The goals of monetary policy are to promote maximum employment, stable prices and moderate long-term interest rates. By implementing effective monetary policy, the Fed can maintain stable prices, thereby supporting conditions for long-term economic growth and maximum employment. (
To be clear I'm arguing against Hanson's prediction markets and in FAVOR of NGDP targeting. The prediction markets would work fine in theory but only if the central bank was willing to really commit to absolutely following the rules and honoring arbitrarily large purchases of these predictive securities.
In this case we were talking about something like Hanson's proposal which I understood to be something like the central bank's choice of interest on reserves, value of purchases/sales by FOMC etc.. was set in an algorithmic way from the market value of certain instruments ISSUED BY THE GOVERNMENT which pay out in certain ways based on future events.
If the government retains the ability not to set their policy in the algorithmic fashion the prices of these securities dictate you lose many benefits. If not and the total value is limited you have a risk of large investors influencing future policy by manipulating the market.
Also, I worry is that in practice the government would only be willing to issue some very limited value of these special instruments and has a substantial chance of paying them off early based on some other formula when they are deemed to weird.
Market manipulation doesn't seem to be a problem in other monetary regimes that peg a price, such a fixed exchange rates. In any case, under my "guardrails" approach the central bank would presumably ignore the attempt of one or two traders to manipulate the market
I'm not sure any of these apply to your proposal in particular as I was responding to Robin's suggestion but to elaborate.
Doesn't it need to be able to *in principle* support large trading to ensure that no one can manipulate it in their economic interest? As long as everyone is convinced that if there were an attempt to manipulate the prediction market there would be sufficient liquidity to frustrate this then that's fine but don't investors need to be convinced that counterfactually there won't be a problem. In particular, it seems like one would run into trouble if the instruments Robin is suggesting were given some prior limit on the total dollar value issued allowing a single actor to buy all the assets of a particular type up.
As far as fiat money debts you are certainly correct but for Robin's proposal to work the government has to issue a bunch of instruments which payoff based on the future state of various economic indicators. I'm not worried that the government will default but that investors will guess that they will be unwound prematurely and thus their market price won't truly reflect the actual prediction but some kind of hybrid of the actual prediction and how these instruments might be unwound early.
The definition of the outcome to be maximized would choose between those goals. But my understanding of this area is that this choice is not in fact that main area of disagreement in monetary policy.
You said yourself that promotion of long term growth is one of the goals.
Moderation of the business cycle isn't necessarily the same thing, or even compatible with that goal.
I'm no macroeconomist (either) but it seems plausible to me that vigorous downswings in the business cycle are needed to flush out inefficient practices in the economy, and so are *necessary* to long term growth.
"That is, monetary policy consists of clear public and discrete choices, such as on short term interest rates. Call each discrete choice option C. And it is widely agreed that the point of this policy is to promote long term growth, in part via moderating the business cycle. So some weighted average of real growth, inflation, unemployment, and perhaps a few more after-the-fact business cycle indicators, over the next decade or two seems a sufficient summary of the desired outcome. Let’s call this summary outcome O."
Scott has argued for NGDP level rising about 5% a year as a good target for monetary policy. Now choosing that kind of target might be a complicated issue and maybe you would prefer a market procedure to also determine a good target, but that sounds more complicated to implement and might not lead to as good stability if market participants aren't quite sure of the optimal target. It also still relies on choosing a good weighted average of the variables you mention and that might not be easier than to determine whether a 5% NGDP growth is a good target (with level targetin). At the very least Scott's idea might be easier as a first step before moving to the kind of system you're proposing.
You could argue that the guardrails proposal is much less risky than things we've already experimented on such as QE and forward guidance (or the Eurozone experiment with negative IOR.) The proposal envisions that the plan start with very wide guardrails, which are not at all risky, and gradually narrow them over time as we learn more.
So yes, there are good arguments against jumping in feet first with a highly risky new regime in the US, but that's not what I'm proposing.
In theory I suspect you are correct. As a practical matter I think the problem with decision markets here is the general resistance plus the fact that you've got to introduce new financial instruments that must support huge markets (to allow the appropriate counters to manipulation). Moreover, to work correctly investors need to be sure that these new instruments really will pay off as stated rather than being cancelled early by the next fed chair or administration who thinks they are 'weird' or morally dubious.
The political problems implicit in making such a commitment make it seem desirable to find existing instruments to base policy on if one can which NGDP targeting seems like it aims to do in some sense (at least according to proponents).
This seems strictly correct to me. There may still be hidden assumptions embedded in "And it is widely agreed that the point of this policy is to promote long term growth, in part via moderating the business cycle. " though. Among Homo Hypocritus, not everything that is widely agreed to be the case is in fact the case, especially in domains touching upon money and power. If monetary policy has other motives, this might not serve them, which might be good or bad for us.
If the choices you posed to decisions markets were different variations of weights on inflation, RGDP, etc., it would give you its favored choice from among those options.
NGDP targeting implicitly puts equal weight on inflation and RGDP growth---two of the variables that you suggested might be used to measure social welfare. Is there some way that decision markets could determine whether something other than equal weights is optimal? Or whether employment is a better real variable than RGDP? I'm unclear on all this, which is why I view NGDP as a reasonable compromise. I don't see a method for determining a better policy objective, although I suspect that better objectives do exist.
To be clear I'm arguing against Hanson's prediction markets and in FAVOR of NGDP targeting. The prediction markets would work fine in theory but only if the central bank was willing to really commit to absolutely following the rules and honoring arbitrarily large purchases of these predictive securities.
In this case we were talking about something like Hanson's proposal which I understood to be something like the central bank's choice of interest on reserves, value of purchases/sales by FOMC etc.. was set in an algorithmic way from the market value of certain instruments ISSUED BY THE GOVERNMENT which pay out in certain ways based on future events.
If the government retains the ability not to set their policy in the algorithmic fashion the prices of these securities dictate you lose many benefits. If not and the total value is limited you have a risk of large investors influencing future policy by manipulating the market.
Also, I worry is that in practice the government would only be willing to issue some very limited value of these special instruments and has a substantial chance of paying them off early based on some other formula when they are deemed to weird.
Market manipulation doesn't seem to be a problem in other monetary regimes that peg a price, such a fixed exchange rates. In any case, under my "guardrails" approach the central bank would presumably ignore the attempt of one or two traders to manipulate the market
I'm not sure any of these apply to your proposal in particular as I was responding to Robin's suggestion but to elaborate.
Doesn't it need to be able to *in principle* support large trading to ensure that no one can manipulate it in their economic interest? As long as everyone is convinced that if there were an attempt to manipulate the prediction market there would be sufficient liquidity to frustrate this then that's fine but don't investors need to be convinced that counterfactually there won't be a problem. In particular, it seems like one would run into trouble if the instruments Robin is suggesting were given some prior limit on the total dollar value issued allowing a single actor to buy all the assets of a particular type up.
As far as fiat money debts you are certainly correct but for Robin's proposal to work the government has to issue a bunch of instruments which payoff based on the future state of various economic indicators. I'm not worried that the government will default but that investors will guess that they will be unwound prematurely and thus their market price won't truly reflect the actual prediction but some kind of hybrid of the actual prediction and how these instruments might be unwound early.
The definition of the outcome to be maximized would choose between those goals. But my understanding of this area is that this choice is not in fact that main area of disagreement in monetary policy.
You said yourself that promotion of long term growth is one of the goals.
Moderation of the business cycle isn't necessarily the same thing, or even compatible with that goal.
I'm no macroeconomist (either) but it seems plausible to me that vigorous downswings in the business cycle are needed to flush out inefficient practices in the economy, and so are *necessary* to long term growth.
I'd say choosing outcome weights is much easier, as we expect a broad flat peak, within which variations don't matter that much.
"That is, monetary policy consists of clear public and discrete choices, such as on short term interest rates. Call each discrete choice option C. And it is widely agreed that the point of this policy is to promote long term growth, in part via moderating the business cycle. So some weighted average of real growth, inflation, unemployment, and perhaps a few more after-the-fact business cycle indicators, over the next decade or two seems a sufficient summary of the desired outcome. Let’s call this summary outcome O."
Scott has argued for NGDP level rising about 5% a year as a good target for monetary policy. Now choosing that kind of target might be a complicated issue and maybe you would prefer a market procedure to also determine a good target, but that sounds more complicated to implement and might not lead to as good stability if market participants aren't quite sure of the optimal target. It also still relies on choosing a good weighted average of the variables you mention and that might not be easier than to determine whether a 5% NGDP growth is a good target (with level targetin). At the very least Scott's idea might be easier as a first step before moving to the kind of system you're proposing.
You could argue that the guardrails proposal is much less risky than things we've already experimented on such as QE and forward guidance (or the Eurozone experiment with negative IOR.) The proposal envisions that the plan start with very wide guardrails, which are not at all risky, and gradually narrow them over time as we learn more.
So yes, there are good arguments against jumping in feet first with a highly risky new regime in the US, but that's not what I'm proposing.
"And the federal government in the US does not default on fiat money debts, so I doubt investors would be concerned about that hypothetical."
With all due respect, do we really want to try this experiment on one of the globe's largest and most important economies first?
Under my "guardrails" proposal the market does not have to be "huge", indeed there need not be any trading at all if the Fed is on track.
And the federal government in the US does not default on fiat money debts, so I doubt investors would be concerned about that hypothetical.
In theory I suspect you are correct. As a practical matter I think the problem with decision markets here is the general resistance plus the fact that you've got to introduce new financial instruments that must support huge markets (to allow the appropriate counters to manipulation). Moreover, to work correctly investors need to be sure that these new instruments really will pay off as stated rather than being cancelled early by the next fed chair or administration who thinks they are 'weird' or morally dubious.
The political problems implicit in making such a commitment make it seem desirable to find existing instruments to base policy on if one can which NGDP targeting seems like it aims to do in some sense (at least according to proponents).
Yes, I agree, its a problem here if monetary policy actually has other goals besides moderating the business cycle.
This seems strictly correct to me. There may still be hidden assumptions embedded in "And it is widely agreed that the point of this policy is to promote long term growth, in part via moderating the business cycle. " though. Among Homo Hypocritus, not everything that is widely agreed to be the case is in fact the case, especially in domains touching upon money and power. If monetary policy has other motives, this might not serve them, which might be good or bad for us.
If the choices you posed to decisions markets were different variations of weights on inflation, RGDP, etc., it would give you its favored choice from among those options.
I'm pretty sure Scott has pitched his proposals to any who he could get to listen.
NGDP targeting implicitly puts equal weight on inflation and RGDP growth---two of the variables that you suggested might be used to measure social welfare. Is there some way that decision markets could determine whether something other than equal weights is optimal? Or whether employment is a better real variable than RGDP? I'm unclear on all this, which is why I view NGDP as a reasonable compromise. I don't see a method for determining a better policy objective, although I suspect that better objectives do exist.