One government, one money

An obscure economic dispute between Metallists and Cartalists reveals an unnoticed difficulty with Emu. Charles Goodhart, a leading financial economist, says that crises in the foreign exchange markets could simply shift to the bond markets
March 20, 1997

The conventional economic analysis of Emu compares the balance of savings in transactions costs-the cost in money, convenience and risk of dealing in more than one currency-against a potential worsening of difficulties in macroeconomic adjustment. After all a country's declining competitiveness can be mediated, and the blow softened, by a declining exchange rate. If there is no insulating national currency, declining competitiveness must translate directly into unemployment. Proponents of Emu tend to be both more optimistic about the size of the transaction cost saving, and more sanguine about any worsening of adjustment difficulties. Those more sceptical about Emu stress the absence of a trans-European mechanism to aid adjustment by transferring wealth from rich areas to poor areas via taxes and benefits, as occurs now within nations. They also doubt the magnitude of the transaction cost savings, and contest the claim of Emu supporters that without a single currency the single market is threatened. Both sets of arguments quickly progress from the economic to the political.

But the conventional view may be too narrow. It ignores the relationship between the fiscal, tax-raising, authority on the one hand and money creation on the other. And by so doing it fails to recognise what may be some of the most significant-and dangerous-of Emu's side effects. Specifically, there is a danger that national governments within Emu will find their fiscal positions more fragile, and that crises in the foreign exchange markets will simply shift to the bond markets. In order to avoid the heightened risk of default and financial contagion national governments will be forced to adopt a more deflationary stance than desirable.

There has been a continuing debate among academic economists between those who argue that the use of currency is based essentially on the power of the issuing authority (Cartalists) and those who argue that the value of a currency depends primarily or solely on its backing (Metallists). This argument becomes, by extension, one between those who see the evolution of money as a private, market-led response to overcoming the transactions costs inherent in barter (Monetarists), and those, again, who see it as largely dependent on the role of the state.

This apparently dry academic dispute may seem remote from policy debates over Emu. But the Metallist/Monetarist (M form) view has conditioned much of the thinking about not only the merits, but also the institutional conditions under which Emu can succeed. For that reason it is an important theory, but also a questionable one.

The relevant application of the M form view in this context is the theory of optimal currency areas (OCA theory). If transaction cost minimisation can explain the use of money in general it can also explain why any particular currency extends across a particular geographical space-why, for example, it is more efficient to have a single currency for the whole US rather than separate currencies for different regions. From this perspective there is no reason why currency domains need to be coterminus with sovereign states. Yet the one to one correspondence between countries and currencies is one of the most robust regularities of monetary economics. If Emu proceeds in Europe ahead of political union it will be an unprecedented event.

The alternative, Cartalist, school sees the power to create money as intimately bound up with the stable existence of government in general and its ability to raise money through taxation in particular, and consequently has no difficulty explaining or predicting the almost universal empirical observation of "one government, one money." The break up of existing federations into separate states (as with the USSR, Czechoslovakia, Yugoslavia, or the Austro-Hungarian empire after the first world war) or, on the other hand, the unification of smaller states into a larger federal state (as with the US, Germany, or Australia on their foundation), provide perhaps the starkest contrast of the predictive power of the two theories. In none of these historical events ought there to have been any substantial change in the OCA calculation, yet in every one the scope of the monetary domains followed that of the political.

The M form view has the stronger following among professional economists, but its popularity reflects its theoretical neatness more than its predictive strength. The interrelationship between the governance structure of the economy, the form and usage of money, and the power to tax is essential. The attempt to model the evolution of money without government (or taxation) may be intellectually rewarding, but is historically and practically invalid.

Does all this matter? From a Cartalist point of view the link between political/fiscal powers and money creation is central and this link will, uniquely, be broken by Emu: money creation will be the responsibility of the European central bank, bound by statute to act independently of the member states, which retain the principal responsibility for taxation. Under what circumstances will this separation work?

On the conventional view, to provide for the independence of the monetary authority, it is only necessary to constrain the fiscal deficit (government debt). In a sovereign state with its own money but an independent central bank the fiscal authorities are constrained to be prudent-not to spend more than they can raise-by the need to maintain demand for the government bonds they have to sell. But if in the end they are not sold, the central bank is bound to give way and bail them out by printing money. This understanding is part of the backdrop to the criterion in the Maastricht treaty (and the provision of the Waigel pact) which restricts the deficits of the member states as a proportion of their respective national incomes: it provides for the independence of the central monetary authority by enforcing prudence among the national fiscal ones.

But the analysis supporting this provision does not take sufficient account of the effect that the divorce between fiscal and monetary authorities will have on the former. The Cartalist view suggests that the fiscal authority of the member states will be significantly weakened. Participating member states will lose their ability to inflate away the real value of their national debt, which is, after all, part of the purpose of the excercise. They will also lose the power effectively to tax the banks by raising their reserve ratios, thereby increasing the money base and raising nominal interest rates. Traditionally a significant part of many governments' revenues, this "seignorage" has generally fallen to a low level in most EU countries. But more importantly they will give up their unchallenged, absolute ability to pay off their debts in their own currency whatever may happen to demand in the bond markets. There can, by definition, be no credit risk (risk of default) with domestic currency sovereign debt, and no liquidity risk (risk of a temporary shortage of buyers) either. The government always has the ability to step into the market to buy or sell its own bonds.

But what happens after Emu when there is a drop in the demand for the bonds of a particular national government? What if, for example, bond investors start to see Belgium as a worsening credit risk: it has an ageing population, a declining tax base, and a government which lacks the political will to trim its spending programme. There is a risk of a rapidly self-reinforcing run in the bond market: higher interest rates which worsen the government's current deficit, which in turn both increases its need to sell bonds and reduces demand for them. Such a run would be the equivalent, within Emu, of the periodic foreign exchange market crises which have plagued medium sized open economies (such as Britain) over the years. But these bond market crises could be much worse than their foreign exchange market forebears if the contagion threatened to spread from the debt of governments to that of the banks, and the financial system itself was thereby undermined.

There seems to be something of a vacuum in conventional thinking when it comes to the prospective markets for the Euro-denominated debt of participating member states. Analysts and commentators have looked to two types of precedent-the foreign currency debt of sovereign states, and the domestic currency debt of subsidiary states (such as those in the US)-and have drawn false comfort from both. For example, it has been suggested that the foreign currency debt ratings of the member states are a good indication of the likely ratings for their Euro debt. But the ratios of these countries' foreign currency debt to GDP will jump by an order of magnitude when domestic currency debt is added to the equation.

Similarly, others have suggested that states participating in Emu can afford to have less constrained debt levels than the US states because their continuing ability to tax is much greater. But what matters is the ratio between a state's debt service requirements and its ability to increase its cashflow via taxation. These ratios are much lower for the US states than for the EU nations. The US state with the worst credit rating in 1994 (Louisiana) had a ratio of debt service to revenues of 11.5 per cent. The equivalent figures were 17.8 per cent for Germany, and 50 per cent for Italy.

Nor do the markets themselves seem to have focused on the potential dangers of fiscal fragility for participating Emu states. The main feature of European government debt markets at present is a "convergence play," whereby interest rates are revised downwards towards German rates according to the perceived probability of a state's joining Emu in 1999. If the arguments above are correct this convergence play has been massively overdone, if it is not altogether wrong headed.

Thus the effective removal of central banks from participating member states within Emu may expose them to serious fragility and credit risk in their bond markets. There is, in effect, a once off loss of creditworthiness. So the fiscal constraints incorporated in the Maastricht/Waigel criteria are perhaps the least strict necessary to retain full confidence in the bonds of participating countries. Yet it looks as though these constraints are about to be relaxed to allow some would-be members (for example Belgium) to join.

All this suggests that Emu may not vanquish the demons of the foreign exchange markets after all, but simply move them on to torment the bond markets instead; that the net effect on macroeconomic management may be a bias toward deflationary policy; and in any case that there are important aspects of the whole enterprise for which both the theory and empirical research are dangerously weak.