Some G20 leaders are calling for an internationally implemented tax on financial institutions to create a fund for use if there is ever a need for more bank bail-outs. This is fairly popular in countries like the U.S., where people resent the fact that taxpayer money was used in past rescues.

Our finance minister, Jim Flaherty, is resisting this idea.

Flaherty makes two points. The first is that he doesn't think that Canadian banks, which did not require bail-outs like those in some countries, should be punished for the sins of others.

To me, that is a bit too complacent and self-righteous. Our government did give some support to our banks by guaranteeing bank debt and purchasing mortgage-backed securities. And we need to think for the future. There is no guarantee that the next crisis, whatever form it takes, will find Canada as well prepared as this one did.

The U.S. used to have a well-regulated financial system, too, with few bank failures for many years. But then they elected an administration with an extreme free market, anti-regulation ideology just as the financial sector was going wild dreaming up new “innovative financial instruments” to get around existing rules, and the rest is history.

Canada is not immune to something like that in the future.

Jim Flaherty's other objection to the bank tax is more reasonable. He says that the creation of such a fund would give financial institutions more incentive to make risky bets because they would know that they would be bailed out if something goes wrong.

It certainly could do that if it is not implemented right, although large financial institutions already know that governments are very likely to help if they get in trouble. But there would be a way to implement such a fund so that it does not create that incentive. And this brings up an important point that is hardly mentioned in all the discussion about bank regulation.

When the financial crisis hit, doing nothing was never a reasonable option. Simply letting financial institutions fail would have had devastating effects on the real economy, creating horrible unemployment in many countries for many years. A return of the great depression was a very real possibility.

But while preventing that required rescuing the institutions, it did not require rescuing their management and their shareholders. This was a mistake made in the U.S. and some European countries. Instead of rewarding the people who created the mess, governments should have taken over those companies, forced shareholders to absorb much of the loss, kicked out the management, injected the money needed, and re-sold the companies.

This is similar to what was done in Sweden during their financial crisis of the early 1990's, and it worked very well. (Sweden did not kick out the bank managers, but it did the other things listed.)

So I see no problem with using a bank tax to build up a fund for future bank rescues as long as there are clear rules about what happens if a bank ever needs that fund. It needs to be a painful process, especially for the bank management.

A bank tax could also be made to serve another purpose. Many people have suggested a financial transaction tax to reduce the sort of frantic, non-productive, speculative trading that produced this mess.

Mr. Flaherty is right when he argues that it is more important to regulate financial institutions against taking excessive risk. But regulation can fail if the regulating authority falls into the wrong hands, as we have just seen. We still need a contingency plan for such a failure, and a properly designed bank tax could reasonably be part of it.