This article seems uniquely inappropriate for the point it is making as to the importance and uses of Presidential charisma. Not that its assessments of FDR or Obama necessarily are wrong, but the main anecdote seems to make precisely the opposite points.

Specifically, bank runs (as shown in It's a Wonderful Life) are indeed a self-perpetuating crisis of confidence, and can stop as suddenly as they start once confidence is restored that claims on the bank will be honored. A great Presidential speech, such as the one given by FDR, can be useful in creating such confidence. But it helps even more if it can be conclusively demonstrated that such claims would, in fact, be honored. And that is exactly what Congress did prior to FDR's speech, in creating federal deposit insurance [UPDATE - I erred here, see below].

The article doesn't deny that the Congressional action helped, but it gives most of the credit (via the Robert Caro quote) to FDR's speech and then uses that assignment of credit to build an argument as to the uses and limits of Presidential charisma generally. That seems strange to me, given that the specific problem of bank runs (a) is more susceptible to matters of confidence than most other issues (does anyone think that more confidence in our health care system would solve its problems?) and (b) were solved by a legislative action that did more to guarantee the necessary confidence than any Presidential speech possibly could.

Maybe it was a "you had to be there" moment. But those of us who weren't there often have a leg up in analyzing such moments, for that very reason.

UPDATE: It also would have helped to get the facts straight. The "legislation securing the banks" cited by the article's author was not the creation of federal deposit insurance, as I erroneously stated, but the Emergency Banking Act, which infused enough capital into the banking system to avert the crisis. I think the general points remain true, namely that (a) FDR's famous speech only came after legislation fixing the problem (albeit not in as permanent a fashion as the FDIC) and thus may be given more credit than it deserves, and (b) bank panics are more susceptible to these type of fixes, making them of limited utility for assessing the usefulness of Presidential charisma. But I did err, and thanks to commenter Spencer for pointing it out.