This paper is written for Michelle, who couldn't think of a good reason why we need central banks but thought it was "silly" not to have them. So this is an attempt to anticipate the arguments for a central bank and counter them.
Firstly there is the argument that the government should determine the rate of interest. Why? Interest rates are a price, the price of future good and services relative to present goods and services. We don't let the government set the prices of petrol, shares, rice, sex, books or chocolate because they set it too high/low and cause shortages/surpluses.
Why would the correct price of getting something earlier be more suitable to government calculation than the price of anything else? The market-clearing price is the price at which as much is produced as is consumed. Interest rates are the same, the correct one is where as much money is availible to borrow as people want to borrow at that price. Any greater and more immediate satisfaction is provided for future reward than people want, any less and the opposite. The correct rate depends on people's personal preference for immediate versus future value. There is no way for the government to know personal preference. That's why they can't efficently decide how much petrol, shares, rice, sex, books or chocolate should cost.
Stability is a common argument for central banks, which is surprising considering that central bankers themselves have admitted to causing the greatest example of instability in economic history. Well perhaps that's unfair, perhaps the great depression wasn't as big as some of the many hyperinflations that central bankers have not admitted to causing. The problem is that central bankers caused all of those too. The extremes of financial and monetary variability are the direct product of the actions of central banks. To argue that they create stability you'd have to find more examples of financial crisis before the advent of central banking than after it. The trouble is that there's only been one real bubble where the government didn't have control of monetary policy and that's "Tulip Mania" (say it with massively overpriced assets). There's been more bubbles than that in America alone in one Fed chairman's reign.
Of course perhaps central bank is more stable but the instability comes in greater doses when it comes. The problem with that theory is 1) there's no evidence of it and 2) there's is good theorectical reasons to disbelieve it. Financial and monetary instability is based in part on perceptions of instabiility itself. If there is a perception that a central bank will lead to large instability then that will by itself cause small instability. We know that there certainly is a perception that central banks contribute to extreme inflationary instability (hyperinflation) because anyone who studies it notices that they're the only known cause of it. And financial markets are comprised of people of people who know at least that much economics. Since at least Bernake's admission that "We [the central bankers] caused it [the great depression] they know that central banks cause extreme deflation too. So why would market be more stable knowing that fundamental decisions are being made by an institution that, every so often, causes massive instaiblity?
Of course there is more to stability than perception. Individual shares, debts and assets values vary depending on individual factors, but the market as a whole varies according to underlying economic facts. So which monetary system would cause these facts to change unpredictably and quckly? The one where a stroke of the pen can increase or decrease money supply by an infinite amount or where it can only increase after years of searching and expensive extraction.
Keynesian theories of economic instability are based on "irrational exuberence" and systematic stupidity of market players. Such stupidity is extremely expensive so why would it persist? Those who resisted such insanity could simply sell assets during the expansionary phase, buy more during the contractionary phase, rinse, repeat. This process would stabilise the market by itself. These "countercyclical investors" might not be rich enough to significantly stabilise the market, but they get rich every cycle. They also get access to greater control of other people's funds as there successful record grows.
Assuming a 4 year cycle with alternative 10% over and under valuations makes a 5.5% return a year on top of normal investment returns. Anyone who gets those returns for say 8 years in a row can easily borrow more to leverage their money. Leverage of only $1 of debt to $1 of your own money gets you 11% p.a. return above market. Such a system working since say 1913 would pay 2,021,543% plus normal investment returns. In other words if you had invested in such a system when America founded it's central bank and borrowed a modest 50% of the funds, you would have 20,215 times your what the average investor would have. And that's assuming you actually buy and sell things, rather than derivatives, which reap higher rewards when you guess right*. The "Mystery of banking" would therefore be, why don't these people own everything? They have had the chance to use this strategy since the first stock exchanges, why weren't they rich enough to stabilise the market by the time America got it's central bank? There must be something actively working against stability. I contend the central bank is it.
It has been contended that central banks should try to set the inflation or interest rate for the common good. This is impossible because, as with all prices there is no common good. That which is good for buyers is bad for sellers. In the case of the inflation and interest rate any movement that is good for those that borrowed money (inflation up, real interest rate down) is bad for those that lent it.
Balancing the interests of these groups is essentially a political decision. In fact it's a class warfare decision since the richer you are after all the more money you usually borrow (the poor can't afford loans). Why would anyone expect that it would be make objectively, if that's even possible? Any government organisation that controls interest rates is either going to be democratically accountable or not. If not then it's going to be used to benefit the ruling class. If it is then it's going to be used to benefit that portion of the population that is most able to intimidate their representatives. In neither case is it likely the good of the country will be the deciding factor.
Without a central bank, who would print the money? Well anyone who wants to and can convince you to accept it. There is no particular reason why printing money should be a government activity, let alone a government monopoly. After all what do you want from your money that only government can provide?
Firstly you want it to be accepted as money worth a predictable amount. You don't want to go to the store and find it's worth less than you expected it would be when you got it. When that happens people's economic plans are thrown into chaos because they cannot properly value things over time. Similarly you don't want it to be worth more than you expected, otherwise those who owe $100,000 suddenly find they need to pay back with goods and services they thought were worth $105,000. Either way people simply can't operate efficently. When people hear that the metric standard for mass is losing micrograms of mass they're rightfully alarmed, they should be just as alarmed when the standard of value unexpectedly changes.
Note that this does not mean that the value of money can't change, just that it should be relatively easy to predict when and by how much. This is not the case in central banks, which even their supporters claim are secretive and operate under principals that most people don't understand (quick what's M3? Is it better than M1? Why?).
Commodity based currencies on the other hand are a constant relative to the difficulty of producing the commodity. If the $US is defined as worth 1/20th of an ounce of gold then any someone can produce an ounce of gold for less than $20 they will. The increased supply of money relative to other goods will push up prices until it costs $20 to produce it again. The only way there could be a change in the value of money is if it became a lot harder or easier to produce the commodity relative to other goods. This happens slower and more predictably than arbitrary decisions of government officials. If a currency is based on a basket of commodities, with each unit entitling them to set amounts of each commodity, it would move even slower and more predictably. I'm not sure the world's ready for the McDollar based on the Big Mac though. These currencies can be issued by anyone who people trust to actually fork over the underlying commodity.
This brings us to the issue of trust. How do we know that private issurers of currency will actually honour their promises? Well the same way we know that the mechanic won't steal our car and that the child minding centre won't sell your kids to white slavers. In any case it's a moot point since we don't know that government will honour it's promises. The American government had promised to provide 1/20th of an ounce of gold per dollar, it then decided to only provided 1/35th. That's better than the British though who only a few years previously had refused to provide anything of value at all for the pound. Various hyperinflations under fiat money make such thefts look like small potatoes.
Secondly money should be transferrable with the minimum of cost. This includes being acceptable in as many places in as wide an area as possible to reduce the costs of changing money into something acceptable to the seller. Governments only advantage here is that it can threaten people with violence if they do not accept their notes. This is somewhat unfortunate to those that don't want to accept them, often with good reason. There is no reason why the free market can't design widely accepted money just like they design widely accepted credit cards. If there is a demand for money that can be spent from Bagdad to Cordova the market will provide it**.
Attempts by governments to provide international currencies depend on unifying fiscal and monetary policy across many nations. That hasn't worked out as well as it should. In the mean time fiat currencies vary in value relative to each other making international trade needlessly risky and therefore expensive. When money was gold and silver people didn't care what country your coinage was from as long as it had the weight of metal***. Imagine getting off the plane anywhere in the world and not changing your money. Well actually you don't need to imagine it, credit card companies already provide that service.
Thirdly money should be hard to fake and easy to differentiate from fakes. That is not only should it be hard to make copies of the money, but when people do it should be relatively easy to detect them. Again there is no particular reason why governments should be better than private enterprise at doing this. It is possible that government monopoly would mean that money is more familiar (since you don't need to remember what several competing firms money looks like) remember there is only a monopoly within the territory. Free market provision of money could easily lead to less issuers of currency.
* Derivatives are financial instruments like puts and calls where you don't actually buy or sell things, you buy or sell the right to buy or sell things. For instance a put is the right but not the obligation to sell a thing at a certain price during a certain time period a call is the opposite, the right to buy. If you buy a put for say 20,000 tonnes of X at a price of $100/tonne and the price goes down to $90/tonne during that period you've got something worth $200,000. If the price is stays above $100/tonne you're not obliged to buy or sell anything, but you've wasted the money you paid for the put. Depending on how likely people thought the price was to go below $100/tonne it could be very cheap. Given that in the "irrationality' theory prices are quite likely to be much higher/lower the market thought they would be profitabe derivatives will often be cheap.
** Those that get the reference will should have a good profit. ;>
*** With some exceptions, the priests at the Jewish Temple insisted on shekels, hence the presence of money changers.