In his comment, Erik Feyen picks up on an important point: “state-owned” banks tend to have broader stated objectives than maximizing a measure of profit. Much of the evidence alluded to by Charles Calomiris is based either on profitability or on closely related measures. I have also written a paper a few years ago using this kind of comparison between the Landesbanken (owned by German states) and private banks, showing that the Landesbanken were a rather bad deal for the taxpayers. That seemed a fair comparison given that, Landesbanken have been essentially doing the same that a privately-owned wholesale bank would do. But such comparisons are trickier for other “non-private” banks such as (to stay with the German example) Sparkassen, i.e. savings banks. These also have relatively lower profitability (and related measures), but have historically performed functions that private banks did not. Comparing the “performance” of state-owned and privately-owned banks requires taking account their different objectives. Doing it properly is not trivial.