewer debit transactions and cancelling automatic recurring withdrawals. This indicates that the act of taking the survey is itself a kind of treatment, one that is effective at changing overdraft-related behaviour.