# Debates in modern monetary macro …

Yesterday, regular commentator JKH wrote a very long comment where he/she challenged some of the statements and logic that modern monetary theorists including myself have been making. While I don’t want to elevate one comment to any special status – all comments are good and add to the debate in some way – this particular comment does make statements that many readers will find themselves asking. In that sense it is illustrative of more general principles, points etc and so today’s blog provides a detailed answer to JKH and tries to make it clear where the differences lie. Some of these differences are at the level of nuance but others are more fundamental.

To also make it a little more interesting my mates Warren Mosler and Scott Fullwiler who were mentioned in the comment by JKH also decided to send me their version of a reply which I dutifully acknowledge throughout. To make it clear who is saying what, all comments other than my own are in blockquote format with the notation: JKH, Warren or Scott on the first line before the statement.

Scott started his reply with a personal message to JKH:

Scott F: Dear JKH Thank you for the comments. You are certainly one of the more thoughtful commentators of all those who post to this and related blogs, and I appreciate the fact that you don’t just “buy” what we’re selling but rather subject it to very careful thought. Of course, as you probably expect, I think your entire critique either misinterprets or otherwise misunderstands the points we’ve made. So, respectfully, I have made the following rejoinders. That I and others took your comments seriously enough to respond at length hopefully shows that we have a good deal of respect your views and your analysis, continuing disagreement or points to be discussed further notwithstanding. Finally, I should note that overall I am in far more agreement with you than I am with neoclassicals … and even the points of continued disagreement are more easily discussed/debated given that we have for the most part a common language and common understanding of accounting.

Okay, pleasantries over lets see where the discussion takes us.

JKH started with the statement:

JKH: My general impression however is that in making the case you and the others are combining the correct explanation of operational capability with a near-rejection of the real world importance of balance sheet “policy” restraints or constraints.

My view is that this is not what I do. I first consider that what is held out as a financial constraint is usually not that at all. Typically, in macroeconomic policy the constraints are political and voluntarily imposed. The sophists then dress these political constraints up as financial constraints using gold standard type macroeconomic models which appear throughout the literature to avoid addressing the real issue.

My assessment is that if the general populace was better educated in these matters – that is, understood the actual operational capabilities of the national government it would be very difficult for the politicians to conflate their own ideological desires with the concept of a financial constraint. In that context, telling us that we had to have 5 or 8 per cent unemployment and rising underemployment because the government cannot afford to purchase all the labour and even if it did it would be inflationary, takes on a different slant.

We would know that they could afford to fully employ the available workforce as long as their were sufficient real resources available to provide the extra food and other things the higher employment levels would invoke. This would then require a higher level of sophistication in the public debate. Are there the extra resources? How close to real capacity are we? That would then promote new research that focused on the nub of the problem rather than the array of dishonesty that parades as knowledge out there in the form of academic papers – which say the government has a financial constraint and will cause higher interest rates, higher taxes, higher inflation if it bucks against it.

Businesses would also have to justify their opposition to true full employment in more sophisticated ways because we would all know that the usual reasons they give – again relating to government budget constraints – are all deeply flawed.

I also would not use the term “balance sheet policy restraints” because that implies that there are financial matters that are at play. By your own admission there are not. I will deal with this in more detail later because I know you think there are “financial issues”.

Warren said at this point that:

Warren M: I for one criticize these policy constraints as counterproductive all the time. That’s hardly rejecting their importance- in fact, that’s pretty much the entire problem!

While Scott added (note Scott provided several points rather than a line by line commentary – so I am threading them in where I can and some of this comments apply to later points as well):

Scott F: We don’t reject “policy” restraints or constraints as you call them. The point is that the intellectual foundation of such restraints/constraints is most often either a misunderstanding of the operations OR the restraints actually do exist but only under different monetary regimes. Get rid of those foundations, and we can then have an honest discussion about the interest burdens you mention. Further, and related, we have ALWAYS and REPEATEDLY acknowledged the inflationary potential of either debt service burdens or just overly large deficits in general. Our point is that THIS should be the issue to debate in regard to macro policy, NOT the stuff that comes from a flawed understanding of operations or the monetary system. Again, our point is that unless you get rid of the flawed intellectual foundations of current policy debates, you can’t have an appropriate discussion of policy or of the “practical constraints” that definitely do exist and which we acknowledge exist.

JKH then continued:

JKH: E.g.: It’s quite clear the government has the operational capability to spend ’till the cows come home, creating broad money and bank reserves in the process. There is no operational limit. But there is a real policy concern in terms of practical balance sheet debt and the interest burdens associated with it. And in the real world, that translates to the notion of a policy limit. And a policy limit is a constraint of sorts. My general impression is that your associated school of thought tends to demote the importance of such practical constraints in order to emphasize the correct operational understanding.

Okay, so there are no financial constraints. We all agree on that. So it would be good jettison all the macroeconomic theory that construes the government budget constraint as an ex ante financial constraint instead of seeing it as an ex post accounting statement, with no operational relevance.

Then we also agree that the public perception of rising public debt arising from the fact that we have been erroneoulsy conditioned (relentlessly) to equate the household budget with the sovereign government’s budget presents a problem to a government who wanted to increase spending. A political problem rather than an economic problem.

I make that distinction because I do not accept the relevance of the Ricardian equivalence theorem. For non-economists – this piece of neo-liberal dogma says that the non-government sector (consumers explicitly) having internalised the government budget constraint will negate any government spending increase whether the government “finances” its spending via taxes or borrowing. So if the government spends and borrows, consumers will anticipate higher future taxes and spend less now offsetting the stimulus).’

I also reject the Policy ineffectiveness proposition from Sargent and Wallace which says that because private agents have rational expectations (RATEX) and truly understand the exact nature of the economic model, the government is powerless to influence real output levels. Private agents will anticipate perfectly what is going on and adjust their behaviour to offset the public policy. I recall Arthur Okun once saying that if we all knew the exact structure of the economy then there would be no need for economists – we would all be forecasters with 100 per cent expectational accuracy! I have noted that none of the RATEX exponents ever turned in their highly paid cosy academic or other positions in response to Okun’s challenge.

I think Modern Monetary Theory (MMT) shows that in a technical sense (after understanding how the operational matters work) that deficits will drive down interest rates unless there is some offsetting central bank action.

The fact that the government voluntary decides to constrain itself by issuing debt $-for-$ into private markets using auction systems which allow the final bidder to determine the bond yield for that issue doesn’t alter that fact. The rising yields might reflect the perceptions of the markets that the deficits are too large – conditioned by the mainstream economics – just as they may reflect a desire to diversify investment portfolios towards higher risk positions. Whatever, the rising yields have nothing to do with the deficits in an intrinsic sense.

That is the problem though. Our mis-education on these matters and the dominance of the mainstream profession in the public debate allows us all to be hoodwinked and see rising yields and rising deficits as in some way substantiating the loanable funds doctrine of classical thought. All the rest of the spurious conclusions follow from these voluntary (neo-liberal) structures that the government erects to obscure there real capacity to advance public purpose.

So far from ignoring these voluntary contraints I am always writing about them – to emphasise the political nature of the decisions taken and the way the options are presented. But as an educator I think it is also important to provide as much detail as I can about the way the system actually should function from first-principles so that the public can make better conclusions about the way the political process deviates from the intrinsic.

In that regard, I hope that as citizens become more and more informed they will actually engage in debate and ask questions of the politicians and commentators in general. The result might be that pressure is brought bear on these characters to really explain why they will not increase public employment, or why they are selling off excellent public enterprises at huge discounts only to see service delivery fall and private wealth more skewed than before. And the rest of it.

Warren’s response:

Warren M: Yes, and that’s exactly the problem. Policy makers who don’t understand the monetary system and therefore fail to recognize the policy options that are wide open operationally.

And the policy limit is definitely a constraint of sorts.

For me, those ‘practical constraints’ are public enemy #1.

JKH continued:

JKH: The constraint shows up in this example in the following ways. First, the operational capability described requires the freedom for the government to run an unlimited overdraft at the central bank. Overdraft is equivalent to credit on demand. Both are equivalent to monetization on demand.

This is the way you are presenting it. The government is the central bank. The fact that we decide to split the treasury and the central bank up into two entities is misleading in terms of understanding the operational side of the government and non-government interaction. It has also been an ideological venture to claim that the central bank is independent and therefore somehow inflation-first monetary policy is above the political process. Which in my view means we degrade our democracies if the cabinet (whatever executive arrangement is in place) can avoid responsibility for stupid monetary policy by saying some non-elected officials have forced this on the economy.

Monetisation is also a term that is ideologically loaded and not very helpful. It is in the ranks of “printing money” as ill-suited terms when describing how governments spend.

Warren elaborated further on this:

Warren M: Depends on your definition of ‘monetization.’ Years back it was a gold standard term.

JKH then says:

JKH: So here’s the policy reality: When was the last time you saw such a government overdraft of any size on a central bank balance sheet?

What exactly is the point? In Australia, we have established processes where the Australian Office of Financial Management (a division of Treasury) manages all government debt issuance. By law it invokes an auction system to raise private funds exactly equal to the net government spending projection. In other words, imposing on the government a voluntary arrangement that eliminates the need for the central bank to purchase treasury bills prior to government spending.

Before 1982, a tap system was used and the bond issue was regularly undersubscribed and the gap was filled by the central bank buying the bonds. The change to the auction system was designed to impose “fiscal discipline” on the elected government – totally voluntary and totally unnecessary.

As Warren points out you will never see this in the US because “it’s prohibited”. A legal constraint not an economic one.

JKH then says:

JKH: I’m sure you can read my mind for the rest of the argument from there. Suffice to say, there are decisions to be made about converting that overdraft into debt issued by the government, etc. Those are policy decisions that act as constraints on the operational freedom that is the foundation of this line of thinking.

We all agree on that. As Warren says “The problem is allowing the Treasury to spend directly without first issuing treasury securities.” This can also be seen in the voluntary decision in Australia to move from a tap to auction system.

Scott F elaborates further:

Scott F: Your point about overdrafts again misses the point we are making. The deficit always comes first unless the non-government sector borrows to get reserve balances to buy the Treasuries. You can’t buy a Treasury without reserve balances (that’s how they clear), and the reserve balances exist either because of a previous deficit, previous open market purchase (which purchased Treasuries from a previous deficit), or (as mentioned) borrowing from the central bank. Our point is that a deficit for the currency issuer is NEVER about borrowing, and bond sales are NEVER finance operations. Also, the distinction you refer to b/n bond sales and monetization is a false one that I explained in detail in “interest rates and fiscal sustainability” on the CFEPS website.

JKH then goes onto to discuss interest rates and seems to be saying that the central bank doesn’t have control of the short-run rate:

JKH: The idea that debt is an interest rate control mechanism in this sense is quite nifty. I like it as a coherent model of how government expenditure relates directly to choices about interest rate levels. Ground zero for interest rate control is the Fed funds rate. But when investors buy government bonds, they make some judgement about future policy for such interest rate control. But the interest rate issue isn’t entirely endogenous to the fact of control over issuing debt; at the end of the day, it is also a function of the market’s expectations for policy – including the Fed funds rate. That influences the yield curve for government debt. So policy is a big factor once again.

I think that we agree on most of this. As I noted above the bond yields will be influenced by the market demand which in turn will be influenced by perceptions of policy. But this is just a product of the political arrangements that are in place. The government is in effect allowing bond yields to rise. It could avoid issuing the debt altogether and the central bank would be able to structure whatever rates it chose.

Warren elaborated further on this:

Warren M: Agreed, under current institutional arrangements and biases. But operationally the Fed can set a fed funds rate out to 40 years if it wants to and make a two way market at its desired rates, and that’s where interest rates will be. The Fed in fact sets the term structure of rates one way or another.

Scott F provides this response which covers this and later points:

Scott F: Regarding debt and interest rates, you just repeat the point we ALWAYS make (and which I made in ‘interest rates and fiscal sustainability’), which is that the interest rate on the national debt (and thus the size of debt service and by extension the potential inflationary effects of debt service) are monetary policy variables (or in the case of long-term bond sales, related to expected monetary policy). I don’t see where you’ve found anything here that contradicts our arguments here. So, this is another reason why your overdraft argument misses the point … if there are no overdrafts, then the interest rate on the debt sold is a policy variable. Our overarching point has been that interest on debt is NOT set by markets, but rather is essentially a monetary policy variable. This is also the opposite of the neoclassical view and the perspective taken in policy debates, which constantly worries about markets “downgrading” Treasuries or China “dumping” them or refusing to buy them.

JKH then continues to provide specific examples of statements I have made which apparently suggest the government is more powerful than it is:

JKH: My sense also is that your blog and the others seem to be driven by concerns mostly about the policy errors of excessive government surplus tendencies. That’s fine. But I don’t think it’s ideal to blend that concern into the description of operational capability. My impression is that as a result some operational capabilities are exaggerated while others are dismissed.

The examples then follow. A fundamental modern monetary principle is that the government “has no more or less capacity to spend today because there were surpluses in the past than it would have if there had have been deficits in the past”, which we all agree on.

But for JKH the agreement is only at the “pure operational sense” (based on JKH’s professed undestanding of the way “the balance sheets that are involved work”), which then leads to:

JKH: I disagree in a real world policy sense. There is obviously an implication from these debt decisions for sustainable or unsustainable interest on debt burdens.

Once again, what is the real world policy sense other than a political one. If you are saying that eventually the interest burden becomes so great that the government loses its capacity to service it then I disagree entirely. That would be a “pure operational constraint” and suggest that the sovereign government is not 100 per cent solvent (technically) in its own currency. I am sure you do not think it is exposed to insolvency risk.

If you are saying that at some point the interest payments as a % of GDP become so large and private sector spending is such that there is less non-inflationary room available for other discretionary spending then fine that is what taxation is for – to reduce private spending and/or the government can reduces its own spending somewhat. But before that happens the current account, tax revenue (from higher activity) and saving will be taking up a signifcant

But this is just saying that prudent government net spending is limited by the available real resources in the economy left by non-government saving desires.

There is also a certain irony that the voluntary decision to issue debt $-for-$ to match net spending then increases spending towards the inflation threshold.

If you are saying that the public and the commentators etc are so conditioned that they will invoke political consequences on a government that has a debt ratio above some “acceptable” level (read what News Limited deems in their ignorance to be acceptable!) then that is a political constraint – which doesn’t reflect any financial reality or any physical reality (real resource capacity).

Warren elaborated further on this:

Warren M: These are not constraints for policy makers who understand monetary operations and reserve accounting. They become inflation issues rather than solvency issues when policy makers learn how it actually works. Solvency fears fall under “false constraints” or whatever it’s best to call them.

JKH then suggests the following:

JKH: Alternatively, there is an implication in terms of sustainable or unsustainable inflation consequences if governments and central banks choose to keep their liabilities as excess broad money and bank reserves rather than term bond debt.

So you are saying that the government which wanted to reduce its debt exposure could just stop issuing debt (I agree) and keep net spending (I agree). And I think it should do that more often. What are the implications of that? It depends to some extent on what else the central bank does. Lets assume it pays no support rate on excess reserves although the tendency at present is for that those payments to be made which renders monetary policy (expressed as a desired short-term interest rate target) independent of the level of bank reserves.

What if the government continued to net spend but didn’t issue debt to drain the resulting reserve add? Well the net spending would still occur – we all agree on that. So it makes the point that the debt issuance doesn’t fund the spending but rather drains the reserves. That is, the funds that are “borrowed” by the government come from the government.

But moreover, the interbank interest rate would drop to zero as the banks competed to lend out the excess reserves. So monetary policy would be forced to reflect that.

And consumers might decide to spend the excess reserves on consumption goods (run down their bank deposits) and/or buy other financial assets available in the markets. The former choice will increase aggregate demand and the automatic stabilisers will reduce the budget deficit accordingly. Further, if nominal demand was pushing up against the inflation barrier the government could “ratify” this increase in consumption spending and reduce its own discretionary spending to balance nominal demand with the real productive capacity.

Which brings us to Warren’s response:

Warren M: No, that’s not the source of ‘inflation’ but there are ‘inflation’ issues for excess demand from ‘excessive’ deficit spending.

That is you have to distinguish between the way in which aggregate demand enters the spending stream and the consequences of too much nominal spending relative to real capacity. Government net spending doesn’t have a monopoly on being able to push the economy into the inflation zone. Excessive private investment or burgeoing net exports will do the same.

Scott F also says:

Scott F: Your point about inflation consequences in the case a government/central bank “keep their liabilities as excess broad money and bank reserves rather than term bond debt” is mistaken. Indeed, it is bond sales that are more inflationary, if anything, as they bring an additional interest payment that reserve balances (assuming here they are without interest payment) don’t. Again, see “interest rates and fiscal sustainability.”

JKH then suggests the following:

JKH: While there is no capacity effect in terms of pure operational capability, there is in terms of policy freedom at the margin. The fact that there is complete operational freedom in theory doesn’t mean there is complete policy freedom in practice. That is, unless you think governments can issue money or debt without limit and without other financial consequences.

Who ever said there was complete policy freedom. When I am giving a public presentation I always follow up the statement “the government is not revenue-constrained and can spend whatever it likes” with the next statement “but that doesn’t mean it should spend whatever it likes”. The point is obvious – there are definite economic limits on the ability of governments to spend and they are defined by the real resources that are available at any point for sale and are not being utilised (or purchased). Beyond that you get inflation. Note I use the term economic limits not financial limits. There are no financial limits on a sovereign government – only economic and political limits.

The problem is that the political limits that ideology imposes on government spending have in the past 30 or so years meant that net spending is well below these economic limits and the consequences of that are clear – persistently high labour underutilisation.

Warren’s reaction to this is:

Warren M: Yes, you’ve again defined the problem with our policy makers. The consequences are ‘inflation’ and distributional issues. But they never even get that far. In fact, they never get past the solvency issues and ‘intergenerational’ issues that don’t exist operationally.

JKH continues in this vein:

JKH: Governments face these policy decisions about the amount and cost of their debt and the inflation consequences of the money they produce from their expenditures.

The only concern the government should have is working out how it can achieve full employment and price stability. Having bright minds calculating debt ratios and the rest of it is a waste of time in my view. They should really get to understand how the inflation process operates and also the real capacity of the economy. They should invest in human skill development to improve productivity and make more room for inflation-free production and income generation.

The fact that they have spent the last 30 or so years managing underutilised labour and running down our educational and training systems reflects a wrong set of priorities.

Warren notes that he “wishes they did settle on the inflation issue and do some real work on that. But, as above, they never get that far.”

JKH starts getting bogged down on the inflation point:

JKH: They may not understand how their balance sheets work, but nevertheless these policy questions are real. As Mosler seems to say a lot, it is scary that the people making the policy decisions don’t understand how the government balance sheet and the monetary system work. Fair enough. But this doesn’t negate the fact that these policy decisions are issues for somebody to face. You can’t just run deficits without limit and let expenditures create money without limit, even if the operational capability would allow it. And I don’t believe this line of thinking should dismiss policy concerns as a sort of corollary to demonstrating the correct operational understanding. Policy is a necessary reality.

We all agree that policy is a necessary reality. But what is the point other than if nominal demand grows beyond the real capacity of the economy to absorb it you will get inflation.

Further, a sensible government would be happy to achieve full employment and leave it at that. If it desired a greater share of real resources in the public sector then it doesn’t have to expand net spending (nominal demand) beyond the inflation barrier. It can increase taxation and choke of private access to real resources. Yes, that introduces political issues which are clearly real.

But these political issues are indelibly conditioned by the state of the public debate. When we abandoned full employment and went down the neo-liberal line we abandoned a notion of collective will and replaced it with a heightened sense of individualism. The political rhetoric in those early days of the neo-liberal takeover was designed to engage us as individuals – “there is no such thing as society” (Margaret Thatcher) – “the government budget is just like the household budget” (Malcolm Fraser) and more.

The media vilified the unemployed – and placed the blame for the system-failure (lack of aggregate demand) onto them and reconstructed it as an individual failure – bad skills, bad attitudes etc.

So political issues are malleable by the media and the broader debate. If people (back to my earlier point) were better educated then the way the rhetoric engages with us would be different.

Warren’s reaction to the point is:

Warren M: the limits are your tolerance of inflation and any distributional issues. I agree that policy is a necessary reality … and the problem!

Scott F also says:

Scott F: Again, we don’t think governments “can issue money or debt without limit” without there being consequences, but the consequences of excessive deficits or debt service are inflation, not (involuntary) default. To repeat yet again, we’ve ALWAYS acknowledged this. And we have NEVER proposed running deficits without limit or suggested this was a good idea, and we have repeatedly explained how large a deficit should be without going over.

After exhausting the excessive aggregate demand leads to inflation arguments (which ultimately we all agree on), JKH then turns to a more fundamental disagreement, I sense. This is in relation to another fundamental proposition of MMT that – “it is a nonsensical notion thinking that a sovereign government would ‘save’ in its own currency.”

JKH: This one is different, because it’s an example an operational statement with which I strongly disagree. And I think it’s a case where a policy bias against government surpluses has influenced the operational description. I will argue that the fact that a government issues its own currency has nothing whatsoever to do with its capacity to save. Here I’ve reversed roles in the sense that I’m talking about pure operational capability before turning to the question of policy restraints or constraints. First, consider (an operationally feasible case) where the central bank issues currency by purchasing private sector financial assets, rather than by purchasing government debt. And suppose the government has run a balanced budget up to this point. Now suppose the government decides to run a surplus. Now suppose the government decides it wants to manage its books “in paradigm” … So in preparations for a surplus operation, the government “prepares” its balance sheet as follows. The government purchases existing non-government assets (probably financial assets) in an amount equal to the size of surplus it intends to run. At this stage, the government has simply acted as extension of its central bank, putting more money into the system by acquiring existing assets. So the government balance sheet, separated from but connected to the central bank balance sheet, shows non-government assets acquired and an overdraft at the central bank.

Well both Warren and I agree that the private sector now would have new reserve balances at the central bank equal to the financial asset purchases and that their net financial asset position hasn’t changed.

JKH then continues the example:

JKH: The government then taxes the equivalent of the overdraft. Considering this flow effect for its balance sheet consequences, the government has created its own equity position (logical and economic, if not in actual bookkeeping). I.e. the government balance sheet (ex the central bank) now includes non-government assets supported by an equity position on the right hand side. This is the canonical example of something that I’ve argued elsewhere, particularly with occasional Mosler commenter “Winterspeak”. i.e. taxation is a flow that produces government equity at the margin. Equity of course is a balance sheet representation of saving, which is exactly what I’m arguing happens in this example.

We have to be clear on the terms we use. When economists talk about saving they are typically relating it to a household decision to defer consumption (now) to provide for greater future consumption. There are elaborate models which describe the so-called intertemporal decision making environment and its consequences in this regard. But the point is directly related to the fact that households which use the currency that the sovereign government issues are revenue-constrained.

This means that a household has to finance every $of spending either by working (earning income); borrowing; running down assets previously accumulated which include prior saving. So the act of saving increases the future spending possibilities but, other things equal, reduces the current spending possibilities. Applying this logic to a government which issues the currency is tantamount to saying that at each point in time that government is financially (or revenue) constrained and has to save now to spend more in the future. This is the basis of the intergenerational myths etc. It is nonsensical because the sovereign government is not revenue-constrained as you acknowledge so whatever accounting gymnastics it performs on its balance sheet, the government is no better able to spend tomorrow as it is today in financial terms. Its spending today might condition its spending room later if the economy is closer to full employment but that is once again the economic (that is, non-inflationary) limits on spending and has nothing to do with financial constraints being eased by prior saving. Warren adds: You can call it that and account for that as ‘equity’ but it is of no practical concern. It doen’t make government any ‘richer’ and it does remove net financial assets from the private sector. It’s an accounting entry. And you can call it anything you want. But it isn’t something that alters the government’s ability to spend at will. It is nothing but record keeping – an entry that “accounts” for what was done but has no causal function. Scott F says: Scott F: Bill’s point that it is “nonsensical” to think of a sovereign government saving in its own currency is NOT about balance sheets. Throughout your critique, you equate “operational” with “accounting”. But they are not always the same. Accounting is a subset of operational realities, not the equivalent. Operational abilities is a broader term that includes rules (sometimes), laws (sometimes), and logical imperatives of system behavior or function resulting from (for instance) alternative monetary regimes. (Much more could be said here on the differences, but I’ll leave that for another time.) None of us has never said that a government can’t “save” in the accounting sense or that it can’t have “equity” on its balance sheet (indeed, it does already). Our point is that a government that does these things do not build up or otherwise enable some additional or otherwise improved “operational ability” to spend now or later as would be the case for a non-currency issuer. Such additional ability is how a government surplus or saving is normally understood (also for the central bank, as the recent debates over the Fed’s equity demonstrate) by economists and in policy debates, and that is what we are critiquing. JKH then continues to push this equity point: JKH: This type of equity is the pure right hand balance sheet type, in this case not represented by an equity financial claim, as is the case with a corporate entity. It is more comparable to household equity as a representation of cumulative net saving. In both cases, equity can be marked to market at one’s choosing, where cumulative net saving undergoes an additional valuation adjustment. But this is just a detail for purpose here. The government in this example has saved. To which Warren replies: Warren: Except that household equity gives households something needed to spend and collateral to possibly use for loans. As before, this government “equity” does nothing for government’s ability to spend at will. Your definition. Functionally, nothing has changed regarding government’s financial position, as above. Government never “has” or “doesn’t have” it’s own currency any more than a football stadium has or doesn’t have points to put on the scoreboard. Once again we need to keep this point in the perspective of the analogy that mainstream macroeconomics draws between the household and the sovereign government. That is the context in which I make the points about a government saving. Scott F notes: Scott F: Regarding accounting and saving, we use the term ‘net saving,’ which isn’t the same as ‘saving’ since the latter usually includes private capital spending (and in that case, the ‘saving’ offset is only the accounting record of the capital spending). And by accounting identity, the non-government sector cannot ‘net save’ unless the government runs a deficit. Again, nobody said the government can’t save in the accounting sense … but rather that it is nonsensical to think that when it does, it enables more spending now or later . . . any other interpretation of Bill’s quote seems to me to be off and conflating ‘accounting’ with ‘operational.’ Certainly, there is a difference between surpluses and deficits in terms of the ‘constraint’ of inflation, but we’ve always acknowledged this. JKH then moves onto public surpluses: JKH: This government surplus position creates a net non-government deficit in the sense of the Kansas City sector financial balances model. The SFB model is defined relative to netting out investment from the non-government sector balance sheet. So the government surplus in this case serves as an offset and partial match for investment, equivalent to the shortfall in non-government saving. (I’m afraid to use the word ‘funding’ here, for fear of being excoriated for false causality. That’s another unfortunate perception problem I won’t get into here – the idea that words previously associated with an improper operational model retain a patina of ineptitude if not criminality when used in association with the correct operational model.) That partial offset to investment again proves the existence of saving. A red herring threatens beneath all this. It doesn’t matter at all that the government acquisition in this case is a stock (i.e. existing asset) as opposed to a flow (i.e. real GDP component) purchase. I.e. it doesn’t matter that it is simply an exchange of newly created money (or bonds asset) for an existing real or financial asset. All that matters is that I have created the capacity for a government balance sheet equity position, and I’ve even done so by remaining “in paradigm” First, the language issue. Paradigms reinforce themselves with a particular language set and the cogniscenti imbue themselves with “understanding” each time they use the terminology. A professor once said to me that the process of university education is to take an otherwise bright person with common sense and intuition and turn them into a confused, jargon-babbling ideologue. The more serious point is that if you are desirous of breaking away from a dominant paradigm then you should avoid using their terms. They are loaded. Warren concluded: Maybe not quite as you are assuming there is some operational further purpose to this entry you call ‘equity’ JKH continues JKH: I.e. The acquisition of assets has created its own (money) finance, and taxation has withdrawn money and reserves from the system, in the process creating the actual equity position. And finally we can ignore from an economic perspective the fact that governments tend to misclassify outlays as between expenditures and investments or financial assets. Such misclassification is a non-issue in terms of the true economics, and the true identification of government saving or dissaving. Moreover, all this has absolutely nothing to do operationally with the fact that the government issues its own currency when it spends. To be sure, the issuance of currency allows expenditures and deficits to be offset operationally with money creation. But money creation doesn’t require deficits. Just look at my example again where the central bank creates money by acquiring non-government assets and the government acquires additional non-government assets in generating an ultimate surplus. Similarly, when the government runs a budget surplus and pays down pre-existing debt (representing cumulative prior deficits), it saves. Once again saving to a household reduces current consumption and increases future consumption within an intertemporal household budget constraint. This constraint is not relevant to the sovereign government and so the terminology is not applicable. Warren repeats this point: Warren: By your definition, which is fine. just that there are no operational consequences from this entry. Scott F has another point here: Scott F: Regarding the point that ‘money creation doesn’t require deficits’ … none of us ever said otherwise. We are endogenous money proponents, after all. First, best to get rid of the term “money” altogether when we discuss this stuff in such detail. “Money” is always someone’s liability, so best to just say whose liability you are talking about to avoid confusion. Anyway, what we say is that private sector credit creation (which creates deposits or other “near monies” on private balance sheets) is effectively a short position in or leveraging of government liabilities. But the amount of “shorts” that can be privately created are obviously never limited by the quantity of government liabilities. JKH continues JKH: The fact that somebody has a net liability position has nothing to do with whether or not it’s possible for them operationally to save. The fact that governments tend to run chronic net liability positions similarly has nothing to do with their operational capacity to save. From a policy perspective, that saving affects a governments ability to borrow in the future. There is less debt on the balance sheet, and less debt servicing interest costs. Obviously this isn’t the case from an operational perspective, which is your ongoing point. A past surplus has never improved the government’s ability to borrow in the future. Further while lower debt levels mean lower interest servicing costs that doesn’t have any operational implications for the sovereign government. Another point I made was that “There is no storage shed in Canberra or Washington or anywhere else where the surpluses are saved up and available for the government to drive a truck down and pick up some dollars to spend.” To which JKH replied: JKH: Neither is there when I save to pay down my credit card debt. Warren responds: Warren: Yes there is, in the sense that operationally you are more able to spend when you have savings and/or credit etc. The governmnet is never more or less able in any case. To understand why JKH is in error here you have to go back to the analogy between the revenue-constrained household and the sovereign government. If the household (you and I) pay down our credit card we now have more “revenue-capacity” than we had previously. This is functionally equivalent to saving. Paying down our cards means we are consuming less now to allow us to spend more in the future. So the conception of putting more “spending capacity” into the shed for the household is apt. But it is never so for a sovereign government. Scott F says: Scott F: Bill’s statement about the storage shed refers to the fact that most think there is something like this. Again, he wasn’t suggesting otherwise > for the non-government sector, but merely critiquing those who think that government revenues enables more government spending the way that income enables more private sector spending. Another fundamental proposition of MMT is that: Surpluses destroy financial assets that were previously in the hands of the non-government sector and these assets are gone forever. In reply, JKH then says: JKH: (With all due respect) so what? The same goes for my credit card company when I save to pay down my balance. Is this an operational point, a policy point, both, or neither? It appears to be a truism operationally, and from a policy perspective what’s the problem with paying down a little debt once in a while? Well we are generally respectful here! And my reaction to the point depends on the state of the business cycle, the state of the external sector, and the extent of underutilised labour resources. If you are at full employment (2 per cent official unemployment, zero hidden unemployment and zero underemployment) then you might want to run surpluses if net exports are strong enough to fund the domestic saving desires. You would not want to run a surplus if the non-government sector was desiring to save and you were not at full employment. Warren says: Warren: There is nothing wrong with paying down debt, if you are trying to cool aggregate demand. But there is a lot wrong with it if there’s a shortage of agg demand Scott F offered this insight: Scott F: Regarding surpluses destroying non-govt financial assets, you’ve missed the point again. When you pay down your credit card, there is no net change for the non-govt sector in terms of net financial assets … the asset and liability were privately held and are both destroyed. But when the goverment pays down its debt, the private sector asset is destroyed, but no private sector liability is destroyed. It’s an accounting point, but it also counters the fact that most think paying off the national debt increases the private sector’s ability to finance capital, when in fact (by accounting definition) it reduces net financial wealth of the non-government sector. Of course, this doesn’t affect the “operational” ability of the non-government sector to finance more capital, but the likelihood of this could be greater with more net financial wealth than with less. The operational ability to finance private capital isn’t the point, though, but rather the facts that with a surplus (1) there is less net financial wealth for the non-government sector, (2) there is certainly no more ability for the private sector to finance capital spending than before the surplus, and (3) most in the policy debates think the reality for (1) and (2) is the opposite. JKH then says: JKH: And from a policy perspective, where’s the logical and mathematical dividing line between running a deficit that is less than infinity and running one that is occasionally less than zero? Who made up that rule and exactly what is it? That is very clear. At all times the public purpose role of fiscal policy should be to ensure that aggregate demand is sufficient to generate output that will create enough jobs to fully employ the labour force. If a budget surplus can do that then fine (for exammple, when there is a low domestic saving desire and significant net exports). But typically the fiscal drag coming from a budget surplus will undermine full employment. It is also unlikely that an infinite (close to it) deficit will be required. Policy is about balance – but at least we should set the goals correctly and make unemployment a policy target again rather than a policy instrument as it has become under the neo-liberal obsession with surpluses and inflation-first monetary policy. Scott F says: Scott F: Our concern about surpluses has to do with two things. First, they reduce net financial assets of the non-govt sector, and in a Minskyan sense, this is ultimately problematic since the non-govt sector is obviously not the currency issuer. Second, because the non-govt sector is a currency user and not the issuer, the non-govt sector will over time (not necessarily every year, but on average) want to net save. Persistent surpluses, paying off the national debt, or basic “sound” fiscal policy that we hear about in the policy debates are all inconsistent with these points. Note that I haven’t even mentioned “operational capacity”, which you seem to think is the overarching foundation for our argument here. Conclusion Long comments invoke long responses. Exhausted! Spread the word ... This Post Has 47 Comments 1. Lefty says: My brain is full to overflowing. I think I will go get a beer and forget about economics for this evening. 2. Good move Lefty. 3. Alan Dunn says: Great so it looks like things have ben reduced to an ideological stalemate. In one corner we have the Modern Monetary theorists who believe that anyone who wants a job should be given the opportunity to have a job. In the other corner we have the Neo-conservatives who think that we need a pool of unemployed to use as a shock absorber to control inlfation. Peas in a pod. 4. Lefty says: And so the war goes on. Endlessly. Economics and politics cannot help but go hand in hand. Even if you understand the functioning of the modern fiat monetary system, I think the broad right will always lean towards neo-liberalism because it naturally favours capital over labour. I have sworn allegiance to the modern monetary corps. 5. In the other corner we have the Neo-conservatives who think that we need a pool of unemployed to use as a shock absorber to control inlfation. WITH THE JG PROPOSAL WE USE A POOL OF EMPLOYED RATHER THAN A POOL OF UNEMPLOYED. AN EMPLOYED POOL IS A FAR MORE EFFECTIVE BUFFER STOCK THAN AN UNEMPLOYED POOL. BUSINESS MUCH MORE READILY HIRES THOSE ALREADY WORKING OVER THOSE WHO ARE UNEMPLOYED. AND THE LONGER THE TIME OF UNEMPLOYMENT THE LESS LIKELY IT IS BUSINESS WILL HIRE YOU. TO BE EFFECTIVE A BUFFER STOCK HAS TO BE ‘FLUID’ 6. Ah, “he said, he said”. Alan, are ready to get a full time job as a reporter for the mass press … just as the mass press is rapidly losing the ability to hire. There is more than the policy argument at hand, though … the neoliberals bolster their policy argument about the desirability of forcing people into unemployment against their will and beyond their ability to do anything about, with a fallacious model of the economy in which that is desirable because of fictitious economic consequences of giving those people jobs, and by assuming non-existent constraints on the way that that employment may be provided. Provided a neolib acknowledges that they are arguing in favor of unemployment on behalf of sectoral interests rather than adopting the pretense that it is for the general good, I for one am happy to have the policy debate in public forums. Indeed, the need to develop and promulgate fallacious models of the macroeconomy in order to support the neolib policy argument would seem to be one indicator of how well they think that policy argument would go if it was pursued on a fair and level playing field. 7. Vinodh says: Good debate. It helps clarify a lot of issues for a person with new found interest in economics. 8. Alan Dunn says: The Neo-liberals have / are going to great lengths to stay in power so to speak. One of my favourite things in the world is to ride BMX (yes yes kids bikes I know – but I love em). On all of the many BMX sites I attend there is always, without fail, a character who has never talked about BMX, never posted a pic about BMX, or indeed contributed anything even resembling BMX. The sole purpose of these people is to influence and form opinion as to the benefits of the neo-liberal ideology. They constantly refer to bogus studies and statistics and through theses actions they slowly but surely build their army for the future by enlisting teenagers who will one day become tax payers and voters. Hence, the truth just doesn’t matter when it comes to these bastards and I’m pretty sure the solution to getting rid of them does not reside around academic debates over how a modern money economy functions. You are fighting the battle on their terms and given the time they have been in control and the size of their army you probably won’t win. 9. JKH says: Bill, Thank you for a very detailed response. And thanks also to Scott and Warren directly and/or indirectly for their feedback. I appreciate it. Scott says, “Of course, as you probably expect, I think your entire critique either misinterprets or otherwise misunderstands the points we’ve made.” The crazy thing is that I agree with most of the points made, except that I don’t think they necessarily reflect what I’ve tried to say. There is far more agreement I think than is first evident. How can this be? Let me try and close the gap. A lot of the difference may be due to the interpretation of language – especially the words “policy” and “constraint”. E.g. Bill, you say, “I first consider that what is held out as a financial constraint is usually not that at all. Typically, in macroeconomic policy the constraints are political and voluntarily imposed.” I agree. That’s how I’ve intended the meaning in this case. Let me elaborate by outlining how I view policy more generally. Much of this may seem obvious and mundane, but it’s important to how we are communicating on the subject, and why I wrote what I did. I tend to view the financial world in terms of balance sheets. People are responsible for managing those balance sheets, explicitly or de facto. I think there are valid comparisons between the roles that people play in managing non-government and government balance sheets. Part of the underlying theme on your blog(s) is that government and Congress don’t understand the broad form government balance sheet well enough to formulate policy effectively. The management of these balance sheets in my view includes policy, strategy, and operational aspects. Banks for example tend to have high level policy committees where business managers go and present their strategies for approval and then go away and implement them. In approving strategy, the committee sets policy, including parameters such as limits relating to discretion in implementation. This is all done against a background of presumed and/or demonstrated operational understanding. To simplify this view, we can blend the idea of policy and strategy, and consider strategy just the implementation of an approved policy. So the balance sheet management world then consists of policy and operations. And policy includes constraints that are “political and voluntarily imposed”. Broadening the application of politics to include “corporate politics”, the government and non-government worlds operate roughly according to similar policy processes. It is very easy to visualize a presentation to a Congressional Committee by the Fed Chairman or the Treasury Secretary and transform it to a similar policy process that happens in the corporate world. So the government, broadly defined, encompassing both Treasury and the Federal Reserve, operates according to this type of policy process. I’m not sure where people stand on the issue of Fed independence exactly, but it seems natural to me to think of a “Monetary and Fiscal Policy Committee” formulating policy on the basis of a correct operational understanding as well as the portfolio effect of monetary and fiscal policy. Whether such a committee should be a creation of Congress or otherwise is something for a different debate. As Dylan said, “You gotta’ serve somebody.” I’ve used the word “constraint” or “policy constraint” in an unfortunate way. I do not mean it in the sense of a financial constraint. I mean something that is self-imposed, for example as in such a policy committee process. In the case of a private sector bank, it includes risk limits. In the case of government, it becomes political by definition, although informed political judgement should include intelligent assessment of economic risks such as inflation. In a sense, risk limits are what Congress deals with now, on occasion. That was the case with TARP for example. Sometimes a limit isn’t really a limit in the sense that the implementer can always come back as ask for more, if he dares. That was the political risk with TARP. I presume the budgeting process is framed implicitly or explicitly in the sense of limits. We agree there are no operational or financial limits to government expenditure, just as we know there are no operational or financial limits to central bank asset acquisition. Suppose you, Bill, or Scott or Warren were responsible for presenting recommended monetary and fiscal policy to such a committee. The ground work would include the educational process of explaining how in fact there are no operational or financial constraints in fiscal and monetary policy. The Committee then says, “We think we understand that, now, thanks. What’s next?” Such a committee is responsible for becoming comfortable with the recommendations for action presented to it, and for sending the proposers away with marching orders and constraints that reinforce that comfort. The committee process is such that the proposer should also present the approval framework which will allow the committee to attain this comfort level. This allows you more control over the process and the outcome. You recommend not only the substance of your proposal, but the format in which you would like them to respond to your proposal. Therefore, you must recommend the nature of the constraint will be self-imposed (on all, but also by them on you) when you go away to implement. The alternative is “trust me”. That doesn’t quite work, as Paulson found out when he was on the other end of the approval process than was the case at Goldman, apparently. Politically imposed constraint, necessarily formulated in economic terms, is a reality, whether that politics occurs in government or non-government. The trick is somehow to get the political process properly informed in economic understanding as groundwork for the thinking that is required to proceed intelligently and dynamically with policy formulation on an iterative basis. More specifically, how would you deal with the question of the deficit and the debt at such a committee? Scott said: “We have ALWAYS and REPEATEDLY acknowledged the inflationary potential of either debt service burdens or just overly large deficits in general. Our point is that THIS should be the issue to debate in regard to macro policy, NOT the stuff that comes from a flawed understanding of operations or the monetary system.” I agree. Scott further: “Again, our point is that unless you get rid of the flawed intellectual foundations of current policy debates, you can’t have an appropriate discussion of policy or of the “practical constraints” that definitely do exist and which we acknowledge exist” I agree. I would also say that a solid intellectual foundation is a necessary but not sufficient condition for good policy. I guess that’s obvious. Bill: “So far from ignoring these voluntary constraints I am always writing about them – to emphasise the political nature of the decisions taken and the way the options are presented. But as an educator I think it is also important to provide as much detail as I can about the way the system actually should function from first-principles so that the public can make better conclusions about the way the political process deviates from the intrinsic.” I understand now that in that sense I’ve overlooked your acknowledgment of the necessity for voluntary constraints. “JKH: I’m sure you can read my mind for the rest of the argument from there. Suffice to say, there are decisions to be made about converting that overdraft into debt issued by the government, etc. Those are policy decisions that act as constraints on the operational freedom that is the foundation of this line of thinking.” Bill in response: “We all agree on that.” Good. Bill: “The point is obvious – there are definite economic limits on the ability of governments to spend and they are defined by the real resources that are available at any point for sale and are not being utilised (or purchased). Beyond that you get inflation. Note I use the term economic limits not financial limits. There are no financial limits on a sovereign government – only economic and political limits.” Agreed, and I guess this summarizes a misunderstanding in my communication. This is quite apt for how I view the correct policy process against the backdrop of the correct operational understanding. Now some further comments on operational aspects: “Scott F: Your point about overdrafts again misses the point we are making. The deficit always comes first unless the non-government sector borrows to get reserve balances to buy the Treasuries. You can’t buy a Treasury without reserve balances (that’s how they clear), and the reserve balances exist either because of a previous deficit, previous open market purchase (which purchased Treasuries from a previous deficit), or (as mentioned) borrowing from the central bank. Our point is that a deficit for the currency issuer is NEVER about borrowing, and bond sales are NEVER finance operations.” I’m not sure what point I was missing in referencing the overdraft aspect. It may be a point I understood, but wasn’t intending to hit with that reference. I believe I understand your point above, and haven’t contradicted it with my point. My point is that somebody has to make a policy decision about issuing bonds. It was not focused on operational mechanics at the reserve level. That said, I will make a further operational point on this now. Nobody can buy anything without reserve settlement, where the buyer and seller have different bankers. If I buy a bond from you, my bank will settle with your bank in reserves. If I buy a bond at auction, my bank will settle with the government’s bank (the central bank) in reserves. I recognize that the central bank is settling the transaction uniquely using its own “liability” as the means of payment. I think that’s the critical point. From that, in the case of the government and the central bank, flows the necessity for the consolidated entity to make an active or passive policy decision on the nature of the ultimate monetary base/debt mix. And while I’m on this point, let me say something a little more controversial, with which you may or may not agree. I notice a tendency in general (but not always) for people to conflate transactions undertaken by non-banks (e.g. households and businesses) as if they were actually transacting directly in reserves. Non-bank system participants do not settle their transactions with banks and each other in terms of bank reserves. Banks settle their transactions with other banks in this mode. A bank’s net reserves due or owed position is a mirror profile of the balance sheet mismatch that results from the random pattern of asset and liability transactions throughout the day between that bank and its non-bank counterparties, as well as with other banks. I’ve never been comfortable with references to non-banks having their own effect on the deployment of reserves, just as I am not comfortable with the language that banks “lend reserves” to non-banks. The non-bank system does not interface directly with the reserve system (other than through the convertibility of deposits for central bank notes.) Banks reserves are not accessible to non-banks. Agree or disagree? Do people that should know better sometimes get sloppy with language here? Or am I wrong on this point at an operational level? “JKH then goes onto to discuss interest rates and seems to be saying that the central bank doesn’t have control of the short-run rate.” I didn’t intend to say this. I think the central bank controls the short rate. I think the market has something to say about pricing the long rate, using expectations for the short rate as it chooses. Bill: “I think that we agree on most of this. As I noted above the bond yields will be influenced by the market demand which in turn will be influenced by perceptions of policy. But this is just a product of the political arrangements that are in place. The government is in effect allowing bond yields to rise. It could avoid issuing the debt altogether and the central bank would be able to structure whatever rates it chose.” I agree, assuming the alternative is the Fed funds rate. That would amount to a multi-trillion dollar bank reserve position with interest paid at something close to the Fed funds rate. Warren: “Agreed, under current institutional arrangements and biases. But operationally the Fed can set a fed funds rate out to 40 years if it wants to and make a two way market at its desired rates, and that’s where interest rates will be. The Fed in fact sets the term structure of rates one way or another.” I defer, but must do more thinking to work this through. For some reason, control of the overnight rate is more obvious to me, and less so for the term rate. To be honest, I really haven’t thought about it yet. But it’s an important point, I think. Scott F: “Regarding debt and interest rates, you just repeat the point we ALWAYS make (and which I made in ‘interest rates and fiscal sustainability’), which is that the interest rate on the national debt (and thus the size of debt service and by extension the potential inflationary effects of debt service) are monetary policy variables (or in the case of long-term bond sales, related to expected monetary policy). I don’t see where you’ve found anything here that contradicts our arguments here. So, this is another reason why your overdraft argument misses the point … if there are no overdrafts, then the interest rate on the debt sold is a policy variable. Our overarching point has been that interest on debt is NOT set by markets, but rather is essentially a monetary policy variable. This is also the opposite of the neoclassical view and the perspective taken in policy debates, which constantly worries about markets “downgrading” Treasuries or China “dumping” them or refusing to buy them.” Again, I don’t know what point you expected me to hit with the overdraft reference. I viewed it as a simple operational fact and the prelude to a necessary policy choice, active or passive. Also, you refer to “expected monetary policy” in one sentence and then say “interest on debt is not set by markets” in the next. This makes no sense to me. Whose expectations are involved in formulating expected monetary policy if not the market’s? Perhaps you can shed some light on this. But combined with Warren’s 40 year point, I need to think about this area more. Bill: “JKH: I disagree in a real world policy sense. There is obviously an implication from these debt decisions for sustainable or unsustainable interest on debt burdens. Once again, what is the real world policy sense other than a political one? If you are saying that eventually the interest burden becomes so great that the government loses its capacity to service it then I disagree entirely. That would be a “pure operational constraint” and suggest that the sovereign government is not 100 per cent solvent (technically) in its own currency. I am sure you do not think it is exposed to insolvency risk. If you are saying that at some point the interest payments as a % of GDP become so large and private sector spending is such that there is less non-inflationary room available for other discretionary spending then fine that is what taxation is for – to reduce private spending and/or the government can reduces its own spending somewhat. But before that happens the current account, tax revenue (from higher activity) and saving will be taking up a significant part of the adjustment.” I agree with all of this. The self-imposed policy constraint I referred to earlier clearly has to do with economic and inflation risk, not operational financing risk. This is becoming clearer now and I think we are agreeing more. “Warren M: These are not constraints for policy makers who understand monetary operations and reserve accounting. They become inflation issues rather than solvency issues when policy makers learn how it actually works. Solvency fears fall under “false constraints” or whatever it’s best to call them.” I agree that solvency fear is a “false constraint”. Conversely, inflation issues should warrant intelligent policy constraints. Bill: “What if the government continued to net spend but didn’t issue debt to drain the resulting reserve add? Well the net spending would still occur – we all agree on that. So it makes the point that the debt issuance doesn’t fund the spending but rather drains the reserves. That is, the funds that are “borrowed” by the government come from the government.” Agreed – the debt issuance certainly “drains reserves”. The commercial bank settles with the central bank using the central bank’s own “liability” as the means of payment. The commercial bank is paying the central bank what it owes it in settlement. That said, it occurs to me that another way of thinking about this could be to view net spending as creating reserves, with bond issuance constituting the “refunding” of that “liability”. It strikes as slightly artificial to be completely dismissive of the language “borrowing” or “funding” in this sense. The important point is that it is not operationally necessary, and it is a policy choice. Perhaps the conventional language becomes slightly more palatable when viewed strictly in this conditional sense. That’s just a thought off the top of my head. “Scott F: Your point about inflation consequences in the case a government/central bank “keep their liabilities as excess broad money and bank reserves rather than term bond debt” is mistaken. Indeed, it is bond sales that are more inflationary, if anything, as they bring an additional interest payment that reserve balances (assuming here they are without interest payment) don’t. Again, see “interest rates and fiscal sustainability.”” I need to do more work in this area. Bill: “The point is obvious – there are definite economic limits on the ability of governments to spend and they are defined by the real resources that are available at any point for sale and are not being utilised (or purchased)). ‘ I agree. This is the basis for the correct policy framework that I envisage in positioning it relative to the operational understanding. And: “The problem is that the political limits that ideology imposes on government spending have in the past 30 or so years meant that net spending is well below these economic limits and the consequences of that are clear – persistently high labour underutilisation. Warren’s reaction to this is: Warren M: Yes, you’ve again defined the problem with our policy makers. The consequences are ‘inflation’ and distributional issues. But they never even get that far. In fact, they never get past the solvency issues and ‘intergenerational’ issues that don’t exist operationally.” And: “Warren M: the limits are your tolerance of inflation and any distributional issues. I agree that policy is a necessary reality … and the problem!” I think this is consistent with how I see the juxtaposition of policy and operations. “Scott F: Again, we don’t think governments “can issue money or debt without limit” without there being consequences, but the consequences of excessive deficits or debt service are inflation, not (involuntary) default. To repeat yet again, we’ve ALWAYS acknowledged this. And we have NEVER proposed running deficits without limit or suggested this was a good idea, and we have repeatedly explained how large a deficit should be without going over.” I understand now. I think you’re saying that there should be (policy) limits in some sense, but that the right policy must be formulated on the basis of analysing the right economic variables (e.g. inflation risk) and on the basis of correct operational understanding. Bill: “We have to be clear on the terms we use. When economists talk about saving they are typically relating it to a household decision to defer consumption (now) to provide for greater future consumption … the sovereign government is not revenue-constrained as you acknowledge so whatever accounting gymnastics it performs on its balance sheet, the government is no better able to spend tomorrow as it is today in financial terms. Its spending today might condition its spending room later if the economy is closer to full employment but that is once again the economic (that is, non-inflationary) limits on spending and has nothing to do with financial constraints being eased by prior saving.” Here I would say that if the government has a soundly constructed policy view of debt, as per the previous point, then it’s spending and deficit capacity from the same perspective will be seen in a different light, depending on the starting point for outstanding debt. Pure operational capability is constraint free and evergreen. But freedom to manoeuvre according to policy over time will be affected by the pattern of utilization of limits under that policy. “Warren adds: You can call it that and account for that as ‘equity’ but it is of no practical concern. It doesn’t make government any ‘richer’ and it does remove net financial assets from the private sector. It’s an accounting entry. And you can call it anything you want. But it isn’t something that alters the government’s ability to spend at will. It is nothing but record keeping – an entry that “accounts” for what was done but has no causal function.” I agree. Equity is a device I use occasionally for my own interpretation of the implied complete balance sheet position. “Scott F: Bill’s point that it is “nonsensical” to think of a sovereign government saving in its own currency is NOT about balance sheets. Throughout your critique, you equate “operational” with “accounting”. But they are not always the same. Accounting is a subset of operational realities, not the equivalent.” I agree. I have a tendency to conflate operational and accounting issues. It is a bad habit that I must grow out of. Scott: “None of us has ever said that a government can’t “save” in the accounting sense or that it can’t have “equity” on its balance sheet (indeed, it does already). Our point is that a government that does these things do not build up or otherwise enable some additional or otherwise improved “operational ability” to spend now or later as would be the case for a non-currency issuer.” I agree as an operational observation. “Warren: Except that household equity gives households something needed to spend and collateral to possibly use for loans. As before, this government “equity” does nothing for government’s ability to spend at will. Your definition. Functionally, nothing has changed regarding government’s financial position, as above. Government never “has” or “doesn’t have” it’s own currency any more than a football stadium has or doesn’t have points to put on the scoreboard.” Again, I agree entirely as an operational point. However, my extreme example shows the government with a position net long private sector assets, in addition to those held as the source of currency and reserve issuance. From a policy perspective, I think the net asset position changes the all-in constraint on future government debt and deficits, insofar as this relates to earlier points about limits to debt levels as affected by economic and inflation risk. Scott: “Certainly, there is a difference between surpluses and deficits in terms of the ‘constraint’ of inflation, but we’ve always acknowledged this.” Again, we are in agreement. Bill – fair point on the language issue. “Warren: By your definition, which is fine. just that there are no operational consequences from this entry.” Agreement again, I think. Bill: “To understand why JKH is in error here you have to go back to the analogy between the revenue-constrained household and the sovereign government. If the household (you and I) pay down our credit card we now have more “revenue-capacity” than we had previously. This is functionally equivalent to saving. Paying down our cards means we are consuming less now to allow us to spend more in the future. So the conception of putting more “spending capacity” into the shed for the household is apt. But it is never so for a sovereign government.” Once again, I agree (except for the part that suggests I’m in error) at the operational level. But it makes a difference relative to a self-imposed policy guideline, based on a proper analysis of economic and inflation risk. “Warren: There is nothing wrong with paying down debt, if you are trying to cool aggregate demand. But there is a lot wrong with it if there’s a shortage of agg demand” Agree again on relative policy orientation. “Scott F: Our concern about surpluses has to do with two things. First, they reduce net financial assets of the non-govt sector, and in a Minskyan sense, this is ultimately problematic since the non-govt sector is obviously not the currency issuer. Second, because the non-govt sector is a currency user and not the issuer, the non-govt sector will over time (not necessarily every year, but on average) want to net save. Persistent surpluses, paying off the national debt, or basic “sound” fiscal policy that we hear about in the policy debates are all inconsistent with these points. Note that I haven’t even mentioned “operational capacity”, which you seem to think is the overarching foundation for our argument here.” I understand the policy concern. Also, it’s a fair criticism that in all cases (Bill, Scott, and Warren), I am probably guilty of understating or under-acknowledging the depth of the policy work that accompanies the operational understanding. My point here is that this was my “first impression” of the way in which the operational argument was being positioned – as necessary and sufficient, if you will, rather than necessary and necessarily combined with some sort of policy process that in effect puts properly conceived self-imposed policy limits on an otherwise unlimited operational capability. This may seem only like common sense to you, but this was my impression. I see now that what I believed to be missing on the surface is obviously there in force along with the operational emphasis. There is much there to agree with, both at the operational and policy level. And there are some nooks and crannies I need to explore further on the operational side, and hopefully I can do that with more clarity now as a result of this discussion. Thanks again for the very thorough and illuminating responses. P.S. I have rarely seen the printed page actually frown at me as much as in the case of my use of the word “monetization”, and earlier on Warren’s blog, with “sterilization”. I know there must be very good reason for this, and shall try and keep it in mind. The good news is that I don’t believe I’ve ever used the phrase “printing money”. 10. Ramanan says: JKH, Lots of things – and like your attitude of not accepting things without careful logical analysis and its relation to empirics. Few things from my side in random order. One thing caught my attention – control of long term rates. If you agree with Bill that government bonds need not be issued then it becomes not to difficult to see that. Let us say that in an economy there are no bonds and that the Treasury decides to issue bonds for the public. Lets for simplicity assume that its a perpetuity which pays$1 every period. The treasury can insist that the yield is 2% per annum. Take it or leave it.

Answer to “So what?” – If Bill, Scott and Warren were to be asked this question by policy makers after they have understood Modern Money, they would make great policies for reducing unemployment and various other things for increasing the quality of life. Such schemes are already in place at some level in some countries

On your question on reserves – Yes banks do not lend those “reserves” (though cash transactions can create problems with terminologies). If the banking system has $100b in reserves, a loan of say$1m to say, a farmer does not change that $100b by a penny. Thats what I have learned in this blog. Going to policies, I think I understand your concerns (that is what everyone new to this blog has). If I may summarize your views in the form of a question – (using some words in your comments) – between the deficit of infinity and zero, what level of deficit is the correct one ? The answer to that question is – like it or not – “depends”. In countries like Japan people want to save more (lots!) and the government has to run huge deficits. In countries like the US, it may not be high (except in 2009/10). It also depends on the past – remember just flows are not enough – stocks (of assets like bonds, real estate, mutual funds, stock/equity) have to be looked at too. If there is a massive fall in the stock market, people will try to save more. People consume not only out of their disposable incomes but also the wealth accumulated in various forms. In effect, the level of deficit a government is not the most trivial question to answer. In addition, the government also has to worry about price stability. The trade deficit is also another thing to look at – if the private sector does not want to finance the savings of the Rest of the World, the government has to. (Please note this kind of logic is opposite of the traditional neoclassical economists who would say that the ROW is financing). 11. Bill, I was wondering if you could clarify a couple of points. The first is in relation to deficits pushing down interest rates. In the scenario discussed, reserves are not drained by bond issuance and then you say: But moreover, the interbank interest rate would drop to zero as the banks competed to lend out the excess reserves. So monetary policy would be forced to reflect that. I would have thought that in this scenario, aggregate reserve balances would become sufficiently large that is would be unlikely on any given day that a commercial bank’s reserve balance would do into debit (or fall below the minimum reserve amount in countries where that applies). Since banks only lend overnight to other banks which need to top up their reserve requirements, I would have thought that rather than competing to lend and pushing the overnight rate to zero, there would generally be no interbank lending overnight at all. The second question arises from the following comment: And consumers might decide to spend the excess reserves on consumption goods (run down their bank deposits) and/or buy other financial assets available in the markets. The former choice will increase aggregate demand and the automatic stabilisers will reduce the budget deficit accordingly. Since reserve balances can only change as a result of transaction between the government and non-government sectors, I don’t see how consumers can “decide to spend the excess reserves”. Note that JKH’s comment about the distinction between commercial banks and the non-bank private sector is an interesting adjunct to this query. 12. Dear Sean Point 1: It is an empirical issue but each bank is more or less acting in its own interests. If the system-wide surplus is so great that every bank is in surplus then maybe the interbank market would dry up. But I doubt that is a typical situation. Point 2: I was referring to the fact that deposit-holders (non-bank private sector) might decide they wanted to spend more and while it won’t alter the overall level of reserves in the commercial banking system it will further increase aggregate demand (if their preferred cash holdings are lower) and reduce the deficit accordingly. Hope that clarifies matters. best wishes bill 13. JKH says: Ramanan, On the bonds, the starting point is excess reserves in the banking system, which must be matched by bank liabilities of some sort, mostly deposits. Suppose people just pass on the bonds and leave their money in the bank. They may well do so as a result of their attitude toward interest rate risk. In this case, they may choose to leave their money short term rather than take a chance on a 2 per cent term bond yield, particularly if they don’t buy into that yield as their expectation for the equivalent path of future short rates. So the result is that even though the government fixes the price, the bonds don’t sell. Warren Mosler’s point on 40 year fed funds is a bit different, because the quantity of bank reserves held at the Fed is controlled by the Fed. There are two different excess reserve regimes – the normal one where the excess reserve position is equivalent to the marginal supply required to control the funds rate, and the quantitative easing mode where the excess reserve position is greater than normal, and where the Fed must pay interest on reserves to control the lower bound for the funds rate. Thinking about it now, I guess a 40 year rate could work in theory. If 40 year yields drift up, the Fed brings it down by doing 40 year OMO’s and pays interest on the excess reserves created. That brings the 40 year rate down as desired and prevents the overnight rate from being pushed down as a casualty of the 40 year intervention. That’s a lot of excess reserve creation if the entire term structure of interest rates behaves in such a way – much more so than required just for the overnight rate, with a much larger Fed balance sheet – but I guess that’s the theory of it. And I suppose it would work for tightening as well along the full curve. My point on “lending reserves” was motivated by reading the same thing as Sean references above relative to consumers “spending reserves”. I don’t know if this is a slip in language, or “acceptable” language. I notice Bernanke used the phrase “lending out reserves” in a recent speech. I’m fairly certain he’s aware of the mechanics around these matters. But there are several different nuances here. One is that we know that banks are not “reserve constrained” in the sense that the Fed determines the supply of reserve balances held at the Fed. The classification may change as between excess and required, but the Fed controls the total at the margin. In this sense, loans create deposits, and the total level of reserves is essentially bypassed in that process. And banks do not “lend out” reserves in the sense that lending “uses” reserves or causes reserves to subsequently “circulate” out to non-bank borrowers. The second nuance to which I’m referring is a little different (I think). It has to do with the correct way to describe the redistribution of existing reserves as a result of lending activity, notwithstanding the fact that the total quantity of reserves hasn’t changed and that reserves haven’t been “used up” by lending at the margin. E.g. it is clear that banks can “lend reserves” to each other in the Fed funds market. It is also clear that non-banks have no direct access to reserves held at the Fed, other than their ability to demand conversion of their commercial bank deposits to currency. That said, the Fed reacts to currency demand at the margin in such a way that it still controls the total level of bank reserves held at the Fed. We also know that if I lend you money directly, my bank owes your bank an equivalent amount of reserves in order to settle the transaction in the form of bank reserves. Our transaction reflects lending and borrowing. The banks’ transaction reflects payment or settlement of an immediate debt due (I think). Some people talk about this as if I have lent you bank reserves. My question is whether or not that is categorically wrong (or in some cases a slip of language), or whether the correct way of thinking about the monetary system actually acknowledges that sort of language. If the former is the case, then somebody involved in the discussion above may have to go to the woodshed. :) On the “depends” issue, I agree. But policy makers want to be able to box this. Otherwise you’re in Paulson TARP land, somewhat hobbled in your credibility by having presented 3 pages for about a$ 1 trillion in requested authority. Another way of viewing the policy formulation challenge is to structure it as a risk management issue – perhaps more on that another time (and I don’t mean value at risk!). I suppose my theme through this has been that the proof of operational flexibility is necessary, but so is some sort of structured approach to dealing with the policy question and the risks around it. Otherwise, the proof of operational flexibility is insufficient to address the problem at hand. It may even be damaging if taken out of context – i.e. lead one to being thrown out of the room. Again, that’s not to diminish the depth of the policy thinking by Bill, Scott, and Warren. I’m just thinking about the optimal positioning of the operational module here, given that the ultimate goal is to influence policy. With that, I suspect my degrees of repetition on this theme approaches the 3 figure level. But that’s the message.

14. Scott Fullwiler says:

Dear JKH

Looks like we mostly agree and were talking past each other a bit before. Just a few loose ends to tie up, which I’ve attempted (perhaps successfully) below.

JKH: “Politically imposed constraint, necessarily formulated in economic terms, is a reality, whether that politics occurs in government or non-government. The trick is somehow to get the political process properly informed in economic understanding as groundwork for the thinking that is required to proceed intelligently and dynamically with policy formulation on an iterative basis.”

Scott: Agree completely. Of course, our publications, seminars, blogs, etc., are in the pursuit of “properly informed” public debate you speak of. Any thoughts on how to better go about this are welcome and encouraged.

JKH: “My point is that somebody has to make a policy decision about issuing bonds.”

Scott: Yes, of course. We’ve always said issuing bonds is a choice. The alternative is to let the overnight rate fall to zero or pay interest on reserve balances.

Scott: Regarding your paragraphs on “Banks reserves are not accessible to non-banks,” agree completely. Some do mess this up, particularly neoclassicals, but I haven’t seen any of us do it.

Scott: Regarding “expected monetary policy” and “markets don’t set rates,” let me clarify because I can see where my wording was troubling for you. My point that “markets don’t set rates” refers to the fact that default risk, inflation expectations, crowding out, etc., aren’t (normally) factors affecting Treasury yields (aside from the slight exceptions that Bill refers to) . . . this is what people normally mean when they discuss the dangers of government borrowing (we often hear that “investors or China won’t want to hold the bonds”). My point was that these factors largely don’t influence the rates on Treasuries, but rather current and expected monetary policy dominate.

Scott: Regarding Warren’s point about setting the fed funds rate out to 40 years, consider a central bank that not only sets an overnight rate but also issues its own “debt” at quarterly maturities out to 10, 20, 30, or even 40 years. There is no question that (1) the cb can set the yield at which it will sell these securities or time deposits (as the alternative to purchasing them is holding overnight liabilities), and (2) this will set the term structure for risk-free rates. Further, in this scenario, whether the Treasury issues debt or not is irrelevant in the sense that if not, the cb is able to drain any balances necessary to hit this desired term structure, and if so, the Treasury issues are nearly perfect (if not perfect) substitutes for the cb’s issues, so again the cb has control over the risk-free term structure. That our real-world cb’s choose not to do this (or are legally prohibited from doing so in some cases) doesn’t mean they don’t have the operational ability to do so (not that you argued otherwise). So, in the absence of this, what we have in the real-world is just one step away from this, where the current and expected overnight rates set by cb’s mostly set the term structure of risk-free rates.

JKH: “I understand now. I think you’re saying that there should be (policy) limits in some sense, but that the right policy must be formulated on the basis of analysing the right economic variables (e.g. inflation risk) and on the basis of correct operational understanding.”

Scott: Yes.

JKH: “However, my extreme example shows the government with a position net long private sector assets, in addition to those held as the source of currency and reserve issuance. From a policy perspective, I think the net asset position changes the all-in constraint on future government debt and deficits, insofar as this relates to earlier points about limits to debt levels as affected by economic and inflation risk.”

Scott: It might lead uninformed policy makers to incorrectly think they have more fiscal “space” for the future, and yes, by definition you now can run a larger deficit without going beyond economic capacity constraints, but the reasoning behind such a policy is ridiculous (if not immoral given the lives disrupted by such deflationary policy). In other words, if we understand the operational side (again, most don’t but assume for a moment they do), then deflationary policy to hold back aggregate demand well below capacity (raise involuntary unemployment, reduce incomes) and create space for future deficits is just stupid and even immoral. Since the history of most wealthy economies (and probably most non-wealthy ones, too) is chronic lack of capacity utilization and mostly low inflation (aside from a few supply shocks, mostly), and since over-utilization can readily be avoided in a job guarantee-type of program (and other functional finance-related initiatives) and control over the interest rate on debt service, the more appropriate “constraint” to be placed on policy is to avoid under-utilization. I do understand that you were laying out an extreme case here, and not making a policy proposal.

Scott: Regarding “monetization,” the problem is this is overwhelmingly used to imply more inflationary impact than otherwise and even a shirking of normal “responsibility” to issue bonds. Neither is true, except under a gold standard. “Sterilization” is problematic because, outside of a gold standard, use of the term misunderstands modern monetary operations . . . if the central bank attempts to alter the exchange rate, it adds/drains reserve balances in the process, and it MUST (assuming no payment of interest on reserve balances) then reverse the effect on reserve balances if it is to hit its overnight target—that is, there is no choice but to “sterilize,” just as with any other change to the central bank’s balance sheet. Further, as with “monetization,” not “sterilizing” would not be more inflationary, either, but would affect the overnight rate (or, with interest payment, would simply raise excess reserve balances earning interest, which again, is not necessarily more inflationary than draining the reserve balances).

Best,
Scott

15. Bill,

Thanks for the responses.

It is certainly not typical at the moment for all banks to be in surplus, but the question was in the context of a scenario in which the government does not issue any bond but simply credits reserve balances to spend. At the moment, aggregate reserve balances are around A$2 billion while total commonwealth government bonds on issue are around A$100 billion. So, hypothetically, if the government bought all of those bonds back and henceforth issued no debt, aggregate reserve balances would be in excess of A\$100 billion and I would say it would be very likely that every bank had more than it needed in its reserves. I am not sure what would happen to the short rate in that situation.

As to the second point, I didn’t think that you (of all people) would have thought that the non-bank private sector could unilaterally change aggregate reserves. So thanks for the clarification: it was just the original wording (“spending reserves”) that confused me.

16. Scott Fullwiler says:

Dear JKH

Regarding “lending” reserve balances in the fed funds market, I’m not entirely opposed to it if it’s clear that the person using the term understands what this does/doesn’t mean in this instance. The lending bank doesn’t require prior reserve balances to make the loan, as it will receive an overdraft automatically if its balances are insufficient. Also, technically what happens is that the lending bank has balances debited and the borrowing bank has balances credited, all on the Fed’s books, so it’s not as if there’s some fixed quantity of balances sloshing around as there would be with paper currency. Because so many can’t use the phrase precisely enough to avoid other more serious mistakes, I prefer to avoid it altogether, just like I try not to say “money” when I’m trying to get precise about central bank operations or banking even though it’s not necessarily wrong to do so.

Regarding “boxing” the size of the deficit, I’m with Ramanan on this. Consider the New Consensus view of the real interest rate. There’s no particular rate of real interest that we must always be at . . . it’s a moving target over time; nobody really cares if it results in a nominal target of 10% or 1% as long as it is the rate most consistent with the central banks policy targets. The deficit in our modern monetary approach is the same way . . . since the government can set the deficit and since it can also (via the central bank) set the size of debt service, there is no reason to worry about the size of the deficit beyond whether it is consistent with full employment and price stability (that’s essentially the definition of functional finance). That someone thinks there should it be “boxed” more than this means they are bringing in other constraints (political or otherwise) inconsistent with operational realities (which isn’t to say these constraints don’t exist, of course). But if at some point policymakers could think of a deficit as nothing more than a policy tool in much the same way they think of the interest rate today, we would be much better off.

Indeed, with a job guarantee-type of functional finance policy in place and operating effectively, achieving full employment and price stability is much easier, since you don’t need some group of policymakers (perhaps) isolated from politics setting the policy dials (and inevitably imposing their individual views on the structure of the economy that are often ideologically based); all you do is offer a job at a living wage to everyone willing and able to work, and the deficit is then always at the “right” size. I’m simplifying here because there is no one “perfect” way to run the “job guarantee” and the proposal is consistent with any number of other policy initiatives on the political left or right, but hopefully the point is clear anyway.

Best,
Scott

17. JKH says:

Scott,

I think my emphasis on “policy” reflects a struggle for personal context from which to view the operational elements. Most of it turns out to be common sense, on which we can agree.

You’ve made the monetization/sterilization issue clearer, thanks.

Agree that current and expected monetary policy dominate on Treasury rates. And your explanation on the 40 year rate is very helpful. I do have a couple of questions further, relating to the rate aspect and the surplus issue, if you have the time to respond.

On the government surplus issue, I think I’m getting a better feel for your fundamental policy point. But have you looked closely at Canada enough to have a view on it in that case? Following a decade of surpluses, the economy seems to be well positioned, outperforming most through the crisis. Now, very large deficits are in the works, and the argument is certainly being broadcast that the low debt to GDP ratio is an advantageous starting point. That is a moderate case of my extreme example, I think.

18. Scott,

Regarding Warren’s point about setting the fed funds rate out to 40 years, consider a central bank that not only sets an overnight rate but also issues its own “debt” at quarterly maturities out to 10, 20, 30, or even 40 years. There is no question that (1) the cb can set the yield at which it will sell these securities or time deposits (as the alternative to purchasing them is holding overnight liabilities), and (2) this will set the term structure for risk-free rates.

The effectiveness of central bank market operations is predicated on being a dominant sized player in the market in which it is operating. Other market participants will respect the rates the central bank targets to avoid standing in the way of the proverbial freight train. I am not so sure that a central bank would be able to maintain this size advantage across such a wide range of maturity points (as many as 160 in the case of quarterly maturities out to 40 years). If interest rates started rising across the yield curve, there would be a lot of buying that the central bank would have to do, so much so that reserves may run out. What would happen then? I see some analogy with the market operations that Australia’s reserve bank has in relation to the Australian dollar. It does enter the market as a large participant to influence the price of the dollar if it sees that it has moved to extreme highs or lows, but it is does not target a specific rate as it is simply not a big enough player to force the issue in the currency market.

19. JKH says:

Sean Carmody,

If I may chip in my two cents on your question:

Regarding the 40 year rate, etc., your scepticism was my first instinct as well.

Thinking about it though, there’s nothing (operationally) preventing the central bank from buying any amount of assets of any type at any maturity, in order to bring down market yields. Bank reserves and deposit liabilities will expand accordingly. With mass purchases, excess reserves are created, on which the central bank must pay interest in order to set a floor on the chosen rate structure, presumably across a range of fixed rate maturities. There is no theoretical limit to this, other than the size of the market for potential purchases. At that theoretical limit, the banking system becomes the sole intermediary in the financial system. The entire term structure of interest rates is set by the central bank and intermediated through the banking system.

There would be a minor problem of interest rate risk on the banking system balance sheet though! There would be no guarantee (and little possibility) that the banks could hedge the enormous interest rate mismatch between the interest rate sensitivity of bank reserves (e.g. 40 year fixed rate) and the natural short term liability profile of bank deposit liabilities. On the other hand, any fixed rate liabilities that are offered by the banking system no doubt would be determined (and hedged) according to the rate structure embedded in banking system assets, which in turn is determined by the central bank.

I think though this is a clear example of the difference between theoretical operational flexibility and pragmatic policy reality.

20. Alan Dunn says:

Sean,

It sounds like you are discussing an endogenous rate of interest? Is this correct ?

Has something changed recently ? I thought the RBA increased (decreased) liquidity in exchange settlement accounts by either buying interest bearing alternatives or selling them ?

Cheers, Alan

21. JKH: you are right, there are no problems buying securities (which would simply result in increasing reserves). But couldn’t selling them eventually become problematic?

Alan: Current RBA open market open market operations do involve buying and selling interest-bearing securities, but this is done by means of repurchase agreements (“repos”) or reverse repos. The involvement of the interest-bearing securities is purely to serve as collateral to the transaction to mitigate credit risk between the counterparties. The only rate that these transactions affect is the short rate of the term of the repo/reverse repo (usually an overnight rate), not the outright yield of the interest-bearing security.

22. JKH says:

Sean,

We’re talking about a very theoretical, extreme case of central bank intervention.

There’s really no operational requirement for the central bank ever to sell their assets back to the public in this extreme case, provided they pay interest on reserves. That allows them to retain control over interest rate levels, which is all they need to do for monetary policy. The interest rate effect is transmitted through the banking system to the public, via the rates on reserves and then on bank liabilities. Banks will set their liability rates in order to manage interest margins against the rates they receive on their reserves.

23. JKH: I suppose I’m thinking about the bond market rather than lending intermediated via banks. If there were no 10, 12, 15, etc year bonds in the market, I struggle to see what the government yields at these time horizons actually mean.

Even looking at bank lending rates wouldn’t help much either. The vast bulk of bank lending to business is on a floating rate basis. Most consumer lending is via mortgages and the majority of this is also variable rate (loosely linked to the overnight rate…not that banks always pass on the cuts) and even fixed rate mortgages tend not to go any longer than 5 years.

The US is a different case, but even when the Fed was buying up Treasuries last year, the 30 year mortgage rates were stubbornly refusing to budge. When we are able to observe both the 30 yr mortgage rates and the yield on Treasuries, we can say that interest margins were widening. However, if there were no bonds in the market while may still be a 30 yr mortgage rate to observe, it would be hard to say what margins were doing.

24. Alan Dunn says:

I thought when the RBA purchased securities they destroyed them? Hence the term retiring debt ?

25. Ramanan says:

JKH,

The yield curve is constrained by forward rates and the movements are not so independent. The yield can be written in terms of forward rates. Formally,

y(T)= T^{-1} \int_{0 }^{T }f(t)}\;dt

So the Fed can either control the bond yields or as Mosler says, set the Fed Funds rate for the next 40 years. It can, for example, say that the Fed Funds rate in the year YYYY is MOD(YYYY,5) which is not smooth (but consistent). Or the Fed can just put an excel sheet of the overnight rates for the next 40 years with the FOMC minutes at the Fed website!. This will change the yield curve because of the equation I wrote (where you use the overnight rates instead of forwards).

26. Alan: the RBA open market operations do not involve retiring or destroying debt.

Ramanan: there is certainly a relationship between spot rates and forward rates and yields at different parts of the yield curve are not independent. I don’t see why that means that the central bank can simply publish target rates for each point on the yield curve and the market will follow them. Central bank rate targeting works because there is the belief that they can defend those rates. Relationship or no, I am yet to be convinced that a central bank could defend so many points on a yield curve.

27. Ramanan says:

Sean,

Let us say that today is 1st Jan 2011. The Fed announces early in the morning that the Fed Funds Target will be 25bps in the years 2011-2020, 50bps in 2021-2030, 75bps in 2031-2040 & 100bps in 2041-2050. It also announces that the discount window rate will be 50 bps higher than the target.

Some of the averages of these rates are as follows (rounded):

2011-2020 – 25 bps
2011-2025 – 33 bps
2011-2030 – 38 bps
2011-2040 – 50 bps
2011-2050 – 63 bps

Now let’s analyze this. If a bank has excess reserves, and assuming a zero coupon treasury of 2050 exists and is yielding 200 bps the bank will surely buy it. The discount window hits a maximum of 150 bps in the years 2041-2050. This is the trivial upper bound.

(Trying to think if 63 bps + 50 bps is the 2050 coupon yield)

28. JKH says:

Ramanan, Sean:

On occasion, the central bank must do OMO to defend the ON rate. This means injecting reserves if the overnight rate trades above target.

There is no reason the same wouldn’t apply if the central bank set a spectrum of target rates. This could include a range of term rates, or a range of forward rates, or a combination of the two, or in the theoretical limit – all such rates, term and forward.

The question is whether the market believes the central bank can defend such an ambitious interest rate control program. The market may bet on inflation, believing that bank will fail, in the same way it could fail in defending a gold standard, for example.

In any event, the central bank must be prepared to defend its set of target rates through corresponding market intervention. This means potential intervention directly at any point along the curve. This means it must be prepared to buy assets with a 40 year term for example. Or make markets in forward or futures markets. Either way, it must be prepared to accumulate term assets. This is a theoretically open ended limit on balance sheet expansion.

Even if the central bank buys up all the treasuries, and risky yields remain high, the bank in theory could start buying up risky term securities, as they have done recently to a limited degree with credit easing.

There are two effects from such central bank intervention. The first is that to the degree the bank accumulates assets, it will pay interest on the corresponding reserves at rates that are consistent with the term structure of its interest rate targeting. The commercial banks will then transmit this same term structure to their own liability rates, because they want to preserve their interest margins where possible. The second is that the central bank’s accumulation of assets reduces the supply of term rate assets that remain freely floating in the market. This increases their prices, other things equal, and prods convergence of the remaining market term structure toward the term structure the central bank is targeting.

This all happens more efficiently than the theoretical limit, because the market is quite aware of the unlimited firepower of the central bank in accumulating assets. It can literally buy everything and transmit its desired interest rate structure through the commercial banking system via the excess reserve mechanism.

The central bank sets the target rate(s) in “spread sheet mode”. But it may have to expand its balance sheet in spread sheet mode as a result, if the market is initially betting against those targets otherwise.

29. Ramanan: keep in mind tha there’s a lot of interest rate risk on a 2050 zero coupon bond. If a bank buys one of these bonds and the next day the central bank changes the target rate for 2050, increasing it by a mere 5bps, if the yield on the bond held by the bank increases by this 5 bps the bank will immediately lose 200 bps. So, it’s not clear to me people would rush to buy these bonds just because they offer a bit of a yield pickup.

JKH: you are right when it comes to buying to keep yields down although the potential size of purchases would be staggering (and I suspect would also break the overnight market as no bank could conceivably need to borrow in increase reserves). But what if the central bank thinks yields are too low? If they keep selling bonds, they’ll get to a point where reserves are all drained. Presumably this would be a scenario in which the central bank wanted to slow the economy, so having the government generate more reserves by increasing spending may not be a good idea.

30. Ramanan says:

JKH,

The present credit easing policy of the Fed is very different from what Warren talked about when he said that “…the Fed can set a fed funds rate out to 40 years if it wants to and make a two way market at its desired rates, and that’s where interest rates will be”

The Fed need not really increase its balance sheet to achieve setting the term structure. If you check this paper by Scott http://www.cfeps.org/pubs/wp-pdf/WP34-Fullwiler.pdf you will get to know that the daily Fed intervention is just a few billion dollars. The present scenario is completely different and you can read about Bill’s analysis of the Fed in another blog post.

Also, central banks use “policy duration” effects to control the rates. They appear frequently in the media and influence short term rates (though without as much commitment as the FOMC meeting minutes). The example I gave makes it really explicit. Plus the Fed has tools like paying interest on reserves and the discount window to control overnight rates.

Inflation does not enter the picture if the Fed Funds rates are set out to 40 years. There are just two choices in front of banks when they look at excess reserves – lend it everyday in the Fed Funds market or buy at the auction (or another bank).

The Central Bank does not have to do forwards, futures or outright purchases of treasuries. Banks will quickly jump at any opportunity of making more money than what the excess reserves make! The essential point Warren was trying to make is that worrying about bond yields when you think of an economy is not really worth it! Easily controllable.

The idea that Warren outlined is really different from quantitative/credit easing. In fact, if I understand him correctly, he is trying to say that the Fed has unnecessarily worried so much about Treasury yields – the main thing for recovery are increased government spending and super-aggressively bringing down unemployment and host of other fiscal measures.

31. JKH says:

Sean,

If the central bank thinks rates are too low, it will raise both the policy target rate and the rate paid on reserves. The reserve rate puts a floor on the actual behaviour of the policy rate against target – e.g. the fed funds rate, or in our discussion, the set of target rates. That will cause the commercial banks to raise their liability rates via competitive bidding, due to fatter interest margins from a higher reserve interest rate. That will cause other non-bank rates to rise, because otherwise people would prefer to put their money in the bank. That rate arbitrage happens quickly, because the non-bank market in aggregate can only sell to the banks, who will only bid for new assets at the new level of interest rates.

At the same time the central bank can start to sell or mature its excess assets. This will drain reserves. It’s not entirely necessary for interest rate control, given interest paid on reserves. But it returns the system to more normal functioning with less central bank intermediation of risk.

32. Ramanan says:

Sean,

Yes I understand the risk of the zero coupon bond. However in the example I gave, the Fed commits to those rates for the next 40 years, come what may. At no point during those 40 years does the Fed change the announcement a bit.

No central bank in the present world would ever take such a step. This is because they are heavily influenced by the “New Monetary Consensus”. However in Warren’s world this is possible. In fact Warren has mentioned in moslereconomics.com and Bill in this blog that the central banks permanently set the overnight rates at 0% !

I think Bill, Scott and Warren would tell you that the effect of monetary policy – though sells well – is really weak. Developed countries have low inflation because of moderate growth and the high amount of competition and host of other factors. Developing nations need to control price stability and improve standard of living for everyone through fiscal policies.

However, I am with JKH when he says that it will be difficult to convince policy makers. They will never come out of the gold standard thinking. Insecurity will prevent them from opening up for debate. Increasing awareness amongst people is a very good idea to achieve implementation of such goals.

33. JKH says:

I’d be interested if somebody could explain to me how it is that commercial bank loans do not create commercial bank deposits in a gold standard or fixed exchange rate system.

34. Alan Dunn says:

Dear Sean,

Why don’t the RBA destroy securities after they purchase them ?

cheers, Alan

35. Ramanan: I see what you are saying now, but I think that this is a bit different from the yield curve interest rate targeting that was being discussed. In your example, once the central bank “targets” a forty year rate than in, say, five year’s time that is the rate they are forced to target for the 35 year point on the curve. This is very different from having the central bank target one level for each point in the yield curve and then in, say, a month’s time being free to change it’s target for every point on the curve.

Alan: I suppose it depends on what you mean by “destroy”. If you definition of destroying securities is “cause to cease to be held by the non-government sector”, then yes they are destroyed. But if you use that language, then a repo also involves a commitment by the central bank to “create” the securities the very next day. In this case, though, the language doesn’t seem particularly useful to me. Basically, the central bank lends a commercial bank money for one day and, during that time, holds securities owned by the bank as collateral against default. It just so happens that they are structure a a sale and repurchase (and there is no price risk on the repurchase: it is set ahead of time to work out to the bank just paying one day’s interest at the agreed rate).

36. Ramanan says:

Sean,

I understand your point about – after 5 years in my example, the forty year coupon becomes a 35. But will you agree that an announcement is sufficient ? Of course they have to do one more thing – control the overnight rates everyday though Open market operations. They will hardly require controlling other rates – that would be AutomatiC. Also, the announcement for 40 years will completely remove most dynamics of the yield curve. The future is almost fixed. Traders would have to trade in a very narrow and tight range. However there are really no worries about bonds crashing etc.

Would like to believe that was what Warren was saying.

The other way of targeting the yield curve is targeting points instead. e.g on random or announced dates, the central bank decides and announces the term structure. In the next meeting or some surprise date, they again decide a new term structure. Not sure how to achieve this but I believe it can. If there is just one perpetuity, its easy to control. The CB/Treasury announce that the yield is x – some people would take it. If they don’t, do not issue! If you seller is not able to find a buyer, the CB buys it. No need to worry about bid/cover ratio. The fear of monetization is irrattional! Else issue 1-day T-bills – which is effectively paying interest on reserves or a Fed Bill (hinting that the CB and the Treasury are effectively one and the same, how much ever one talks of the central bank ‘independence)

Couldn’t catch what Scott is saying. Am with you – there is still an interest rate risk for bond holders.

37. Alan Dunn says:

Thanks Sean, that’s exactly what I was referring to.

38. Ramanan: I still am not convinced that simply publishing rates means that’s where they will be. The central bank has to have sufficient firepower to buy and sell at those levels (buying ok, just building reserves, selling trickier). Otherwise, it would be just as easy for a central bank to target currencies. While a central bank can buy as much foreign currency as it likes (simply increasing local currency reserves), it’s ability to buy local currency is limited by its stock of foreign currency. Also, I am almost certain that your scenario is not what Warren had in mind. I could, of course, be wrong.

Alan: Fair enough. To me using the word “destroy” in this context seems melodramatic without adding much insight into the market operations, but as long as it’s understood that they’re always “reborn” the next day, that’s fine.

39. Ramanan says:

Sean,

The CB just has to support/defend the overnight rates everyday. e.g., on Oct 2 2013 (a random day), they have to target the overnight rate of that day not the 2050 coupon of my example. Its basically stretching the “policy duration” effect to as maximum as you can.

Instead of looking at 40 years, one can look at 3 months. If the overnight rate is 25 bps and the three month bill is yielding 75 bps (using daily compounding), banks will rush to buy it. You know that the overnight rate will remain at 25 bps in the next three months. They may also use up their required reserves to buy more and get a loan through the discount window to maintain the required reserves and/or daily settlements. Its an arbitrage. Hence 75 bps for the 3m T-bill is high. Here the central bank didnt have to intervene, its all interbank competition. Doesnt matter if there is an interest rate risk – in the end there are more reserves. You can keep applying this logic to 6m, 1y … 40y

40. Sean Carmody says:

Ramanan: of course they defend their target rate every day. And there’s nothing too special about the overnight rate. The central bank of NZ used to target a 90 day rate. My point is that central banks are sufficiently big to have the credibility to defend a single interest rate. I am just not convinced they are big enough to defend 10s of different rates at once. It’s a matter of scale not process.

41. Ramanan: perhaps I should add that my view is merely one of some skepticism. I understand the mechanisms, which are simply an extension of the current process. I would see the degree of difficulty sitting somewhere between targeting a single interest rate (which works) and targeting currency (which often fails). So, it may well work but I have my doubts. I am open to being convinced but simply describing the process will not do that.

42. Ramanan says:

Sean,

I understand your point. It sounds reasonable on paper but very difficult to implement. e.g. banks may for some technical reasons, not want to do exchange reserves for treasuries and non-banks generally do not have access to the overnight market and yield like 150 bps for 30y may not appeal to them. And hosts of other issues you may point out.

I will try to come back to you after a bit of thinking.

43. Sean Carmody says on Saturday, August 8, 2009 at 9:37,
Ramanan: of course they defend their target rate every day. And there’s nothing too special about the overnight rate. The central bank of NZ used to target a 90 day rate. My point is that central banks are sufficiently big to have the credibility to defend a single interest rate. I am just not convinced they are big enough to defend 10s of different rates at once. It’s a matter of scale not process.

The ability to set an interest rate target is not a credibility question … its a question of capacity. Within the structural parameters of the institution … between a nominal rate of 0% and a nominal rate that shuts down the interbank landing market, the Central Bank can set and maintain an interbank reserve lending rate target even if the commercial banks think that they can’t.

They can do so with less work because commercial banks understand that they have the power to enforce it, certainly, but unlike the supply of money, the Central Bank can manhandle the supply of reserves until the reserve lending market submits to the target.

However, with deregulation of commercial bank lending and removal of institutional barriers between different types of credit markets, setting ten different interest rate targets in credit markets is not a direct capacity … indeed, in many cases, the distinct interest rate targets in different credit markets will require incompatible policy actions. There are simply not as many policy instruments for most central banks to deploy as were available to them in, say, the 1950’s.

44. Ramanan says:

Sean,

Here is an attempt to convince you. If I were a bond trader at a bank, and if I were given a choice to get one time series “from the future”, I would pick up the overnight target rates. Not inflation data, not the components of GDP, not deficits, not supply. I would then compute

pie*(1+f_i) – see http://mathurl.com/nqcu8q for exact equation

for each horizon and compare it with the (respective) daily compounded yields of bonds in the markets.

Back to the example I gave, the central bank would have to intervene neither in the bond market nor in the money market. Don’t look at it as targeting different rates. The open market operations would simply be the operations it does in the usual case of targeting overnight rates only.

In the case of the Reserve Bank of New Zealand, I couldn’t really find a good document yet for the methodology for targeting the 90 day rate. However I can only conjecture the following. The RBNZ meets every six weeks. Within the 90 day period, there is always one policy meet. Chances are, at this meeting, the policy rate is changed. Also, here the Reserve Bank is targeting two rates. In my example, they are operationally targeting only the overnight rates. The promise of keeping the overnight rate at a level in the future is almost sufficient.

45. Ramanan: there’s one technical point you are missing: the future path of the overnight cash rate does not completely determine the prices of government bonds because there is a spread between the cash rate (which is an interbank rate) and the rate on tradeable government debt and this spread is not constant. You can see this clearly if you look at the price of Treasury bills with a maturity that occurs prior to the next central bank meeting. During this period you can safely expect the target cash rate to be unchanged and yet this does not determine the yield on a T Bill.

That aside, I still think that the idea of fixing rates out to the future is not what was originally intended in the discussion. For example, in the case of RBNZ targeting the 90 day rate, this did not mean that they tried to keep the rate constant for 90 days. Rather, it meant that one day they would set a target for where 90 day bills should trade that day, bit in 6 weeks time they may dramatically change the target, depending on economic conditions. There would be no attempt to keep the 48 day rate (90 days minus six weeks) at any level consistent with where it had previously been set. Note also that the RBNZ now targets the overnight rate.

Since the whole point of monetary policy is to respond to economic conditions, I cannot seriously believe that anyone is suggesting locking in a fixed pattern of overnight rates for 40 years into the future! Rather, I believe that what Warren and others were referring to is the possibility of the central bank targeting current spot rates out along the yield curve. For example, the central bank may target 2% for the cash rate, 3% for the 10 year bond rate and 3.5% for the 15 year rate. In a few months time, the central bank could then target a completely different set of rates, no longer being tied into previous targets.

46. Ramanan: thinking about it further, even if there was no spread between government securities and the cash rate and this future path of cash rates was indeed what the central bank was trying to target, I would still doubt the effectiveness. Imagine that, sometime in the future, inflation began to accelerate well above the level of cash rates (set some time in the past). It would seem very likely that there would be enormous pressure on the central bank to reassess that path of cash rates in order to fight inflation. Given that likelihood, it is hard to see that a path of future cash rates set today can be credible very far into the future.

47. Ramanan says:

Sean,

I think Warren said that the Fed could set the rates for 40 years. Quoting from Bill’s post:

Warren M: Agreed, under current institutional arrangements and biases. But operationally the Fed can set a fed funds rate out to 40 years if it wants to and make a two way market at its desired rates, and that’s where interest rates will be. The Fed in fact sets the term structure of rates one way or another.

I haven’t seen the data on T-bills expiring before a policy meet vs. the overnight target. With this constraint, I can only conjecture that it could be because the overnight rate fluctuates around the target. The fluctuations depends on the penalty for overnight rates, discount window etc. There is an uncertainty at this level. e.g., a bank wouldn’t buy the T-bill from another bank just because it looks cheaper compared to the overnight target. Chances, that the bank may have to use the discount window on every day till the maturity and it could see itself in advance that in hindsight it may realize that it had not bought those bills.

However given a permanent policy for overnight rates, the penalty rates, discount window rate compared to the target etc. can set bounds on money market rates and bond yields. And the range for trading will be very narrow.

I think there are two worlds here. First – New Consensus world where central banks respond to inflation expectations and change rates accordingly. You see bond yields move up or down when there is a news release of GDP/inflation because that changes imply that monetary policy will react accordingly because Central banks think that monetary policy has an impact. There is a stigma on the usage of the discount window and it is relevant in my T-bill argument.

In Bill’s world however, monetary policy has a very weak impact. Right Bill ? I have seen Randy Wray and Scott writing about this in detail. Overnight rates can be set to zero permanently. Inflation control is beyond monetary policy and is more effective through the fiscal route. So in this world one can set the Fed Funds rate out to 40 years, the Fed can lend unlimited reserves and term structure is under control and trades in a narrow range.

Back to the present world, one may ask – how can this fit in ? Since, the monetary policy anyway has a weak effect on the economy, the central bank can just ditch the Neoclassical world and adopt the modern monetary theories.

One may also ask why the bonds in my example trade with negative real rates. My answer to that would be – government debt is not debt. Its an alternative to holding cash/bank account.

Basically I agree with you 99.99% – in one of my posts above I had mentioned that no central bank would take such a step in line with what you are saying. And I would agree with you if you say that if central banks suddenly realize how modern money works (as articulated in Bill and his collaborators work) and adopt it instantaneously, it could catch people by surprise. In that transition period, the central bank may not be able to control the term structure. However I would imagine that this transition period would not last long.

Controlling spot rates at various points in the yield curve is also possible. I understand this is discussed in the posts above. For the government and the central bank to achieve this, the self-imposed constraint that the government has to run balanced budgets has to be removed. Once this is done, the central bank could everyday announce the price of every government security in the market. The central bank should be willing to accept the securities for cash at that rate. A mutual fund M would not sell it to a bank B at a lower price, since the central bank is the better buyer. The overnight rates can be prevented from hitting zero by paying interest on reserves equal to the target.

This site uses Akismet to reduce spam. Learn how your comment data is processed.