ESSA

ESSA

Economics in the pub: money supply and inflation


Yannis Goutzamanis

By

September 13th, 2014


Yannis Goutzamanis recounts a passionate pub debate about money supply and inflation. He then examines the empirical basis for the popular quantity theory of money.


It was the summer of 2012 and I was at the pub enjoying myself with friends when I encountered an interesting young gentleman sporting a black t-shirt stamped with the slogan “911 was an Inside Job”. He was an acolyte of former US congressman Ron Paul and an avid listener to Alex Jones’ radio show. When it became clear that I had no time for his bizarre views on the September 11 terrorist attacks, our conversation turned to monetary policy.

On monetary policy, this man’s views were not quite so peculiar, but just as incorrect as his views of September 11. He was highly critical of the US Federal Reserve’s policy of quantitative easing, which he believed would lead to massive inflation. As we now know this did not occur (perhaps one of the reasons for the collapse in the gold price). Whether he knew it or not this man was invoking the quantity theory of money (QTOM) which posits a positive relationship between the growth in money supply and inflation.

What is QTOM?

The QTOM was popularised by Nobel laureate Milton Friedman during the 1950s and has featured in countless macroeconomics textbooks ever since. A crude version of the economics behind the theory is as follows:

MV ≡ PY

Where M = Money Supply, V = Velocity of money, P = Price level and Y = GDP.

If we assume that V is stable (or only changes gradually over time) and that Y tends towards full capacity/natural level of output then it must be changes in M that cause changes in P. This was the basic logic behind Friedman’s argument. Friedman reworked the above in growth terms to conclude that money supply growth would lead to inflation.

What is the evidence?

Friedman eloquently made the empirical case for the QTOM by reference to the Japanese inflation of the 1970s. He pointed to a clear correlation between inflation and monetary growth with a 12-18 month lag as he predicted in his theoretical work. However, as the adage goes, correlation does not imply causation and as any student of history will tell you the 1973 Oil Crisis led to widespread inflation across the globe. With this in mind, Friedman’s empirical case for the QTOM appears weaker.

In 1985, when the Reserve Bank of Australia (RBA) formally abandoned money supply targeting, it concluded that money supply does not lead to inflation and moreover that there existed no systematic relationship between money supply and inflation.

Study after study found that the other macroeconomic variables such as wages dominate money supply in terms of explaining inflation. A recent RBA discussion paper on modelling inflation revisited the topic and concluded, “the inclusion of money in our inflation models does not impart significant benefit” precisely because of the lack of hard empirical evidence for a causal relationship.

Why does QTOM fail in reality?

One view is that the empirical failure of the QTOM is a fulfilment of Goodhart’s law. That is, when the central bank attempts to target money supply to control inflation, investors try to anticipate changes in money supply and invest so as to benefit from it. This destroys any existing relationship between the variables.

Proponents of the QTOM might simply assert that the lag between money growth and inflation is longer than Friedman predicted. For example, former Bank of England Chairman Mervyn King contends that while the 2-5 year correlation between money growth and inflation is not impressive, over periods of 5-10 years, “the correlations become almost perfect”. For this reason, quantity theorists often castigate the Fed for being myopic and using short-term models.

Whilst the jury is not out on this view, it can seem underhanded at times. It is as if people wish to cling to the QTOM and thus postulate a longer lead-time. One of my economics lecturers once joked that if the USA experiences a burst of inflation in 30 years time quantity theorists will blame it on the Federal Reserve’s current monetary policy. The other sneaky tactic quantity theorists employ is talking in terms of correlations (scatter plots, regression analysis, etc) and ensuring that the 1970s features prominently in their datasets.

Conclusion

As the RBA carefully notes, one should not dismiss the role of money in monetary policy. Whilst it is not a strong explanatory variable of inflation it is nevertheless an important “information variable” on the state of the economy. However, I am left wondering why the QTOM is still featured in so many macroeconomics texts since central banks around the world dismissed it decades ago? The problem arises when students only take an introductory economics unit and are left believing that the QTOM is gospel truth. Whilst there is a legitimate need not to overcomplicate things in introductory economics courses, textbooks should go beyond the neat theory and at least acknowledge some of the empirical work on the QTOM.

The views expressed within this article are those of the author and do not represent the views of the ESSA Committee or the Society's sponsors. Use of any content from this article should clearly attribute the work to the author and not to ESSA or its sponsors.

  • Robert Greco

    Always an interesting read Yannis. To play devil’s advocate, I have four points to make:
    Firstly, in contrast to the RBA abandoning money supply targeting in 1985, is this view strongly held by other reserve banks around the world? For example, in a speech in 2002, Ben Bernanke conceded that the fed caused the Great Depression and then went on to approve Milton’s work and say “thanks to [Milton Friedman], we won’t do it again”. Does this imply that, at least when Bernanke was at the helm, the US fed targeted money supply? (Unrelated, but there is a strong case to argue that poor fiscal policy was at least in part responsible for the GFC)
    Secondly, what should people make of Milton’s work on the Great Depression? It has appeared to have been approved by the fed as explaining the great depression. Does this mean money supply targeting can theoretically work but it’s too hard to do so?
    Thirdly, while the RBA has rejected the view that it targets money supply, but does it do so through another guise, namely through that of OMO. Although this is done primarily to keep inflation in check, isn’t that what money supply targeting sets out to achieve anyway?
    Fourthly, somewhat facetiously, Bitcoin, if it takes off (I believe it will for various reasons), has been mooted as a libertarian utopia (Milton himself envisaged something very similar in the 90s). Until the supply of Bitcoin ends sometime in the next century (2140 seems to be the consensus), the system would’ve have generated a fixed amount of currency before another ecurrency presumably takes over. Could this be the type of thing that accurately identifies the perfect amount of money supply? Again, this is a rather facetious point but I have heard opinions been thrown around to this effect.

    As always, would love to hear your thoughts.

    • Robert Greco

      poor monetary* policy was at least in part responsible for the GFC

    • Yannis Goutzamanis

      Hey Rob, you raise some interesting points.

      With regard to your first point, most major central banks
      around the world, including the US Federal Reserve, have ditched money supply
      targeting in favour of interest rate targeting. This is because they have concluded
      that targeting the shortest term interest rate in the economy as their
      intermediate target is a better way of achieving their ultimate target
      inflation and GDP growth rates. With regard to Ben Bernanke’s personal view I
      am unsure. In the macroeconomics textbook that he co-authors he seems to have
      time for the QTOM. I would guess that he certainly thinks money supply is an
      important economic variable that can have an effect of GDP (given he is an
      advocate of quantitative easing) but I am unsure what his precise view on the
      relationship between money supply and inflation is.

      On Friedman’s work on the Great Depression and Ben
      Bernanke’s acceptance of it I would agree. The Fed certainly played a role in
      causing and/or exacerbating the Great Depression via contractionary monetary
      policy at a time when the economy was weakening. However, this again goes to
      the relationship between money supply and real GDP/output not money supply and
      inflation. Also, as you note, there is a case to be made that money supply
      targeting can work theoretically but is too difficult in reality. One of the
      major reasons the RBA cited for abandoning money supply targeting, along with
      the econometric work on the QTOM, was that it is impossible to control money
      supply in a deregulated financial system.

      On your third point, the RBA uses OMO to control the cash
      rate not the level of money in the economy. The RBA will buy or sell securities
      each day to maintain the cash rate at its target. If you picture this on a
      MD/MS diagram it would effectively mean that the MS becomes horizontal instead
      of vertical. This is because it doesn’t matter what the MS is as the RBA will
      use OMO to maintain their target interest rate. So yes, this achieves the same
      aim as that which money supply targeting seeks to achieve, namely controlling
      inflation. However, it does so through controlling the cash rate as opposed to
      controlling money supply as the QTOM would have us do.

      On your BitCoin point that is very interesting. I had not
      considered the implications of BitCoin. It could actually be a rare chance for
      an economic experiment on money supply and inflation. Presumably, if the QTOM
      holds, then once the amount of BitCoins is fixed, the prices of products that
      can only be purchased in BitCoins should not rise to erode the purchasing power
      of the BitCoin. Although there might be some confounding variables that would
      need controlling.

Founding sponsors

 

 

Partner

Gold sponsors

 

Silver sponsors

 

 

 

 


Affiliates