One of the ways in which mathematics might have helped to cause the financial crisis hasn't received much attention.
As economics developed over the last several decades, the technical requirements -- the math requirements in particular -- have increased substantially. In order to fit these new mathematical tools and techniques into the graduate curriculum and give them the attention they needed, it became necessary to find other courses to cut.
As economics departments looked around for courses to eliminate, eventually, if not initially, they turned to courses on US economic history and the history of economic thought. This was a mistake. Part of the reason we missed the crisis, I think, is that we forgot what crises look like. We forgot how they happen, even in some cases the fact that they happen at all. In short, we forgot the lessons of the past. What happened this time involved a different type of asset, and a new promise about how this time is different, but for the most part it really wasn't that different from what had happened many times before.
As we looked around to find courses to cut to make room for the new technical tools that were coming into economics, our mistake was thinking that US economic history and the history of economic thought were the least valuable courses that we offered. In fact, they may be among the most important. There's no way to know if more attention to and knowledge of the past would have made a difference, but I think it could have, and we need to rethink the elimination of these courses from graduate studies.