The Retreat from Monetarism

At the end of the Seventies, having received both a Nobel Prize and the still greater accolade of his own TV series, a diminutive retired professor from Chicago became for the moment one of the most influential private individuals in the world. Fêted by financiers, mobbed by the media, patronised by presidents and prime ministers, Milton Friedman had at last arrived. The doctrine for which he had fought – initially almost single-handedly – for a quarter of a century had become the New Orthodoxy. Monetarism had overthrown discredited Keynesianism, and nowhere was its victory more warmly welcomed than at Numbers 10 and 11 Downing Street.

Yet today things look a little different. Despite more than six years’ declared commitment to monetary discipline (even longer if we include the half-hearted efforts of Denis Healey), the UK inflation rate remains stubbornly above that which the Government wishes to see. Meanwhile unemployment, which was only expected to rise temporarily, has been stuck at pre-war levels for four years or more, and shows no real sign of falling. Moreover measures of the money supply – supposed indicators of the monetary stance of the Government – are all over the place. Some over-predict the inflation rate; some under-predict it. In the United States, on the other hand, a theoretical commitment to monetarism has gone hand-in-hand with a fairly permissive monetary policy and an outrageously lax fiscal policy. This combination should, in Friedman’s framework, have produced disaster. However, the performance of the American economy continues to make those on this side of the Atlantic turn green with envy.

Small wonder, then, that the Chancellor has back-pedalled. In his Mansion House speech this autumn he effectively abandoned the sterling M₃ target which had been the central indicator of monetary policy since the Thatcher Administration came to power. And behind the smokescreen of the receipts from privatising everything that isn’t tied down, it is suggested that he intends to produce an old-fashioned reflation timed nicely for the runup to the next election (much to the disgust of his erstwhile supporters at the Centre for Policy Studies, who have recently denounced him as a turncoat).

But this is not simply a case of politicians running away from the task of administering a harsh but necessary medicine to an increasingly fractious and disagreeable patient. For it’s clear that the doctors clustering around the patient’s bedside are less and less confident that the Friedman medicine is the panacea they had been led to expect. Some are muttering that they had always believed monetarism to be a quack remedy anyway and offer their own nostrums instead. Others, while still holding that the medicine is of therapeutic value, now argue that it can only be a part of the recovery regimen.

So what’s it all about then? In order to get some idea of the reasons for this retreat from the pristine certainties of monetarism, we have first to clarify what the doctrine entails.

Friedman’s monetarism, remember, is a modern restatement of a very old idea: the Quantity Theory of Money. This theory, clearly present in the work of David Hume and David Ricardo, was formalised in the work of Alfred Marshall, A.C. Pigou and Irving Fisher at the beginning of this century. With the rise of Keynesianism, however, the primary lesson of the old-time religion – crudely, that price inflation is caused by an increase in the money supply – fell into disfavour. This was because Keynesians thought that the velocity of circulation (the frequency with which money changes hands) would alter in response to interest-rate changes brought about by changes in the money supply. If this occurred, there was no need for the price level to alter. Friedman accepted that changes in the interest rate on bonds could in principle affect the velocity of circulation. But he pointed out that this is also true of the rate of return on all other assets, and that there is no reason to suppose that interest rates alone will have all that significant an impact. In any case, this is an empirical question, and Friedman has always been a keen advocate of testing hypotheses. His econometric work, and that of his colleagues over many years, culminated in his monumental Monetary Trends in the United States and the United Kingdom, co-authored with Anna Schwartz and published in 1982. This body of work suggested that the velocity of circulation, although varying over the business cycle, is in the long run fairly stable. In particular, it is more stable than the value of the Keynesian multiplier. This implies that monetary policy is more powerful than the fiscal policy espoused by Keynesians, and that monetary disturbances are more likely to destabilise the economy than are independent changes in consumption or investment.

Friedman has been expounding this view to anybody who would listen since the late Fifties. After years of neglect, why did it catch on so rapidly in the Seventies? One very obvious reason was the dramatically higher rate of inflation in this period. The Keynesian orthodoxy had little to say on this: Keynes’s General Theory had been conceived in a climate of stable or falling prices. Insofar as Keynesianism had an answer for inflation it was in terms of wage restraint, and repeated attempts at incomes policy in Britain, the United States and elsewhere had been only partially successful. By contrast, Friedman offered a simple theoretical explanation of rapidly increasing prices (‘inflation is always and everywhere a monetary phenomenon’) and an apparently clear remedy – monetary restraint.

Moreover, it was increasingly realised in the Seventies that inflation was not something which could be tolerated in return for lower levels of unemployment – a belief, popular in the Sixties, which was to be disavowed by Jim Callaghan at the 1976 Labour Party Conference. The unemployment-inflation ‘trade-off’ was embodied in the famous Phillips curve. As early as 1967, in a Presidential Address to the American Economic Association, Friedman had offered a theoretical explanation for the existence of a short-run trade-off, but argued that this option would disappear in the long run. This prediction of the eventual breakdown of the Phillips Curve seemed to be vindicated in the Seventies as unemployment and the inflation rate rose together.

His argument was couched in micro-economic terms. Individuals and firms reacted rationally to what they perceived to be real variables: money wages and prices adjusted for anticipated inflation. If people correctly anticipated inflation, Friedman showed there could be no trade-off with unemployment (which would settle at a ‘natural’ rate determined by labour market phenomena such as trade unions, minimum wage laws, social security, provisions). If a worker was unemployed at a wage rate of £x per week, such a worker would remain unemployed if wages rose to £2x per week but prices were also perceived to have doubled.

However, if inflation was incorrectly anticipated, changes in behaviour would occur. If workers noticed only the increase in money wage offers they received, and not the increase in prices, they would think real wages had increased and be more willing to take jobs. If employers only noticed that the prices they could charge had increased, but failed to notice all their costs were rising at the same rate, they would think production was more profitable in real terms and would employ extra labour.

Thus unanticipated inflation could lead to unemployment falling below its natural rate. But Friedman argued that in the course of time expectations of inflation would adapt to changed conditions. When this occurred, people would revert to their original behaviour patterns and unemployment would return to its natural rate. Unemployment could only be kept permanently below this rate by continually accelerating the rate of inflation, a potentially disastrous policy which Friedman saw as breaking the relationship of trust which ought to exist between governors and governed. Keynesianism, in this view, was a fraud and a deceit, as well as being ultimately pointless. The monetarist sees himself as a moralist, whatever bishops may think.

All this led Friedman to conclude that variations in output, employment and (crucially) prices will be minimised by the adoption of a firm ‘monetary rule’. This involves a clear and irrevocable government commitment to expand the money supply at a constant rate reflecting the underlying rate of growth of the economy plus any predictable change in the velocity of circulation as a result of long-term institutional change. However, a consequence of the way in which inflationary expectations are held to lag behind current experience is that an attempt to reduce monetary growth to its long-run path very quickly may lead to a considerable temporary rise in unemployment. Workers who are expecting a high rate of inflation will hold out for high money wages which employers, seeing that they cannot increase prices very rapidly, will be unwilling to grant. Thus Friedman has always advocated a gradual disinflation rather than a short sharp shock. His reasoning lay behind the adoption in 1980 of the British Government’s Medium Term Financial Strategy. Under the MTFS the monetary targets were planned to fall steadily until inflation was reduced to a very low level. And in line with MTFS targets, the Public Sector Borrowing Requirement (thought to be the major influence on the rate of growth of the money supply) was gradually to be reduced.

So much for the background to the UK monetarist experiment. What about its outcome so far? Well, it can’t be denied that inflation has fallen from the high rates reached in the Seventies. But it has not fallen as far or as fast as anticipated. Moreover, rather than monetary restraint leading to greater macro-economic stability, we have experienced a deeper recession than any experienced since the war, and one which is more marked than in other developed countries.

The Public Sector Borrowing Requirement has been considerably reduced. Indeed, on a cyclically-adjusted calculation (reflecting the fact that the PSBR automatically tends to rise in recession, and fall in periods of recovery), we are now in the position of effectively running a budget surplus. Other countries, such as Japan and the United States, have run or are running massive budget deficits and yet have inflation rates no worse than ours and unemployment records which are much better.

Despite controlling the PSBR, however, the Government has rarely managed to keep its main money supply indicator, sterling M₃, within the target range. Its recent performance in this respect has, as we noted earlier, led the Chancellor to play down this indicator of monetary stance. Now the mumbo-jumbo of monetary targets, with their differing definitions of money – unproblematic to the man or woman on the Clapham omnibus – is too obscure to go into here* At its simplest, there are different views as to what constitutes an appropriate definition of money in modern conditions. Different measures of the money supply may behave differently, and early monetarists argued that the indicator of money supply to focus on was that which was most highly correlated with money income. This was believed to be sterling M₃, a ‘broad’ money measure consisting of notes and coins, and current accounts and time deposits with commercial banks. More than 90 per cent of this measure is made up of bank deposits, and can only be influenced indirectly by the manipulation of interest rates by the Bank of England. It is clear that the significance of sterling M₃ has changed in recent years as the financial market has altered (for instance, with the increasing competition offered by building societies), and the effectiveness of government influences on this aggregate has diminished. This has persuaded monetarists, led by Friedman, to advocate targeting M₀, a very narrow measure of money (mainly notes and coins) which the Government in principle controls directly. However, severe restrictions on M₀ would involve much greater fluctuations than previously in interest rates, which would probably have a destabilising effect on investment, and would conflict with the Bank of England’s traditional function as lender of last resort.

In any case, the difficulty of controlling money supply is only part of the problem. For in the last few years the supposedly strong relationship between money supply indicators – whether broad or narrow – and money income has broken down. In Britain, the United States, Japan and other countries there have been large unexplained changes in the velocity of circulation. Moreover, serious doubt has been cast on some of the earlier evidence on which the monetarists’ policy proposals were based. British econometricians David Hendry and Neil Ericsson, for instance, have subjected Friedman and Schwartz’s work to detailed criticism. They claim that the data were massaged to fit Friedman’s hypothesis, and that the long-run velocity of circulation in the UK and the United States is in fact highly volatile.

Faced with these problems, economists have been discussing alternatives to monetary targets à la Friedman. One suggestion is that targets should be set for the level of money national income in the economy, the restraint being that excessive monetary expansion would mean that a planned rate of growth of money income would then mainly come about through an increase in the price level, and that the rate of growth of real output would have to be small or even negative. A major problem with this approach is that data for national income only appear after a considerable time-lag (unlike money supply figures, which appear each month), too late for monetary restraint to have an impact.

A more plausible suggestion, and one with which the Chancellor appears to have some sympathy, is that the exchange rate should be targeted. The idea here is that an excessive rate of monetary expansion causes a fall in a country’s exchange rate (a point which was not stressed in early monetarist writing). Thus keeping to a target exchange rate will involve monetary restraint without the need to rely on any particular monetary indicator. It is now widely realised that the pre-1972 system of fixed exchange rates was a powerful means of enforcing monetary restraint, and was largely responsible for the relatively low inflation rates of the Fifties and Sixties. There is a considerable irony here, for Friedman was one of the earliest advocates of scrapping the system of fixed exchange rates.

In practice, then, the links between PSBR and money supply, money supply and money income, money income and prices, have proved to be less robust than Friedman and his supporters confidently predicted a decade or more ago. This has led governments – even Mrs Thatcher’s rather rigid crew – to modify their stance.

Changes have also taken place in the economic theory connecting inflation to unemployment, where the Friedman ‘story’ has increasingly been challenged. One line of criticism has come from the ‘New Classical’ economists, whose most noted adherents come from a younger generation – in the United States, Robert Lucas and Thomas Sargent, and in the UK, Patrick Minford. These writers tend to see Friedman’s arguments as outdated, tarred with the same brush as the Keynesianism he previously denounced. New Classicals build their analysis on three principles. First, they argue that inflationary expectations are formed on a ‘rational’ basis: people use all the information they possess, including knowledge about government policy, to predict inflation rates. New Classicals therefore deny that people can be fooled by governments in the way Friedman suggested. Secondly, New Classicals argue that markets continuously clear, and that supply and demand can therefore never be out of line – an assumption which again goes considerably further than Friedman would go. Thirdly, aggregate supply depends only on the ‘natural’ level of output – corresponding to the natural rate of unemployment – unless there are unanticipated price shocks which disturb the economy. This means that debates about the relative effectiveness of fiscal and monetary policy are irrelevant – neither has any effect on real output, even in the short run.

In this approach, Friedman’s economics seem old-fashioned and outmoded – and also econometrically naive, since he has always worked with small-scale partial-equilibrium models, rather than with the large-scale macro-economic models favoured by the New Classicals.

An interesting offshoot of New Classical thinking is renewed interest in the business cycle as the centre of macro-economic analysis, rather than the determination of some static-equilibrium level of output, which was the focus for Keynesians and monetarists alike. This emphasis recalls the pre-war work of older writers such as Hayek, and it’s worth pointing out that Hayek and Friedman (although often classed together as Thatcherite gurus) differ considerably over a number of policy issues. For instance, Hayek has in recent years been very sceptical of monetary targets, to which he believes governments will never stick. Far better, he argues, to denationalise money and leave its creation to the free market. This may be an extreme and impractical view, although it does in a way seem a more logical outcome of Friedman’s thinking than the proposal for a monetary rule. Friedman has always been a determined advocate of deregulation in product and labour markets: why then does he insist on government regulation of the financial and money markets?

The other main line of reasoning – that associated with neo-Keynesians such as the late Arthur Okun – takes on board Friedmanite criticisms of Keynesianism, but has built up a new approach which concentrates on disequilibrium situations where wages and prices do not automatically adjust to bring supply and demand into balance. This situation is explained in terms of micro-economic reasoning: wage and price stability bring benefits to workers and firms in a world where information is imperfect, suppliers and purchasers, employers and employees, are heterogeneous, and there is not the continuous, costless search implied in, say, the New Classical approach. A model based on these assumptions may have the ‘Keynesian’ property that an increase in effective demand can increase real output.

So where does all this leave us? Whereas in 1969 President Nixon could assert ‘we are all Keynesians now’ and a decade later just about everybody could assert ‘we are all monetarists now,’ the position today is much less clearcut. Monetarism is no longer the state of the art in macro-economic theory, and governments have realised that the apparent simplicities of its policy recommendations are illusory. Nevertheless there will be lasting legacies of monetarism’s brief reign. I find it difficult to believe we will ever go back to the uncontrolled monetary expansionism of, say, the Barber Boom period in the early Seventies. Governments will continue to be aware of the need for monetary restraint, even though their focus may be on the exchange rate rather than on measures of the money stock as an indication of monetary stance. They are unlikely to fall again into the trap of placing all their anti-inflationary eggs into the incomes policy basket, as so many administrations did in the past.

Despite the partial retreat from monetarism, nobody should conclude that we are heading back to the consensus politics of the Fifties and Sixties in this country. Although inflation is now at tolerable levels, the Thatcher government is clearly determined not to go back to the old routine of recklessly stoking up demand in an attempt to reduce unemployment. To the extent that there has been some monetary ‘overkill’ in the last year or two, the current mild easing in monetary restraint will probably not do much harm and may slightly reduce unemployment. However, even anti-Thatcherite pressure groups such as the allparty Employment Institute envisage only a small fall in unemployment – perhaps 1½ to 2 percentage points – as a result of reflation. They regard micro-economic policies – tax and national insurance reductions, enhanced retraining programmes – as the key to larger-scale reductions in unemployment.

The Government, too, sees supply-side policies as the way back to fuller employment. But measures such as weakening wage councils and employment-protection legislation, restriction of trade-union influence and reduction of social security benefits are the elements which make up their preferred package for the labour market, together with increased privatisation and deregulation in the product market. The retreat from monetarism is not a return to muddled consensus masquerading as sweetness and light. The economic climate of the mid-Eighties is too harsh for that.