Axel Leijonhufvud: The Perils of Inflation Targeting

With a Fed meeting and rate decision today, this seems topical. Axel Leijonhufvud argues that inflation targeting can cause overly expansionary monetary policy when there are conditions - in this case foreign central banks absorbing liquidity to prevent exchange rate movements - that suppress the true inflation rate:

The perils of inflation targeting, by Axel Leijonhufvud, VoxEU: To control the price level, Patinkin demonstrated many years ago, you need control of one interest rate and one nominal asset for which the private sector cannot produce a close substitute. Although the theory did not say so, in practice it was obvious that this nominal stock had better not be very small. Just the copper coinage would not do, for instance. Thus Monetarism relied on 1) control of the base, 2) the stability of the base-multiplier which was ensured by reserve requirements on banks, and 3) the only very slowly changing habits of the public with respect to the use of paper currency.

Monetary policy, in this context, was thought of as operating on the determinants of the equilibrium price level to which the actual level would adjust, albeit with “long and variable lags.”

All that is long gone, of course. Reserve requirements are no longer enforced, the private sector is busy producing ever more substitutes for the use of currency -- and the base is now demand-determined. We are now in a Wicksellian world of pure inside money in which no determinate equilibrium for the price level exists. Monetary policy then becomes the art of using the federal funds rate to control the rate of change of the price level. Ideally, the Central Bank should hit Wicksell’s “natural rate” on the nose so that the price level would stay constant. But it does not know what that rate is. It has to behave adaptively, therefore, watching the inflation rate and countering any movements in it by moving the interest rate in the opposite direction.

This is a High Wire Act! The long and variable lags are presumably still with us. The feedback, when it arrives, is not always unambiguous. What prices belong in “core inflation” if the CPI does not provide the best signal?

If this is so hard in theory, why does it seem to be so easy in practice? Interest-targeting is widely regarded as an almost unqualified success. But consider American monetary policy. Does the absence of inflation since the turn of the century show that the Fed has smartly kept the interest rate in the near neighbourhood of the “natural rate”? Obviously not. More than a dozen quarter-point hikes of the short rate to which the markets paid basically no attention (the long rate not reacting) mean instead that the Bank had engaged in such an extraordinarily expansionary policy that it had lost all contact with the markets.

If the Fed was flooding the world with dollar-denominated liquidity, why was there no significant inflation? Part of the answer is that in the early-going, the Fed was fighting deflation in the wake of the ITC bust. But the more significant part of the answer lies in the determination of other central banks to prevent their currencies from appreciating against the dollar. With their exchange rates more or less fixed, the elasticity of their exports has kept American inflation in check. This short-circuits the feedback loop on which an adaptive inflation-targeting policy relies. The behaviour of the price level provided no clue to the Fed that its policy was far too expansionary.

If you run a very expansionary monetary policy and the historical conjuncture happens to be such that you get no inflation, what do you get? The answer, of course, is asset price inflation and deterioration of credit. This is the troubling legacy of policy that we are now left with.

How dangerous is it? Judgements vary and they do so because no unconditional answer is possible. What dangers will actually materialise depends on how the current financial imbalances will eventually unravel, on where inflationary pressures will first become serious, and on how the policy-makers of the major countries will respond to unfolding events. Some plausible scenarios are more reassuring than others.

The sanguine view is that securitisation and credit derivatives have made the world of finance a safer place than it used to be and that, besides, liquidity is ample all around. But it is not likely that the world will stay awash in liquidity forever. At some stage, central banks will have to mop it up or see inflation do it for them. Securitisation and credit derivatives have certainly dispersed risk through the economy and away from the banks where it used to be concentrated. But by the same token, the system has taken on more risk and we know less about where large concentrations of risk-bearing may be located. Risk spreads have narrowed in part permanently because of these new risk-sharing technologies, but in part transitorily because of the extraordinary level of liquidity. Narrow spreads have in turn induced some institutions to assume high leverage in search of yield.

A number of very large failures – LTCM, Enron, Amaranth – have occurred causing nary a macroeconomic ripple, and this is frequently cited as proof of the resilience that recent financial innovations have imparted to the system. It may be, however, that the more appropriate conclusion to draw is that macroeconomic developments are more likely to trigger trouble in financial markets than vice versa.

Comments

[anne]

Mark Thoma:

"Axel Leijonhufvud argues that inflation targeting can cause overly expansionary monetary policy when there are conditions - in this case foreign central banks absorbing liquidity to prevent exchange rate movements - that suppress the true inflation rate."

The problem here is that the true inflation rate is either the inflation rate measured by consumer price indexes, or the inflation rate measured by just what the Federal Reserve long ago abandoned, some measures of money supply. This is a sort of imaginary economics, in which the Fed imagines what inflation really is as opposed to using direct measures of inflation.

The quoted passage perfectly describes the self-contradiction of this column.

[Meh]

anne: but all the estimate of M3 do rather suggest there is inflation out there, which is not officially measured, because M3 is no longer an official measure...

[kharris]

Meh,

Depends on what the definition of "inflation" is. Fed policy makers seem to be focused over the long term on consumer prices, rather than any other prices. M3 growth can (and often does, apparently) show up most strongly in asset and commodity prices, but less reliably in consumer prices. If our present circumstances tend toward a slower pace of consumer price increase at any given pace of asset and commodity price increase, then we risk misleading ourselves by focusing on asset and commodity prices or on M3. So M3 "suggesting" inflation depends on one's judgement about whether M3 is any longer a reliable "suggester".

This is the very debate that goes on in the financial press and over drinks and in central bank meetings over and over. Traditionalists insist that M3 => inflation, or that asset price inflation => consumer price inflation. Those looking at recent evidence are not convinced. Is the difference between recent behavior and historic behavior due to mismeasurement? Is it a temporary? Has there been a structural shift? Can't tell.

The Fed insists that historic relations between M3 and everything else have broken down, and so has stopped publishing the series. ECB officials say otherwise. I am not wise enough to judge who is right.

[anne]

Alan Greenspan had become convinced between 1985 and 1990, the money supply measures do not reflect or project inflation characteristics. Short term money flows can be reflected in short term asset price changes, but little else about money supply change appears to effect inflation or finally even expected inflation.

KHarris comments well, as always, and I will think through the issue again, but aside from noting short term money flows from unusual changes in monetary policy, I would watch the bond market.

I am fairly convinced, for instance, as I was at the time, that the short term money supply increase in the transition to 2000, fed or allowed for the capping of an unfortunate rise in the prices of information technology stocks, but this sort of influence is an unusual and generally reversable occurence.

[robertdfeinman]

It seems that whether inflation should be tackled and how to best do this are topics of never ending debate. To me this indicates that all approaches are problematical. If there was a clear winner then there would be no debates.

It's like in medicine. When a new blockbuster drug really comes along it gets adopted in short order. With those of dubious efficacy claims get made (and refuted) for long periods of time.

What would world governments do if they had to acknowledge that they don't really know how to deal with the inflation rate in a surefire way? It would be like having a disease with only doubtful treatment options. Do you tell the patient to try this or that anyway just to give them some hope or do you just try to make them comfortable and admit the lack of knowledge?

I think the strong convictions by those promoting monetary policies are the biggest impediment to developing better techniques. Each side tends to overstate the effects of its preferred policies and ignore the contradictory evidence.

As I've said previously, it is not even clear to me whether (modest) inflation is worth discussing. The results of inflation seems only to affect the rentier class which has strong influence in the halls of congress. Why was Greenspan meeting frequently with Cheney and others in the white house. Isn't the Fed supposed to be independent?

[Outside the Box]

The Fed is furiously expanding the money supply to drive up the price of domestically produced items fast enough to compensate for falling import prices. Some of the extra liquidity will find its way into asset prices, and low credit quality loans. It follows that the more foreign central banks act to lower the price of foreign products, the faster the Fed has to increase domestic prices to compensate.

This policy results in the price of domestically produced items increasing in price faster than the CPI. Since real GDP is calculated using the CPI, it will overstate real GDP growth. Many Americans are switching from making items that can be sold on the international market to costly labor intensive services that cannot be legally imported.

If foreign central banks change policy, some Americans will have to switch back to producing some of the items now imported with borrowed money.