Instrument independence vs target independence under flexible exchange rates

Sebastian Edwards (HT Mark Thoma) says that monetary policy independence under flexible exchange rates is an illusion. This conflates instrument independence with target independence. Instrument independence is always an illusion, given the target. Target independence is not an illusion.

Suppose that Canada were a completely closed economy, with no international trade, no international finance, and no international communication whatsoever. The Bank of Canada can choose any monetary policy target it likes (within reason) regardless of the monetary policy of the US Fed, or any other foreign central bank. If the Bank of Canada wants to target 2% inflation, it can.

Given that 2% inflation target for monetary policy, the Bank of Canada has no choice whatsoever over the setting of the monetary policy instrument. The 2% inflation target, plus the shocks that hit the Canadian economy, completely determine the nominal interest rate it needs to set to hit that target. (Almost) any shock that hits the Canadian economy (strictly, any information the Bank of Canada has about any shock that hits the Canadian economy) will force the Bank of Canada's hand. Once I have chosen the road to drive, the bends in that road force my hands to turn the steering wheel.

Now go to the opposite extreme, where Canada is a completely open economy, with free trade and free capital mobility. Does anything change? No.

The Bank of Canada is still free to choose any monetary policy target it likes, regardless of the monetary policy of the US Fed, or any other foreign central bank. If the Bank of Canada wants to target 2% inflation, it can. (If it fixes the exchange rate to the US dollar, then that fixed exchange rate is the Bank of Canada's target.)

And given that 2% inflation target for monetary policy, the Bank of Canada still has no choice whatsoever over the setting of the monetary policy instrument. The 2% inflation target, plus the shocks that hit the Canadian economy, completely determine the nominal interest rate it needs to set to hit that target.

The only difference is that the set of shocks that hit the Canadian economy now includes shocks coming from outside Canada, including the interest rate decisions of the US Fed (which are themselves responses to shocks hitting the US economy).

Of course Canadian and US interest rates will be correlated (though the degree of correlation depends on whether the shocks are symmetric or asymmetric, and permanent or temporary, and what the two central banks are targeting). Just like the Canadian and US prices of wheat will be correlated. But that has nothing to do with monetary policy independence under flexible exchange rates. Monetary policy is not interest rate policy. The target is not the instrument.

Comments

[Kenneth Duda]

NIce post, Nick. I only read the first part of Sebastian Edwards' paper, and instantly had the same reaction: "of course a central bank trying to stabilize its own nominal economy will have to respond to external shocks, including those due to the monetary actions of other central banks." Only you put it so much more nicely then I ever could have.

Kenneth Duda
Menlo Park, CA

[Nick Rowe]

Thanks Ken! Just by chance, I happen to be on the committee of a very good Poli Sci(!) student whose PhD thesis is on whether the Bank of Canada is independent from the hegemonic US Fed, and had been talking about this distinction with her. So when I saw Sebastian's post, I already had my argument all lined up and ready to go.

[josh L]

I'm not sure I understand the distinction you're trying to make. Yes, the BoC can target what it wants, but if it can't implement it, isn't that the same thing as saying the target is not independant?

Ie, the BoC wants to 2% inflation. However, say the US/Canada are high correlated, and the Fed wants 1%. It seems possible that the actions required for BoC to get to 2% might be so astronomical that it's beyond what they have the resources for.

To use a recent example, the Swedish banking system wanted to target a certain level of inflation of the franc. But it simply got swamped out by QE, and they were forced to give up.

I don't see the difference in distinguishing between the BoC being able to target 2% but can't actually implement, instead of just saying it can't target 2%.

Also, i'd have to reread Sebastian's post, but it seems like he's saying that if the Fed increases interest rates, rather than dealing with it as a shock, emerging markets will partially adjust their interest rates higher, rather than dealing with it fully as a shock. Sure, the emerging markets could try to stay at say, 2%- but if the Fed goes from 2-2.5%, they may simply have to accept part of that, and increase their rates to say, 2.25%. They could stay on the 2% road, but Sebastian seems to be saying that for the most part, the banks are instead saying "this road is too difficult, we'll take something in between". Which in practice means that they aren't independent. There's probably some subtleties in certain situations, where making the road more difficult is not the same as forcing a change in targets, but in many cases they seem functionally identical.

[Nick Rowe]

josh; I though I was being perfectly clear!

The Bank of Canada CAN implement a 2% inflation target. We know this, because it has been doing it (plus or minus random shocks, because it doesn't have a crystal ball).

There are some targets it cannot implement, like a full employment target, or too low an inflation target. that would be true even if Canada were a totally closed economy.

And it was Switzerland, not Sweden, that tried to target a low inflation rate. Now the idiots are trying to target an inflation rate even lower. Read my post a couple of days back, about central banks that want to shrink.

[Scott Sumner]

Nick, The title of Edwards' article says it's an illusion, but not the article itself. The article says there is not completely policy independence under a managed float. And that claim seems correct.

I completely agree with your post, but I'm not sure it contradicts anything Edwards said, except the title of his article.

[Nick Rowe]

Scott: people talk about "managed float", but I have no idea what it means. "flexible exchange rates" is also almost meaningless. It means "not fixed", but that leaves 1,001 different possible monetary policy targets.

In the textbooks (e.g. Mundell-Fleming ISLMBP), "flexible exchange rates" really means "fixed money supply", but that is only one of those 1,001 different possibilities. It could also mean "target inflation", or "target NGDP", or "target the price of oil", etc.