MONEY, INFLATION AND THE ECONOMY

 M

ost people think economics is the study of money. But there is a paradox in the role of money in economic policy, which is this: the attention actually paid by central banks to money has declined, whereas in fact, price stability is recognised as the central objective of central banks.

2It is no accident that during the “Great Inflation” of the post-war period, money, as a causal factor for inflation, was ignored by much of the economic establishment. In the late 1970s, the counter-revolution in economics – the idea that in the long run money affected the price level and not the level of output – returned money to the centre stage in economic policy. As Milton Friedman put it, “inflation is always and everywhere a monetary phenomenon”. If inflation was a monetary phenomenon, then controlling the supply of money was the route to low inflation. Monetary aggregates became central to the conduct of monetary policy. But the passage to low inflation proved painful. Nor did the monetary aggregates respond kindly to the attempts by central banks to control them. As the governor of the Bank of Canada at the time, Gerald Bouey, remarked, “we didn’t abandon the monetary aggregates, they abandoned us”.

3So, as central banks became more and more focussed on achieving price stability, less and less attention was paid to movements in money. Indeed, the decline of interest in money appeared to go hand in hand with success in maintaining low and stable inflation. How do we explain the apparent contradiction that the acceptance of the idea that inflation is a monetary phenomenon has been accompanied by the lack of any reference to money in the conduct of monetary policy during its most successful period? That paradox is the subject of my talk.

4Of course, some central banks, especially the Bundesbank and the Swiss National Bank, always paid a good deal of attention to monetary aggregates. But when the European Central Bank acquired responsibility for monetary policy it adopted a reference value for money growth as only one of its two pillars of monetary policy, with an assessment of the outlook for inflation as the other. And the Swiss National Bank recently replaced its target for the monetary aggregates with one for inflation. In the United States, the Federal Reserve, at its own request, has been relieved of the statutory requirement, imposed in 1978, to report twice a year on its target ranges for the growth of money and credit. As Larry Meyer, a Governor of the Federal Reserve Board explained earlier this year, “money plays no explicit role in today’s consensus macro model, and it plays virtually no role in the conduct of monetary policy.”

5The decline in the importance of money in policy formation can be witnessed by the reduction in the number of references to money in the speeches of central bank governors. So much so that over the past two years, Governor Eddie George, of the Bank of England, has made one reference to money in 29 speeches, Chairman Greenspan, of the Federal Reserve, one in 17, Governor Hayami, of the Bank of Japan, one in 11, and Wim Duisenberg, of the European Central bank, three in 30.

Money and inflation: the evidence

6Let me begin by looking at some of the historical evidence. Chart 1, which extends the results of McCandless and Weber (1995), shows the correlation between the growth of the monetary base and of inflation over different time horizons for a large sample of 116 countries. Countries with faster growth rates of money experience higher inflation. It is clear from CHART 1 that the correlation between money growth and inflation is greater the longer is the time horizon over which both are measured. In the short run, the correlation between monetary growth and inflation is much less apparent. Understanding why this is so is at the heart of monetary economics and still poses problems for economists trying to understand the impact of money on the economy. I shall return to this later.

Chart 1

Few empirical regularities in economics are so well documented as the co-movement of money and inflation. Chart 2 shows that this relationship is true for broad money as well as the monetary base. The other side of the coin to this close relationship between money and prices is the absence of a long-run relationship between money and output growth, shown in Chart 3. Over the thirty year horizon 1968-1998, the correlation coefficient between the growth rates of both narrow and broad money, on the one hand, and inflation, on the other, was 0.99. Correspondingly, the correlation between the growth of narrow money and real output growth was – 0.09 and between broad money growth and output growth was – 0.08.

Chart 2

Chart 3

Correlation, of course, is not causation. The essence of monetary theory is trying to understand the structural relationship between money growth, money demand, output and price movements. Stable structural relationships can give rise to unstable short run correlations between any of these variables. It is, therefore, somewhat surprising that some economists have argued that the instability of observed short run correlations casts doubt on the long run importance of money growth in the inflationary process. Chart 4 shows the behaviour of the price level in the UK and its relationship with the ratio of money to real income over the period from 1885 to 1998. Short-run movements in the velocity of money are apparent, as well as the long-run link between money and inflation.

Chart 4

7The view that money does not matter has been encouraged by those who point to regressions of inflation and output on monetary growth, and find that the influence of money is either insignificant or unstable. But these results tell us little about the significance of money in the transmission mechanism of monetary policy. They are based on what economists call reduced-form equations, the coefficients of which will be complex functions of the true structural parameters of the economy, as well as expectations of future policy responses by the monetary authorities. There is no reason to expect a simple relationship between inflation and output and money growth in reduced form estimates.

8This last point was clearly grasped by Friedman and Schwartz in their classic 1963 study of money in the United States. They took great care to identify periods in which there was an exogenous shock to the money supply, such as moves on to and off the gold standard, and changes in reserve requirements imposed on banks. More recent studies, such as Estrella and Mishkin (1997), Hendry (2001), Gerlach and Svensson (2000) and Stock and Watson (1999) produce conflicting and unstable regression results for the influence of money growth on inflation.

9To understand the true role of money, a clear theoretical model is required and that model must allow for the central role of expectations. The key role of expectations is best illustrated by considering extreme cases of high inflation, known as hyperinflations. In hyperinflations the effect of expectations on money and inflation is amplified relative to other influences, such as the business cycle.

10Chart 5 shows the link between money and prices in four hyperinflations. Two of these are drawn from the inter-war period, namely the hyperinflations in Austria and Hungary, and two are post-war hyperinflations, in Argentina and Israel. At their peak, these hyperinflations involved annual inflation rates of 9,244%, 4,300%, 20,266%, and 486% respectively. All four hyperinflations illustrate the importance of expectations. In the case of the two inter-war hyperinflations, large government deficits were monetised, leading to rapid money growth and inflation. The public tried to economise on money holdings, and so real money demand fell. Announcements of credible fiscal stabilisations changed inflation expectations and led extremely quickly to a rapid fall in inflation. Lower inflation encouraged real money demand to rise again, and so nominal money growth continued to rise for some time after inflation had fallen. Inflation was, therefore, stabilised ahead of the slowdown in money growth, although the causation ran from the credible announcement of monetary contraction to lower inflation. The vertical lines in the charts indicate the announcement dates of stabilisation packages. In Argentina, inflation expectations were stabilised by the convertibility plan of 1991 which established a currency board to back the local currency in terms of the US dollar. Inflation expectations fell and, as in the earlier cases, the fall in inflation preceded the slowdown in money growth. The case of Israel is somewhat different in that the absence of any delay between the announcement and the implementation of the stabilisation programme in 1985 meant that the gap between the fall in inflation and the contraction of monetary growth was shorter than in the other cases shown in Chart 5. Although hyperinflations are extreme examples, they do illustrate the fact that, even when monetary contraction is evidently the cause of a fall in inflation, the rapid response of expectations means that inflation may fall before signs of a slowing of monetary growth itself.