Central banks cannot ”do nothing”
Some commentators have suggested that central banks should ”do nothing” in the present crisis, but even though that on the surface sounds appealing it is in fact nonsense to say a central bank should do nothing. Central banks in fact cannot “do nothing”. Let me explain why.
The first thing to ask is what “doing nothing” means. Often people talk about monetary policy as manipulating interest rates up and down and doing nothing is taken to mean that the central bank should keep interest rates “unchanged”. However, what we really are talking about is that the central bank is intervening in the money markets to keep the price of overnight credit fixed at a given level. So imagine the demand for overnight liquidity spikes for some reason then the central bank will have to increase liquidity to keep the market interest rate from rising. Hence, even a central bank that is “doing nothing” in the sense of keeping interest rates fixed might end up doing quite a bit. Central bank credibility might reduce the need for actual intervention to keep the interest rate fixed, but that does not change the principle that ultimately the central bank will have to actively manage things.
The story is the same for a central bank that has announce a fixed exchange rate policy. Here “doing nothing” is normally taken to mean that the central bank buys and sell the currency to ensure that the exchange rate indeed remains fixed. So again “doing nothing” might involve doing quite a bit – even though again credibility might indeed reduce the need to doing something on a daily basis, but even the most credibility fixed exchange rate regimes like the Denmark’s peg to the euro or Hong Kong’s peg to the dollar from time to time (quite often in fact) would require the central banks to buy and sell their currency.
In fact all central banking involve controlling the money base. The central bank can use different operational targets like interest rates or exchange rates, but the central bank is never doing nothing. George Selgin who (indirectly) inspired this blog post would of course say that if you want central banks to do nothing then you should abolish central banking all together, but that is not the purpose of this discussion.
An example of the fallacy that a central bank can do nothing is the debate about “quantitative easing” (QE). There is really nothing special about QE as it basically just means to increase the money base. This in someway is seen to be “dirty” or dangerous and it is getting a lot of attention, but some central banks are doing QE all the time, but it is getting no attention at all. Lets say a country has a fixed exchange rate policy and the demand for its currency for some reason increases – then the central bank will have to sell it own currency to curb the strengthening of the currency. But what does it mean to “sell the currency”? In fact that means to increase the money base. That is QE. So central banks with fixed exchanges could in fact be “doing nothing” and at the same time be engaged in QE on a massive scale – just ask the good people at People’s Bank of China about that.
“Doing nothing” in monetary policy is not really as simple as it is often made up to be. There is, however, another way of looking at things and that is to differentiate between rules and discretion.
NGDP Targeting is as close to “doing nothing” as you get
After the outbreak of the Great Recession a lot of central banks have been conducting monetary policy on a discretionary basis – jumping from one crisis to another without defining the rules of engagement so to speak. An obvious example is the Federal Reserve which have implemented QE1 and QE2 and even the odd “operation twist” without bothering to state what the purpose of these policies are and under which circumstances to scale them up and down. Interestingly enough the Fed has been criticised for doing what central banks do – “playing around” with the money base – but there has been little criticism the discretionary fashion in which US monetary policy has been conducted. Even most of the Market Monetarist bloggers have failed in clearly stating this (sorry guys…).
Imagine instead that there had been a NGDP level target in place in the US when the Great Recession started. A NGDP target would have been a clear rule for the conduct of US monetary policy. It would have stated that if NGDP expectations (either market expectations or the Fed’s own forecast) drops below a certain target then the Fed should take actions to increase the money base (without any restrictions) until NGDP expectations had returned to the target level. That likely would have led to a significant increase in the money base, but within a very clearly defined framework and the increase in the money base would have been completely automatic (as would have been the “exit” from the boost in the money base). Very likely there would not have been any debate about whether this increase in the money base or not if the NGDP target framework had been in place. In fact the Fed could have said it was “doing nothing” – even though that would as demonstrated above, but it would not have done anything discretionary. The real problem with QE is not that the money base is increase, but that is done in a completely random fashion without any clear framework. So the best thing the Fed could do was to very soon implement some rules of engagement – preferably a market based NGDP level target.
PS Those of my reader who are in favour of a true gold standard should know that the central bank can easily end of doing quite a bit of manipulation of the money base within the framework of a gold standard.
PPS Just came to think of it – why did nobody debate the increase in the US money base prior to Y2K (that was actually quite insane a policy) or after 911?
Lorenzo from Oz
/ October 25, 2011the central bank can easily end of doing quite a bit of manipulation of the money base within the framework of a gold standard Yes, as the Bank of France and US Fed did in 1928-1932 (pdf).
QE seems to be what you announce when you do not actually have a coherent monetary policy regime: a case of trumpeting mechanics over managing expectations.
JKH
/ October 25, 2011“but within a very clearly defined framework and the increase in the money base would have been completely automatic .. very likely there would not have been any debate about whether this increase in the money base..”
Perhaps more debate with Scott Sumner’s version, including transparent NGDP futures market intervention?
JP Koning
/ October 25, 2011What was the matter with the Y2K increase in currency outstanding? The public was demanding more precautionary cash balances, the central bank simply provided this cash.
Alex Salter
/ October 25, 2011All good points. I think the scariest part of “do nothing” policies is the door it opens for “countercyclical” fiscal policy. There is a clear demand to “do something”; if we shut off the monetary channel, we’re going to get another ill-conceived fiscal stimulus package. Ultimately the choice is between corrective measures which facilitate a market-oriented approach or a politically-oriented approach. You already know which one I’m in favor of!
Lars Christensen
/ October 25, 2011JPK, I am not sure that demand for base money was there. Nonetheless the base increase was relatively fasted scaled back.
Alex, I agree on that point. Milton Friedman has a similar point regarding in his famous “The Case for Flexible Exchange Rates”. If you don’t allow for exchange rates to float freely then policy makers will use other instruments to reduce “imbalance” such as tariffs and capital control. Douglas Irwin makes also makes this point in his new book “Trade Policy Disaster: Lessons from the 1930s” (review coming up soon!).
JP Koning
/ October 25, 2011I know this isn’t the main point of your post, but regarding Y2K demand… In late 1999, people wanted more cash because they were worried, so they withdrew it from banks. Banks converted some of their reserves into cash to meet this demand. As a result, there were less reserves in the banking system, so there was upwards pressure put on the Fed funds rate. The Fed created reserves to keep it on target.
So the demand must’ve been there. This is demonstrated by the huge rise in cash in circulation around that time, which reflect a simple portfolio decision on the part of Americans.
Lars Christensen
/ October 25, 2011JPK, you have a point. I must admit that I did to much thought when I wrote it. But you way of describing the Y2K event in fact demonstrates my point quite well – the Fed was “doing nothing” in the sense of keeping the Fed funds rate fixed by increasing the money base.
Mike Sproul
/ October 25, 2011Think of silver spoons. If the government established a ‘central silversmith’, then ‘doing nothing’ might mean that the central silversmith tries to keep spoon prices constant, or stamps all silver brought to it into spoons, or stamps a set number of spoons each year, etc. In every case, the silversmith is actually doing something.
The only way I can see that the central silversmith would really do nothing is if it were abolished. Of course, there is a presumption that the government would not interfere with private silversmiths, and that the invisible hand would be left free to determine the amount of silver that gets stamped into spoons, the number of spoons that get melted back to bullion, etc.
This is the most coherent definition of ‘doing nothing’: Get the government out of the money/banking business, and allow the invisible hand to determine the quantity and type of money that is issued, just like the invisible hand determines the quantity and type of silver spoons.
Lars Christensen
/ October 25, 2011Mike, that I agree with. If you like the central bank to “do nothing” the only option is Free Banking. That said, central banks can do more or less – or in a more or less transparent way. The question therefore is IF central banks do exist what should they do and how?
Mike Sproul
/ October 25, 2011Lars:
That’s easy. They should stop restricting the activities (mainly issuance of money) of private banks. If free bankers are right and private bankers can out-compete the central bank, then the central bank will die quietly from a lack of customers. If the government has some advantage at producing money(which I doubt), then the central bank will survive, and then we’d have to come back and ask your question about what the central bank should do. History and logic say that the central bank should stand ready to issue new money to anyone who offers assets of at least equal value in exchange for that money, and to also stand ready to buy that money back at the same price, less its transaction costs.
Lars Christensen
/ October 25, 2011Mike, I am not going to disagree with you on that one…
Benjamin Cole
/ October 26, 2011In utopia perhaps, central banks could “do nothing.” I like transparently targeting NGDP by rules, btw.
But stuff happens. Wars, economic collapses in other nations. Huge gigantic frauds, like Long Term Capital Management. Perhaps the housing bust (though that may have been caused by bad central banking). Nuclear plant meltdowns. Suppose a major earthquake hit the American Midwest again, and incredible amounts of rebuilding was needed?
It was only in the 1960s and 70s we had riots in the streets of the USA. Imagine if the unemployed today rioted in the streets and shut down commerce. Obviously, we would have to spur the economy pronto, and worry about inflation later.
In utopia, theories work. On the ground, we need central banks.
Blake Johnson
/ October 26, 2011@Benjamin: The implicit assumption you make is that central banks actually have a significantly better track record at avoiding recessions than the pre-central bank era. There is an increasing amount of work challenging this claim, the best of which I have come across is White, Selgin, and Lastrapes working paper Has the Fed Been a Failure?. A similar argument to the one you made has helped lead us to the sovereign debt crises. We can worry about who will pay for all of this later, what we need is more public works now.
@Lars: I am particularly interested in the extent to which central banks were able to manipulate the monetary base under a gold standard, and for what period of time. I recently read Douglas Irwin’s paper on the French and US hoarding of gold as one of the causes of the Great Depression (which I believe I found via a link on your blog), and was curious to what extent has this happened historically, and for how long central banks may be able to hoard gold in this manner before the price-specie flow mechanism might correct their behavior. If Irwin is correct, and I believe he is, it would seem that even a period of only a few years was enough to cause great macroeconomic harm. I have to wonder how likely it would have been to occur again though, given the drastic consequences and limited upside. I suppose that depends on a great number of factors, including whether or not the connection was recognized. Irwin’s research does seem to suggest that at least in England, they recognized what the French were doing and what the implications were for the monetary base of other nations.
Lars Christensen
/ October 26, 2011@ Blake, happy to hear that you have read Doug’s paper and you should be happy to hear that he will be visiting GMU soon (as far as I know…).
I believe in tightening the hands of central banks to certain rules and a gold standard is such rule. In periods the gold standard has worked very well, but clearly in the second half of 1920s especially France, but also the US did not play by the book. A number of economists also have questioned whether it is really a price-specie flow mechanism that was working in the more successful periods of the gold standard. I am myself not sure about that.
Btw the White, Selgin, and Lastrapes working paper you mentioned to Benjamin is excellent. Leaving aside whether or not you are a Free Banking proponent or not the paper is an very interesting study of central bank failure. We can not just assume that central banks do the right thing. However, the paper says very little about the extremely successful period in US monetary US and that is the Great Moderation – particularly the 1990s where US monetary policy provided a very stable nominal anchor. I happen to believe that that period was a period of implicit NGDP targeting. Obvious White and Selgin would argue that US monetary policy became to loose and that that helped cause the Great Recession. I think that is partly right (US monetary policy probably was too loose from 2003-4 and until crisis hit).
Blake Johnson
/ October 26, 2011Yes, Professor Irwin will be presenting his paper at next weeks Public Choice seminar here at Mason, and I plan to be in attendance. I look forward to the chance to discuss his paper with him a little more in person.
I’m afraid I am not familiar with what the alternative mechanism that would be functioning during the more successful periods of the gold standard would have been. If you could point me in the direction of something discussing the alternative, I would be grateful. At the least, I believe some sort of commodity standard would be beneficial in at least forcing central banks to be constrained at least in some sense, and it resulted in almost no cases of hyperinflation (though the German hyperinflation comes to mind, although it started while they were still off the gold standard). At times I am rather fascinated with the idea of a Brick standard proposed by C.O. Hardy, and in some ways endorsed by Milton Friedman. Of course, as Hardy himself noted, the biggest practical obstacle would be to get people to think of bricks as money. In this manner however, the production of bricks would be elastic enough that it would be very difficult for one or two central banks to increase their brick holdings at the expense of others.
I think the White, Selgin, and Lastrape paper does not focus on the great moderation because the paper was more about recessionary periods both before and after the advent of the Federal reserve. I think they do address it at least partially, in their discussion of Cristina Romer’s work, and of benign deflations, which the Fed eliminated during the great moderation. They also argue that we would have expected volatility of output to decrease as the economy became more diversified, and less vulnerable to large supply shocks in a single sector, i.e. a failed harvest when the economy was largely agricultural.