The Cato Blog has a great post based on an excerpt from Larry White's piece on Free Banking in History and Theory.
Free banking is the view that money and banking are best left to the invisible hand of market forces - no need for a central bank, lender of last resort or bank regulation. Free bankers argue that these things tend to destabilise banking systems and economies. They point to historical episodes where monetary and banking systems were relatively free of interference by the government to show the efficacy of free banking - episodes where banks issued notes redeemable for a precious commodity. Of course, no historical system is perfectly free of government interference, but Australia in the second half of the nineteenth century probably comes closest.
I've dabbled in free banking history in the past - see my article on Australia here and my article on early joint-stock banking here. I find that free banking worked when there were credible constraints on the ability of bankers to risk shift (e.g., unlimited liability) and failed when such constraints were absent e.g., Australia had a large banking crisis in 1893 - free banking went awry.
Do we need central banks? In my new book Banking in Crisis, I suggest that free banking was not possible in a world where governments needed access to the emergency war finance which could only be generated by central banks. However, one has to ask whether this is still the case for most modern economies. If it is not the case, what then is the rationale for central banks?