I used to think that the most important objective for monetary policy was to stabilize the price level, and that it mattered less which particular price level was stabilized than that some price level be stabilized. I thought that really bad things happen when there is inflation or deflation, but if the price level — any reasonably broad price level — could be stabilized, whatever fluctuations occurred would be of a second order of magnitude compared to the high inflation or substantial deflation liable to occur without an explicit commitment to price-level stabilization. I also thought, having learned Friedman’s lesson of the natural rate of unemployment a little too well, that aiming for price-level stability would made it unnecessary to worry about preventing unemployment, because high and long-lasting unemployment could not occur without falling prices. That seemed to imply that if you could just ensure that monetary policy would keep prices broadly stable, unemployment would take care of itself. In other words, deliberately trying to reduce unemployment would only get you a temporary reduction in unemployment at the cost of a permanent increase in inflation; a bad bargain, or so, at any rate, it seemed to me.
That was one of the main messages of my book Free Banking and Monetary Reform in which I advocated stabilizing the expected wage level (via a mechanism invented by Earl Thompson). Although I pointed out that stabilizing an output price index could have undesirable effects in case of a supply shock that raised input prices, in retrospect I don’t think that I took that contingency as seriously as I should have. If you try to keep the level of prices constant in the face of such a supply shock, you will succeed only if you can force down nominal wages or the return on investment. Either one is a recipe for a major recession. If you allow output prices to rise to reflect the increased cost of inputs, real wages will fall without a reduction in nominal wages, avoiding the costly adjustment (i.e., reduced output and employment) associated with trying to effect a reduction in nominal wages.
But the problem goes even deeper than that. Suppose you have a central bank that is credibly committed to stabilizing the price level or to stabilizing the rate of inflation at some target level, say, just to pick a number at random, about 2% a year or slightly below that. Then suppose that there is a supply shock, so that the central bank has basically two choices.
The first would be for the central bank to allow the supply shock to work its way through the system, enabling producers to pass through their increased input costs to consumers by providing enough monetary expansion to allow increased input costs to be added to output prices without forcing any other inputs to absorb a nominal reduction in their nominal incomes. Thus to avoid a recession, you would probably need a slightly higher rate of NGDP growth than the rate corresponding to to the one that would have met the inflation target had there been no supply shock. In other words, I am suggesting, though I could be wrong about this, and I invite others to weigh in on this point, that accommodating the supply shock requires a slight loosening of monetary policy relative to what it was before the shock. But if the central bank accommodates the supply shock, it will overshoot its inflation target, undermining its precious inflation-fighting credibility.
The second option of course is to resist the supply shock in order to maintain the precious inflation-fighting credibility of the central bank. But this requires the central bank to tighten its policy, because unless policy is tightened some part of the unexpected increase in input prices will get passed forward into the price of output, forcing the realized rate of inflation to rise above the target rate. The tightening of policy therefore necessarily results in reduced nominal incomes to other inputs (i.e., labor and capital) causing a decline in real output and employment.
At least in broad outline (though I (or we) perhaps have to do some more work on the details), there is nothing really new in this discussion. But I think that there is something else going on here that is not so well understood. The part that is not so well understood is that if the public understands that the central bank cares more about its precious inflation-fighting credibility than about causing a recession, the public will anticipate that the central bank will tighten monetary policy, which means that the public will immediately increase its precautionary demand for money, which means a spontaneous demand-induced tightening of monetary policy even before the central bank lifts a finger.
I have no doubt that something like this was going on in the spring and summer of 2008 when the FOMC kept making periodic and downright scarry statements about how increases in headline inflation caused by rising commodity prices (Oh, Lord, protect us from those rising commodity prices!) were threatening to cause inflation expectations from becoming unanchored even as the economy was rapidly going down the tubes long before the Lehman debacle (and don’t forget that it took the FOMC three whole weeks after Lehman collapsed — during which time there was an already scheduled FOMC meeting at which the status quo was reaffirmed! — to reduce the federal funds rate to 1.5% from the 2% rate at which it had been held since March).
And I also believe that something like this has been going on over the past few weeks as inflation and money-printing hysteria has increased, fueled, among other things, by an unexpected 0.5% increase in the July CPI. I am suggesting that the 0.5% increase in the CPI caused the markets to attach an increased probability to a tightening of monetary policy, causing an increased precautionary demand for money.
In chapter 10 of my book (pp. 218-21) in the section “the lessons of the Monetarist experiment 1979-82,” I described the perverse expectational reactions triggered by the Fed’s attempt to follow the Monetarist prescription for targeting the growth rates of the monetary aggregates. I introduced that section with the following prescient quotation from Hayek’s Denationalization of Money.
As regards Professor Friedman’s proposal of a legal limit on the rate at which a monopolistic issuer of money was to be allowed to increase the quantity in circulation, I can only say that I would not like to see what would happen if under such a provision it ever became known that the amount of cash in circulation was approaching the upper limit and therefore a need for increased liquidity could not be met
The perverse response under inflation targeting when there is a supply shock is, I think, more or less analogous to the one so clearly foreseen by Hayek, which I documented in my book for the 1979-82 period (with intermittent recurrences in 1983-84 as well). Who says history never repeats itself?
Uneasy Money 
David. They are perverse all right. And coupled with interest rate targeting the “cocktail” is explosive:
“The September (08) meeting was just 48 hours after the Lehman collapse and a few hours after the release of the industrial production figure for August, which showed a 1% drop MoM.
By not “looking”, let alone targeting AD, FOMC members confused a negative supply shock from rising oil and commodity prices, which would not require “tightening” measures if the goal were to maintain AD on its target growth path, with an increase in AD! And the FOMC decisions and statements were taken in the context of an extremely hostile environment for nominal spending, an inexcusable mistake that proved very costly.
http://thefaintofheart.wordpress.com/2011/05/01/insisting-on-targeting-the-fed-funds-rate-can-be-bad-for-the-economy%C2%B4s-health/
LikeLike
David:
I just deleted my previous comment.
Do you mean that a supply shock would require an increase in NGDP growth rate relative to its past trend, or do you mean that the NGDP growth rate would have to be higher than what would be necessary to prevent the inflation?
I think it is obvious that a price level target requires a reduction in the growth path of NGDP in the face of a negative supply shock.
If we have an fully unanticipated surprise supply shock, and we have inflation targeting, so that the increase in the price level is ignored and the price level just goes up from whever it happens to be, then I don’t think NGDP growth has to change.
On the other hand, when we anticipate an adverse supply shock, then the Fed would need to restrict NGDP growth to prevent it from causing an increase in the inflation rate. (And if we believe what members of the FOMC say, they seem to take what look to be current supply shocks, project them into the future, and take action to dampen them. Oil prices are heading up, that will cause more inflation, we have to prevent inflation expectations from being unachored…)
In my view, NGDP targeting, which I support, definitely implies that adverse supply shocks will cause a higher growth path for prices, and so inflation. The particular rule I favor has no trend inflation, so the result is a higher price level, and inflation to get there.
Now, does this cause “tight money?”
Imagine a pure price shock. Firms raise prices and would like to sell the exact same output. This reduces the real quantity of money. Given the demand for money, there is an excess demand for money. Real expenditures fall. Firms sell less and produce less and lower their prices from the intial level. The result is that the price level rises and real output falls in inverse proportion. This story includes an excess demand for money. A central bank could avoid it by accomodating the increase in the price level. The price level would be unanchored.
Now, suppose that firms keep their price the same, but for some reason, they can’t produce as much. The reduction in real output is a reduction in real income. The demand for real money balances falls. Given the quantity of money, there is an excess supply of money. The excess money is spent, so real expenditures rise. Firms raise their prices and perhaps produce more at the higher prices–more than the initially lower level of output. In the end, real output would be lower and the price level would be higher, in invers proportion. This story involves an excess supply of money. The central bank could avoid it by decreases in the quantity of money, leaving the price level unchanged.
Now, in the real world, supply shocks are specific. And they generally result in higher prices and lower production for particular goods. By a matter of arithmetic, the price level is higher and aggregate real output is lower.
Bad harvest. The supply of wheat is lower. The price of wheat is higher, and the quantity of wheat is lower. If nothing else in the entire economy changed, the price level is higher and aggregate output is lower.
If the demand for wheat were unit elastic, then NGDP would be unchanged. Expenditure on wheat is unchanged, and expenditures on everything else could be unchanged. Obviously, there are substitutes and complements in production and consumption, but this is the first pass.
If the demand for wheat is inelastic, the the price of wheat rises more than in proportion to the decrease in the quantity of wheat. Expenditures on wheat rise. Expenditures in the rest of the economy must fall if NGDP is to remain stable. We have something like the excess demand for money described in the first scenario, with the generalized price hike.
If the demand for wheat is elastic, then expenditures on wheat fall. Expenditures in the rest of the economy must rise to keep NGDP on target. We have something like the excess supply of money in the second scenario, where output fell and firms kept their price the same.
My understanding of the orthodox criticism of NGDP targeting is that the proportional trade off between the price level and real output in the face of supply shocks may not reflect the optimum trade off in the social welfare function.
My own view is that NGDP works perfectly in the face of supply shocks only if the particular good with the shock has unit elastic demand. Otherwise, the rule causes collateral damange.
On the other hand, the “collateral damange” is plausibly related to the need for resource allocation. If the demand for wheat is inelastic, we really do need to more more resources to wheat production. And if the demand is elastic, we need to shift resources from wheat to the rest of the economy.
I support it, because it is the least bad option. It creates the least bad environment for macroeconomic adjustment.
LikeLike
David,
I agree that taking the Fed’s objectives and the way that it thinks about the economy as a given, then within those parameters the Fed made a mistake in not easing policy more aggressively after July 2008.
But had they at that point immediately cut rates to zero would it have done much good?
And one can argue (with Marc Faber) that a premature easing in late 2007 actually triggered a blow-off final move in commodities and the dollar that could have been avoided had the Fed kept a steadier hand.
My personal view is that a framework whereby one accounts for the entire extent of the bust as being due to a failure in monetary policy _once the bust has started_ is lacking in perspective.
I say this having sat down in late 2004 with the then chief US economist for HSBC and put forth the case that we would have ultimately first a near-depressionary episode, and then a reflationary attempt likely ending eventually (after some years) in roaring inflation. Many others diagnosed the situation and identified the magnitude of the mess that was coming.
So are you really so sure that by the time the credit bubble had been inflated and had begun to burst that it would have been possible to avoid the pain by more aggressive and faster monetary easing?
LikeLike
There is a similar distortion in the opposite direction: say there is 7% growth in output, due to a dramatic oil discovery/nanotech/return to trend after a dip/etc., in an economy with a 5% NGDP trend of growth that tends to split 3/2 in favour of output. To keep inflation at 2%, the central bank must inflate NGDP up to 9%. In doing so, it will boost the money supply dramatically, distort the financial system and sow the seeds of an inflationary period when the rate of output growth falls.
“But obviously no inflation-targeting central bank would be so dogmatic!”
If inflation-targeting only works if it’s ignored when inconvenient, then it’s obviously not a very good system. In practice, inflation targeting banks almost never target their headline inflation; they target some derived index, NGDP, output gaps etc.
The main consequence of inflation-targeting in the UK has been to reduce the mobility of action of the Bank of England in pursuing optimal monetary conditions.
LikeLike
I find the notion of a supply shock as the underlying cause of the Great Recession puzzling. What is this shock supposed to be? The increase in oil prices?
More fundamentally, in the world you are describing, prices accelerated and the Fed raised rates, limiting the increase in NGDP. In that In the world I’m observing, prices decelerated, the fed lowered rates, and NGDP still fell. In a world where the Fed has been following an unprecedentedly accomodative policy for years now, I don’t think you can blame (fear of) rate increases for depressed output. Not everything is the Fed’s fault; the reality, as Keynes wrote 75 years ago, is that in a capitalist economy aggregate demand (especially investment demand) is liable to larger fluctuations than can be offset by any feasible change in the supply of liquidity.
You have a really solid argument here … against Paul Volcker in 1979-82.
LikeLike
Fascinating blog.
You know, I wonder if a few bad episodes–Wiemar Republic, Zimbabwe, Argentina, perhaps the USA of the 1970s–has not made most central bankers and policy wonks far too fearful of inflation.
After all, how many nations prosper through moderate and varying inflation, ala China, S. Korea, and the USA of the 1980s and 1990s?
And then we have Japan, no inflation and little prosperity.
According to the Hong Kong Monetary Authority, the “revealed preference” for mainland China’s central bankers is toward growth, rather than inflation fighting. We may assume Japan has taken the opposite tack.
What often annoys me is that there is some unspoken premise that static prices are a great moral virtue (set aside that even measuring prices is controversial, see the CPI).
Asceticism is a common human trait. Religious groups sometime practice self-flagellation. Suffering for a higher moral cause is captivating at times.
And the money fetishists have agreed that monetary asceticism is divine–that we should suffer to protect the purchasing power of the dollar, even for decades, even forever.
Baloney. Really, varying inflation in the three to six percent range probably is excellent for economic growth. Certainly the US track record of 1980-2000 suggests that.
I suggest that fighting inflation makes less sense in an open economy such as the USA, where labor, capital, services and goods easily cross borders. If something rises in price in the USA, it is imported more. Printing more money probably leads to long-term growth, and then maybe some inflation, although ever there is global competition holding down prices. And productivity.
The US private labor force has just about de-unionized.
Please Bernanke, print a lot more money.
LikeLike
Time ago I have discussed that with Marcus & Scott. Both deny that the financial shock was the cause of the crisis. It is unacceptable for me, in spite of agreeing with that Bernanke was too late to react in 2008.
But, for me, He was too late because of the financial crunch.
Only an monetary mistake and a supply shock? I can´t think that monetarism would be so narrow.
To deny the financial shock (that explain the slowness to go out of the recession) is not see that the fall in asset prices is much more destructive and violent that the deflation in goods markets.
To disregard the financial problem as the main root of monetary misadjustment, hummm! Ther is not money and real markets only: ther is a big black hole that some times explodes.
LikeLike
David, isn’t all this really about the same people talking to the same people about how we will save the world? I just did a presentation in front of 50 economists in the Danish central bank and I am pretty that they had no clue about what the hell I was talking about…
LikeLike
Marcus, Thanks for the link. I will read it with great interest. You have done some really important work and we are all indebted to you for it.
Bill, I think that what I mean is that to prevent the prices of inputs that are complementary to the one experiencing the supply shock from falling, you would need to increase NGDP. For simplicity, let’s do the analysis in terms of fixed levels of NGDP or or price or wages. When we have solved for the zero rate of growth case, then it should be straightforward to extrapolate to any particular growth rate. You go through a number of different scenarios, but I don’t see a general solution to the question yet. So, we need to keep thinking about it. Thanks for your putting your mind to it.
Cantillon, I agree that it would not have been appropriate to cut rates to zero in July 2008, but a cut to 1.5% percent might have been. Have you read the FOMC statements? When I said they were scary, I was not kidding, and I think that the markets were scared. You raise some very good questions which underscore the difficulty of the situation in which monetary authorities found themselves. Because conditions were so dangerous, it may have been that any course of action would have led to a bad outcome. Nevertheless, it is hard to imagine an outcome any worse than the one we got.
I think that I will have to respond to the remaining comments tomorrow.
LikeLike
W. Peden, You are correct to point out the symmetrical nature of the problem. I agree that we don’t want to make it too easy for a central bank to ignore its target, but isn’t an inflation rate unnecessarily restrictive. That seems to me to be an important argument in favor of price level targeting as opposed to inflation targeting.
JW, I didn’t mean to suggest that a supply shock was “the underlying cause” of the Great Recession. We were already in a recession before the supply shock. But the interaction of the supply shock with the recession produced a perverse tightening of monetary policy in the midst of a recession which helped cause things to spin out of control. Something similar happened in 1973-74 when monetary policy was tightened just as a supply shock was occurring. I don’t think that the world that you are observing corresponds to what was happening in the spring and summer and early fall of 2008. The economy was deteriorating and the FOMC was explicitly refusing to cut rates further, indeed was signalling a desire to do the opposite, because of rising commodity prices and increasing headline inflation. Go read the FOMC minutes.
Benjamin, I agree we need to rethink views about inflation. The criterion that I am toying with is that low inflation or even deflation is OK as long as interest rates don’t fall below 2%. However once you get interest rates below 2%, that may be a signal that higher inflation is needed. That’s one reason I think it has been so misguided for the Fed to try to sell QE by promising that it would reduce interest rates when what we really want is higher interest rates. And the way to get higher interest rates is to increase inflation expectations.
Luis, I am not sure if I get your point. But I don’t discount (I can’t speak for Marcus and Scott) the financial aspect of the crisis. In the current environment with a very steep yield curve, it is really very difficult for financial intermediaries to function because they typically borrow long and lend short. That is one more reason why it is important to increase inflation expectations and get the yield curve to flatten out.
Lars, You are right, but what else can we do?
LikeLike
David, I have never denied that the financial problem requires an expansive action the central bank. On the contrary, I think that heightens the need.
What I say is that I do not think that following a rule NGDP type or PL, is sufficient when the banks collapse, as happened in 2008.
I think for normal times these rules do apply, but not for the 2008 crisis as
I also think that these crises leave an imprint on the mindset that takes time to disappear. But that is not denying the need for central bank actions. Simply i´m saying that entrepreneurial spirit need long to recover.
Is that keynesian afirmation? I don´t know, but I don´t believe in fiscal policy.
LikeLike
David G-
How can deflation be good when an economy is over-leveraged? I prefer the “Triple Nickel”: Five years of five percent inflation and five percent real growth.
LikeLike
Luis, Don’t you think that inflation would speed up the repair of balance sheets and the recovery of entrepreneurial spirit?
Benjamin, You misunderstood. I wasn’t supporting deflation now. My point is that an economy that is growing rapidly with a high real interest rate can absorb some level of deflation as long as the real interest rate doesn’t go down to much, say, below 2 percent. Under current conditions even a low rate of inflation (i.e., 1-2 percent, isn’t nearly enough). So I would front load the inflation, but not necessarily commit to keeping it going for 5 years.
LikeLike
Lars, is it possible to see the notes/slides to your presentation somewhere? I am a Danish grad student who has been following the quasi-monetarists, (David, Scott Sumner et. al.) and it would be interesting to see how the argument would be presented in a Danish context.
LikeLike