It is a cliche (among economists) to say that expectations matter. It is even more of a cliche to say that expectations matter for asset prices. But how much do expectations matter? I'm going to do a quick and dirty back of the envelope calculation to show that expectations matter a lot. And that the lower are interest rates, the more expectations matter.
Then I'm going to remind people that money is an asset. Then I'm going to remind people that recessions are always and everywhere a monetary phenomenon, and that recessions are an excess demand for money. So expectations of future monetary policy matter a lot for recessions. Especially when interest rates are low.
Take an infinitely-lived bond that pays a coupon of $1 per year. If the market rate of interest is 5% per year, the market price of that bond will be $20. Suppose the current annual coupon payment is lost, due to an accident that people expect will never be repeated. The value of the bond drops to $19. Only 5% of the value of the bond depends on what happens this year. The remaining 95% of the value of the bond depends on what is expected to happen in future years. (The proportion would rise to over 99% if coupons were paid monthly and we were talking about future months.)
Take the same bond, but now assume the market rate of interest is 1% per year. The market price of the bond will be $100. The loss of the current year's coupon, if expectations of future coupons are unaffected, will drop the price of the bond to $99. Now 99% of the value of the bond depends on expectations of future years.
Expectations matter a lot for the demand for long-lived assets, and hence for the price of long-lived assets. The lower the rate of interest, the more expectations matter. What happens this year doesn't matter very much at all (unless it affects expectations of future years), and it matters even less as the rate of interest gets lower.
Money is an asset. It's also a very long-lived asset (unless inflation is high). The demand for money, and the equilibrium price of money, depend, like all assets, much more on what is expected to happen in future years than on what happens this year. That is even more true when interest rates are very low. Interest rates are very low now.
Recessions are always and everywhere a monetary (medium of exchange) phenomenon. Recessions are an excess demand for money (the medium of exchange). The demand for money is the demand for an asset. Since the demand for money, like the demand for all assets, depends very much on expectations, especially when interest rates are low, recessions depend very much on expectations too, especially when interest rates are low.
This is for Chris Dillow, who says "@ Nick – you've raised one of my beefs with NGDP targets – that relying upon expectations to do work is to rely upon a weak lever." (with a big HT to Left Outside who also has a good example to show how monetary expectations matter). (And it's also for the people of the concrete steppes).
Worthwhile Canadian Initiative 
Nick,
yes I would see the tax/subsidy as being being used purely to adjust the money supply. The subsidy would be funded by newly created money and the tax would lead to money being destroyed.
I need to research a bit to understand the “dead-weight loss” issue – but if the policy was applied to all final transactions (including perhaps transactions involving labor purchases) its not obvious what distortions this would cause.
I like this approach better than increasing the money supply via asset purchases or interest rate policy because it neither distorts asset prices nor leads to an increase in bank lending – both of which appear to carry some risk.
There are pros and cons to monetary policy as well.
Look, it sounds like our macroeconomic goals are similar, broadly speaking (economy near full capacity and price stability). And there is a lot of similarity between your comment that there is an excess demand for financial assets and the way post-keynesians say it. The big disagreement is how do we get there, and many of the post-keynesians don’t really buy the monetary policy solution. But I am all for trying it and being proved wrong.
OK, Nick, you can stick with a denial of the second law of thermodynamics if you want. But decelerating some large process in a complex dynamical system is not the same thing as decelerating the causal factor that originally accelerated it.
I see nothing wrong with being a “one-way” Keynesian, since dynamic processes have a preferred time direction and a preferred spatio-temporal direction of causal propagation.
If the Kingdom of Horsemen is impoverished, and the husbandmen who breed and raise the horses that supply the cavalier armies are do not have enough feed to breed and raise horses, then you address the problem by getting them more feed. If you overshoot and there are now wild horses stampeding the fields and rogue horsemen pillaging the peasant, you don’t address the problem by taxing feed from the husbandmen. You have to round up the extra horses and the pillagers, while adjusting the rate of horse breeding to get it just right. Stimulating the economy at the ground level of consumers and small business may lead eventually to a growth of prosperity that develops into a flow of wealth toward the top of the economy, a glut of luxury goods and speculative lending and bubbles among those with a lot of surplus capital. At that point, to slow things down, you have to go after the surplus, not the ground level.
This idea that the economy has some sort of karmic balance that must be preserved, and that Keynesian demand stimulus in one time has to be balanced off by an opposite demand de-stimulus at a later time, and that every deficit now must be matched with a surplus later, is encountered in a lot of the things Krugman says – and it drives me crazy.
Re “This is part of a general rule: fiscal policy has lots of microeconomic objectives too. If fiscal policy is doing those micro jobs well, it can’t also do the macro jobs well.”
Why not ? If you use tax/subsidies to optimize AD there is nothing to stop you using other fiscal policies to accomplish other objectives. You could for example have a sales subsidy of 3% to increase NGDP, while at the same time taxing and spending in other ways for other things with no obvious conflict, unless I have missed something.
Of course if merely stating that you will use fiscal policy to stabilize AD is sufficient to prevent it being needed then other fiscal policy can continue unaffected.
Ron: this Wikipedia page is an OK start, though it’s not ideal. The area of the welfare loss triangle increases with the square of the tax rate, so ideally you want to have that tax rate constant over time.
Any fiscal policy will have the same issues. The optimal setting from the microeconomic perspective won’t be the same as what’s optimal from the macroeconomic perspective, and it can’t be in two places at once. Lump sum taxes would seem to be an exception, since they don’t cause deadweight cost triangles, but they have their own distributional consequences.
wh10 @01.50pm. Agreed. And in practice, in the Canadian context, I supported the use of fiscal alongside monetary, provided we tightened fiscal policy up again as soon as we were able (which seems to be happening). Because, even though I think that monetary policy, done properly, can do the job alone, I didn’t want to take the risk: that it would be done properly; that I might be wrong. Belt and braces; give the bear both barrels at once; choose your metaphor. I would be less sanguine about fiscal policy in the US, given the politics and the higher Federal debt/GDP ratio and prior deficit. Canada had a surplus going into the recession, so taxes would need to be cut and/or spending raised at some future point in any case.
“Expectations matter a lot for the demand for long-lived assets, and hence for the price of long-lived assets.”
You mean, more so for things like houses, factories, cars as compared with paper plates, plastic spoons, frozen foods, children’s clothes, etc?
In other words, expectations matter more DEPENDING ON THE LENGTH OF THE STRUCTURE OF PRODUCTION INVOLVED.
Wonder it that might have anything to do with the cycle of boom and busts and all that goes with it.
Hmmmm.
Ya think?
Nick,
Thanks for the wikipedia reference – I did learn from it but I still don’t really see how it applies here.
the article provides an example “If the price of a glass of beer is $3.00 and the price of a glass of wine is $3.00, a consumer might prefer to drink beer. If the government decides to levy a beer tax of $3.00 per glass, the consumer might prefer to drink wine. The excess burden of taxation is the loss of utility to the consumer for drinking wine instead of beer, since everything else remains unchanged.” But if the tax was applied equally to beer and wine then where is the loss of utility ?
I can see that if we were in equilibrium (relative prices including the price of money correctly aligned to reflect underlying preferences) then even a general sales tax/subsidy would distort and move the economy away from equilibrium. However if the demand for money has increased but supplier are unable to quickly reduce prices to reflect this then an increase in the money supply would be beneficial. I believe a sales subsidy would be the less distorting than any other means to achieve this increase because it would very directly get “marginal buyers” back into the market and lead to new equilibrium that would be very similar to what would have been achieved had prices been able to fall immediately following the change in the demand for money.
Greg: Thank God! Someone who doesn’t need to be convinced that expectations might matter, and have this explained to him!
Ron: Yep, if truly all final goods were taxed at the same rate, this wouldn’t be a problem. But we don’t tax: leisure; the joys of a pleasant job; home production; etc. And then there’s tax evasion.
A subsidy is exactly the same as a negative tax. It has a welfare-loss triangle just like the tax triangle, only on the other side of the supply and demand curves.
“But we don’t tax: leisure; the joys of a pleasant job; home production; etc. And then there’s tax evasion.”
Good point – there would indeed be some market distortions but would they not be less than those from monetary policy that would either(if the policy is carried out via lowering interest rates) increase banks lending and cause the kind of problems the Austrians and post-Keynsians are always highlighting, or (if the policy is carried out via asset purchases) cause asset-price inflation , and changes to the structure of demand to reflect the preferences of those whose assets were bought ?
Ron: suppose i were in charge of monetary policy and you were in charge of fiscal. I would be keeping NGDP on track regardless of what you did. If at some point you said “Hmmm, I think interest rates are too low and I want to loosen fiscal policy to increase interest rates” I would shrug my shoulders and say “Do whatever you want to do, for whatever reasons you want, but just don’t think you will shift the AD curve by doing that, because I will make sure it doesn’t shift”.
Yes, at this point, you might say the argument is becoming semantic. Are monetary and fiscal policy doing different jobs, or are they sharing the same job? Depends how you define “monetary policy”. Does it mean setting NGDP, or does it mean setting a rate of interest?
wh10,
“Totally lost you as usual Peden, and as usual, you twist and read my words with inaccurate, loaded assumptions.”
You say that taxes should be used to regulate AD. The original context was talking about state handouts to boost AD, that is Mike S’s comment “Instead of giving some mental telepathy to get spending going by increasing lending, why not just give actual poor people money?”
Now, if you disagreed with what Mike S posted and if you want to do it entirely through tax cuts/rises, why didn’t you say so?
Nick Rowe,
I was actually thinking of something even more mindlessly automatic: the government spends as before, but funds its expenditure differently.
Of course, in practice that’s quite unrealistic: governments facing recessions have tremendous pressure to win popularity, and a very interventionist fiscal policy is a way to do it. Hence things like “cash for clunkers”, Green Jobs and loan guarantees for small businesses. Whatever the fiscal multiplier of such spending, it has a vastly higher political multiplier (the important multiplier for a democratic government) than simply switching from selling news bonds to getting loans from banks.
You’d obviously have to make sure that monetary policy and fiscal policy were aligned – as the govt ultimately controls both then in theory that should be achievable.
I’m assuming here that NGDP targeting is the desired goal and I’m arguing the view that fiscal policy (in this case defined as stabalizing NGDP via a variable tax or subsidy on all transactions) seems more direct and less distortionary than monetary policy (defined as attempts to stabilize NGDP by either setting interest rates or “quantitative easing”-type asset purchases).
In my model there would in fact be no monetary policy as fiscal policy would do all the work, and fiscal policy would consist of just calculating the level of tax/subsidy needed to achieve an optimum NGDP target. Any other fiscal policy the govt chose to implement would be to achieve non-economic social and/or political goals and be outside the scope of the GDP-stabilizing policy envisaged here. These additional polices would ideally have to operate on a “balanced budget” basis.
Ron: “In my model there would in fact be no monetary policy as fiscal policy would do all the work,…”
What does “no monetary policy” mean? What does “doing nothing” mean?
This is getting a bit zen.
I defined monetary policy as ‘attempts to stabilize NGDP by either setting interest rates or “quantitative easing”-type asset purchases’. I can see you could just as easily define it more broadly as ‘polices that alter the money supply and therefore affect NGDP”.
However that seems irrelevant. I’m really just trying to compare different techniques for stabilizing NGDP in terms of effectiveness and non-distortionary effects.
If expectation are the key to stabilizing he demand for money then what matters is not if a policy will actually work if tested but whether it will be perceived by the public as going to work. If the public believes it will work then it will never actually need to be tested.
Perhaps then economists, rather than debating which policies are optimal, should actually be conducting market-research on which policies the public would most likely believe in ?
OK. You are the government. You control lots of different fiscal levers, which affect lots of different things in lots of different ways. You care about: NGDP; and a lot of other things as well.
I am your humble servant monetary policy. I only control one lever (ultimately, the size of my balance sheet). So I can only target one thing. One restriction is that the thing I am targeting must have $ in the units. I note that NGDP has $ in the units. What would you like me to target? Would you like me to target NGDP, so you can delegate that responsibility to me, freeing you up to move your levers to try to attain as many of your other objectives as possible? You will never have as many levers as you want to have, so it would be unwise for you not to delegate one of your objectives to me, or tell me to “do nothing”.
Would it be possible for a central bank not to target some macroeconomic quantity, but simply set a ceiling of some kind and otherwise get out of the way of the rest of the government? Couldn’t they just say: “We’re going to stay out of the way of the politicians, unless inflation appears poised to exceeds 5% (e.g.), in which case we will then raise interest rates.”
For example, if the fire department says “No more than 362 people in this club at any time” we wouldn’t say that the fire department was “targeting” an attendance of 362 people for the club’s nightly attendance. My beef is with the idea that the central bank can actually deliver the kinds of economic performance we need.
Peden, I meant some combination of fiscal policy more generally, but I was responding to your comment that one day you might have to take “money back.” That sounded like you were referring to taxing. Of course, you could alternatively withdraw govt spending.
wh10,
I see. In practice, of course, I expect it to generally be easier to raise taxes that cut spending, although in some political situations (like the US recently) it seems that cutting taxes can become just as difficult as cutting spending.
Dan: so you are saying I can do whatever I feel like, as long as inflation is less than 5%? “Cool!” replies the crazed deflationista, “I think I will tighten monetary policy and aim for a repeat of the Great Depression!”
Nick, I think this is the best example of the power of expectations: [link embedded here NR]
A clear target and a credible commitment, and the exchange rate immediately changed to reflect it. I don’t think there’s any other way to interpret this than as an expectation driven event.
Dan: so you are saying I can do whatever I feel like, as long as inflation is less than 5%? “Cool!” replies the crazed deflationista, “I think I will tighten monetary policy and aim for a repeat of the Great Depression!”
How did you get that? I suggested that the CB’s policy should stay out of things unless inflation gets too high. Not that it should do whatever it likes. Saying that it will accept whatever inflation rate happens to occur as long as it is not above certain number is not the same thing as targeting a specific rate of inflation or spending. And it is certainly not the same thing as interfering to cause deflation.
I thought you guys had this new argument to the effect that fiscal policy wouldn’t work because people would fear that the Fed would step in and counteract the effects of fiscal policy. So their policy should just be not to do that, unless inflation gets really out of control.
What I want is for people to get out of the mindset of thinking that the central bank can make lots of good things happen. They can’t. The bank can instead just determine stay out of the way as the folks with the real ability to make things happen get to work.
“You control lots of different fiscal levers, which affect lots of different things in lots of different ways. You care about: NGDP; and a lot of other things as well.”
What if you think that if you have stable NGDP then the market can sort out everything else for itself – you wouldn’t have any other policy objectives and then you could choose the best policy to achieve that one goal.
A clear target and a credible commitment, and the exchange rate immediately changed to reflect it. I don’t think there’s any other way to interpret this than as an expectation driven event.
That’s like the Fed Funds rate curiouseconomist. If the Fed makes a commitment to control a rate in a market whose rates they actually do control, then people in that market will jump to that rate right away. Same with foreign exchange. With a rate like the total volume of nominal spending in the entire economy, not so much. It’s like expecting them to control the tides.
curiouseconomist: I think that’s a good example. But I was just literally 5 minutes ago reading your excellent long comment on Chris Dillow’s blog. And I think the evidence you marshal there is more important. And I was wondering who you were, and then returned here to find your comment!
And I was just wondering whether it would be possible to repeat what you did there, only using NGDP forecasts instead of RGDP forecasts, and comparing countries? If there are RGDP forecasts and inflation forecasts, it should be possible to construct NGDP forecasts. Then repeat what you did in your comment, to argue it should have been quite possible for central banks to have loosened monetary policy to prevent forecasted NGDP falling like that?
Dan: “How did you get that? I suggested that the CB’s policy should stay out of things unless inflation gets too high. Not that it should do whatever it likes.”
OK. I knew (of course) that you thought it’s not OK for the central bank to do whatever it likes, as long as inflation is less than 5%. But what does “staying out of things” (or “stay out of the way”) mean? Since you are not indifferent among all the things the central bank might conceivably do, is there one out of that large set of things you would most prefer it to do?
“If the Fed makes a commitment to control a rate in a market whose rates they actually do control, then people in that market will jump to that rate right away. Same with foreign exchange. With a rate like the total volume of nominal spending in the entire economy, not so much. It’s like expecting them to control the tides.”
If Chuck Norris works for the overnight rate, and the exchange rate, and if nominal spending depends on interest rates and exchange rates, and on expected future nominal spending, which in turn depends on expected future interest rates and expected future exchange rates, then why can’t Chuck Norris work for NGDP too?
Ron: “What if you think that if you have stable NGDP then the market can sort out everything else for itself – you wouldn’t have any other policy objectives and then you could choose the best policy to achieve that one goal.”
If I really didn’t care about taxes and government spending, except insofar as they influenced NGDP, then you would be right. I would have 3 policy levers (G,T, and M) and only one target (NGDP) so I could ignore any 2 of those 3 levers. But I do care about G and T, quite apart from their effects on NGDP.
@Nick
If Chuck Norris works for the overnight rate, and the exchange rate, and if nominal spending depends on interest rates and exchange rates, and on expected future nominal spending, which in turn depends on expected future interest rates and expected future exchange rates, then why can’t Chuck Norris work for NGDP too?
I think because nominal spending depends on a lot more than just interest rates and exchange rates. If Chuck Norris says “I intend now to use my patented spin kick maneuver to break 15 bones in Dr. Chan’s body. Take bets now on how many broken bones Dr. Chan will have when I’m done,” then a lot of people will bet somewhere around 15. But Chuck Norris might say, “I expect 75,000 additional broken bones in America between now and the end of the month. Place your bets.” I don’t think this will have much impact. And you can count on one hand the people who will go out and buy broken bone insurance in response to Chuck’s statements.
Now I admit it’s a little iffier to try to predict the effect of a central banker’s statements. Since fewer people have heard of Ben Bernanke than have heard of Chuck Norris, and since even experts seem to be in some doubt about what it is exactly Bernanke does, then it is harder to say what number of people would modify their behavior in response to some statement he made.
Dan: “I think because nominal spending depends on a lot more than just interest rates and exchange rates.”
Agreed. Monetary policy will never be able to control actual NGDP precisely, because it depends on a lot of things. It’s going to be a question of how precisely, and how quickly it can react to new information on those other things, and what information it has. Just like trying to keep the speed of a car constant when that speed depends on a lot of things, not just the gas pedal, especially if there’s a lag in the speedometer.
But controlling expected future NGDP is going to be easier. That’s because other people don’t know what actual future NGDP will be either. And presumably the central bank knows all the information they know, and can make as good a guess as they can, and can guess what they can guess. Plus, it can even ask people what they expect future NGDP to be. Or watch a futures market to learn what people expect in real time (a la Scott Sumner).
It’s the same as targeting inflation. Central banks can’t do it exactly, but they can do it roughly. Targeting expected future NGDP should be easier.
@nick
“Suppose there were a bit of paper, which gave the owner:
1. The (exclusive) right to live in a particular house for one day,
2. The right to sell that bit of paper to whomever,
3. The right to exchange that bit of paper for a second bit of paper dated the next day.
ownership of that bit of paper constitutes ownership of that particular house. The bit of paper looks like a one-day asset, but given 3 it lasts as long as the house.”
Congratulations. You just reinvented the repo, which you seem to think is money. I’ll accept that.
Peter N: No, I think I just reinvented the title deed to a house! But I don’t think it’s money. Individual houses don’t work very well as money. It costs so much to do the home inspection, pay the realtor’s fees, the lawyer’s fees, etc.
Individual houses don’t work very well as money.
true, but mortgage backed securities might.
@Nick
“that would work (if it’s money-financed), except: a VAT also has microeconomic consequences. It causes a deadweight loss (roughly) proportional to the square of the tax rate, so that having a varying tax rate causes higher average deadweight costs than having a constant tax rate that collects the same average revenue. Also, it has distributional consequences too, that may or may not be desired.”
You have to be careful here. The micro level deadweight loss is embedded in an economy. Income tax in the US not only has the deadweight loss of a tax, but also has a whole set of extra losses that are systemic. It’s the systemic losses you want to minimize, not the immediate losses from ideal collection.
Suppose at the micro level you include costs of collection for both supplier and customer. The actual tax loss at least generates revenue. If this loss is deadweight, the other losses are deader than dead. These deaderweight losses plague modern tax systems.
I suppose one may care about G and T because one wants to redistribute income or change the supply and demand curves of certain industries (for whatever reason).
You would do this via laws , taxes and subsidies.
The combined effect of these laws, taxes and subsidies would give you either a balanced budget, a deficit or a surplus. If it was a deficit you could fund this via either borrowing or printing money, if a surplus you could pay down govt debt or destroy money. Either option would have some effect on NGDP. If you were also targeting NGDP then you may also have to make further adjustments to achieve the target. One would then want to use the most efficient lever to achieve this.
One could define “fiscal policy” as those policies that aim for income redistribution or other kinds of market intervention and “monetary policy” as those policies that allow the NGDP target to be met. As long as the latter are neutral in respect to their effect on the former then one should just choose the most efficient lever. I have not seen any argument that dissuades me from my view that tax/subsidy (on all transactions) is more efficient than influencing the money supply via interest rates and/or asset purchases.
@Dan Kervick,
You wrote:
“I thought you guys had this new argument to the effect that fiscal policy wouldn’t work because people would fear that the Fed would step in and counteract the effects of fiscal policy.”
I’ve been staying out of this but that particular statement really irked me. There’s nothing new about this at all. This is pretty old and very conventional macro. I think it’s safe to say nearly every economist by graduate school has been exposed to a demonstration using the IS/LM model of the circumstances under which fiscal and monetary policy trade places in effectiveness (fixed versus flexible exchange rates).
Under a flexible exchange rate regime fiscal policy is generally considered impotent since it is subject to the reaction function of monetary policy. It’s only the fact that some economists think we are now in a liquidity trap that fiscal policy may have traction, but that is of course subject to enourmous disagreement.
P.S. There are other channels for the monetary transmission mechanism than the interest rate and exchange rate. In particular, if I remember correctly Mishkin’s intermediate textbook lists seven others: Tobin’s q Theory Channel, Wealth Effect Channel, Bank Lending Channel, Balance Sheet Channel, Cash Flow Channel, Unanticipated Price Channel and Household Liquidity Effects Channel. This of course implies that monetary policy’s ability to influence nominal spending has numerous pathways.
I simply don’t see a mechanism for the Fed to induce wage inflation without help from fiscal policy, or at least for fiscal policy to stop pulling in the wrong direction. Monetary policy is important and works well for some situations. Monetary policy works well to balance economic growth when the economy is within an area close to full employment and close to the inflation target.
Monetary policy is not very good for addressing speculative bubbles. Regulatory is the best and most appropriate tool for bubbles.
Fiscal policy is the best mechanism to intervene during periods of slack demand and labor underutilization whether from structural or cyclical reasons.
All 3 policies work best when they are pulling in the same direction. Monetary policy, no matter how good, cannot compensate for truly awful regulatory or fiscal policy.
You’re not really done with a definition of monetary policy until you say what the central bank can buy, how much of it and what you expect the effect to be.
This paper discusses channels, central banks and the business cycle. I thought it was pretty good. Much too good to try to summarize.
Click to access Ivanov-BESi-2009-Paper-proofread.pdf
Under a flexible exchange rate regime fiscal policy is generally considered impotent since it is subject to the reaction function of monetary policy. It’s only the fact that some economists think we are now in a liquidity trap that fiscal policy may have traction, but that is of course subject to enourmous disagreement.
Mark, is there a model that explains how monetary policy is also impotent because fiscal authorities will always act to cancel out the effects of monetary policy, too?
The Fed is a creature of the US Congress, created by an act of Congress, which has all of its powers by virtue of a delegation of powers constitutionally assigned to congress. Congress can claw back as much of that power as it wants at any time. They could even move the central banking function to the Treasury – or to a committee in Congress itself. Short of that, if a fiscally energetic government of the future is concerned that the monetary authorities are preparing to act in a way which will neutralize the economic impact of their polices, they can write and pass a simple law: “Don’t”.
It seems strange to regard these highly contingent institutional and political arrangements, and the policy fancies of the neoliberal era, as though they are some kind of hard-wired economic laws.
@Dan Kervick
“Short of that, if a fiscally energetic government of the future is concerned that the monetary authorities are preparing to act in a way which will neutralize the economic impact of their polices, they can write and pass a simple law: “Don’t”.”
Congress has intervened in monetary policy before. They did it in the 1930s.
Right Peter N. So many of these economic models seem to set up the world as a very simplified game with a very limited number of permitted moves – which just happen to be the only moves the model-builders want us to consider.
@Dan Kervick,
This is less a political matter than an economic matter. You may not like the fact that the earth revolves around the sun, but there it is, and no matter how many acts of congress you pass nothing will change that fact.
That same model shows that monetary policy is ineffective and fiscal plicy is effective with a fixed exchange rate regime. But with the collapse of Bretton Woods in 1971 I think it may be too late to put that genie back in the bottle. (And, prior to the 1951 Accord, the Fed was largely subservient to the Treasury.)
P.S. I sometimes think that the main problem with Post Keynesians and MMTers is that they don’t acknowledge the change in exchange rate regimes. Much of what they say might have been appropriate 40 or more years ago, but now it is largely irrelevant.
Mark, the central bank reaction function is derived from rules that are expressions of policy preferences. How can you regard these constructions as in any way to the revolution of the earth around the sun?
All you are telling me when you say that a central bank behaves in a way represented by a reaction function according to which it will offset fiscal expansion by raising interest rates is that central banks in the past have followed a Taylor rule or something like it. But aren’t NGDP targeters among others explicitly proposing that the central bank not follow the customary rules? The reaction function isn’t a law of nature – it is just a mathematized coding of recent standard operating procedure. If the procedure changes, then the mathematical models that presuppose it collapse.
On the issue of MMT and exchange rates, Warren Mosler’s demand-side framework for full employment and price stability is built explicitly on a floating fx model:
http://www.epicoalition.org/docs/exchange_rate_policy_and_full_em.htm
Dan,
The link you provided points out some of the potential hazards of maintaining a fixed exchange rate in the post Bretton Woods environment (and a good argument can be made that thanks to the Triffen Dilemma that system was itself unsustainable). It also seems to suggest that fiscal stimulus is compatible with a floating exchange rate regime. But much of what it says in that regard is counterintuitive and contradicts simple conventional macroeconomic models.
Let me explain why in the context of the basic IS/LM model.
Let’s assume we have flexible exchange rates, perfect capital mobility, and we attempt a fiscal stimulus. The increase in Government spending means there’s an increase in total expenditures, therefore the IS curve shifts to the right and output and real interest rates increase. Because of the increased inerest rates there are capital inflows as foreign investors seek to purchase higher returning domestic assets. These investors are exchanging their currency for the more desirable domestic currency. This increased demand for the domstic currency causes the value of the domestic currency to rise on foreign exchange markets (i.e. it appreciates, e decreases). As e decreases, net exports decrease as domestic goods become relatively more expensive on international markets. As net exports decreases, total expenditures fall and the IS curve shifts to the left. The exchange rate will continue to appreciate, and the IS curve will continue to shift to the left until the capital inflow is halted (i.e., until the domestic interest rate equals the foreign interest rate). The new equilibrium is at the same level of output as the initial level.
The only alternative is for the money supply to be increased simultaneously in such a way that the exchange rate is maintained. And if you do that what was the point of doing the fiscal stimulus in the first place?
You’re much better off using monetary policy to stabilize short run aggregate demand so that you can free up fiscal policy to be focused on long run aggregate supply issues.
Nick
“Exactly. It is very easy to advocate a fiscal policy that increases AD. But unless someone who argues for a particular fiscal policy is willing to argue for the exact same policy in reverse when we need to reduce AD, you know they are just “One-way Keynesians”, and using AD arguments insincerely to hide some other agenda.
Similarly, some people argue we should halve debts to increase AD. I haven’t heard any of them say we should double debts when we want to reduce AD.”
These are both cheap shots. When you want to cool down the economy, there are fewer poor people to give money to. You don’t have to increase debts, because higher interest rates do the same to effective demand for you. The liquidity trap, really is a special case.
“….there are fewer poor people to give money to”
Hence my point in another post about the possibility of having bigger automatic stabilisers.
Mark,
isn’t your argument with capital inflows self contradictory. If capital inflows completely offset fiscal policy, and capitalists know that, then the capital inflows won’t happen.
This sounds like a wonderful paradox – fiscal policy is ineffective, because it is effective.
But I think the key point is here:
“The only alternative is for the money supply to be increased simultaneously in such a way that the exchange rate is maintained. And if you do that what was the point of doing the fiscal stimulus in the first place? ”
Because it is a DIRECTER way of doing what we actually want to do (give the unemployed jobs).