Tuesday, November 9, 2010

Martin Wolf demolishes the gold standard

Could the world go back to the gold standard?

Martin Wolf, FT
During any period of monetary disorder — the 1970s, for example, or today — a host of people calls for a return to the gold standard. This is not the only free-market response to the current system of fiat (or government-made) money. Other proposals are for privatising the creation of money altogether. But the gold standard is the classic alternative to fiat money.
It is not hard to understand the attractions of a gold standard. Money is a social convention. The advantage of a link to gold (or some other commodity) is that the value of money would apparently be free from manipulation by the government. The aim, then, would be to “de-politicise” money.
The argument in favour of doing so is that in the long-run governments will always abuse the right to create money at will. Historical experience suggests that this is indeed the case.
So why choose gold? It is, after all, an impossibly inconvenient means of exchange. But gold has a lengthy history as a widely-accepted store of value. If one is looking to reinstate a pre-modern monetary, gold is the obvious place to start.
After the experience of the last three decades the monetarism of Milton Friedman is no longer a credible alternative. It was abandoned for two simple reasons: first, it proved impossible for monetarists to agree on what money is; and, second, the relation between any given monetary aggregate and nominal income proved unstable.
Again, recent experience suggests that we can no longer be so confident that delegation to independent central banks protects against severe monetary instability. That system permitted a gigantic increase in credit, relative to gross domestic product. It is equally clear that governments do not wish to see this edifice collapse, for understandable reasons. This being so, the ultimate solution may be to increase nominal incomes, via inflation. Indeed, several economists recommend this. If that did happen, it would support those who argue for abandonment of the modern experiment with fiat money.
So would the gold standard be the answer? We would need to start by asking what a return to the “gold standard” might mean.
The most limited reform would be for the central bank to adjust interest rates in light of the gold price. But that would just be a form of price-level targeting. I can see no reason why one would want to target the gold price, rather than the price of goods and services, in aggregate.
The opposite extreme would be a move back into a world of metallic currency. But money in circulation will continue to be predominantly electronic, with a small quantity of paper, as today. That is the only convenient way to run a modern economy.
Finally, a return to the Bretton Woods system, in which the US promised to convert dollars into gold, at a fixed price, but only for other governments, would lack any credibility, since there would then be no direct link between gold stocks and the domestic money supply.
With these possibilities eliminated, the obvious form of a contemporary gold standard would be a direct link between base money and gold. Base money — the note issue, plus reserves of commercial banks at the central bank (if any such institution survives) — would be 100 per cent gold-backed. The central bank would then become a currency board in gold, with the unit of account (the dollar, say) defined in terms of a given weight of gold.
In a less rigid version of such a system, the central bank might keep an excess gold reserve, which would allow it to act as lender of last resort to the financial system in times of crisis. That is how the Bank of England behaved during the 19th century, as explained by Walter Bagehot in his classic book, Lombard Street.
So what would be the objections to such a system? There are three: difficulties with the transition; instability; and lack of credibility.
The biggest transition problem is the mismatch between the value of official gold holdings and the size of the monetary system. The value of gold held by central banks is apparently about $1,300bn, while global deposits of the banking system were about $61,000bn in 2008, according to the McKinsey Global Institute. To survive the slightest financial panic, the ratio of gold to bank money would need to be perhaps an order of magnitude higher.
One obvious objection is that this would generate huge windfall gains to holders of gold. More important, if policymakers set this initial price wrong, as they certainly would, they could unleash either deflation or inflation: the latter is far more likely, in fact, because private holders would start selling their gold to the central banks at such a high price. Apparently, about 90 per cent of gold is now privately held. So the expansion in the monetary base could be enormous.
Moreover, gold reserves are distributed quite erratically around the world. So some currencies would have to experience inflation and others severe deflation. A similar problem explains why it was impossible to recreate the gold standard after the First World War: too much of the world’s gold reserves were then held by the US.
What, then, about the problems of the steady state? One obvious point is that we would be back to the world in which the balance of payments would be settled by physical shipment of gold or, as it was later, by movements within central bank vaults. That would, at the least, be absurd.
A far more important problem is that of financial stability. Economists of the Austrian school wish to abolish fractional reserve banking. But we know that this is a natural consequence of market forces. It is wasteful to hold a 100 per cent reserve in a bank, if depositors do not need their money almost all of the time. Banks have a strong incentive to lend some of the money deposited with them, so expanding the aggregate supply of money and credit.
The government might seek to impose narrow banking: banks would have to back any deposits with notes or reserves at the central bank. But entrepreneurs could then create quasi-banks (let us call them “shadow banks”). These would hold deposits in the safe narrow banks and offer higher returns to customers, because they lend out surplus reserves for profit.
Such a system is unstable. In good times, credit, deposit money and the ratio of deposit money to the monetary base expands. In bad times, this pyramid collapses. The result is financial crises, as happened repeatedly in the 19th century. To prevent this one would have to move into the world of limited purpose banking recommended by Larry Kotlikoff, in which no financial institution would be allowed to promise redemption at par unless it held matching assets.* If so, the pure gold standard would require abandonment of the current banking system altogether.
A further danger is that the response to all shocks would have to come via nominal wage and price flexibility. A less obvious point is that the gold standard does not guarantee price stability. Depending on the supply conditions for gold, the price level might move up or down. In the long-run, however, the price level would probably tend to fall (because the supply of gold fails to keep pace with global activity). Such a world of trend deflation is liable to depressions if or when the equilibrium real rate of interest is less than the rate of deflation.
Another and, in my view, even more serious, threat to the stability of any gold standard regime is international. A peg to gold may prove radically destabilising for any currency if other significant countries failed to sustain domestic monetary and financial stability. There could then be floods of gold into or out of a currency that is well managed. The monetary and financial consequences could be dramatic, with severe deflation one obvious threat. This is precisely what happened in the interwar years, with the chaos emanating mainly from the US.
Finally, there is the fundamental problem of credibility - or rather lack of it. As Bennett McCallum of Carnegie Mellon University also notes in the Cato Journal, the forces that now demand inflation from time-to-time would demand a change in the gold weight of the currency as happened in the 1930s. “Historically”, he notes, “the gold standard provided a reasonable degree of price level stability over long spans of time because the population at large had at that time a semi-religious belief that the price of gold should not be varied but should be maintained ‘forever’.”
That faith has perished. Moreover, everybody knows it has perished. So whenever the economy was in difficulty, the only question would be how soon the gold price would be changed or the link abandoned.
In short, we cannot and will not go back to the gold standard. As L.P. Hartley wrote, “The past is a foreign country: they do things differently there.” We cannot live in the 19th century. It is foolish to pretend that we can.

No comments: