Thanks for your interesting reply. I’ve read lots of Sumner (and to a lesser extent Christensen) but their discussion of Britain is fleeting at best.
How do you draw a line in the sand and say “you shall not cross” when it comes to NGDP? That is the whole question. If you are defending a currency peg, you have a clear policy tool (forex purchases) and you can see the effects of your interventions in an actively traded market. The SNB can check the CHF:EUR exchange rate at any time, and intervene in the market to sell CHFs and buy EURs as necessary, until the price adjusts—and reverse course if necessary. Note by the way that currency pegs are much harder to defend in the opposite direction; the SNB cannot print unlimited quantities of Euros to buy Swiss Francs.
However, when it comes to NGDP targeting, there is no clear tool and no immediate way to see the effect of your interventions. The quarterly GDP numbers are backward-looking and already out of date when they are printed. How can you prevent the line in the sand being crossed, when you can’t see if anyone is coming close to it until 4 months later?
Even this might be superable if there was a clear tool. If you want to keep the CHF:Euro ratio low, you print CHFs and buy Euros. If you want to keep the £:NGDP ratio low, you print £s and buy… NGDP? Except you can’t! You have to buy outstanding government bonds and then hope for various “transmission mechanisms” to take effect. But will this in fact do anything with interest rates at zero? Lots of people say otherwise (e.g. “liquidity trap”, “market segmentation”, “creditism” etc). In point of fact, the BoE was buying government bonds furiously and every quarter the NGDP numbers came in low. Meanwhile we had high inflation and a collapsing £, so it’s not like their asset purchases were doing nothing, it’s just that they weren’t raising NGDP.
This is what people like Cochrane mean when they say that they don’t believe the central bank controls NGDP. It’s not that nominal variables don’t have real effects, it’s that transmission mechanisms matter, and are complicated, and the MMs can’t just feign indifference. Why will these purchases raise NGDP, and not just asset prices? Zen masters like Sumner mock this kind of thinking as “people of the concrete steppes” and say that setting the target is enough? But if transmission mechanisms don’t matter, how do the expectations become unmoored? And if transmission mechanisms do matter, then how can you just blithely ignore them?
Now occasionally you’ll hear MMs respond to this with talk of a Central Bank-created NGDP futures market, which is essentially Hansonian futarchy. This might allow realistic feedback (but unfortunately does nothing to solve the lack of a tool). But you’ll note that none of Sweden, Israel and Australia have any such thing, so if this toy market is necessary to the proposal, then you can’t cite them as existing successes.
Four years ago I was a market monetarist (although in those days we just called it mainstream macro). Now I think that the cause of Britain’s problems is supply side. I am less confident about the US, but I would not be surprised at all if NGDP targeting proved a huge disappointment there too.
I was never in doubt about a determined central bank’s ABILITY to PASS a monetary target, not hit it like a bulls eye, but pass it like a tape at the end of a race. The runners keep running and may need sometime to slow down :)
A central bank can create a depression. Increase the rates to a level where no conceivable investment is feasible.
A central bank can create hyper inflation. Announce that you are willing to buy everything and start buying it. Somewhere along that route, measured nominal GDP would exceed the target laid out for it. So, the central bank can hit the target, by limiting its craziness.
My doubt in regards market monetarism was—is this a good thing to do, in the long run?
For supply side vs demand side issues, I guess the question to ask is if the productivity measurement is correct. If TFP is 1 % and you’re aiming at 6% NGDP growth, you’re just hurting the economy. Maybe, (don’t know for sure), britain was aiming at too high a growth rate.
I too have no doubt that the central bank can cause hyperinflation or severe deflation. The question is, can it thread the needle and get levels where it wants them to be? If during a velocity shock like 2007-8, the only way to get 5% NGDP growth is to alternate between 200% NGDP growth and −47.5% NGDP growth, then it seems like the cure is worse than the disease, in terms of providing the kind of stable monetary expectations that MMs agree are necessary for economic growth, investment, etc.
Suppose we have a patient with a fever (or pneumonia). The doctor is unable to stabilise the patient’s temperature. It’s true that if we throw her in an ice floe she’ll be too cold, and if we throw her in an oven she’ll be too hot, but that doesn’t prove that we can stabilise her temperature, and I sometimes feel that Market Monetarists are coming uncomfortably close to making this argument.
I doubt that there will be that much variation since even if you look at it, you are considering a sale of assets or a purchase. There are thousands of such transactions happening everyday. That range of fluctuations doesn’t happen. I’m not great fan of central planning, but I think that NGDP or nominal wage targets are possible to hit without a 100% variation.
Thanks for your interesting reply. I’ve read lots of Sumner (and to a lesser extent Christensen) but their discussion of Britain is fleeting at best.
How do you draw a line in the sand and say “you shall not cross” when it comes to NGDP? That is the whole question. If you are defending a currency peg, you have a clear policy tool (forex purchases) and you can see the effects of your interventions in an actively traded market. The SNB can check the CHF:EUR exchange rate at any time, and intervene in the market to sell CHFs and buy EURs as necessary, until the price adjusts—and reverse course if necessary. Note by the way that currency pegs are much harder to defend in the opposite direction; the SNB cannot print unlimited quantities of Euros to buy Swiss Francs.
However, when it comes to NGDP targeting, there is no clear tool and no immediate way to see the effect of your interventions. The quarterly GDP numbers are backward-looking and already out of date when they are printed. How can you prevent the line in the sand being crossed, when you can’t see if anyone is coming close to it until 4 months later?
Even this might be superable if there was a clear tool. If you want to keep the CHF:Euro ratio low, you print CHFs and buy Euros. If you want to keep the £:NGDP ratio low, you print £s and buy… NGDP? Except you can’t! You have to buy outstanding government bonds and then hope for various “transmission mechanisms” to take effect. But will this in fact do anything with interest rates at zero? Lots of people say otherwise (e.g. “liquidity trap”, “market segmentation”, “creditism” etc). In point of fact, the BoE was buying government bonds furiously and every quarter the NGDP numbers came in low. Meanwhile we had high inflation and a collapsing £, so it’s not like their asset purchases were doing nothing, it’s just that they weren’t raising NGDP.
This is what people like Cochrane mean when they say that they don’t believe the central bank controls NGDP. It’s not that nominal variables don’t have real effects, it’s that transmission mechanisms matter, and are complicated, and the MMs can’t just feign indifference. Why will these purchases raise NGDP, and not just asset prices? Zen masters like Sumner mock this kind of thinking as “people of the concrete steppes” and say that setting the target is enough? But if transmission mechanisms don’t matter, how do the expectations become unmoored? And if transmission mechanisms do matter, then how can you just blithely ignore them?
Now occasionally you’ll hear MMs respond to this with talk of a Central Bank-created NGDP futures market, which is essentially Hansonian futarchy. This might allow realistic feedback (but unfortunately does nothing to solve the lack of a tool). But you’ll note that none of Sweden, Israel and Australia have any such thing, so if this toy market is necessary to the proposal, then you can’t cite them as existing successes.
Four years ago I was a market monetarist (although in those days we just called it mainstream macro). Now I think that the cause of Britain’s problems is supply side. I am less confident about the US, but I would not be surprised at all if NGDP targeting proved a huge disappointment there too.
Thanks for your patience and civility.
I was never in doubt about a determined central bank’s ABILITY to PASS a monetary target, not hit it like a bulls eye, but pass it like a tape at the end of a race. The runners keep running and may need sometime to slow down :)
A central bank can create a depression. Increase the rates to a level where no conceivable investment is feasible.
A central bank can create hyper inflation. Announce that you are willing to buy everything and start buying it. Somewhere along that route, measured nominal GDP would exceed the target laid out for it. So, the central bank can hit the target, by limiting its craziness.
My doubt in regards market monetarism was—is this a good thing to do, in the long run?
For supply side vs demand side issues, I guess the question to ask is if the productivity measurement is correct. If TFP is 1 % and you’re aiming at 6% NGDP growth, you’re just hurting the economy. Maybe, (don’t know for sure), britain was aiming at too high a growth rate.
I too have no doubt that the central bank can cause hyperinflation or severe deflation. The question is, can it thread the needle and get levels where it wants them to be? If during a velocity shock like 2007-8, the only way to get 5% NGDP growth is to alternate between 200% NGDP growth and −47.5% NGDP growth, then it seems like the cure is worse than the disease, in terms of providing the kind of stable monetary expectations that MMs agree are necessary for economic growth, investment, etc.
Suppose we have a patient with a fever (or pneumonia). The doctor is unable to stabilise the patient’s temperature. It’s true that if we throw her in an ice floe she’ll be too cold, and if we throw her in an oven she’ll be too hot, but that doesn’t prove that we can stabilise her temperature, and I sometimes feel that Market Monetarists are coming uncomfortably close to making this argument.
I doubt that there will be that much variation since even if you look at it, you are considering a sale of assets or a purchase. There are thousands of such transactions happening everyday. That range of fluctuations doesn’t happen. I’m not great fan of central planning, but I think that NGDP or nominal wage targets are possible to hit without a 100% variation.