The Bank of Canada and the Fed: rules vs. discretion

I just can't get interested in the Bank of Canada right now. I can only think about the Fed. The Bank of Canada is just too boring. Which it should be.

I'm not alone. Simon van Norden has had breakfast (and lunch and dinner) at the Bank of Canada many times, but has only blogged about his breakfast at the Fed.

The Bank of Canada raised the overnight rate to 1.00% on Wednesday. I think that was a mistake. If I were sure it were a mistake, and if I could prove it, or if I thought it mattered a lot, I would write a post on it. But I don't, can't, and don't; so I won't.

It's not just that Canada seems (touch wood) to be recovering from the recession and the US isn't. Nor just that Canada is boring and America exciting. It's because of a very old debate in monetary policy: rules vs discretion.

The "rules vs discretion" debate has gone through many different interpretations of what we might mean by the distinction. But at its simplest, it's the debate between those who believe that monetary policy should follow some sort of fixed rule, and those who believe that monetary policy should be left to the discretion of those in authority to do what they think is best.

The battlefield kept changing over the years, but was eventually fought to a draw, with most protagonists agreeing on a compromise. The compromise said there should be a clear commitment to some fixed strategic target for monetary policy, like 2% inflation for example; but the tactical decisions taken to try to hit that target should be left to the discretion of an independent central bank.

The Bank of Canada is boring because I already know its strategy. I know what it is trying to do. The Fed is exciting because I don't know what it is trying to do. Non-economists know better what the Bank of Canada is trying to do than expert Fed-watchers know what the Fed is trying to do. The Bank of Canada is targeting 2% inflation; everyone knows that. Fed-watchers don't know the Fed's target. It's a bit like Kremlinology in the old days.

The Bank of Canada is following a rule; the Fed isn't.

Why does it matter?

The classical liberal argument for rules vs discretion (or rules vs authority) was articulated by Henry Simons (pdf). A free enterprise economy requires the rule of law, not the rule of men, in monetary policy just as much as in property rights or contract law. The "rules of the game" must be clear, and not depend on personalities. We should not be talking, for example, about the Greenspan put.

"An enterprise system cannot function effectively in the face of extreme uncertainty as to the action of monetary authorities or, for that matter, as to monetary legislation. We must avoid a situation where every business venture becomes largely a speculation on the future of monetary policy."

We want those businesses that make the best use of resources to succeed, not those that make the best guesses about future monetary policy. We want businesses to invest in something other than Fed-watching.

The rules vs discretion debate took a wrong turn during the arguments over the "monetary policy ineffectiveness proposition" in the 1970's. It was really a debate over whether simple rules could do as well as more complex feedback rules, not over rules vs discretion. That became clearer with hindsight. But very complex feedback rules, where it is hard to specify in advance what you would do in every possible circumstance, look a lot like discretion.

Kydland and Prescott (1977) re-invigorated the debate by re-interpreting rules as precommitment. If you can make credible promises you can usually do better than if you can't, because by promising you can influence others' expectations of your future actions, even if you subsequently have to do something you didn't want to do, because you had promised to do it. It's the same for central banks.

Inflation targeting wasn't something dreamed up in academic seminars. It came out of the central banks themselves, and their interactions with governments. But it turned out to fit right in with the rules vs discretion debate. Twenty years ago the new Prime Minister of New Zealand was going around all the government departments insisting they announce some sort of target they would be held accountable for meeting. When asked for his target, the Governor of the Reserve Bank of New Zealand said "Ummm, inflation?". The Bank of Canada had been quietly working on implementing an inflation target at the same time. It announced its target with the government's backing.

The announced commitment to an inflation target is the rule. How best to implement that rule is left to the discretion of an independent central bank, and its authority on that question should not be challenged by the government of the day, or by anyone else, just as we shouldn't challenge the authority of the Supreme Court to interpret the law.

But that does not mean that a central bank should have the independent authority to do whatever it feels like doing at the time. That's exactly what Henry Simons didn't want.

And one of the strongest arguments for rules is that it is a lot easier for a central bank to hit its inflation target (or price level path target, or nominal GDP target) if it tells people what that target is. The first reason comes from the supply-side, or expectations-augmented Phillips Curve. Expected inflation is one of the biggest influences on actual inflation. The second reason comes from the demand side. Expected inflation, and expected GDP, are two of the biggest influences on aggregate demand and hence on actual inflation and actual GDP. It's easier to get the economy to go where you want it to go if the economy knows where you want it to go. Just like herding cattle.

Because people know where the Bank of Canada wants the economy to go, and believe it will get it there eventually, small tactical mistakes (if that's what they are) by the Bank of Canada really don't matter very much. They have very little effect on people's expectations. They aren't really worth blogging about. David Beckworth shows how expected inflation in the US has been falling a lot over the past 8 months. People don't know where the Fed wants the economy to go, and keep on revising their expectations down. Fed-watchers don't know either.

That's no way to run monetary policy. We can argue about what the Fed's target should be, and about what tactics would best hit that target. But it ought at least say what its target is.

Of course, when the Bank of Canada acts as lender of last resort, rather than doing monetary policy in the narrow sense, it's just as bad as the Fed. There are no rules; it's all discretion.

[Update: This interview (H/T Mark Thoma) with James Bullard, President of the St Louis Fed, is an example of what I mean. I used to think that the Fed had an implicit inflation target of around 2%-2.5%. Now I learn that one man who votes on Fed decisions has a lower bound of somewhere just below 1% on his personal inflation target.

'He also suggested a renewed vigilance on a threshold that, if crossed, would cause the Fed to act. Referring to the annualized gains in core consumer prices of around 1%, Bullard said, “That’s not extremely low. But if it would start to go down from there, to 50 basis points or down to zero, then I would start to get concerned we’d be in a Japanese style scenario, so I think we want to defend our inflation target on the low side, and take aggressive action if necessary.”'

Though the analogy is imperfect (and the fact that it is imperfect says a lot), we do not interview deputy governors of the Bank of Canada to find out their personal inflation targets. They would all give exactly the same 2% answer, and think you were silly for asking. Which is not to say that some of them don't think the target should be changed; but until it is changed, and the new target is announced, it's 2% for all of them. Period. Rule of law vs rule of men.]

Comments

[Simon van Norden]

For the record, I have never had a one-on-one breakfast, lunch or dinner with a current or former Bank of Canada Governor. (But then, the president of a regional fed is much less important than the governor of a real central bank.)

[Simon van Norden]

"When asked for his target, the Governor of the Reserve Bank of New Zealand said "Ummm, inflation?". The Bank of Canada had been quietly working on implementing an inflation target at the same time. It announced its target with the government's backing."

Are you sure that the public inflation targeting regime was an initiative that came from the BoC? Do you have a source for that? I've heard so many conflicting stories that I can no longer keep them straight in my head....

[Simon van Norden]

There's a fun discussion on rules versus discretion in another context that is still pretty active: financial regulation. Basel II was, in some ways, a rules-based system on steriods. There were clear rules on how much risk banks and other institutions could take on. If they developed new products, they could provide their own models to prove to regulators that they still met the rule-based limits on risk.

In the wake of the financial crisis, some regulators complained that these rules became a straightjacket; they were unable to take action against risks they saw accumulating in the system because financial firms were playing within the rules. As a result, some regulators argue that they have an essential need for discretion as a tool to respond to a financial market that evolves to exploit regulatory weakness and loopholes.

[Determinant]

Is that not just an instance of the Rule of Conservatism, that is if it's new, innovative and untested, it is assumed to be risky until proven otherwise?

Or put another way, there is always Black Swan risk.

[Nick Rowe]

Simon: the nearest equivalent to a regional Fed President I could think of is a Deputy Governor. And I was trying to imagine you asking those same 5 questions (or anything vaguely similar) of a deputy governor at the BoC over breakfast. And the image was just too incongruous. If you asked them what they were targeting, they would ask you if you were feeling OK after last night.

My source on the origins of inflation targeting in Canada? My fuzzy recollection of chats with Chuck Freedman. But I had better check with Chuck, to see if I got it roughly right.

Maybe the big difference between financial regulation and monetary policy on rules vs discretion is this: in monetary policy, the cattle basically want to follow the rules, because they want to stay with the herd. In financial regulation, the individual cattle are always looking for a loophole to get one step away from the herd.

[Jim Rootham]

In monetary policy there is only one cow (possibly a bull, but that has an undesirable association). In financial regulation there is a herd of cattle.

The problem with central banks just targeting inflation is that no-one is responsible for unemployment. The ethically critical task for economists is to figure out how to get the NAIRU down.

[Stephen Gordon]

Are you sure? Because there's a really cheap and easy way to do that: abolish Employment Insurance and social assistance. Be careful of what you wish for.

[Gregory Sokoloff]

Cf. Greeks, Romans, and Financial Reform

[Angelo Melino]

In his book, "Making Money", John Crow says that the initiative for inflation targeting came from the Dept of Finance, but it fit in with some of the things the Bank was thinking about.

[Simon van Norden]

"And I was trying to imagine you asking those same 5 questions (or anything vaguely similar) of a deputy governor at the BoC over breakfast. And the image was just too incongruous."

I suspect that I *have* asked the competence/incompetence one to a deputy governor (Pierre Duguay?) in the context of a discussion of central bank accountability. My vague recollection is that he enjoyed sparring with me. As for the others...

"1. Faced with conflicting expert opinion on inflation and growth, how does he justify his focus on inflation risks? or more generally, how should monetary policy be made in the face of such uncertainty?"
They probably would have given me a technical literature survey, with references to Bayesian model averaging and robust control theory, and concluded that it was difficult in practice.

"2. Given lots of experience with globalization and its effects on monetary and fiscal policy in countries like Canada and New Zealand and the UK, why does he think this is poorly understood or different in the case of the USA?"
Yep, incongruous.

"3. How much is his thinking influenced by the experiences with financial and banking crises in other developed economies? By some of the research that has come out of the IMF or World Bank or BIS over the past decade?"
They would have given me a summary and mentioned the latest papers from the BIS conference they had just attended.

"4. Knowing what he knows now about the events of 2008, does he regret his push for tighter monetary policy during a recession and just before a massive systemic banking crisis? Or, given the behaviour of oil and commodity prices at the time, would he do the same again?"
Yep, incongruous.

I think this shows that the Bank of Canada clearly has the caliber of people required to run a regional FRB. I mention this only because, in informal discussions over lunch (NOT with Mr. Fisher) in Dallas, the subject of whether Canada should adopt the US dollar came up. I suggested that this might be a good idea provided that
1) Two Canadian Federal Reserve Banks be added to the system; one in Vancouver/Whistler and one in Montreal/Mt. Tremblant.
2) We move two of the less interesting Federal Reserve Banks to more interesting cities like New Orleans and Boulder.
3) They host more conferences for macroeconomists.

[Simon van Norden]

P.S.: I think Nick is probably right about the cattle.

[K]

Stephen Gordon [re lowering the NAIRU]: "Are you sure? Because there's a really cheap and easy way to do that: abolish Employment Insurance and social assistance. Be careful of what you wish for."

Why? Throw in a citizen's dividend instead, and there's nothing to be scared of. Keeps the tax structure nice and flat and incentives strong at the bottom. Finance it from Pigovian (carbon? land?) taxes and throw in some new money (unbacked BOC balance sheet debasement!) if/when in liquidity trap and we can even use it to solve the greatest problem in monetary policy.

[Dan]

It's all nice, but until the issue of whether there's a bubble in housing in Canada is settled. the 'rules vs. discretion' point is moot in this instance. It might well end up being the case where one country tapped the 'sub-prime' demand for housing to temporarily prop-up the economy one business cycle earlier than another.

[Andy Harless]

The Fed does have long-range projections, which it encourages people to interpret as targets. The problem is that it has failed to give a clear explanation of why it does not expect to hit those targets over the next few years. But in principle, the same issues could exist even if the targets were official and mandated. (1) Is it possible to hit the target? (2) Would a policy designed to hit the target as a mean or median outcome imply an unacceptably large variance around the target? (3) Would a policy designed to hit the target over a shorter horizon imply missing the target over a longer horizon? (4) Would policies designed to hit the target present unacceptable costs or risks?

[Just visiting from Macleans]

Q from a layperson.

I'm curious about the structure of BofC and the Fed. Simon had breakfast with the President of the Federal Reserve Bank of Dallas, and in your update you quote President of the St Louis Fed.

In Canada, do we have regional deputy governors? If not, let's suppose we did. Would we get two differing views from say the Deputy Governor of Alberta vs the Deputy Governor of Ontario regarding inflation targets etc? It seems to me that funding/fueling the O&G sector in Alberta would have differing needs/concerns that funding/fueling a manufacturing sector in Ontario.

Could that account for the differing opinions down south?

[Patrick]

"Keeps the tax structure nice and flat and incentives strong at the bottom."

That's great! 'Cause theres a huge queue to join the bottom end and we really need to keep it under control. And who needs equal marginal sacrifice anyway?

[Nick Rowe]

Andy: yep. I had forgotten those long range projections. When they came out a year or two back, we all thought they might be the first step towards inflation targeting. Now nobody seems to talk about them any more, and instead we listen to interviews of regional Fed presidents on what their personal target might be. Announcing a target still leaves a lot of questions open. But right now we can't tell if the Fed is on target, or if it's off target and doesn't know how to get to it, or if it's planning to slowly return to target but doesn't want to take drastic policy actions to get there more quickly.

JVFM: The BoC does not have regional deputy governors. It does have some small regional offices to try and get a better sense of economic conditions in different parts of the country. Those regional offices are important when it comes to the Bank's quarterly surveys, for example. The Bank is always wary of the accusation that it is too central Canada focussed.

I can't see any reason to argue that the best long-term inflation target for (say) Alberta would be 3%, and for Ontario would be (say) 1%. But there will be times when one region is in recession with falling local inflation, and another is booming, with rising local inflation. So monetary policy should ideally be temporarily looser in one region, and temporarily tighter in another. Which can't be done with a single currency of course. This is the old "Optimal currency area" question.

My guess is that the regional Feds were originally set up in part to make sure that the Fed wasn't too DC-centric. I would be interested to hear if divergences of opinion between regional Fed presidents could be explained by regional divergences of unemployment and inflation. My guess is that it can't, but I don't know that. And at least, if there were a clearly announced long term target, we could distinguish between regional differences, differences of opinion on what that target should be, and differences of opinion on how to hit that target. As it is, without a target, it's just a confusing mess.

It would be good to hear US Fed-watchers giving a better answer to your question than mine.

Jim and K: Look, it is an interesting question to ask whether or not a different target for monetary policy might have an effect on the average unemployment rate. But:

1. If you want to talk about that question, it is a very bad idea to frame it as using a different target to reduce the NAIRU. Because the whole NAIRU/natural rate concept *presumes* that you can define some sort of long run equilibrium unemployment rate that is independent of the long run monetary policy target. So you totally confuse the debate, by saying something that is prime-facie self-contradictory.

2. Regardless of whether a different monetary policy might reduce average unemployment, and recognising that other (non-monetary) policies (like EI etc.) will almost certainly affect the NAIRU and average unemployment rate, it is still an open question of whether we would want to implement those policies. Jailing the unemployed would bring down the NAIRU (and not just statistically). But I don't want to do it. That was Stephen's point.

3. This post is not about *which particular* monetary policy target is best. It's about why it's best to have an announced target.

[Nick Rowe]

Simon: yes, I think I'm right about the cattle too. But one thing makes me pause: Don Drummond of TD said at that Carleton conference, IIRC, that they saw the hole in the fence, knew it was risky, and so didn't go through it. But he doesn't know what would have happened in Canada if there had been more banks, and one of the banks had gone through the hole in the fence. Would all of the other cows have followed, because if they didn't follow they would lose customers?

[himaginary]

"This interview (H/T Mark Thoma) with James Bullard, President of the St Louis Fed, is an example of what I mean. I used to think that the Fed had an implicit inflation target of around 2%-2.5%. Now I learn that one man who votes on Fed decisions has a lower bound of somewhere just below 1% on his personal inflation target."

I was shocked to read this paragraph. Because Bullard himself was the one who warned against such a behavior.

From his paper w/t In-Koo Cho:

"In this paper we have developed a theory of near-zero nominal interest rates, as observed in Japan in the 1990s and the U.S. in the 1930s. Our theory is that the economy inadvertently "slides down a Fisher relation" because of misunderstanding concerning the nature of the government’s inflation target. The theory is based on the existence of a self-confirming equilibrium in which inflation and nominal interest rates are relatively high. Our dynamic system can make sudden departures from that equilibrium towards a persistent low inflation, low nominal interest rate outcome which looks like observed “liquidity trap” episodes in major industrialized countries.
...the third element needed to generate escape dynamics in this model is the feedback from private sector beliefs to the inflation target. If the government could credibly commit to a constant long-run inflation target, there could be no escape from the unique self-confirming equilibrium in this model. A number of central banks have, in recent years, begun to state their inflation target more explicitly, although not the Bank of Japan or the Federal Reserve."

[Nick Rowe]

Angelo and Simon: It's a real pity you didn't post those comments before I went off to the grad students' reception yesterday. I could have asked Paul Jenkins for his recollections. I'm going to ask Paul and Chuck. They are both at Carleton now.

To me, it's an open question whether we interpret the joint 2% target as the government tying the BoC's hands, or the BoC tying the government's hands. Or both, for that matter. John Crow leans towards the former, as I read him.

[Nick Rowe]

himaginary: Good find! This paper is of course directly related to the "Kocherlakota controversy". I'm reading it through.

[K]

Nick: Was narrowly replying to Stephen's suggestion that lowering the NAIRU required dumping support for the most disadvantaged. Was about as OT as Stephen, but I apologize.

Patrick: if you're attacking me from the left, I don't think you understood what I was saying. Anyways... OT. Over and out.

[Jon]

The operative phrase in us law is price stability (And full employment). Viewed in that way, the fed has spent the past thirty years trying go move us from a high inflation to a stable price regime while maintaining full employment.

Perhaps they are following a rule, just a much more literal one than you thought.

[Nick Rowe]

himaginary: I have just skimmed that paper. I agree with you. That paper shows exactly the dangers that come when you don't have a clear nominal anchor, and you let expected inflation get too low. My guess is that what James Bullard himself would want, and what he realistically thinks the other members of the Fed will do, aren't the same. And he's speaking of what he hopes the Fed would eventually do, not what he would do right now. Kremlinology again.

Jon: do you mean that the Fed interprets "price stability" to mean zero inflation? It's taken them an awful long time to get to that target, if that's what it is. And why don't they tell us?

[JP Koning]

The paper you linked to by Simons calls economists' preference for central banking over unmanaged banking the "first serious heresy among liberals." The rules vs discretion argument presupposes that this "heresy" has been accepted. If, on the other hand, central banking is shown to be unnecessary, even destructive, and unmanaged banking therefore being preferred, then the rules vs discretion argument is reduced to a simple matter of consumer preference for certain competing banking policies. But we're stuck with the central banking beast, and rules are vital to control it and the people who manage it.

I don't have as much confidence in the permanency of the BoC's inflation targeting rule as you, Nick. As far as I know there are no legislative obligations in the Bank of Canada Act to set price targets. As such the BoC can break its commitments without contradicting any laws or facing any legal challenges/penalties; it would only be breaking with a 15 year tradition; not costless in terms of reputation, but doable.

In this sense, one might say that the BoC's rules are still fairly discretionary. Especially when compared to the Reserve Bank of New Zealand, which explicitly sets out the necessity of setting price targets in its constituting act, and lists the penalties to be faced for breaking those targets. Of course central bank acts can be changed or subtly evaded, so even then there is a touch of discretion.

Anyways, I see rules vs discretion as a continuum... the closer to the rules side the better. But a shade of authority (discretion) will always be present, that's the nature of the beast.

[Patrick]

Seems to my inexpert eye that the Bullard paper is similar to the George Evans model/video from a week or so ago...

[Determinant]

JP:

As the sole shareholder of the Bank of Canada, the Government of Canada has full authority to make such a demand of the Bank. Furthermore, the Minister of Finance can issue policy instructions to the Bank in "exceptional circumstances" under the Bank of Canada Act. The Governor may also be dismissed by the Cabinet.

The Bank of Canada isn't the Fed.

[himaginary]

Patrick:"Seems to my inexpert eye that the Bullard paper is similar to the George Evans model/video from a week or so ago..."

Yes, but this model incorporates a new mechanism which makes economy slip down from stable area to deflationary trap area. That mechanism is the feedback loop in which CB cuts down target inflation rate because the private sector thinks CB has cut down target inflation rate.

And Bullard's remark quoted by Nick suggests, perhaps unintentionally, that Fed has cut down target inflation rate from 2%-2.5% to somewhere just below 1%.

[Lorenzo from Oz]

You North Americans, you forget us Downunder. The Reserve Bank of Australia is every bit as boring as the Bank of Canada! And that is a good thing.

[Nick Rowe]

JP: I don't think the 2% inflation target is permanent either. The BoC itself says it's renewable. It's quite possible it will change it next year, maybe for a price level path target (which would not be a big change de facto from what has actually happened over the last 15 years). But it will announce any such change, and won't do it on a whim. I think custom and tradition are just as important, maybe more important, than any formal legislation. Unwritten quasi constitutions.

Determinant: there's what the laws say, and what the custom is. IIRC the Governor of the BoC is appointed for 7 years "on good behaviour", which means the government can't just switch him mid-term because they disagree with his policies. And is appointed by the Board of Directors with Cabinet approval. (Presumably that means that Cabinet can veto, but not appoint). And the Minister of Finance can issue a written directive, but this is a sledgehammer, and one the government would be very unwilling to use. Though there's "constructive ambiguity" from the BoC on how it would respond to this nuclear option, it's likely the Governor would resign if that happened. The Coyne Affair established the expectation that the government would not mess with the BoC without thinking very carefully about doing so. The government really doesn't want to go there.

Patrick: yes. But what himaginary said.

Lorenzo: I didn't forget! I think about Oz a lot (was just talking about Sydney and Windsor house prices on the phone with an Australian 30 minutes ago, funnily enough). Yep. Cross out "BoC" and "2%" and replace it with "RBA" and "2.5%", and my post would be just the same. Or with any number of central banks. What's the Banco de Mexico trying to do? Go on the web and you get the answer. 3% inflation, + or - 1%. And you can ignore the "+ or - 1%" bit, because no central bank can shoot straight enough to keep within those bounds unless it aims for the centre. Only the tactics are uncertain, not the strategy.

[Adam P]

"And Bullard's remark quoted by Nick suggests, perhaps unintentionally, that Fed has cut down target inflation rate from 2%-2.5% to somewhere just below 1%."

That's another of the sort of interesting things about Kocherlakota's infamous comments. Everyone focused on the apparent silliness of the part that essentially said, if the Fed wants higher inflation just raise rates.

Nobody really paid attention to the other implication of that remark which was that Kocherlakota thinks the current roughly 1% inflation rate is what the Fed wants.

[Stephen Gordon]

While I was reading Jim Hamilton's latest post on QE, it occurred to me that if the Bank of Canada were in the position that the Fed now finds itself, the question wouldn't be *if* it should be buying more long-term, more risky assets. The question would be 'which ones?'.

The Bank showed in the beginning of the 1990s that it was willing to provoke a recession in order to hit its target. I don't think anyone thought it was bluffing when it asked for and obtained the means to do some serious QE if push came to shove.

[Determinant]

Nick:

I'm well aware of the custom of Independence at the Bank of Canada and the Coyne Affair. My point was that the Government of Canada had ample authority to request an explicit inflation target from the Bank of Canada and the Bank was well within the letter and spirit of the law to oblige. JP Koning seems to disagree with this point.

As you point out, there are established customs for what happens if the Bank and the Government have a fundamental disagreement over monetary policy, which has also led to the more normal informal channels for resolving those issues.

[edeast]

The humphrey-hawkins act provided explicit goals. 4% unemployment, and interim 3% inflation. The government employer of last resort policy was stripped out of the bill. So just goals, no policy.

What is a quick and dirty answer to why can't we have low inflation and low unemployment? people's utility decreases, sticky wages?.

[Nick Rowe]

edeast: Let me re-phrase your question: what is the quick and dirty answer to why monetary policy alone can't create low unemployment? Superneutrality of money. Monet doesn't affect anything real (like unemployment) in the long run, only price and inflation.

[edeast]

Ok, then what's the problem with Japan and now the fed seeking 1% essentially 0% inflation according to this summary.(cpi variance of 1%).. Acknowledging the pole on a stick, cow herder problem. Can't the 'other' QE nethods protect against deflation?

Is business cycle smoothing worthwhile? If money has no effect in the long run.

Tyler Cowen had a post couple months ago showing the US productivity is back up to where it was prerecession, with far less people working.

[edeast]

no need to respond; AS curve horizontal then vertical in the long run. Finally reading an into text.

[JP Koning]

"The Bank of Canada isn't the Fed."

Let me rephrase: "As such the BoC, in consultation with the department of finance (perhaps even at its initiative), can break the BoC's commitments without contradicting any laws or facing any legal challenges/penalties." My point continues to be that BoC inflation targeting is entirely discretionary.

Nick, I'd agree with you in one sense about traditions; in the context of the formal common law, they are every bit (perhaps more) durable as legislation. Societal norms are very strong and have penalties for breaking them.

But the game changes when the institution in question is a monopoly backed by the coercive power of the state. Tradition becomes moot; what possible punishment might the BoC face when there aren't any competitors for Canadians to turn to, or formal written procedures to prove that power has been abused?

Say one day the governor of the BoC gets a directive from the government he disagrees with, but rather than resigning he implements it with the government's blessing. As one of a handful of intellectuals knowledgeable of the oral tradition being contravened, the ball might fall in your court to uphold it. What mechanism would you use? Without an actual law to point to, the courts would be off limits. Maybe a war in the press, hoping the BoC/government back down? I'm pretty sure that a readable law, easily verified by the public, press, and courts, would give you much more strength than tradition in what would otherwise be a fairly lonely battle against a very powerful opponent.

[Jim Rootham]

To go back to the thread drift for a moment.

OK, I said that wrong. What I would really like is a method of controlling inflation that does not involve kicking the marginal members of the economy in the teeth. The current techniques do not qualify.

[K]

Nick: "do you mean that the Fed interprets 'price stability' to mean zero inflation?"

No. It means "as much surprise disinflation as we can get away with"

Nick:"And why don't they tell us?"

Cause then they wouldn't be able to get away with it.

[david]

Hi Nick. Good post. Late to the party but I wonder if that Simons quote wouldn't also be true even if it was modified as follows:

"An enterprise system cannot function effectively in the face of extreme uncertainty as to the action of [fiscal policy] or, for that matter, as to [fiscal] legislation. We must avoid a situation where every business venture becomes largely a speculation on the future of [fiscal] policy."

It seems to me that fiscal policy, as much as monetary policy, can be a source of uncertainty and discoordination.

[Phil Koop]

"My guess is that the regional Feds were originally set up in part to make sure that the Fed wasn't too DC-centric."

The purpose of decentralization (which was present in the original Aldrich formulation) was to reduce the power of NY, not DC. It was private ownership that was seen as the bulwark against government. This was a concern of both Republicans and conservative Democrats - the progressives wanted full public ownership. Public control of the board of directors (rather than merely public representation) was the principal difference between the Aldrich proposal and the Federal Reserve Act as finally passed. This compromise with the progressives was necessary because the final vote was split on partisan lines. (Wherefore is this day different from any other day?)

[Nick Rowe]

Adam: "Nobody really paid attention to the other implication of that remark which was that Kocherlakota thinks the current roughly 1% inflation rate is what the Fed wants."

Yep. I admit I didn't pay any attention to that either. I should have done.

Stephen: I think I agree. Plus, the BoC has always been very clear that its inflation target is symmetric. It doesn't like inflation either above or below 2%.

Determinant and JP: I'm enjoying your points, but don't have much useful to add.

edeast: "Is business cycle smoothing worthwhile? If money has no effect in the long run."

That's a good question to ask. I think "yes!". First, even if average output is the same, we don't want it to fluctuate (the benefits from booms aren't as big as the costs of recessions, even if they are the same size). Second, I think of superneutrality as only a first-order approximation to the truth. For starters, the variance of monetary policy might affect average output, and long run growth.

"no need to respond; AS curve horizontal then vertical in the long run. Finally reading an into text."

But I want to respond! Because that raises an interesting puzzle. Why does it *look like* Japan's LRAS curve is reverse-L-shaped? Why isn't it vertical everywhere? Brad DeLong thinks there's an absolute downward money wage rigidity, even in the long run. He might be right.

Jim: agreed. I want that too. But really, the question you have asked there is just a re-framing of the question "How can we reduce the NAIRU?". Some people frame the question as "How can we control inflation without creating so much unemployment?", and others frame it as "How can we reduce the NAIRU/natural rate?". But ultimately it's the same question. Which way you frame it depends on whether you think unemployment causes disinflation, or whether you think disinflation causes unemployment. When you have a vertical LR Phillips Curve, and you want to reduce unemployment permanently, you have to re-frame the question in terms of reducing the NAIRU, because there's no long-run equilibrium relation between inflation and unemployment.

K: Unfortunately, I fear your answer may be correct.

david: Thanks! Yes, it would be. But:
1. Right now, I think the biggest uncertainty is from monetary, not fiscal policy.
2. One of the whole points of inflation targeting was that monetary policy would take full responsibility for AD, leaving fiscal policy to concentrate on micro stuff, which would mean much less uncertainty about fiscal policy, because there would be negligible macro changes in fiscal policy.

Phil: good comment. Nothing to add.

[Determinant]

JP Koning:

At the level of interaction between the Bank of Canada and the Department of Finance, the distinction between discretion/rules is irrelevant. If Finance wants something policy-wise it will get it, either by asking the Bank or by changing the law. The kind of law which would bind the Bank and its Governor with penalties is meaningless in an environment where the Minister of Finance is a phone call or a short walk away, those kinds of meetings are supposed to happen, and the Minister is expected to update economic policy yearly through the budget process.

"Policy Laws" have the same level of enforceability as balanced-budget laws do on provincial legislatures: none.

The same assertion hold with respect to New Zealand too.

I do however hold that the Bank itself is bound to its inflation policy through its relationship with the Minister of Finance (ownership of shares, plus removal powers). It's politically impossible for the Bank to swim against the tide of the government's wishes for any length of time, even without reaching the level Coyne Affair.

The Government may have a light hand with respect to the Bank, but ultimate authority and responsibility for economic policy as expressed in an inflation target rests with the Government, not the bank.

Your post seems to imply that you prefer rules to discretion as you don't trust human nature. A common enough belief, but at the level of Bank/Government relations hard and fast rules are simply impossible to implement.

[edeast]

Nick: Thanks, more questions; You know how you mention a 'what the heck is keeping the keeping the economy going' like the pre-darwinian economics post. Why no stone age for us. Can you explain the accelerator-multiplier lower limit?

"When output plummets downward rapidly, the acceleration principle calls for negative investment. But gross investment in plant and equipment can hardly be negative for the whole economy, so this puts a floor on how far investment can fall."

Why can't investment be negative for the whole economy? can't consume buildings... or does machines getting used for something else, count as investment, can there be a barter of purpose. There has to be examples of economies failing.

"So the slump contains the source of its own recovery. Once investment has hit the basement, it must stop falling; but then so must output. At this point, then, firms may need some replacement investment; so gross investment turns up again, and a new cycle can begin."

I should say this is from a 25 yr old Samuelson text, I'm trying to get a state of knowledge before adding FTPL, RATEX, NewKeynes. Sumner, McCallum etc. All the stuff, that you guys are killing each other over.

[Nick Rowe]

edeast: If you could convert capital goods into consumption goods, eat the machines, as it were, then gross investment could be negative. The capital stock would fall faster than depreciation. We could imagine a world with zero employment, zero production, positive consumption and negative investment. The implicit assumption of the multiplier-accelerator model is that you can't eat the machines. You either use them, or let them sit idle, rusting.

[JP Koning]

Changing central bank acts is not a costless and/or instantaneous process. If the New Zealand government wanted the Reserve Bank of New Zealand to end price targeting altogether, they'd have to change entire sections of the RBNZ Act, since the obligation to target prices is hard-coded into the act. Obscure changes to bank acts can go through in maybe 12 months. But laws governing something as significant as price targeting would be incredibly difficult and time consuming to revise. Some of the nastiest and most drawn out monetary controversies have erupted over significant changes to banking laws - Peel's Act of 1844 comes to mind.

Your point seems to be that in the end, monetary policy is always based on authority, not rule of law. I see a spectrum between the two extremes, and while authority can never be entirely displaced, one nation's central bank will always approach that ideal better than others. New Zealand and the Federal Reserve get closer to it than Canada. So does Japan. The ECB definitely does. Even with its price targets, the BoC is all about authority, not rules.

[edeast]

I'll push my luck with another question;
Dealing with the money multiplier and reserves. Andolfatto's text, treats it as gov /private moneys, which twigged the remembrance of your what makes a central bank central post. In that it's the redeemability of the issue.
Is a central bank the only one who has 100% reserves, based on their one client the gov? They aren't a real bank, more like an exchange? For instance, the liquidity crisis, was when a number of private moneys failed. Banks wouldn't trade with each other. The government stepped in with tarp, and then in january the fed with mbs purchases. I used to be annoyed with Morgan Wrastler on Sumner's blog, but right now the fed's hands are tied with the mbs, which has the fed in charge of two money supplies. I know I'm probably bastardizing it, but the reason they can't engage in OMO's is because of the value of the mbs.