Trevor Chow sent me the following questions, as a follow up to my recent post: I was wondering what your thoughts are regarding the following:1. I agree that the thought experiment shows monetary policy can set nominal aggregates as high as they would like. But I’m not entirely sure this solves the machete-scalpel problem.2. What I mean is the following. Suppose we start in a world where we’re on some nGDP growth path. Suddenly, there’s a recession (V falls/k rises) and expected nGDP falls. The central bank tells everyone about this helicopter thought experiment and persuades them that it can credibly maintain the nGDP growth path. Consequently, it needs to do very little actual money creation in order to keep nGDP futures on the nGDP growth path. And so expected nGDP
Scott Sumner considers the following as important: Uncategorized
This could be interesting, too:
Scott Sumner writes Schrödinger’s inflation
Scott Sumner writes Random articles
conversableeconomist writes Was Vaccine Development an Example of Successful “Industrial Policy”?
conversableeconomist writes Some Economics of Sawdust
Trevor Chow sent me the following questions, as a follow up to my recent post:
I was wondering what your thoughts are regarding the following:
1. I agree that the thought experiment shows monetary policy can set nominal aggregates as high as they would like. But I’m not entirely sure this solves the machete-scalpel problem.
2. What I mean is the following. Suppose we start in a world where we’re on some nGDP growth path. Suddenly, there’s a recession (V falls/k rises) and expected nGDP falls. The central bank tells everyone about this helicopter thought experiment and persuades them that it can credibly maintain the nGDP growth path. Consequently, it needs to do very little actual money creation in order to keep nGDP futures on the nGDP growth path. And so expected nGDP doesn’t actually fall. This is basically what you were saying.
3. However, suppose instead that for some reason it screws up. And for some reason, people no longer believe that the thought experiment is enough i.e. they believe in the machete-scalpel problem. Perhaps this is because they mess up for a period of time or because everyone misdiagnoses what tight and loose policy are. Regardless, it is no longer credible. Then it can’t just say it wants an nGDP path and hope that people will respond accordingly, and in fact might have to do a lot just to get close.
4. This may be particularly difficult because people’s belief in a central bank depends on whether they think other people find the central bank to be credible. And in that sense, it would seem likely that while what you’re saying is true most of the time, it is possible to get stuck in bad equilibria where you need a lot more than the usual show of force to even begin regaining control of the scalpel?
5. What I’m trying to say, albeit somewhat poorly, is that the machete-scalpel problem seems to be true if everyone believes it to be true, and to be false if everyone believes it to be false.
I’m going to disagree with this:
This may be particularly difficult because people’s belief in a central bank depends on whether they think other people find the central bank to be credible.
In my view, people tend to doubt central bank commitments mostly when central banks are not in fact committed to doing whatever it takes. Yes, you can imagine a hypothetical case where a central bank was committed to do whatever it takes to boost inflation, but the public for some reason still did not believe the central bank. But I suspect that that’s pretty rare. It is most likely to happen in a case where a central bank had previously made a commitment that was not sincere, reneged on its promise, and then subsequently made a commitment that was sincere, but not credible.
For instance, Volcker backed off on his first attempt to reduce inflation in late 1980, and then in late 1981 adopted a “whatever it takes” approach that was ultimately successful. Note, however, that it didn’t take all that long to convince people that the Fed was serious, and NGDP growth rates plunged sharply in late 1981 and 1982. And when it comes to stimulus, the central bank’s job is actually much easier in a political sense. Volcker faced doubts because a contractionary monetary policy hurts workers, businesses and borrowers. There was intense pressure on him to back off on the tight money policy. In contrast, a policy aimed at raising inflation from zero to 2% tends to make the economy better in the short run, and thus is less politically controversial.
Central banks are not completely transparent, but they are also not particularly secretive. Top Fed officials frequently describe their policy preferences to people in the media. Thus I doubt that a sincere and publicly announced policy to do whatever it takes to hit an NGDP level target (or price level target) would be non-credible. On the other hand, I could easily imagine such a policy not being sincere.
Consider the following example. Suppose that in 2003, BOJ policymakers announced that they were going to adopt Lars Svensson’s “foolproof” plan to escape from a liquidity trap. They say that they will do “whatever it takes” to hit the target amount of yen depreciation. But privately, BOJ officials decide that if the US complains too strongly about a weak yen, and threatens trade sanctions, they will back off. That’s not a sincere “whatever it takes” policy decision, and the markets would (correctly) find the promise non-credible. They’d see right through it.
In the vast majority of cases, promises are seen as non-credible because they are not in fact sincere, and when sincere they are seen as credible. For that reason, I’m not very worried about the hypothetical raised by Trevor. Nonetheless, let’s consider the outcome of a sincere promise that was seen as being non-credible, no matter how remote that possibility. Suppose the BOJ made the sincere decision to implement Svensson’s recommendation regardless of how much flack they received from the US, but the decision was not credible. What then?
I’d still say the policy was a scalpel, not a machete. Yes, the BOJ would probably have to increase the monetary base dramatically, as speculators would anticipate big profits if and when the BOJ caved in and let the yen appreciate. But in the longer term, the central bank that abandons its stimulus is actually forced to do more money creation than the sincere central bank.
Thus in early 2015, both the Swiss and Danish central banks were under pressure from speculators who anticipated a break in their currency’s euro exchange rate peg. The Swiss did cave in, a decision that in retrospect was clearly a mistake, whereas the Danish central bank refused to give in to speculators. Indeed I’d even go further. I suspect that speculators attacked the Swiss franc in early 2015 partly because they anticipated that the SNB was not sincere, and would soon let the franc appreciate as a result of the false belief that Switzerland was threatened with spillover inflation from the eurozone. Speculators didn’t so much force the SNB’s hand as they anticipated a bad decision that was already in the works.
Denmark got the last laugh, however, as once speculators saw the Danes would not let the krone appreciate, they stopped speculating in their currency. Thus, in the long run, the SNB had to increase their monetary base far more than the Danish central bank, as speculators assumed that the Swiss franc would continue to appreciate over time.
Given the fact that a non-credible central bank might have to dramatically boost the monetary base, why do I insist that this policy is a scalpel and not a machete? Because in this case it is more useful to think of the policy instrument as the exchange rate, not the base. The monetary base might change dramatically, but only in service of a central bank policy aimed at targeting the exchange rate (the scalpel) at a precise level.
Here’s an analogy. Consider a Taylor Rule-type policy in 1991, which sets the fed funds rate at 3%. Then assume a sudden collapse of the Soviet bloc, which leads to massive hoarding of US currency. To keep the fed funds rate at 3%, the Fed must accommodate that demand by printing enough base money to meet the growing demand for US currency in the former Soviet Union. From the perspective of the rapidly growing monetary base, policy looks like a blunt instrument, a machete. But from the perspective of the fed funds target, policy looks like a scalpel, a very precise instrument that keeps the fed funds rate exactly where the Fed wants it.
I favor a policy where the Fed adjusts the base as much as needed to keep its policy indicator right on target. That indicator might be an exchange rate, an NGDP futures price, or a TIPS spread. More likely it would be a hybrid market/internal forecast of something like NGDP or core inflation. But I don’t want the Fed to actually target the monetary base, or to use it as an indicator of the stance of policy.
The base is like the steering wheel for the nominal economy. A steering wheel controls the path of the car, but I don’t “target” the position of the steering wheel. Rather I move it as needed so that my GPS says I’m moving to the place I’d like to reach in the most efficient way possible.
When monetary policymakers are sincere but not believed by the public, the central bank may have to move the base by more than otherwise, until the public becomes convinced that it is sincere. But, in general, when a central bank is truly sincere it won’t take the public long to figure that out. They are like 5-year old children, whose intentions can easily be read in their face. Speculators that are slow to realize the central bank’s sincerity will lose money on their investments.
The key is that the central bank must be sincere. Everything works so much better if you tell the truth.