Saturday, June 15, 2019

MMT and Inflation


One of the key elements of the MMT argument is the claim that their money finance proposal is not inherently inflationary.  They put this forward in contradiction of the monetarist argument that inflation is a result of increasing the money supply faster than the rate at which GDP is increasing.  (Interestingly, David Andolfatto seems to classify MMT as a version of monetarism.)   As in so many instances, what you take as MMT theory depends on which MMT theorist you’re reading – some seem more open than others to the idea that money-financed spending, shifting the AD curve to the right, might at  some point (i.e. when the AD curve reaches the steeply positively sloped segment of the AS curve) create an inflationary process which would have to be countered by using taxes to pull money out of the economy. 


The pure monetarist inflationary position holds when the AD curve cuts the vertical segment of the AS curve.  In that case, a rightward shift of the AD curve with no shift of the AS curve will definitely be inflationary.  To the extent that the AS curve is shifting to the right, because of an exogenous growth process (driven, perhaps, by past investment spending), part of the tendency for the price level to rise will be countered.  If the AS curve is shifting to the right at the same rate as the AD curve, where the shift of the AD curve is assumed to be driven by an expansion of the money supply, there will be no tendency for the general price level to rise.

The fundamental point of weakness in the conventional argument, which the MMT argument draws on, is how is it that exogenous expansions of the money supply are converted into inflationary pressure.  They make the argument that the massive increase in the US Monetary Base associated with Quantitative Easing did not, contrary to the expectation of many mainstream economists, create an inflation, and that the Japanese government has not succeeded in creating inflation with its own monetary expansion.  Yet standard theory holds that expanding the money supply faster than real output is expanding must cause inflation.

The key to understanding how the QE episode differs from the standard inflation examples lies in something that Richard Cantillon pointed out back in about 1730.  To understand the effects of an increase in the money supply on the real economy, you have to pay attention to the way the money enters the economy.  The Cantillon effect is invoked most often to explain how it is that expansions of the money supply could have an effect on real output in the short run even if its only effect in the long run is on the price level (meaning that any short run effect on real output must at some point be undone).

In the argument which we are considering, the key issue is whether the money enters the economy by being given to someone who intends to spend it on goods and services.

Consider the classic examples of monetarist inflations through history: inflation in China in the days of Marco Polo, the episode of the playing card money in New France, the Weimar hyperinflation of the 1920s and the Great Inflation of the 1970s.  In each case the reason for the printing of large quantities of new money was to allow someone to do spending which would otherwise not be possible.  In most cases this was the government directly.  Printing money allowed the government to spend, and hence take ownership of gods and services,  without previously having used its power of direct taxation to reduce the spending power of the private sector.  The government takes the newly printed money and buys goods with it, leaving less for the private sector to buy, even though the private sector’s holdings of the transactions medium (money) have not ben reduced.  The purpose of the printing of the new money is basically to let the government get in first and buy goods. 

Whether this will cause an inflation, of course, depends on where the AD curve cuts the AS curve.  If the intersection is in the horizontal segment of a Keynesian AS curve, the economy can respond to the increase in government demand for goods by producing more goods, and real output will expand at the same rate as demand shifted to the right with no significant increase in prices.  If the economy is operating at potential inflation will result, as in the case of the Vietnam War episode, which caused the Great Inflation of the 1970s.

The Weimar episode fits this story even though Germany was clearly running below its pre-First World War potential level of output.  The first stages of the inflationary process was the result of the Weimar government printing money to make payments to returning German soldiers, the idea being to avoid social unrest.  Unfortunately, the increase in the medium of exchange (on top of the money which had been printed to pay for the German war effort) was not accompanied by an increase in real, physical output.  The German authorities managed to disguise this for a while, making it look as if Germany had come through the war in much better shape than Britain or France had, which in turn encouraged capital inflows from the US (it was in this period that Maynard Keynes lost a large amount of money on currency speculation).  Thus German railway workers returning from military service were notionally re-employed, and being paid, even though there were no trains for them to run.

The Weimar hyperinflation was a result of a deliberate expansion of the money supply with no intent of increasing output.  The money was printed to make payments to Germans engaging in passive resistance to the occupation of the Ruhr by French and Belgian troops in 1923.  Not surprisingly, this produced an inflation – so much so that prices were rising faster than the money supply was growing, as people spent in anticipation of ongoing price increases.  They spent the money on commodities partly because incomes in Germany generally were by then lagging well behind the rate of increase of the cost of living, but also because there were no financial instruments which could be used to shield incomes from inflation.  Patriotic middle class Germans who had put their savings into government bonds during the war (heeding exhortations to exchange their holdings of gold coins for government bonds) found themselves wiped out by the inflation.  Only real goods had any chance of holding their value since their prices rose (by definition) at the rate of inflation. 

The hyperinflation only came to an end with the issue of Rentenmarks, notionally to replace Reichsmarks.  Technically the Rentenmark was a land backed currency but the real reason it brought the hyperinflation to an almost dead halt was because it provided an opportunity for the increase in the money supply to be halted.  So in its origin and in its end the Weimar inflation was a classic monetarist episode.

The recent QE episode in the US is quite a different creature.  The expansion of the monetary base, even though the Fed was using it to buy US government bonds, was not aimed directly at increasing spending.  The initial idea was to ensure the liquidity of the US financial system.  The Fed was buying bonds at auction, in the manner in which it was accustomed to conduct open market operations, but on a much larger scale than usual.  It was able to buy in massive quantity because it could print as much cash as was needed to persuade holders of government debt to sell.

The reason someone holds government debt is to convert it to cash at some point in the future.  If you hold it to maturity you know exactly how much cash you’ll get for it.  Because the Fed can create the cash, it can offer whatever amount it takes to convince holders of government debt that they’d be better off selling now than holding to maturity (or selling at some point before maturity). 

Basically, then, what the fed was doing was convincing savers to exchange a significant part of their holdings of government debt for more cash than they’d expected to get for it.  But the people who had sold the debt to the fed hadn’t been planning to spend it now – they were using it as a saving instrument.  What the fed had done was given them the capacity to buy considerably more financial instruments.  After all, government debt was not the only form of debt available at various maturities.  It’s not surprising if people who had no particular plans to spend their wealth today reacted to having more cash than they had expected by replacing the maturities they’d sold, in the process helping to push interest rates down.

In addition, the Fed started paying interest on excess reserves of commercial banks.  That encouraged the banks (which also increased their holdings of highly liquid government debt) to treat deposits at the Fed as a safe asset.  If they expanded their loans to the private sector, those reserves would cease to be excess reserves and would become required reserves, so the banks would have exchanged a safe interest earning asset for a risky (private sector) one. (Interest on excess reserves also ties in to the argument about whether a group in the US should be permitted to establish a Narrow Bank.  John Cochrane discusses the arguments here.) 

In the QE episode, then, people who sold their holdings of government bonds to the Fed were not doing so with the intent of spending the money on goods, while the banks, in whose accounts with the Fed toe proceeds of those sales would up, were initially very risk averse about expanding their loans – their own liquidity preference had increased – and were then encouraged to keep those funds with the Fed to earn interest payments on excess reserves.  Any stimulus to spending was supposed to come through low interest rates encouraging spending on physical capital, but the inward shift of the Investment function more than outweighed any effect associated with a movement along the in-shifted function.  The MMTers argue, and its hard to dispute them on this point, that monetary policy operating through the interest rate channel to physical investment is not a very effective anti-recessionary policy.

Overall, then, so long as we’re prepared to work in a Keynesian AD/AS context, we can make sense of elements of the MMT argument but also make a case that their use of the QE episode to dispute the monetarist inflation argument does not hold up.  The circumstances are different, but Richard Cantillon’s argument that you have to pay attention to the way new money enters the economy still holds up.

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.