At this point, I hope I've convinced the practical minded that a fixed money supply is both the most pragmatic and the most ideal approach to a functional system of money.
But now, practically speaking, how does one go about setting it up?
There would be a great deal of controversy surrounding the transition, that is for sure, and I do not think I'm up to commenting on such a task. But I do think that it is important to sketch out some idea of what the final system should look like. Over the years, many ways of cheating a system intended to enforce a fixed supply of money have been devised, along with the political salesmanship to get the necessary rule changes passed into law. Recognizing the critical characteristics of such a system helps one to recognize the cheating.
Cold Hard Cash
Probably the first idea that comes to mind for limiting money supplies would be some kind of law to the effect of "Thou shall not produce any more money." This approach is unlikely to be very successful, as any law requires state enforcement and the state is pretty unreliable as an overseer of money supplies, or pretty much anything else that matters. It is far too tempting to ignore the rules on such an issue. It is usually the state itself or some kind of state-banking alliance that is directly responsible for instigating money supply expansion, so it doesn't make much sense to leave the state in charge of keeping supplies contained, at least by this sort of method.
A far more effective safeguard against monetary gamesmanship is to tie money directly to something physical, as in using for money some physical object or substance such that the creation of new money requires the generation of more of it. Usually, the thing is chosen such that it is very difficult to get any more, so that the quantity and value of money is protected by this physical barrier to an increase in its supply. This could be part of a legal regime, but the important aspect is more cultural. When people come to expect money to mean some physical object perceived to be valuable, they are less likely to accept a substitute. Legislation to change money to something more fungible will be unpopular and viewed with suspicion. Unfortunately, with the tradition of paper money long established, that is not the case today.
The most obvious historic example of a substance successfully employed as money is gold. Gold has several important properties that help to protect against changes in supply. The first is that gold is extremely resistant to chemical destruction through processes like corrosion. It is also relatively rare, so that it naturally becomes too expensive to use for most applications that would result in its destruction or loss by other means. Both properties work to create a situation in which whatever gold is physically mined from the ground tends to persist in circulation virtually forever. Also due to its rarity and persistance, the quantity of gold mined each year is dwarfed by the total amount of gold presently in circulation, so that the total supply creeps up only very slowly over time.
Many other substances and items have served as money through the ages, but few can compare with as widespread a record of success as gold. However, silver is one that comes very close. It has the 'advantage' of being far more plentiful than gold, allowing easy coinage in substantial volumes, but this also works against silver in that it is plentiful enough to be relatively cheap. Many industrial processes use silver, and an increase in its relative value can spur increased supply through mining. In addition, mining of base metals yields significant quantities of silver as a byproduct, so there is a constant flow of silver into markets irrespective of changes in demand for silver. These silver 'sources and sinks' cause supply and demand fluctuations that mean that silver's market value tends to be more volatile than gold, and not necessarily for reasons connected with its role as money. For these reasons, I would favor gold as the money system of choice, and most modern, advanced economies have favored gold-based money systems over silver. But to be completely fair, those systems have never been pure gold commodity money systems, at least to my knowledge. Such a system in a modern economy is historically unproven, while systems much closer to pure silver systems have proven workable.
Austrian monetary theory holds that the selection of a monetary unit is a process that naturally occurs in the marketplace, requiring no intervention by government or any other such entity. Barter transactions will naturally sort out which is the most preffered commodity as participants gravitate towards one good or another to serve as an intermediary in acquiring the goods they actually want. This becomes money naturally once it becomes the most sought after good for this purpose. Traders will naturally seek out a good like gold that is durable and exists in limited supply for such a function.
In the present case, it probably is not necessary to go through the whole re-enactment of a barter economy to sort things out 'naturally.' That would be a bit goofy and impractical. It is already an established historical fact that precious metal moneys have performed best in this function and were eventually 'chosen' in some form by virtually every advanced economy in human history. So, it would be much easier and save a lot of trouble to just skip to the best available answer history has to offer. Hopefully, if and when it comes to that, there will be enough of a natural market instinct towards precious metals for historical reasons that it will happen without too much trouble.
Accept No Substitutes
Part of the problem with any commodity money is that, at some point, people do not want to actually physically transport the stuff around. It can be cumbersome and expensive. Many people would much rather use paper or digital accounting transfers, like paper money, checks, and electronic banking. This is all practical and understandable, but it creates dilemmas.
The paper and digits are used in lieu of specie precious metal itself because they are easier to deal with, like poker chips in lieu of paper money in a casino. The important point here is that when something begins to act as money, it is money by definition. If people are seeking digital accounting marks in the marketplace as the preferred tradable good, those digits become money. The casino holds cash on deposit in exchange for chips, and exchanges them back when one is done playing. It doesn't play games with the money in the meanwhile, so that the 'ratio' of chips to money on deposit is always on-to-one. The chips are effectively money in this case, but there are no problems created because the cash money held on deposit is effectively retired for the duration of the use of the chips. That is not the case in a banking system.
Today, most money is of this 'substitutionary' type. When someone writes a check and the check is deposited at another institution, the indicated quantity of the account balance is simply transferred between the two accounts. The balances themselves are the money in such a transaction, not anything in particular that they are supposed to represent. Nothing physical changes hands, like actual paper dollar bills. The deposits, meanwhile, have been lent out and are circulating right alongside the balances that supposedly represent them!
On some level, the use of 'substitutionary instruments' is going to happen in any system one might conceive of. It is inevitable. The important thing for maintaining a stable money supply is to ensure that these instruments do not become money in addition to the specie they are supposed to represent. If gold held on deposit actually gets lent out, or if deposits of money substitutes are allowed to count as if they were actual gold on deposit, or if in general more substitutes are allowed to be put into existence than a holding bank has on deposit in physical gold, you have the beginning of the fractional reserve process even with a supposed gold money system. Both the gold and the digits are now circulating around as separate money. Once that happens, all bets are off and the money supply begins its shot to the moon. The rules of such 'in lieu of' exchanges must be scrutinized very closely to make sure that money supplies are conserved throughout the accounting.
The Hollow Gold Standard
Enthusiasts of banking history will recognize these types of shell games as 'gold bullion standards' and 'gold exchange standards' and the like. Do not be confused by name games. These are not the same thing as a gold commodity money system and are not even the same as eachother. They are dressed up ways of cheating a real gold money system with accounting games.
Do not be fooled either by those touting the 'gold standard' as ideal, or by those bashing the 'gold standard' as proving that gold is useless as money. Both speak either in ignorance or dishonestly. The 'gold standard' is merely an enforced exchange rate between paper dollars and gold, so that some banking entity must redeem paper dollars for gold on demand. That does not in any way imply a one-to-one ratio of dollars to gold, only the legal obligation of redemption on demand. There is a big difference.
Typically it means that the banking system cheats, creating more dollars than can actually be redeemed in order to collect interest on money that should not exist. At some point, redemption becomes impossible and either paper money gets destroyed as the banks fail and depositors get stiffed, or an intervention takes place to allow evasion of the redemption guarantee or alter the exchange rate. Which is to say, there is a systemic default.
A 'gold standard' isn't enough for a stable money supply. Gold must literally be money in order for its properties to fully protect money's value. Merely gold backed money is subject to all sorts of trickery and cheating. The 'gold standard' is only an accounting rule or two away from a fiat money system and is not to be trusted.
Squashing Fractional Reserve Banking
Getting people onto the precious metals bandwagon may come pretty naturally once the fiat system we are presently using really gets out of control. The problem will be purging the system of the accounting games of fractional reserve accounting and central banking. And, frankly, it probably will not happen, so eventually the system will face more such crises down the road. But as long as I'm entertaining the pipe dream of establishing a perfect money system, I might as well talk about the rest of the picture. There are basically two approaches advocated by the Austrian School to keep the accounting games in check.
The first approach is to enforce 100% reserve banking by law, punishing infractions as acts of fraud. Which they are. The problem with this approach is the mechanics of enforcement. An enforcement agency would need pretty sweeping powers to audit the banking system and would probably need to be very aggressive and intrusive to actually keep monetary pyramiding on top of precious metal deposits to negligible levels. This assumes even that the regulators could not be corrupted, which is dubious. Such an enforcement regimen might run into the same kinds of problems that one sees with the War on Drugs and other such vice crusades that create black markets, violence, and corruption in their wake. On the other hand, the act in question is an act of fraud, which really does fall under the authority of the state to prosecute even by the most libertarian standard. To not prosecute is comparable to not prosecuting for acts of theft or violence.
The second approach is basically the "free-banking" approach. Banks are allowed to do business as they see fit. The state only intervenes in the instance of a complaint of violation of contract, like when the bank pyramids up a bunch of money on its precious metal deposits, goes bust, and can't pay back its depositors. The 'check' on phony accounting would be the pain inflicted on bankers who face bankruptcy proceedings, and the depositors who lose their money. It probably wouldn't be all that effective, since it wasn't effective in the past, either. But at least it would get the FED out of the way and enforce a fixed gold exchange rate, even if the rate was being gamed. Long term prices would presumably be fairly stable, at least as long as the 'gold standard' stayed in force.
This approach would basically take the banking system back to more or less what it was before the advent of the FED in 1913. There would still be the booms and busts of the business cycle as in the old days, but no regulatory police state, no massive systemwide financial catastrophes like the Great Depression and whatever the events surrounding 2008 get called, and long term price stability even with the short term price fluctuations of the business cycle. It wouldn't be a Utopia, but nothing ever is.
A Golden Business Cycle?
A couple of very logical questions usually come up when thinking about "Austrian Quasi-Utopia." The first is whether or not gold mining, which would increase the money supply even in a strict gold money system with no fractional reserve lending, would cause a business cycle in the same way that credit money expansion causes a business cycle now.
Of course, nobody really knows the answer because it has never been done before, and as one might guess, the experts come down on both sides of the fence. Some think there would be, some don't. I'm in the first camp. But I do think that it would be a very tepid cycle. The new money would not enter credit markets first (which is the argument the other camp uses to justify its position), so theoretically it would not cause the disparity in prices between capital goods and consumer goods that leads to the investment boom and bust. The gold-mining industry would get the money first, so it would presumably cause a price disparity in markets related to mining. Mining would therefore theoretically attract a disproportionate fraction of available resources. Whether or not that would lead to a full-blown system-wide cycle is anybody's guess.
Whatever the case may be, I think this is a small price to pay for a more honest and reliable monetary system. Milton Friedman opposed the use of gold as money because he understood this effect and saw 'mining for money' as a waste of resources. But for whatever reason, he trusted government with the issuing of money instead, which I always thought was strange for someone who supposedly mistrusted government. If you ask me, I'll take the gold miners over the politicians.
The other question that usually gets asked is how banking will work if the banks are forbidden from lending against deposits. What will banks do if not bank?
The answer is that they will more honestly fill the role that they have always filled -- matching up creditors with borrowers in return for a cut of the action. One typically talks about 'borrowing money from a bank,' but in reality, the ultimate lender is the depositor to the bank, not the bank itself. A bank is really just a middle-man. Without fractional reserve lending, this relationship would be far more straightforward.
A depositors who wished to collect interest on his deposit would be matched by the bank to a borrower. The depositor would become a lender and would forfeit use of their money for the duration of the loan because the borrower would actually receive the borrowed money from the lender (I know -- what a revolutionary concept!). Basically, it would be like purchasing a CD or a bond. This way, deposited money does not circulate alongside the lent out money and the money supply is conserved. Checking and savings accounts would not be elligible for lending, nor would they be elligible to collect interest. Users of these services would probably be charged a periodic fee. But that only makes sense -- it is, after all, a service. "Free checking" should be one of those things that raises an eyebrow of suspicion.
Banks have also traditionally filled many other roles which would not change at all. Only the transactions of lending against demand deposits would be eliminated. So, there would still be plenty of business for banks.
A gold commodity money system with discouragement or elimination of the double-counting of fractional reserve lending is probably the best money system realistically possible. It still would not have money in absolutely fixed supply, but it would be close. Placing authority over money in the hands of government is not a realistic solution and has not worked very well in the past.
This course of action is probably never going to be adopted, however. It is not politically popular, and like all systems of account, is an attractive target for opportunists and prone to erosion and corruption.
On the other hand, the failure of the present fiat money scheme appears to be approaching. The shock of it will probably create a strong popular desire for something more tangible as money, and with a good dose of luck, we might even get it. Should that happen, most likely it would be a return to some sort of gold or mixed metal system a la the 'Gold Standard.' The FED might even be abolished if it manages to be adequately demonized.
Things will be better on the monetary front, for a while. Then the games and the erosion will begin anew. In another century or two, there will probably be another fiat system in place, for awhile. Then another financial catastrophe.
Or it may take several more of these tragedies to shake faith in government enough see even that brief spell of stable money.
Some lessons are never learned.