Standing the Test of Time

The striking feature of Prof. Herbert Giersch's comments on indexation over a 25-year period is how well his arguments have survived the test of time. The 1974 paper can be read today with hardly any amendments.

This is not surprising. For Prof. Giersch has long cherished a vision of an entrepreneurial market economy. In such an economy the role of money would be neutral - or better - transparent. It would facilitate transactions but play as small an independent role as possible.

Indexation has been advocated for several reasons. It has been seen both as an aid to reducing inflation and as a means of living with any temporary or irreducible rate of inflation. Some of the original arguments against indexation during the inflation of the 1970s made me think of a man knocking his head against a brick wall. He refuses to stop knocking because he would prefer the wall to go away instead..

Deflation

While it is extremely unlikely that we can just forget about inflation on the basis of a few years or so so of favourable experience, deflation is also emerging as a danger. Policymakers have to be alert to threats from both directions. Indeed looking at the American economy and stock markets I worry about a burst of inflation followed by deflation in an explosive manner.

Indexation is, however, even more attractive against deflation. For a defect of conventional instruments is that there is no way by which nominal interest rates can fall below zero; and in practice below some small positive amount. The result can be quite high real interest rates in a periodofd slump or recession. It was the inability of real interest rates to fall sufficiently in a slump that lay behind the Keynes liquidity trap and it has been a big factor in the current Japanese long-lived recession. One compensating advantage of the price rises which accompanied the world oil price explosion of the 1970's was that real interest rates became temporarily negative; and this must have tempered the accompanying recession in the industrial world.

It is worth looking again at a few of Prof. Giersch's original statements and see how much they apply to a threat of deflation as well as inflation. In what follows I have followed Giersch's wording apart from changing the signs from inflation to deflation where appropriate.

For instance, if the prospective rate of deflation has been underestimated, escalator clauses protect debtors and borrowers. They also help to avoid a deflation-caused flight from real assets into money. Indexation will make calculation in real terms easier and make the conclusion of long-term contracts more attractive. If wage contracts with two-way escalator clauses are in use deflation will not result in unintended rises in real wages and therefore loss of employment. Layoffs in the structurally weak sectors will not be accelerated by unintended real wage increases.

Kinds of Indexation

There are many aspects to indexation. The obvious main categories are capital market indexation, wage indexation and indexation of social security or pension benefits. Running across this is a distinction between public sector and private sector indexation.

Perhaps a new category is that of exchange rate indexation. This has some subtleties. The Vauxhall motor company of the UK (owned by General Motors) proposed a few months ago indexation against the sterling-euro exchange rate. If sterling were weak against the euro and profitability rose in export markets sterling wages would be adjusted upwards. If sterling rose, nominal wage increases would be less, although I doubt if the firm had the courage to say they would actually be cut.

It is too early to say whether such deals can be successfully negotiated or whether they will catch on. They have to overcome a counterintuitive element. For the stronger that sterling and the greater the international confidence in the pound, the less favourably wage earners can be treated.

Kinds of Capitalism

One complication in someone from the Anglo-Saxon world trying to comment on Prof Giersch's papers is that the institutional background has become so different - more so that when he originally began to write on this subject.

As of 1997, indexed bonds were issued by at least 15 countries. The UK was probably the first western industrial country to do so (1981). It is has been followed by several others, most notably the United States and France. On the other hand such bonds still appear to be illegal - or at least frowned upon - in the German-speaking world.

Another difference is that nationwide pay agreements and even collective bargaining no longer have the same importance in Britain and the US as they do in many Continental countries. The Giersch proposals for two-way escalator clauses would probably make a big improvement in a country which cannot escape centralised bargaining. But they have less relevance where decentralised pay fixing already prevails.

In any case I suspect that capital market indexation is safer than wage indexation. Indexation of public spending, is worst of all - precisely because it takes away the safety valve which occasionally enables public spending to be edged downwards in the way Giersch would like.

The problem with wage indexation is that it is easier to negotiate - or even to get unorganised workers to accept - in an upward than a downward adjustment. Giersch himself refers to the danger ofindexationn being used to protectexcessivee real wages. (Section II, Para 8).In practice indexation contracts are likely to provide an uplift in times of inflation and, at best, stability in periods of deflation. This applies even more to government expenditure on, for instance, social security benefits.

This may conflict with economists priors. But if anyone does not believe me let him try adjusting public pensions or family benefits downwards. To be anecdotal for a second: I adjust the pay of my part time housekeeper upwards whenever the price index rises by a full point. I have never dared to make a downward adjustment, even though there have been some months when the UK consumer price index has fallen; and I do not fancy the task of having to explain to her what has happened.

Indexed Bonds

The biggest advance in indexation has been achieved in the UK in the Government bond market. One of the greatest puzzles is that while governments have gradually come round to issuing indexed bonds, the private sector has resolutely refused to go that way. There are tax obstacles, but I doubt if that explains all the resistance. If you talk to corporate borrowers they tend to say that indexed borrowing would be too risky because they do not know what they would have to pay in interest charges and on redemption. Economists can go on explaining to them until they are blue in the face that the real risk comes in unindexed borrowing, but it has little impact.

At this point rationalisations can go in two opposite directions. Those who are most attached to formal reasoning can denounce the stupidity of the private sector. Those who believe in the profound wisdom of market participants could say they have stumbled across defects of indexation which economists have yet to explain. In fact it is even worse. For while there is some willingness to index wages, at least in an upward direction, there is almost none to borrow on indexed terms, where the case is much more clear-cut.

In practice even indexed government bonds have not been issued on a sufficient scale to protect us from the consequences of past inflation or feared future deflation. (Just under 20per centc of outstanding UK Government bonds). But they have provided instead a bonus for the economics profession. For they furnish estimates of the real rate of interest likely to be more reliable than the traditional habit of subtracting from the nominal interest rate the historical inflation rate realised over anarbitraryy period.

A related use of indexed bonds is to provide some idea of expected future inflation, which can make official policy less dependent of the vagaries of econometric forecasts. The standard central bank objection to this approach is that medium term inflation expectations are themselves based on expectations of central bank behaviour. This objection is only a knock-out one if it is believed that central banks can and do rigidly determine the inflation rate rather than merely have a big influence on it.

Indexed bonds are not yet a perfect instrument for estimating either real interest rates or inflation expectations. Real yields on indexed bonds have recently hovered around 2 per cent in the case of Britain, while they are nearer 3 or 4 per cent for the US and other countries. This requires some explanation in a world of free capital movements.

The Bank of England analysis published in 1997 (Gilts and the Gilt Market) remarks that in most markets (eg the US, Canada and Australia) the increase in the redemption value of the principal due to the inflation uplift is treated as current income for tax purposes. But in the UK., the whole return on index-linked government bonds is taxed on a real basis. This widens the appeal of the bond beyond tax-exempt institutions. The Bank also mentions different degrees of uncertainty over the inflation index to be used and different lags in the computation of the inflation adjustment. The effect of these distortions is in my view exaggerated by the limited issue of such bonds and the thinness of the market in them.

Indexation for Monetary Policy

The most interesting aspect of his paper are the suggestions that Prof Giersch has given us for using indexation not merely as a way of living way inflation or deflation,but as an aid to preventing these monetary disturbances from happening. His first group of proposals involves fiscal policy and the Maastricht treaty. His third refers to wage policy. I leave others to comment on these. The heart of his suggestions lies in the second proposals for a new monetary rule for central banks. They should buy index linked bonds whenever the price level falls; and they should sell them whenever the price level rises.

Giersch is in a distinguished line of economists who have advocated automatic offsets to price level changes. They have the advantage over the technical monetarists in not requiring us either to fix a definition of money in a world of changing financial institutions. Nor do they require putting faith in estimates of the trend of velocity. The rule is operational and the reaction is directly to the price level rather than some intermediate indicator.

In any case nobody need worry that the proposal would deprive central banks of their much loved discretion. For they would still have to decide on many matters, such as how long a price increase or decrease would have to go on before engaging in open market operations, how far to take these operations and how long to wait for results before stepping up their scale. There would also be the familiar problems of which price index to use.

The Giersch proposal can be broken down into three components:-