I ran across this odd debate between Bryan Caplan and Walter Block. Here is Block's argument ("frb" means fractional reserve banking):
Consider this: A deposits 10 ounces of gold in B's bank; B gives A a demand deposit for these 10 ounces. B turns around and lends C 9 of these ounces, giving C a demand deposit for these 9 ounces. Thus, A and C both own full rights to these 9 ounces.
There is now a problem of over-determination or conflict in rights. A and C both have a FULL right to these selfsame 9 ounces of gold. They are both FULL owners of these 9 ounces.
But, one of the essences of the libertarian philosophy we share is that there CANNOT be a conflict in rights. Any seeming conflict is due to a misspecification of one or the other right. Yet, here, with frb, we have a GENUINE conflict in rights. Thus, frb is incompatible with libertarianism.
Note, I am NOT talking about practicality. It might well be (given no bank run) that A and C will not ACT incompatibly with one another; that is, both will not demand that B pay them these 9 ounces, an utter impossibility. No, I am talking about RIGHTS. Right now, before any bank run, there is STILL a rights contradiction.
Caplan just sputters. Here is the problem with Block's argument, as I think any lawyer would immediately recognize.
Block confuses property rights and contract rights. If I give the bank some cash and pay it to put this cash in a safety deposit box, then the bank can't use that cash. It can't lend it out to someone else; it can't list it as an asset on its balance sheet; it can't touch it without my permission. If the bank were to do so, then it would have engaged in theft, and the relevant employees would go to jail. Lawyers call this transaction a bailment.
But if I deposit some cash with the bank, I don't retain my property interest. Instead, I'm making a loan to the bank and I obtain a contractual right to repayment on demand. If I demand my cash (plus interest, if any) and the bank fails to pay me, then I can sue it for breach of contract and demand expectation damages. If the bank were not a bank but just an ordinary borrower, and it was insolvent, then I have to race other creditors for its assets; otherwise, my contract right is converted into a claim in bankruptcy, and I have to share with other creditors. (Since it is a bank, I may well obtain full compensation from the government, but that is not relevant to the debate.)
If you asked the bank whether it might lend out your money, it would most certainly tell you that it would. So it is not lying to you, and there can't be fraud. Nor is there any other contradiction, incompatibility, or problem with the arrangement. Depositors take a risk that the bank will breach the contract but anyone who enters a contract takes the same risk.
Block doesn't seem to have any problem with contract rights per se, but he does have a problem with a person entering a contract that gives another person the right to demand assets that the first person might not have. But all contracts are like this. People enter contracts expecting that they will be able to transfer money, goods, or services when they are due, but everyone understands that intervening events might make the transfer impossible, impractical, or unwise. The other party obtains a right to obtain damages for breach of contract, but if every contract where the probability of nonperformance is greater than zero were considered fraudulent, we would have no economy.