« Are we hard-wired for capitalism? | Main | Only one degree of freedom per Fixed Announcement Date »


Feed You can follow this conversation by subscribing to the comment feed for this post.

The fundamental question is, if you believe in the equation of exchange, is the belief based upon M being base money and PQ being NGDP, or alternatively that PQ is domestic rather that international in character for an open economy. To my knowledge, CBs that believe in the equation of exchange believe that M is broad-money not base money.

Where broad money is any money substitute, e.g. Tbills

OK, in a system where the central bank targets inflation and the political class is agreed to keep the target rate at 2%, doesn't this argue against money-financed deficits since the "price" of such a deficit is inflation?

In my observations the generation that is in power now (Carney may be exception) came of age when interest rates were climbing and so was inflation. They remember 18% rates as nearly criminal. That was their defining moment. Now that we are 180 degrees from that at the opposite end of the spectrum, they can't recognize that they have to turn their received wisdom on its head because inflation isn't the problem anymore.

Jon: but the key question for this purpose is this: what is the demand function for base money? Theory says it must be proportional to P. Theory and evidence says it is also probably roughly proportional to real income. So assuming it is proportional to NGDP is probably roughly correct. Of course, it will depend on other things too.

Determinant: a country that's facing too low aggregate demand, and the threat of deflation, *wants* a policy that will create inflationary forces. If you are too cold, you want to turn the heat up. You don't reply "But that will make you too hot!"

Of course, Canada right now seems (touch wood) to be recovering OK. I wouldn't recommend any more expansionary a monetary and fiscal policy than we have right now. But for much of the world (US, Europe) that's not so.

I agree about aggregate demand and deflation, I wasn't arguing that, but the guard that is in power now seems to be to be full of inflation-phobes.

What I am arguing is that the present crew DOES cry "but it will get to hot!" because that is their experience speaking. It's not rational and its contrary to theory, that's my point.

I am explaining why they are trigger-shy.

The question may be how much this a demand shortfall and how much a supply shortfall, the supply being oil. If stronger growth led directly to $150 oil, it probably couldn't be sustained anyway, and oil is too inelastic in the short run to accommodate even average growth before banging into the ceiling.

If I read you right Nick, you're saying that the folks in charge should print money but either they don't realize this because they are confused, or they are incapable of doing so even though they want to.

In my opinion, if they wanted to, they easily could, but they don't want to because creditors run the show, and creditors don't like money being printed.

The ruling class runs the show. They only care about unemployment when it's too low, since that tends to increase the wage share and reduce the capitalist share of income. They always care about inflation because it reduces the real value of their fortune. So they promote generalized phobia about inflation combined with all sorts of pseudo-rationales for tolerating--indeed, engineering--large-scale long-term unemployment.

Thatcher was the trailblazer for this sort of thing.

Now it's practically second nature among large swathes of economists, the financial world, and the business press. But of course it's basically just class warfare dressed up to look intellectually respectable. The way this is done is to say this is so comlex that olicymakers are genuinely unsure what the right policy as it's not obvious what that policy is. Funny, then, that the 'right' policy tends to coincide with promoting the interests of the ruling class. And over the past three decades, that's been the most obvious thing on the planet.

Commenter: "It's just so simple and obvious; just finance your fiscal programs with new money. Money doesn't pay interest, and doesn't have to be paid back (at least, irredeemable base money doesn't), so isn't really debt, in any sense that matters."

Nick Rowe: "He's basically right. (And the Neo-Chartalist/MMTers who make basically the same point are basically right too.)"

And Henry Ford and Thomas Edison were right, too, eh?

Nick Rowe: "In order to explain why it isn't happening, I've got to make it less simple and obvious; I've got to make it complicated and obscure. Because that's how it will appear to the policy-makers; and explains why they aren't doing it."

Sorry to head you off at the pass, Nick, but this is not a case of Burridan's Ass. (Even if the policy makers are asses. ;)) Should we finance the deficit this way or that way? Gee, I dunno. Oh! I've got it! Let's not run a deficit at all! Of course! No problem!

It is a case of fear mongering. Whether the fear mongers believe their own propaganda is another question. Look over the last several months. The propaganda started with the apparently innocent proposition that sometime in the future deficits would have to be reduced, interest rates raised, quantitative easing reversed, or inflation would result. Bernanke resisted targeting inflation at a higher rate because the Fed would lose "credibility". People were asking whether QE was effective, or simply "pushing on a string". Then you got things like the President of the United States saying that we were running out of money. Absurd, but who said so? Then you got dire warnings that the U. S. and U. K. might default or go bankrupt. Even more absurd, but who said so? (Yeah, Galbraith and Auerback and others, but their voices are being drowned out.) Even the latest "Economist" warns of a "day of reckoning" for the U. S., as though the people who believe that a fiat currency is a Ponzi scheme were right. All of that cannot be explained by, Gee I dunno.

There are echoes from the past. Andrew Jackson paid off the national debt in the U. S., which probably precipitated the depression of 1837. Jackson attributed that to paper money. Bill Clinton ran surpluses, and, to judge by his recent remarks, would also like to pay off the national debt. Clinton has also said-- I saw this a while back on CSPAN, so I do not remember exactly -- that he was talking with Rubin about something he wanted to do, and Rubin told him that the bond market would not allow it. (You can imagine how Jackson would have replied!) There is a powerful idea that modern nations, even the most powerful, are hostage to the bond market. Even that they should be, because politicians are irresponsible and, in a democracy, will just spend, spend, spend unless they are subjected to the discipline of bond traders and bankers. The fact that the traders and bankers might be more interested in turning a profit than in the public good is glossed over. After all, The Market knows best.

Thomas Edison: "It is a terrible situation when the Government, to increase the national wealth, must go into debt and submit to ruinous interest charges at the hands of men who control the fictitious values of gold."

Nick Rowe: "It's not obvious to economists. More importantly, it's not obvious to policy-makers who must choose whether or not to run a deficit. And much more importantly still, because expectations matter for aggregate demand, and matter more than anything else, it is not obvious for ordinary people who make spending decisions."

Expectations matter for aggregate demand, unless you have no choice. That is why it is important to get money into the hands of people who will actually spend it. Like the poor, like people who have been unemployed for a long time, like state governments who have obligations to spend, but lack the funds.

Determinant: OK. I misunderstood you.

Lord: I don't really buy the oil argument. We can change the mix of goods and services produced to use less oil per unit of GDP or employment; the fall in GDP was associated with disinflation, which matches a fall in AD, rather than inflation, which matches a fall in AS; the timing doesn't work, because oil prices fell as we went into the recession. It all looks rather different from 1973.

Declan: given the way central banks' operations are framed today, they can't even think in those terms. They sort of giggle when you say "printing money", then patiently explain the operating procedures of a modern central bank (reconstituting their own social construction of reality).

stunney: there are several implicit assumptions underlying your argument. One is that the "ruling class" is a creditor of nominal assets, and the non-ruling class the debtor. My hunch would be the opposite. Politically powerful people tend to hold equities, not bonds or GICs. My guess is that most nominal bonds are ultimately held by pension plans.

Anyone know for sure?

Min: You lost me there.

If any part of the deficit is financed by creating more money, then the fedfunds falls to zero and remains at zero until the federal reserve sells enough bonds to sterilize the excess reserves. Effectively 100% of any money creation will excess reserves, as the reserves *needed* by banks are determined by the technological methods of payment settlement, and are in any case tiny. 50 billion dollars should be enough to push fedfunds to zero under pretty much any situation, and to keep it there for years.

So as long as central banks would like to have the freedom to set the fedfunds rate to a positive value, the government does not face a choice of financing deficits with money or bonds, they must finance with bonds.

For the same reason, discussion of whether the increase in money is permanent or not reduce to speculations of when the central bank will raise rates. Prior to doing so, the central bank will need to reverse the previous money creation.

So that *should* clear up the confusion as to whether the deficit will be funded by bond sales or money creation. It will be funded by bond sales unless we commit to permanent zero call money rates and vacate most of the central bank staff that are interested in monetary policy. On the other hand, that would free up office space for regulating banks.

However, what is more confusing for me is this entire debate.

I can see that some people (e.g. adherents of autarky) would not believe that a fiscal deficit is stimulative. We should be able to dispense with those people. To the best of my knowledge, 90% of those advocating for the ineffectuality of fiscal policy -- and this includes the IMF and other organizations -- are all on the government payroll. Simply by cutting off their salaries, as per their advice, we can reduce the debate to those who believe that fiscal spending is stimulative. We can require that each person swear an oath "I believe that my salary is a net benefit to the economy that would not be provided more efficiently by the private sector, and this is still the case if the government borrowed to pay my salary". Whoever does not swear the oath can be moved to the private sector.

Those that are left will be the ones who believe deficits are stimulative, but have debates as to how to spend the money. Do we cut taxes, create a government work program to directly employ the unemployed, repair sewers, provide broadband for everyone, or open the phone book and randomly dial businesses, offering to boost their sales by 30% for the next quarter. There are many options, and as a democracy we can debate what would be the best of them, and economists can contribute by discussing which programs would be more stimulative. I even think that there is hard data on this.

For some reason, that was not listed as a source of confusion. The financing of the deficit, which is already pre-determined, was listed as the confusion.

RSJ: "If any part of the deficit is financed by creating more money, then the fedfunds falls to zero and remains at zero until the federal reserve sells enough bonds to sterilize the excess reserves."

I don't think that's correct. That would mean that base money in Canada was constant until just over a year ago, when the Bank of Canada cut the overnight rate to 0.25%. But it wasn't. The Bank of Canada's balance sheet was steadily expanding (mostly currency), at around 5% per year, IIRC. And I can't see any reason theoretically to think it might be true, either.

You are right! I should have said "beyond the usual amounts by which the CB increases the money supply in order to maintain its CB target".

I think in both countries this would grow roughly with GDP, for a seignorage of about 0.4% of GDP/year for the U.S. Not sure about Canada.

But we've already added a lot to CB balance sheet, so it will take a long time to exhaust a doubling of the base and allow for positive rates. In any case, this cannot fund deficit spending for any material amount unless there is a commitment to stop using monetary policy to set call money rates.

Well, I guess we'll just have to agree to disagree, Nick. I just don't think central bankers are that obtuse. As Scott Sumner points out, Bernanke in particular is on the record as understanding quite well that he could print money if he wanted to, and that that was the right thing to do in these circumstances - and I'm sure that folks like Carney around the world are bright enough to follow along if he was to explain it to them (maybe he could use diagrams?).

Maybe I'm smarter than the collective wisdom and knowledge of the world's central bankers on a topic that is their life's work and on which they have access to all the best minds on the planet, and where I have only a passing interest, with most of my information picked up from blogs, but that seems about as plausible to me as the notion that I have a better grasp on global climate trends than all the world experts in that field do.

"And I can't see any reason theoretically to think it might be true, either."

OK, when the central bank increases the supply of currency, this does not cause the amount of deposits to increase, but rather the amount of cash backing those deposits increases.

Banks only need a small amount of cash -- vault cash -- to meet withdrawal needs, and they set aside a bit more for tax and interbank settlement needs. Any excess cash is lent in the overnight market. As soon as there is a general excess amount of cash beyond what the banks need, then they will be unable to find buyers and will drive the rate to zero (or a very small positive number). I.e. the banking system is unable to rid itself of aggregate excess reserves, nor to meet an aggregate reserve shortfall.

In general, adding a small amount of cash pushes the interbank rate down to whatever the central bank wants, at which point that operation is reversed, as leaving the excess cash in place would drive the overnight rates to close to zero. Similarly, removing a small amount of cash lifts the interbank rate to whatever the central bank wants, and then the operation is reversed.

Therefore it would not be possible to finance the deficit by creating more cash while continuing to conduct monetary policy under our current arrangements.

It would be possible to do this with a different financial architecture -- just not the one we have now.


I think the dynamic is more complicated than you lay out (a simple choice between money and bond finance).

First, the issue with money finance is not initiating it, but ending it. Take a $1tr injection of base money through asset purchases to finance a fiscal deficit in year 1. In year 2, exit from QE implies zero deficit financing, leaving the Treasury to find $1tr in new funding from creditors. If real rates rise as a result, the Fed has a dilemma: rising real rates are deflationary, and therefore QE is warranted again. This is the process through which markets begin to perceive that growing money finance will be permanent, and that inflation hedging is warranted. Most economists make the critical assumption THAT STIMULUS WILL WORK. When markets challenge that assumption, inflationary dynamics (one in which the only solution for structural fiscal deficits is permanent money finance) kick in.

Second, there is a relationship between base money growth and velocity. That is, if a promise of base money growth is what creates the expectations that make velocity rise, then, again, what is the implication of removing that base money growth on expectations and velocity? Specifically, if financial markets believe the Fed will be a seller of long-term assets, then will they engage in a carry trade (which the Fed is encouraging) for those same assets? And if they don't, what will happen to long term real rates and to the health of the banking system? Again, the issue is not initiating base money growth, but ending it during a period in which, for real, structural reasons, self-sustaining growth is difficult to spark. By real structural reasons, of course, I mean conditions of high leverage among households and financial institutions.

So I would ask, if the government chooses money finance, and it doesn't result in self-sustaining growth, how do you prevent expectations that long term structural deficits will be financed by money?

Nick writes:

Theory says it must be proportional to P. Theory and evidence says it is also probably roughly proportional to real income. So assuming it is proportional to NGDP is probably roughly correct.

That's ignoring the matter at hand. How much money funded deficit spending do you need to move the domestic GDP indicator you're talking about? When real-rates are below the natural-rate, whether we're talking about broad-money or narrow money or domestic versus global NGDP is irrelevant because there is an amplification mechanism that works at scale.

When the real-rate is above the natural-rate, there is not amplification mechanism. Consequently the details matter quite a bit.

Nick Rowe: "Min: You lost me there."

Maybe I misinterpreted you, Nick. I thought that you were arguing that the reason for inaction at this point was that policy makers did not know which means of addressing the problem is best (among alternatives that are not mutually exclusive). That, OC, is irrational (Burridan's ass). My response is that the inaction (if not deliberate) is the result of debt/deficit fear mongering, aimed at the public and perhaps at them.

David Pearson: "So I would ask, if the government chooses money finance, and it doesn't result in self-sustaining growth, how do you prevent expectations that long term structural deficits will be financed by money?"

If you are going to go that route, then structural deficits should be financed by money. Why should they be financed by borrowing, which leads to more and more borrowing in the future to pay the accumulating interest? If you are going to finance anything by money, it should be permanent expenditures. Borrowed money disappears when the loan is paid off.

As for self-sustaining growth, why should borrowing produce it when money won't?

Michael Woodford is promoting a new better social construction of monetary policy. There are three independent dimensions of monetary policy - interest rates, quantity of reserves and composition of the asset side of CB balance sheet.

More here: http://themoneydemand.blogspot.com/2010/05/michael-woodford-inflation-targeting.html

We really didn't change the production mix to use less oil, we just exported oil intensive work to countries with lower wages so we could afford more expensive oil. Once oil collapsed in the mid 80s, better growth resumed. Oil is still largely irreplaceable since all alternatives are higher, and insofar it is nonconservable, its price directly contributes to all other prices. The timing was reasonable if demand was being lowered to lower the cost of oil and to coordinate consumption with production. Most of all, it is the future expectation of oil prices that would impact the economy. This was market driven rather than OPEC driven but the economy hasn't been able to tolerate more than 6% of gdp to it and if it doubles in a period of months as it did in 2006-2008, its volatility makes it difficult for the economy to manage. After a brief retreat it remains near 2005 levels rather than 2003 and prior levels even with high unemployment making more similar to the early 80s. We could grow faster, but how much faster before oil doubles again? Not that much I expect.

What if the bank of Canada gave every citizen a deposit account (not just the banks). Interest rate policy could be conducted directly to consumers and would no longer be intermediated by a (dysfunctional) financial system. As such it would be more like fiscal stimulus, but without the deficits - and this in a way that even a central banker could understand.

The entire capitalist system would come crashing down....

RSJ: "I think in both countries this would grow roughly with GDP, for a seignorage of about 0.4% of GDP/year for the U.S. Not sure about Canada."

It's about half that in Canada. The main reason is that currency in Canada as a percent of GDP is about half what it is in the US, IIRC. And the main reason for that is probably that foreigners hold US currency, but rarely hold Canadian currency.

So, on average, the amount of any deficit that is money-financed isn't trivial, but isn't massive, either. But it might be a lot bigger than this during temporary periods, depending on a lot of things. I wish I could get my brain to think more clearly about this subject.

"But we've already added a lot to CB balance sheet, so it will take a long time to exhaust a doubling of the base and allow for positive rates."

That depends. If it causes the price level, and hence nominal GDP, to be permanently higher than it otherwise would be, that will create a permanently higher time-path of the nominal demand for the monetary base. It's not that NGDP follows an exogenous path, so we have to wait till the demand for the base exogenously catches up to the supply. In the long run, it's the other way around. The supply of money creates its own demand.

Declan: Maybe. But I don't ever remember seeing any sort of research, or commentary, saying anything like "How much of the current deficit in Canada (or pick any country) is money-financed?" There's just a total silence on this question, or anything like it. About a year ago, I remember David Laidler saying he wanted to see money-financed deficit spending. That's all. And he's the exception that proves the rule, because he thinks in Monetarist terms. I don't think this question is even on the agenda?

Anyone want to confirm or contradict what I just said?

RSJ @10.11 But remember currency. And remember demand adjusts to supply of money. And remember that we're not talking about whether the deficit is *temporarily* money-financed, but whether it's *permanently* money-financed. What happens to the long-run time-path of the demand for base money (or NGDP, as a proxy) as a consequence of today's deficit? Is it permanently higher? how much higher?

David: the dynamics are probably more complicated than in my simple temporary/permanent dichotomy. But trying to think of a time-path for *Nominal* GDP that is permanently higher than the counterfactual, by some fixed percentage, so it has the same growth rate, and so should have the same velocity, seemed to me to be a good start. I did say in my post that the *initial* mix of money vs bond-financing is largely irrelevant.

Jon: "How much money funded deficit spending do you need to move the domestic [Nominal] GDP indicator you're talking about?" (I added the word "nominal").

That's probably the key question. The bigger the fiscal multiplier (dNGDP/dDEF) the bigger the proportion of the deficit that will be money financed. But at the same time, the bigger the proportion that is money-financed, the bigger the multiplier will be (money-financed deficits have a bigger effect on AD than bond-financed deficits). I ducked that conundrum.

You lost me on the rest of your comment. Or were you saying what I just said above?

Min: I think you did misunderstand me. I'm not saying that policymakers are torn between two alternatives (money-finance vs bond-finance), like Buridan's ass. I'm saying they don't clearly understand whether and to what extent deficits will be money-financed.

TMDB: I disagree. I don't think Michael Woodford is promoting a new social construction of monetary policy. I think he's just adding a couple of epicycles on his existing social construction of monetary policy. And he, more than anyone, is responsible for the existing social construction that I would prefer to destroy.

You are close to the solution, man; but not yet. Your last problem is: you think the only way to absorb excess money supply in the future (when/if the economy comes roaring back to its full potential) is for the FED to sell back Gov Bonds to the market.
Flex your mind, man! Think of helicopter money as 100% permanent, uncle Ben (Trichet) burning his Gov bonds out on Liberty Street, Manhattan (ok, bonds are paperless, but ... u get the point). CONSEQUENCES: (1) The Gov budget constraint is relaxed, (2) AD goes up (money goes where it should, Main Street, instead of generating unstable bubbles in Wall Street), the quality of bank loans improves (healin' the financial system)... and you're out of the recession. (3) Too much money around, we risk (big) inflation? But whatever your assumptions on the money demand reaction function, Uncle Ben (assume he can observe money demand behaviour in real time) can reduce money supply as needed at his will, through many channels OTHER THAN SELLING BACK GOV BONDS!, And I don't need to tell you which ones.
Then, the only issues that remain on the table are peanuts (almost): it's hard for the FED to fine-tune the exit strategy, etc...
To your credit: you are the one American economist who came closest to the solution of this crisis. Yes, the solution is: the right coordination of fiscal and monetary policy.

First and foremost, when are we going to conclude that judging monetary policy by the interest rate (real or nominal) is misleading and wrong?

What I am getting at is that in order to know whether monetary policy can do more, we need an explicit, stated goal for policy. Suppose that the Fed announced an inflation target of X%. If the inflation rate comes in at a rate greater than X%, we know that policy is too loose. The level of the interest rate is (almost) completely irrelevant.

The fundamental problem with supporting AD-increasing policies in the current environment is that there are people out there looking at the fed funds rate at (basically) 0% and concluding that monetary policy is too loose. On the flip side, many of those who believe that we need to boost AD don't believe that monetary policy can be effective at the zero lower bound. This obsession with the level of the nominal interest rate is growing very tiresome and is problematic for moving forward.

This post from Mark Thoma, quotes Peter Dornan making basically the same case I am, that solutions aren't on the table because they don't suit the interests of the folks in the ruling class.

The only problem with Dornan is that he hasn't 'flexed his mind' (as PierGiorgio might say per comment above) and still hasn't managed to consider the notion of actually dropping money from the helicopters, rather than borrowing it and paying interest on it for no particular reason.

"And remember demand adjusts to supply of money. "

The "money" that households care about is unchanged when the government replaces bonds with cash. Households care first and foremost about their incomes, some of that income is spent on goods, some held as deposits and some held as financial assets (e.g. bonds and equity).

But households are able to control the ratio of deposits to financial assets independent of what the government does. If households want to hold more deposits and less bonds, or more bonds and less deposits, then the financial sector absorbs the discrepancy in its balance sheet. The financial sector intermediates to allow households to hold the type of financial assets that they want, while allowing borrowers to issue the type of liabilities that they want. In this case, Goverment is a borrower that wants to issue some combination of bonds and cash, and households want to hold some proportion of deposits and bonds. If the government issues more bonds, than the financial sector still supplies households with deposits and holds bonds against those deposits, earning a premium from the spread. If the government only issued cash, the financial sector would hold cash against household deposits, and would not earn a spread. But the actual level of household deposits is a function of household income -- government deficit spending and credit-based borrowing, as well as household expenditure plans. If you keep household income fixed, then tweaking the bond/cash mix is not going to cause a change in expenditure plans or household deposits. Only banks will find themselves holding "excess money balances".

The financial sector will find itself holding excess or insufficient reserves, and then the central bank will supply or remove cash as necessary to maintain the government-set marginal cost of reserves. Therefore a government that uses open market operations to control the marginal cost of reserves does not have discretion to monetize or not monetize debt. Obviously you can impose marginal costs on banks in other ways, and then there would be this discretion.

But even if the government has discretion to control the ratio of government bonds to cash in the economy as a whole, still this would not alter the ratio of deposits to bonds held by households or businesses; and it would not cause NGDP to change. No one other thank banks would find themselves with excess money balances.

What would cause NGDP to change would be the interest rate channel. I.e. when the financial sector holds excess cash, the marginal cost of reserves falls to zero and banks have lower funding costs. That *might* cause the borrowing costs of households to fall, which *might* cause households to borrow more, which would create a flow of income and it would increase household deposits -- e.g. more "money". And this might be inflationary if the money was in excess in relation to goods.

But these interest rate channels are asymmetrical, state-dependent, and complex. Currently, real mortgage rates are not low by historical norms, and the correlation between real interest rates and debt growth is weak. And as fed funds is already zero, there is no reason to believe that forcing banks to hold additional levels of cash will cause interest rates to fall further.

In any case, no sane household cares whether there is 5% cash or 10% cash backing its deposit. Such a shift cannot case NGDP to increase or decrease, whether it is temporary or permanent. The only possible effects are via the interest rate channel, and this channel is not currently functioning, nor should we *want* to stimulate the economy this way -- if anything, households are trying to borrow *less*, because we've been busy stimulating them to borrow for the last 30 years, and have now run out monetary bullets until household balance sheets are repaired.

Nick: I'm explaining why expectations matter. If expected inflation is low, real-rates can be above the natural-(real)-rate such that there is no operative money multiplier.

Once this happens the MB may expand by financing deficits but this is a small proportion of broad money, and thus has little to no-effect unless expected inflation increases. Once it does so, broad-money increases proportionate to MB, and thus inflation will materialize as expected.

The expectations are self-fulfilling in this sense and quite a bit more powerful than base-money.

Government spending is irrelevant. Even if Government spending increases 10% in an recession and is money financed there are too many reasons to believe that this is GDP neutral. The basic problem is that GDP is about the flow of NEW goods. If I sell you my kitchen table that income (to me) is not a part of GDP. In recession one often observes the liquidation of employed capital (both machines by depreciation and more importantly stocks). Money today is preferred to assets today. Consequently, a rise in government spending is just as likely to draw upon and encourage the liquidation as to generate new production--and this is on top of basic problem that transfer payments do not contribute to GDP and government 'production' displaces private production. Its the amount of new production (not whether its public or private) that is restricted when deflation sets in.

To review:
1) real-rates above the natural rate neutralize the money multiplier
2) deflation encourages the liquidation of capital in favor of money, deflation can cause point #1.
3) demand can be satiated by production or liquidation
4) government spending is by no means certain to induce new-production.

In short, it takes a lot of money financed "G" to move NGDP upward.

"I disagree. I don't think Michael Woodford is promoting a new social construction of monetary policy. I think he's just adding a couple of epicycles on his existing social construction of monetary policy. And he, more than anyone, is responsible for the existing social construction that I would prefer to destroy."
What's wrong with his current theory? He has level targeting, he recognizes that bad banks don't create money so he supports credit easing.

PierGiorgio: "you are the one American economist who came closest to the solution of this crisis." Thanks! But Canadian!

"Think of helicopter money as 100% permanent, uncle Ben (Trichet) burning his Gov bonds out on Liberty Street, Manhattan (ok, bonds are paperless, but ... u get the point)"

It took me a little while to get the point, then I got it. It's very similar to a point I had made about a year ago. If you destroy the Fed's balance sheet (my example was having the Fed buy worthless assets), then the Fed cannot afford to retire the newly-issued money in future, so the increase in the money supply is expected to be permanent. Yep. I like your way of thinking about it.

I've found my old post on this topic! Praise be to Google!

I had totally forgotten about my old posts on this subject!

It's similar to the "backing" theory of money. Destroy some of the assets "backing" the money supply, and the Fed must print more money in future than it would otherwise have done, and so the existing money is worth less.

I must do a post on this. It suggests there is hope for the Eurozone after all. We want to hope that Greece partially defaults, because that will harm the ECB's balance sheet, and make the bond purchases unsterilised ex post.

Josh: "First and foremost, when are we going to conclude that judging monetary policy by the interest rate (real or nominal) is misleading and wrong?"

In my case, some years ago. But I have to keep backsliding, to re-frame the question in the way central banks and most other people frame it. So I can communicate. But I did add my little "rant" paragraph. I totally agree with your comment. But I despair.

Declan: "...that solutions aren't on the table because they don't suit the interests of the folks in the ruling class."

Who are the ruling class? If they are shareholders, a money-financed deficit would be very much in their economic interests. Firms typically have nominal debt. Deflation raises, and inflation lowers, the real value of that debt. Plus, recovery would be good for firms' profits. Share prices, in real terms, rise for both those reasons.

RSJ: For 20 years, (until the recent hiccough), the Bank of Canada has been controlling inflation almost perfectly. The only thing the Bank of Canada ultimately controls is its own balance sheet. On the liability side of that balance sheet is the amount of base money. By controlling base money, the BoC controls the price level. Double the size of the nominal balance sheet, permanently, ceteris paribus, and the nominal price level doubles too. The influence of the central bank on the real rate of interest is purely temporary, due to sticky prices. There exists a natural rate of interest that is independent of monetary policy. Theory and empirical evidence support this. It's all in the units. The interest rate has the wrong units. You are deep inside the Neo-Wicksellian social construction of monetary policy as interest rates!

Jon: I am not one of those "true-believers" in fiscal policy, who are sure it must always work. But, in general, I see it as increasing the demand for newly-produced goods, and thus raising the natural rate.

TMDB: A couple of my old posts might explain why I don't like thinking, and don't want central banks to encourage people thinking, of monetary policy as interest rates:



You wrote, "In my case, some years ago. But I have to keep backsliding, to re-frame the question in the way central banks and most other people frame it. So I can communicate. But I did add my little "rant" paragraph. I totally agree with your comment. But I despair."

Sorry, this wasn't a comment directed at you, but rather the debate in general.

"If you destroy the Fed's balance sheet (my example was having the Fed buy worthless assets), then the Fed cannot afford to retire the newly-issued money in future". (Nick)

What the FED cannot really afford anymore is to circulate again the assets it has destroyed. The Public Debt/GDP ratio will fall. (Whereas if the FED/BCE doesn't burn its Gov bonds the Debt/GDP ratio stays higher - in accounting terms - even if ex-post it turns out that they keep them forever). The expectations that will change most will be those concerning the solvency of governments.

As for money supply, I’m not sure at all that “then the Fed cannot afford to retire the newly-issued money in future”, nor I am sure this would be a good thing: for we need expected inflation to rise, but to 5%, not 20-100%, whereas the increase in base money we need right now is very large indeed… The FED/BCE can (when/if a recovery generates the risk of inflation) reduce M supply again by selling other assets, and by increasing the reserve ratio of commercial banks (or similar measures). So in my view the initial increase in money supply is not necessarily "permanent", but there will be a permanent shift in M aggregates (more base money, but not much more M5).

All this would amount to a big transfer operation (seigneurage without inflation!), but this is another story.

Monetary policy today is powerless because the new money goes to the financial sector, and stays there. We have to break the wall. There's too much money around right now, too much pressure (think of the BP oil well: we risk an explosion. I understand the German fears and warnings that the situation "dangerous") is needed just to generate a small leakage to the real sector. I suggest less money out there, but better targeted.
Thanks 4 your reply! We should write a paper on "the helicopter money solution"! By the way, I though Canadians were Americans…

"For 20 years, (until the recent hiccough), the Bank of Canada has been controlling inflation almost perfectly."

That was a special time -- interest rates were in a secular decline, and households were willing to assume debt at a rate faster than GDP growth (to buy real estate). Such a situation is not sustainable, and if you keep using monetary policy in this way, it eventually stops being effective.

I could just as well say that for 20 years, Japan has *failed* to be able to control inflation despite heroic monetary efforts.

When you try to stimulate the economy via monetary policy, you are requiring the private sector (and really the household sector) to self-stimulate by increasing their rate of borrowing. But don't you notice how you can never raise the call money rates to the level they were before? They must always fall lower. You cannot in general count on this form of bubble stimulus to remain effective.

"The only thing the Bank of Canada ultimately controls is its own balance sheet."

No, it doesn't control the size of it's balance sheet. Currency is demand driven. If it tried to do this, it would end up rationing loans, which is very destructive. As a lender of last result, the central bank *must* supply reserves to banks that ask for them who are in a reserve short-fall. If the banking system is in aggregate short of currency, then the CB must either force some of the banks to fail, or supply the currency needed. So the CB, in practice, can only control the price of reserves.

"On the liability side of that balance sheet is the amount of base money. By controlling base money, the BoC controls the price level."

Until it doesn't -- a la Japan. Again, you need to articulate a channel whereby buying a bond for cash increases someone's propensity to spend. The only such channel is the interest rate channel. But if you can think of another channel whereby open market operations can cause households to spend, I'd love to hear it.

"Double the size of the nominal balance sheet, permanently, ceteris paribus, and the nominal price level doubles too."

Again, no -- you need to define a plausible mechanism.

I think the issue here is that you believe that households have some unmet need for deposits that only the government can fulfill. But the banking sector already fulfills this need. If households really wanted to exchange their riskless treasury bonds for cash, they could sell those bonds to the banks, who would be more than happy to earn money from the spread. Of course, households do not want to do this -- there is no unique demand for cash. There is a demand for safe financial assets in general, rather than goods. Everyone wants more money -- they just don't want to buy that money by giving up treasury bonds. They want to either get that money for free, or by giving up goods.

"There exists a natural rate of interest that is independent of monetary policy. "

I *thought* we had agreed that this was the consol rate! But again, amalgamation produces confusion. Why must people confuse the consol rate with the overnight interbank rate?! This drives me nuts.

The FedFunds rate is an administered rate applicable to the interbank market only. It is not "natural", because no investor borrows at this rate or receives interest at this rate. It is just a way for the government to impose marginal costs on regulated banks when they create credit. There is such a thing as a natural fully loaded credit creation cost, of which FedFunds is one component, but we could impose this cost even with fedfunds permanently set to zero -- and the rest of the economy wouldn't notice. E.g. FedFunds is a "hidden variable" of the banking system, and what really matters is the actual marginal cost of credit creation, and to measure this, you need to take into account capital adequacy requirements, regulatory costs, longer term money market rates, and fedfunds.

"Theory and empirical evidence support this."
Yes -- there is a natural consol rate, and a "natural" term structure of rates, and even a natural marginal cost of funds for banks as a whole. But FedFunds is not any of these things. But I agree that lowering fed funds can drive bank cost of funds below the "natural" rate, and this can, in some cases, promote household balance sheet expansion. When this happens, monetary policy is said to "work" to stimulate growth. On the other hand, rasing fedfunds will certainly drive borrowers into default or price them out of the market, and in this case, the balance sheets of households will contract -- the effectiveness of monetary policy is asymmetric.

"You are deep inside the Neo-Wicksellian social construction of monetary policy as interest rates!"

Let's not resort to name calling :P

As a sworn enemy of loanable funds, I am not in the Wicksellian camp. But more importantly, instead of talking about social constructions, why not try to describe actual channels by which open market operations can force people to spend more and therefore increase NGDP. Japan would be interested in hearing how monetary policy, when properly socially reconstructed, can be used to exit a liquidity trap. I think the U.S. will too, and the rest of the nations following the real estate boom/monetary policy approach to output stabilization will soon join them.

RSJ: Wow! Long comment. My brain is getting tired. Just two points:

1. According to the standard Neo-Wicksellian view, the transmission mmechanism from monetary policy a.k.a "interest rates", is via an intertemporal *substitution effect*. Spending today (consumption and investment) vs spending tomorrow. It's symmetric. It does not rest on increased borrowing and debt per se. If r falls, borrowers (dissavers) want to borrow more, but lenders (savers) want to lend less. It's not about balance sheets and accounting. It's not any sort of income or wealth effect. It's a relative price effect.

2. A transmission mechanism and instrument that doesn't work via nominal i? First rule, stop targeting i. Stop talking about i. Strongly discourage anyone from thinking about the central bank targeting the time-path of i. That's what's at the root of the problem.

What to do instead? Pick some *nominal* variable to target instead. Talk about the time path of (say) the TSX (a Canadian share price index). It has the right units, $. Say the BoC will buy or sell the TSX in order to make the TSX follow a defined upward and higher path. Transmission mechanism? All the usual suspects, the substitution effects from real share prices to investment (Tobin's Q), and consumption demand. Plus wealth effects, and even balance sheet effects, if you like. But, above all, expectations.

Sorry for the bombardment. I am waiting on some paper-work now -- banks! -- and have nothing to do!

re 1.

Yup, which is why I took offense at the Wicksellian label. In general, I think the standard view of these things does not take balance sheet expansion into account, yet balance sheet expansion -- even while real interest rates are constant -- is an absolutely key driver of booms and busts. I've only read 5% of the models that you have, but I challenge you to cite one that includes household balance sheet expansion -- it is ruled out via no-ponzi conditions that don't match observed reality, and so the economic problems that we have now are also ruled out.

re 2:
OK, this would require (in the U.S. at least) a constitutional amendment -- only Congress can spend money, and this power cannot be delegated to another branch of government.

But aside from that, when the Central Bank owns my stock, do they have voting rights? Would they only buy stocks of firms that pay dividends? If not, I can see a nice profit opportunity. In that case, such a policy would be stimulative, but not necessarily in a way that you want. Better to get a road or health-care out of the deal, rather than pumping money into the hands of equity issuers (if you want to decrease yields) or short-sellers (if you want to increase them). I would prefer, say, a congressionally funded small business loan program, or even grants to public purpose projects rather than for profit entities.

Think of it fundamentally. The Bank of Canada is a financial institution. Like any FI, the only thing it can directly control is its own balance sheet. So what gives the BoC way more power than any of the other Canadian banks, which are much bigger than the BoC?

The liabilities of all the other banks are redeemable into the BoC's liabilities. They promise to redeem them for BoC liabilities. They BoC makes no such promise the other way around.

The BoC's liabilities are media of exchange. The BoC's liabilities are media of account.Insofar as other banks' liabilities are media of exchange also (chequing accounts), they are redeemable into BoC liabilities. This means the BoC ultimately controls the supply of media of exchange and media of account.

Declines in real GDP due to general gluts are due an excess demand for the media of exchange.

The CPI (or GDP deflator) is the price of goods in terms of the medium of account.

Since the BoC ultimately controls its own balance sheet, and the supplies of media of exchange and account, it affects GDP, and P, and thus NGDP.

Any "causal" transmission mechanism from BoC actions to NGDP is a *disequilibrium process*. Looking at the equations that define the *equilibrium* relationships between M and NGDP, will not show us that transmission mechanism, only the outcome.

There is no single transmission mechanism that can be defined independently of people's expectations, and those depends on the pattern of conduct of monetary policy they are used to, and how it is framed. The Neo-Wicksellian interest rate story is just one story, appropriate when expectations are defined over time paths of nominal interest rates. Change the framing of monetary policy, and everything else may change too.

RSJ: "Yup, which is why I took offense at the Wicksellian label."

Sorry! "Horizontalist"?

"I've only read 5% of the models that you have,.."

maybe not. I'm very good at faking it!

"...but I challenge you to cite one that includes household balance sheet expansion.."

A challenge I will not accept!

I like to think of Central Banks buying the TSX as a thought experiment. Not sure how practical it is. But if they buy government bonds, stretch that to bank bonds (which is effectively what they are doing when they lend to banks), then to commercial bonds, it's not much of a stretch to go on to buying the index of stocks. The voting rights question is a good one. I don't see why dividends matter?? And I don't see any (arbitrage?) profit opportunities, unless the market prices for the TSX and TSX futures deviate from the announced BoC path? But then that's the same sort of arbitrage profits we get now on interest rates, and those arbitrage profits, and their exploitation, are part of the transmission mechanism.

Alternatively, the BoC could buy farmland. But it takes a bit more managing than bonds or stocks. Or things like bricks. But where to store them all? Or futures contracts on real assets.

By all means have it buy new roads and schools. But whether we need to expand the supplies of medium of exchange and account, and whether we want more roads and schools, might be different questions. The TSX is just a good place to park your seigniorage profits until the government decides it would give a higher rate of (social) return to invest in roads and schools instead.

Well, Nick, I guess this boils down whether you believe that institutions and legal frameworks can change the fundamental analysis, or whether they can't.

I believe that they can -- for example, the introduction of money would radically transform an autarky, allowing for the possibility of general gluts due to a desire to hold money.

And the introduction of credit markets radically transforms the process of savings, allowing IOUs to become a form of money, and creating a disconnect between time preference and interest rates. Now, a belief in a return that meets or exceeds the the risk free return will cause the investment to get funded in all cases (or *almost* all cases)

The introduction of the ability to short or borrow to purchase IOUs also transforms interest rate dynamics, flattening out the demand curve with respect to quantity (not *perfectly* flat).

The existence of central banks that are committed to supply member banks with unlimited amounts of cash at specified pre-announced interest rates also changes the picture, as it means that central banks no longer have freedom to control the size of their balance sheet.

The existence of deposit insurance and other government commitments takes bank runs out of the picture, and makes bank deposits as good as cash. Effectively the entire deposit base becomes an off-balance sheet liability of the treasury, which means that deposits are as good as cash.

The existence of fiat money and a government commitment to pay debt takes default risk (but not inflation or market risk) out of government bonds and out of government insured securities.

The existence of a repo market that allows anyone to convert government bonds (and often other types of assets) into cash whenever they want means that cash has no operational or default risk advantage over bonds.


So if you put all of these things together (hopefully in a consistent way), then you get my core position. Not sure whether to call it horizontalist -- institutionalist? observationalist?

But what's more interesting to me is how on earth I could convince anyone else that the above points are the "right" way of modeling our current economy. And more importantly, that economists will always need to change their modeling techniques as new institutions and behavioral options emerge.

If I cite data, then I am hit with a Lucas Type critique which seems to be a get-out-of-jail card for data. If I make a rigorous, axiomatized model that demonstrates this, then would the model be dismissed because the axioms are "wrong"?

If I try to use "reasoning" to argue that, given a certain set of options, it would be irrational for people to behave differently -- would that work? Seriously, I am at a loss.

And none of the above is meant to say that my own specific examples must be right -- I am open to being proved wrong. But there should be a possibility of being proved right, too; there should be a space for models that assume people maximize utility by taking advantage of *all* the operations available to them under the current set of institutional arrangements, rather than requiring that

* banks must wait until a depositor appears before making a loan, and cannot tap the central bank for the funds to make the loan.

* every household that buys a bond must hold it to term, and the funds to buy it must come at the expense of consumption (e.g. the funds cannot be borrowed by shorting some other security, or borrowed from a bank, which itself could get the funds from the central bank).

* an increase in the desire to save must result in people buying more call options and not more put options. And therefore an increase in savings must drive down yields

It seems that the options each individual has in our current system are broader than the options that respectable models allow, and there is no way for me to convince anyone that the larger set of options lead to new dynamics. Arrgh!

Interesting comment RSJ. I will try to give a useful answer in the morning.


The Fed is a Quasi-government organization. The US Constitution stipulates the legislative process for spending tax revenues, but that has absolutely nothing to do with the Fed. The US Government is a depositor to the Fed. The Fed can spend money any way it wants. If it wants to switch to Nick Rowe's Non-Interest Targeting System (TM) it could do so immediately. At most it would require a few legislative changes to the Federal Reserve Act. No constitutional issues here.

Determinant -- this is not the case, but admittedly I am not a lawyer and now we are delving into issues of constitutional law.

You may recall that a few years ago -- or decades? (I am old!) There was an attempt to pass a "line-item veto" bill. The purpose of that bill would be for the president to veto parts of a bill -- the pork parts -- while passing the remainder. This was ruled unconstitutional, and then congress went back and tried to re-write it, and it was again ruled unconstitutional. The reason why is because no branch of government can "delegate" its powers to another branch -- this would break the separation of powers principle which is a constitutional bedrock. All spending must originate in the house of representatives, be passed by the Senate, and signed into law by the president. And congress cannot delegate its spending powers to anyone else. That's the problem here.

Now in a general model, you can roll the federal reserve, congress, the president, and everything else into "G" and then it is obvious that "G" can cause inflation. FDR did it, and Obama could do it. But in practice, the "G" is often controlled by people who have their heads up their [ ], and so politically we are prevented from causing inflation, even though in principle it is easy to do.


Jon: I am not one of those "true-believers" in fiscal policy, who are sure it must always work. But, in general, I see it as increasing the demand for newly-produced goods, and thus raising the natural rate.

We can agree about the monotonicity but not the magnitude. You've offered no argument about the magnitude...

I've grown very skeptical during 2008-2009. The US Government increased spending by about fifteen percent of GDP to no observable effect. Half of that increase was (temporary) money financed.

Its very clear that expectations matter a lot, but once you concede that, its enough to realize that expectations of inflation will move the real-rate down (below the natural-rate), and therefore fiscal policy is irrelevant.

Perhaps the most important idea is that no economist should ever mention or credit the notion of a liquidity trap. Talking about the liquidity trap, believing that it is a possibility, that's what makes the problem...

Jon "Perhaps the most important idea is that no economist should ever mention or credit the notion of a liquidity trap. Talking about the liquidity trap, believing that it is a possibility, that's what makes the problem..."

I think no economist should ever mention or credit the notion that monetary policy is interest rates ;)

RSJ: Long thoughtful comment @9.47. I will do my best. Your paragraphs are my numbers:

1. I think nearly all economists do believe that institutions matter. I do.

2. Absolutely. A minority of macroeconomists (Real Business Cycle theorists) believe you can understand business cycles as if in a barter economy. That's primarily because they believe in perfect price flexibility. I don't, and most don't.

3. Some (not all) IOUs become media of exchange. And many more become more liquid than real physical assets. You lost me on the time preference - interest rate disconnect. I don't see it. In principle, intertemporal trade is like all other trade in this regard. Different people's preferences, at the margin, become equalised to prices (interest rates), and so equalised to each other.

4. You lost me. Are you talking about the slope of the yield curve? It doesn't have to be flat, even with "perfect" credit markets. Maybe everyone wants to retire and spend in 2021, which will cause the yield curve to be different at an 11 year horizon.

5. Some (many) would agree with you; I disagree. The CB can and does change the overnight rate in the light of circumstance. Plus, it can only control the overnight rate by adjusting its balance sheet. Even if you ask an orthodox central banker (not someone like me) how they hit the overnight rate target on a daily or hourly basis, they talk about injecting and withdrawing settlement balances.

6. I partly agree. Some would agree more strongly. People still hold currency, even when they risk getting mugged, and deposits earn interest.

7. Yep. Probably.

8. No. Will the market function? instantly? Zero transactions costs? I really wish i understood liquidity better, but it seems that what look like small differences in liquidity are very important to traders.

9. "Horizontalist" would be fair, and useful, I think.

10. 11, 12, 13: You do the same as the rest of us. You do a mixture of all those things. And, like the rest of us, you can never really prove anything. Your analysis is always open to criticism, and some will always find your model empirical evidence, and appeal to "how the world is" unconvincing. That's life. But come on in; the water's lovely!

*1. Do existing models assume that (implicitly)? I don't think so.

*2. Ditto.

*3 You lost me. The effect of increased desire to save on interest rates needn't (doesn't?) have anything to do with put and call options.

Are you sure that those extra options make much difference? Do they all wash out in aggregate, for example? Models simplify, of course. But if you think a simplification is restrictive, and really does mean that something important is left out, the normal rule is that the onus is on he who wants to complicate things to prove they need complicating.

Good luck, anyway!

Woodford is now saying that monetary policy is:
1. hitting price level target.
2. setting the quantity of reserves and interest rate on reserves.
3. purchasing risky assets during periods of financial stress.

I still don't see what's wrong with that or how is it so different from what you wrote in those two posts you mentioned.


I remember the line-item veto controversy, it was during Clinton's time. It is also irrelevant to the status of the Fed.

As I said, the Fed has a depositor/client relationship with the US Government. It isn't a government department, and it doesn't receive tax appropriations. Since it does not, unlike say the US Treasury or the Department of Justice, it isn't subject to legislative spending oversight and related provisions. Simply because economists like to lump the Central Bank in with the Executive and the Legislature in an aggregate variable called G does not mean that the US Constitution follows the same reasoning. It doesn't.

The Fed can spend its money any way it wants. Witness its "Impaired Assets" facilities. As long as the Fed stands behind its US Government deposits, it's fine.

The Fed does not raise or spend tax revenue. While it may do so indirectly via its ownership of US bonds, that is also fine as millions of private entities do the same thing.

If you want to argue that the Fed and the US Government are the same and the US Constitution's spending provisions cover the Fed, you're on your own out there in left field.


The U.S. Constitution gives congress the right to issue and regulate money, and also to regulate the value of money, to incur debts and to raise revenue. Congress has elected to delegate the "regulate the value of money" as well as other powers, to the Fed by allowing them to purchase government liabilities with notes that they issue. They are still subject to that charter and cannot overstep it anymore than the commerce department can overstep their charter. The Fed is subject to the Federal Reserve act and must abide by it, and that act controls what they can and cannot do. If you think they can do a helicopter drop, then the Fed itself disagrees with you, as they have been very clear to delineate the difference between fiscal and monetary policy. One spends and taxes, and the other buys and sells government liabilities.

The impaired assets are "justified" by the emergency lending rights that congress explicitly granted the fed. But these cannot be used as a matter of course, and are subject to congressional restriction if abused.

The notion that the Fed is an extra-governmental agency, or a private agency, is the left field view. They are a government department with G-x employees, a congressional charter, and presidential appointments of leadership -- just like the EPA, the commerce department, or any other agency.

Wrong. This isn't 1789 and the US isn't on the silver/gold standard anymore. Nor does it have fixed exchange rates. So the "value" of money clause is a dead letter, and has been so since 1973. You're pitching right-wing theories that don't have a basis in reality, legal or otherwise. This reminds me of the "Horse and Buggy" critique of the Interstate Commerce Clause.

The Federal Reserve's website will inform you that it isn't a department of the Executive Branch. Fed employees are not employees of the US Government. Yes, this horse is in fact a zebra.

The Federal Reserve Act didn't stop Quantitative Easing.

If you believe that the US Constitution "value of money" clause implies that the US congress has to fix the total stock of US Dollars at $5 Trillion or some other number, or maintain a formal inflation target, then you are contradicted by 40 years of actual policy and history.

The comments to this entry are closed.

Search this site

  • Google

Blog powered by Typepad