Functional Finance vs the Long Run Government Budget Constraint

(I had been planning to write this post before Steven Landsburg started a whole blogosphere argument about what taxes are for. Honest!)

Functional Finance says you only use taxes if you want to reduce Aggregate Demand to prevent inflation. The Long Run Government Budget Constraint says you use taxes to pay for past, present or future government spending. They sound very different. They aren't.

There's a general principle in economics: first you eat the free lunches; then you  look at the hard trade-offs. Functional Finance says "first eat the free lunches". The Long Run Government Budget Constraint says "then look at the hard trade-offs".

Suppose, just suppose, that if you kept on doing what you were planning to do, you never had to worry about inflation. Not now, not in the future, not ever. Because Aggregate Demand was too low now, and was projected to be too low forever. So you are not worried about inflation. Instead you are worried about deflation. And you were a government that could print your own money. What would you do?

You would print money and spend it. Or print money and use it to finance tax cuts. And you would keep on doing it, more and more, until you got to the point where you did start to worry about inflation. You first eat all the free lunches.

That's the underlying kernel of truth in Abba Lerner's Functional Finance (pdf) [see also his book The Economics of Control (pdf)]. And I don't know of any mainstream macroeconomist who would disagree. You use taxes only so you don't have to print money. When you get to the point that Aggregate Demand is high enough, so you start to worry about inflation, you use taxes to finance past, present, or future government expenditure precisely because you don't want to print more money and make inflation higher.

Suppose inflation isn't a problem right now, because Aggregate Demand is currently too low. Does that mean the government should print money and spend it? Not necessarily. Print money yes, but instead of spending it on goods, or on tax cuts, it might be better to use it to buy back some interest-paying government bonds. Because even though inflation isn't a problem right now, it may be a problem some time in the future. So you can buy the money back in future, by re-issuing the bonds (and save on interest in the meantime) without having to raise future taxes or cut future spending.

Here's an easier way to think about it. Print enough money to get Aggregate Demand and inflation where you want it to be. That's the free lunch the government can eat. Any additional government spending must be paid for, sooner or later, with taxes. The present value of taxes, plus the present value of newly-printed money (seigniorage), equals the present value of government spending, plus the existing debt.

Seigniorage revenue belongs in the government budget constraint. The government can pay for part of its spending by printing money. But don't get too excited. It's not that big, on average. If central bank currency is around 5% of annual nominal income, and if nominal income is growing at 5% per year (3% real plus 2% inflation) then 5% x 5% = 0.25% of GDP. (In the right ballpark for Canada -- it shows up as the profits the Bank of Canada hands over to the government -- but maybe double it for the US). Printing money is a nice little sideline, but we are still going to need taxes.

Let's ask a slightly different question. Why do governments pay interest on their debt? Actually, it sounds like a different question, but it's really the same question. Why finance government deficits with interest-paying debt, when you could use non-interest-paying currency?

The answer is the same: you pay interest on the debt to encourage people to hold it and stop people spending it. If you cut the interest rate on government debt, will people sell it back to the government for money, spend the money, and cause inflation? If not, then Aggregate Demand is too low, and the problem is deflation, not inflation. So there's a free lunch from cutting the interest rate on government debt. And the government should eat that free lunch. And then the Long Run Government Budget Constraint kicks in again.

Now suppose the real rate of interest on government debt is below the real growth rate of the economy. (Or the nominal rate of interest is below the growth rate of nominal GDP -- same thing). And suppose it will be like that forever, if you keep on doing what you were planning to do. The government can run a Ponzi scheme forever. It can borrow and spend, then borrow to pay the interest forever, and the debt grows more slowly than the economy, and the debt/GDP ratio declines over time. The Long Run Government Budget Constraint is undefined. The Present Value of taxes can be less than the Present Value of Government spending.

That's another free lunch that needs eating. The economy is dynamically inefficient. The economy wants a Ponzi scheme. And the government should satisfy that demand. Issue debt until the interest rate equals the growth rate. Then the Long Run Government Budget Constraint kicks in again.

Comments

[Determinant]

Beautiful. Wonderful. Can you run in some Ottawa riding so we can make you Minister of Finance, Nick? Or how about Governor of the Bank of Canada? Both positions come with a complimentary helicopter.

But the problem is that the mainstream media has been angonizing, handwringing and generally worried itself silly about inflation since 1975. It's a bogeyman. Every 0.1% increase in inflation is seen as a calamity. A 1% increase is a catastrophe. Anytime the Minister of Finance or the Governor put their hands anywhere near the monetary or fiscal levers and hint about pushing them in the "wrong" direction the media explodes about inflation.

Inflation? It's not an economic fact so much a collective delusional nightmare.

Have we all forgotten the pain of chronic unemployment? Well, that CSLS paper said we did.

[tjfxh]

MMT would apply functional finance in terms of sectoral balances to affect demand. The government fiscal balance and the nongovernment (domestic private sector and external) balance sum to zero as an accounting identity. Taxes, saving and net imports constitute demand leakage. Functional finance is used to offset demand leakage in order to keep effective demand sufficient to purchase potential supply and maintain full employment. So if the domestic private sector desires to net save and the country is a net importer (like the US at present), then the fiscal deficit has to offset this demand leakage, or the economy will contract and unemployment rise. This way, the government fiscal balance would respond to shifts in the nongovernment balance as savings desires and the trade balance change, using expenditure to inject nongovernment net financial assets to increase nominal aggregate demand and taxation to withdraw NFA to decrease NAD.

[MDM]

Nick, thanks for the post and continued engagement with MMT.

I’m just going to comment on a few points of your post and provide what I think is an overview of some key MMT points. I’ll try to keep my points as brief as possible.

I’d just like to start off with a disclaimer, that I am only an undergrad, and therefore my views on MMT may be incorrect. If I have made an errors please let me know.

Financed spending versus ‘unfinanced’ spending:

I think a key concept which may be missed in this debate is that MMT doesn’t define the choice between either taxing or ‘printing money’, ‘printing money’ always occurs whenever governments spend. Allow me to explain:

1.MMT define money as a credit-debt relation; money is always financial asset with a matching liability.
2.As Minsky states, all entities can create money, the key is getting your liabilities accepted.
3.Some entities get their liabilities accepted by explicitly stating that they will convert it into something else. For instance, banks allow their money to be convertible into state money.
4.States (defined as a set of institutions which have a monopoly on force over a geographic area) have an easier time in getting their liabilities accepted, they can impose a tax upon the population which can only be extinguished by using their liabilities. Of course the ability to do this depends upon the ability of the state to enforce their taxes.

It follows from this the primary function of taxes is to create demand for government currency; they do not finance government spending. In fact, government spending (or central bank ‘loan’) is logically required first before taxes can be paid. Once a tax has been imposed, spending and taxing become separate operations – governments spend by crediting private sector bank accounts or by issuing a cheque, and tax by debiting private sector bank accounts.
The option then isn’t between financed spending versus ‘unfinanced’ spending, as government spending is never financially constrained.

Furthermore as you have pointed out the other function of taxation is to then reduce aggregate demand.

In fact, the ability for the private sector to pay taxes requires that the government spend (or central bank issues a ‘loan’) first, before taxes can be paid.

“Print money yes, but instead of spending it on goods, or on tax cuts, it might be better to use it to buy back some interest-paying government bonds. Because even though inflation isn't a problem right now, it may be a problem some time in the future. So you can buy the money back in future, by re-issuing the bonds (and save on interest in the meantime) without having to raise future taxes or cut future spending.”

I have a few questions regarding this. As I understand it the operation of ‘printing money’ to buy bonds, is essentially changing the composition of financial assets within the private sector, from interesting earning to non-interest earning. MMT is critical of this operation because firstly, it reduces interest income, which reduces demand, and secondly, because apart from lowering the yield curve, it doesn’t do much else.

I don’t understand how the option of issuing bonds in the future is dependent on doing this either, surely the option is always available to government? I also don’t see the how the issuing of bonds is meant to reduce aggregate demand. Wouldn’t the purchasing of the bonds reflect the desired savings of those entities? The government has simply provided a safe form in which that savings make take.

Function of bonds:

"Let's ask a slightly different question. Why do governments pay interest on their debt? Actually, it sounds like a different question, but it's really the same question. Why finance government deficits with interest-paying debt, when you could use non-interest-paying currency?"

Apart from self imposed requirements, MMT states that bonds function to assist central banks in maintaining their target rate. This is how I see it, I don’t believe this is inconsistent with the mainstream view:

1. Central banks have a monopoly on issuing reserves.
2. Central banks can either set the quantity and allow the price to float, or set the price and allow the quantity to float. Central banks target a short-term rate, they therefore allow the quantity to float.
3. From 1 only changes in the central bank’s balance sheet can alter the total number of reserves.
It follows then that the banking system cannot create or destroy reserves, they can only shift them around.
4. Following from 3,
A.if the banking system finds itself deficient in reserves, then there is no way that it can overcome this deficiency, the central bank has the choice of either meeting the demand for reserves and hence maintaining its target rate, or allow the target rate to be pushed up (of course the market will never clear).
B.If the banking system has excess reserves, then the there is no way that it can get rid of these excess reserves, the overnight rate will be pushed down below the target rate.
5. In 4B: The issuing of bonds does two things, firstly, it changes the composition of financial assets in the private sector, from non-interest assets to interest earning. Secondly, allowing the central bank to maintain its target rate – this is why MMT refers to bonds are “interest rate maintenance operations”.
Much the same as taxes, bonds do not function to ‘finance’ government spending. In a system where the government has no financial constraint it doesn’t make any sense to consider bonds as financing operating. The same point re taxes can be made with bonds too, that is, the ability of the private sector to purchase bonds is only possible once the government has spent or the central bank has issued a loan.

-------

As I see it, once the functions of taxes and bonds are understood – that is, once they are no longer defined as financing operations – we can then turn to more important questions such as the appropriate mix of each, as your post alludes to.

[Mario]

I'll need to re-read this again as it's getting late and I'm getting tired,

but as far as I hear Warren Mosler keep saying...cash and bonds are simply an asset swap. The interest rates fluctuate and reveal indifference levels in the market. Savers stay savers essentially forever and so the "inflation risk" (aka demand-pull inflation) of bond interest is essentially zero since those savings will more than likely stay in bonds or if anything move to another savings/investment vehicle. The likelihood of those savings to turn around and become consumptive (buying ipods with bond interest woo-hoo!!!) is very, very slim.

So I guess I am confused with your points...again it's late for me right now, I'm tired, and I could be way off here so chime in and set me straight.

cheers!

[Ralph Musgrave]

In his final para Nick Rowe says “the government should. . . issue debt until the interest rate equals the growth rate.

Well, suppose growth declines to zero because the population ceases to grow, and because technology ceases to advance. Or suppose the economy ceases to grow because we decide to factor in the real environmental costs of our Western lifestyle. Do we take it that the optimum level of govt debt then becomes zero? The reasoning there smells fishy to me.

[Tschäff Reisberg]

I'm glad you are engaging these ideas Nick, this openness to ideas is refreshing and encouraging!

Scott Fullwiler addressed the intertemporal budget constraint in detail at the start of this paper: http://www.cfeps.org/pubs/wp-pdf/WP53-Fullwiler.pdf

[Max]

It is critical to understand that "printing money" is no different from "printing bonds". Much confusion stems from the belief that one is more inflationary than the other. This belief has been debunked both by a technical understanding of monetary operations as well as from a black box empirical point of view (e.g. the non-effect of quantitative easing).

Also, many government (including the U.S. since 2008) pay interest on money (bank reserves). So earning interest is not a difference between money and bonds.

Finally, the interest that the government pays on its money/debt is set by the government. And there is no economic law that it has to be related to inflation or economic growth. It could be a constant, which would eliminate an artificial source of risk from financial markets.

[Ralph Musgrave]

Tjfxh, [Ralph means to reply to MDM -- the name comes below the comment in this blog Ralph -- NR] First, don’t worry about being “only an undergrad”. When it comes to something relatively new (or as is the case with MMT, something that is being “re-born”), fully qualified academics are not necessarily better than anyone else. In fact some of them are so brainwashed by the conventional wisdom, that people new to the subject are better. And more fundamentally, there is Jean Paul Satre’s dictum: “man is condemned to freedom”. I.e. you have no option but to think for yourself.

Second, re your claim that “money is always financial asset with a matching liability”, this is not really the case with monetary base. Of course the £20 notes in my wallet say “I promise to pay the bearer on demand the sum of twenty pounds”. But that’s nonsense. Central banks have no intention of honouring their cash liabilities, e.g. with lumps of gold.

Third, you say that “MMT is critical of” printing money to buy bonds because it reduces interest payments which reduces demand. My impression is that it is only Warren Mosler who has made this claim, and even he admits to not being sure of the point.

This demand reducing effect depends on the institutional customs or arrangements in the country concerned. E.g. if it normal practice for such payments to be funded by the central bank with money produced out of thin air, then obviously there is a demand influencing effect. But if it is normal practice for the interest to be funded out of tax, then the demand effect is much less or even zero.

In any case, the whole argument is irrelevant, in that given deficient demand, the government / central bank machine can boost demand any time it wants.

Fourth, you say “I also don’t see the how the issuing of bonds is meant to reduce aggregate demand.” Nick gave the reason, seems to me, when he said “you pay interest on the debt to encourage people to hold it and stop people spending it.” Of course, as you suggest, some people need very little inducement to buy bonds, so there is no effect on demand here. But others do need the inducement.

Fifth, re what you call the “appropriate mix” of cash and bonds, Warren Mosler advocated a “zero bonds” regime in this Huffington article (see 2nd last para).

http://www.huffingtonpost.com/warren-mosler/proposals-for-the-banking_b_432105.html

I agree with that, but for different reasons. I would say that MMT advocates influencing demand (and inflation) simply by altering the amount of money that the government / central bank machine prints and spends. (That’s the Abba Lerner “money pump”). Given that tool for influencing demand, I don’t see the need for a second tool, i.e. adjusting interest rates.

I will publish a paper on the Munich repec site in the next week or two which gives the whole interest rate adjustment idea good kick in the whatsit. I think the title will be “Monetary and fiscal policy should be merged, which in turn would change the role of central banks.”

[Nick Rowe]

BTW: does anyone know of a better link to Abba Lerner's Functional Finance? I found that link on Ralph's blog, I think. (I had read the book Functional Finance years ago, but didn't know that article existed). The article is a little hard to read (I mean it's blurry -- not that it's ... you know what I mean).

[Nick Rowe]

Determinant: thanks! But I don't think I have written anything that economists at the Bank of Canada or Finance would majorly disagree with.

tjfxh: let me re-write what you said using the language of conventional macro:

"The IS curve incorporates the equilibrium condition that desired savings minus investment = the government deficit + net exports. If that equilibrium condition results in too low a level of income (for preventing inflation falling below the 2% target) the Bank of Canada should lower the rate of interest to move down along the IS curve (increasing investment and reducing savings) and/or lowering the real exchange rate (increasing net exports and/or the government should loosen fiscal policy to shift the IS curve right."

It's a different way of saying the same thing. The only difference is that *maybe* MMTers don't believe interest rates affect desired savings or investment, so want to use fiscal policy. See my last post on reverse engineering.

[Nick Rowe]

MDM: "The option then isn’t between financed spending versus ‘unfinanced’ spending, as government spending is never financially constrained."

Translated, that means the government can finance its spending by printing money (seigniorage). But there's a limit. The first limit is when inflation gets too high, because there's only so much 0% interest money people want to hold. That limit is approximately $2 billion per year on average in Canada, for 2% inflation. That's not much. The second limit is when you hit the seigniorage-maximising inflation rate. I don't know where that limit is (it depends on the shape of the money demand curve), but Zimbabwe went past it.

"I have a few questions regarding this."

Standard theory believes that the IS curve is not vertical but slopes down, because a lower (real) interest rate will increase desired investment and/or reduce desired savings via a substitution effect. If you don't believe this -- if you believe that the IS curve is (approximately) vertical (or slopes up), then you get very different results. See my recent post on "reverse engineering MMT".

People are willing to hold *some* of their wealth in the form of money that pays little or no interest, because money is a medium of exchange, and it's really inconvenient to do the shopping with too small an amount of money. But there's a limit to how much money people are willing to hold. That's why the money/bond mix matters. Go past that limit, and either issue too much non-interest-paying money, or bonds with too low an interest rate, and people want to spend it, and AD gets too big, and you get inflation.

[Nick Rowe]

Mario: "Savers stay savers essentially forever and so the "inflation risk" (aka demand-pull inflation) of bond interest is essentially zero since those savings will more than likely stay in bonds or if anything move to another savings/investment vehicle."

That's the key assumption. If you believe the IS curve is vertical, because a lower rate of interest will neither increase desired investment nor reduce desired saving, then there is little rationale for the central bank to set an interest rate above 0%. See my "reverse engineering" post. Mainstream economists do not believe this. They believe the IS curve slopes down.

Ralph: (assume the real growth rate is 0%, and the real interest rate is above 0%) "Do we take it that the optimum level of govt debt then becomes zero?"

No. If r is less that g (forever), there's a free lunch from issuing debt, so the government should issue it. If r exceeds growth rate, there's no longer a *free* lunch. But it might still make sense to pay for lunch, depending on the costs and benefits. E.g. if there's a real need for high government spending right now (war, for example) but you want to smooth taxes over time to reduce the total distortions, then debt-finance makes sense.

[Nick Rowe]

Tschaff: I think that's the same paper by Scott I have already had a go at reading. (Scott told me about it in comments on Steve Waldmann's blog). This post, my previous post, and my comments on Steve's blog, are partly in response.

Max: "It is critical to understand that "printing money" is no different from "printing bonds". Much confusion stems from the belief that one is more inflationary than the other."

That's where we disagree. People will hold some amount of money because it's a medium of exchange, even when the real rate of return on holding that money is considerably below that of bonds.

All the seigniorage revenue from printing money comes from the interest rate differential between money and bonds. That's where the Bank of Canada's profits come from. It issues zero interest currency, and holds interest-paying bonds. The difference is the $2 billion per year. If we ignore that, then any free lunch from issuing money+bonds only happens when the interest rate on bonds is below the growth rate.

[Max]

"Go past that limit, and either issue too much non-interest-paying money, or bonds with too low an interest rate, and people want to spend it, and AD gets too big, and you get inflation."

The supply of non-interest-paying money (i.e. paper money) is endogenous. The government has no contol over it. It can control the supply of bank reserves, but in that case it must pay interest if it wants to target a >0% rate.

If the government is bidding up prices with its spending it's going to cause inflation even with a high interest rate. The high interest may temporarily mask the inflation problem by propping up the exchange rate, but the problem remains.

[Nick Rowe]

Max: "If the government is bidding up prices with its spending it's going to cause inflation even with a high interest rate."

That, in a nutshell, is a re-statement of my reverse-engineered MMT model in my previous post. MMTers believe that the IS curve is vertical.

[Nick Rowe]

Max: "The supply of non-interest-paying money (i.e. paper money) is endogenous. The government has no contol over it."

That is also the conventional (New Keynesian/Neo-Wicksellian) view. The Bank of Canada sets a nominal rate of interest, and the stock of currency is demand-determined. The LM curve is horizontal. Again, that's my reverse-engineered MMT model. Horizontal LM, and vertical IS.

[RSJ]

" People will hold some amount of money because it's a medium of exchange, even when the real rate of return on holding that money is considerably below that of bonds."

That's a good point -- we should think of different demands for different types of assets, rather than one large savings demand.

However, the public is *already* demanding to hold a stock of MZM equal to its holdings of federal debt.

It's just mostly outside money, and we allow banks to earn the seignorage income instead of the government.

Banks are allowed to supply households with 7 Trillion in deposits, supposedly because households demand it for transactional (or short term storage) use.

On the other hand the government must not create more money because households are not willing to hold more than a small amount. The government must borrow 7 Trillion -- the debt held by the public -- at the long term rate, rather than MZM own rate.

I think the seignorage income is *huge* -- enough to retire the entire debt, for the most part. But we hand almost all of this income over to the financial sector.

[vimothy]

RSJ: Banks earn money because they provide a useful service--intermediation.

[JW Mason]

But isn't there a difference between the economy-wide inflation constraint, and the government budget constraint? Anyone can pay for things with IOUs, there's nothing unique about governments issuing their own currency. Anyone can borrow. The difference is that a private borrower faces rising interest costs, or an inability to increase their debt past a certain point (or an unwillingness of sellers to accept their IOUs) long before their purchases reach the rising part of the supply curve of the goods they are purchasing. You might in some limited formal way be able to make limited demand for government liabilities look like limited supply of private goods. But they're really not the same. And MMT claims that (some) governments only face constraints of the second kind, while the conventional view is they face both.

[Nick Rowe]

RSJ: (BTW, what's "MZM"? And did you mean "It's mostly just *inside* money"?)

Let me re-state your point in my language:

The government issues currency paying 0% nominal interest, and bonds at i%. That's the source of its seigniorage revenue from printing currency (minus costs of paper and ink). iM/P is like a "tax" on holding currency. Or, more controversially, it's the government's monopoly profits, from its legal or de facto monopoly on note issue.

(Milton Friedman said that the optimal tax on currency is zero, because taxes distort. So the optimal rate of inflation should be minus the real natural rate of interest, to drive i down to 0%. That avoids the wasted shoe-leather costs of people holding a suboptimal stock of currency, and running back and forth between the bond market (or ATM) and the supermarket. But notice that Milton Friedman's policy recommendation effectively means we are in the ZLB trap permanently. Money and bonds are effectively the same.)

By "taxing" currency in that way, the government creates an incentive for commercial banks to produce currency substitutes (demand deposits etc.) to help people avoid those taxes. But, *if* the banking system is competitive, the super-normal seigniorage profits of the commercial banks get competed down to zero, either eaten up in operating costs, or by paying interest on demand deposits.

Given that governments generally want a (say) 2% inflation target, and it's difficult to pay interest on currency, we can't eliminate that seigniorage tax. We are in the world of the second best. In a second-best world, it's not a priori clear whether it's better to allow banks unrestricted rights to help people avoid the tax on currency and the shoe-leather costs, or whether it would be better for the government to tax demand deposits (for example by requiring reserve ratios with no interest on reserves). The 100% required reserve advocates take one extreme position. Canada has taken the opposite extreme position, with no required reserves, and interest on reserves.

[vimothy]

"Printing money is a nice little sideline, but we are still going to need taxes."

Incidentally, in the UK, quantity of notes and coin as a % of GDP has a long-run downwards trend. So the amount of "free lunch" on offer here is gradually vanishing.

[vimothy]

MZM = money of zero maturity

[vimothy]

"iM/P is like a "tax" on holding currency."

I think this is the most appropriate way to think of seigniorage. In effect, it's not a free lunch either, which brings us back to Landsburg's point about incidence. Somebody has to bear the costs. The real question is who that is going to be.

[Nick Rowe]

JW: "The difference is that a private borrower faces rising interest costs, or an inability to increase their debt past a certain point (or an unwillingness of sellers to accept their IOUs) long before their purchases reach the rising part of the supply curve of the goods they are purchasing"

That "rising part of the supply curve" is the flip-side of the falling value of the falling value of the IOU's issued, and their downward-sloping demand curve. If I start issuing IOUs, they will not usually circulate as a medium of exchange. That's why there's no "convenience yield" on Nick Rowe IOUs like there is on Bank of Canada currency IOUs. So I can't earn seigniorage, and have to pay interest to persuade people to hold a positive stock of my IOUs, otherwise they will fall in value very rapidly as I issue more. Rather like a bus company paying its drivers in bus tickets.

[Nick Rowe]

vimothy: thanks on the MZM.

There are two ways to think of seigniorage profits:

1. i(M/P) where M is currency, and i the nominal rate on bonds.

2. dM/P = (dM/M)(M/P) where dM means amount of new currency issued per year.

IIRC, the present value of the two flows of seigniorage is the same.

[Ramanan]

Nick,

You have talked of the long run budget constraint.

I think the most important point of Macroeconomics can be rephrased in a similar manner: An increase in Government expenditures ΔG = $1 has an effect given by the Keynesian multiplier but the long run multiplier is 0. And given this, is there a way out ? (and of course there is).

Of course, this conjecture depends on a bit of abstraction and really going into high level of detail in deciding which ceteris is paribus, which stocks/flows are exogenous or endogenous etc. Because looked at another way, the multiplier is very strong.

All sorts of explanations have been given ranging from crowding out to inflation etc by the profession...

The Neochartalists' argument is that the budget constraint is not really a constraint because the so called "sovereign" governments cannot default on their obligations. In fixed exchange rate regimes, it is a constraint since there is a peg to be defended and foreigners' claims can be very high putting a tight reign on the government's fiscal stance.

The 1970s I believe was the second most important period in Economics (first is the present shared with 1930s ?) .. the Keynesian demand management failed and economists were left with the task of explaining this failure.

There was one mechanism which I liked which explains the paradox of the multiplier but which I believe is quite incorrect. It was from the London Business School - Ball and Burns. (Have just read an account of this not the original model.). An initial injection of $100 extra of government expenditures increases demand by a huge amount but over time this effect gets nullified. The reason given is that the increased demand increases wages slowly over time and while the higher demand increases imports, higher wages decreases exports. Though I do not believe in this, I thought it is somewhat enlightening.

Coming back to MMT, I believe they have missed addressing the mix between real and monetary phenomenon - focussing all their energy on the fact that governments do not default - with catch phrases such as "debt is not debt" to the point of being dismissive of ideas that there can be constraints from the external world to arguing that imports are "free lunch" (your terminology)!

[Mike Sproul]

This is a good example of why macroeconomic models are not covered in texts on microeconomics or financial economics. The macro models simply make no sense. In Micro, we put the quantity of apples on the horizontal axis and the price of apples on the vertical axis, and we get the demand for apples. Then macro comes along and puts the quantity of everything on the horizontal axis and the price of everything on the vertical axis and calls it an AD curve. If people object to this meaningless aggregation then just start piling on math until they walk away. After that, you can start talking about how to "use taxes if you want to reduce Aggregate Demand to prevent inflation", and by then the micro folks won't be listening, and you can have a nice friendly discussion with the people who believe as you do, and who actually think they are making sense when they talk about AS/AD, IS/LM, Y=C+I+G, and MV=Py.

[tjfxh]

NIck: "It's a different way of saying the same thing. The only difference is that *maybe* MMTers don't believe interest rates affect desired savings or investment, so want to use fiscal policy. See my last post on reverse engineering."

No "maybe" about it. We seem to be are going round and round on this. MMT flatly and emphatically rejects monetary policy and favors fiscal policy. See your last post on reverse engineering. :)

MMT combines Godley's SFC macro modeling and sectoral balance approach with Lerners's functional finance, among other things, to achieve its aim, full employment with price stability, which is the Fed's mandate, btw. MMT is recommending substituting fiscal policy for monetary policy. No and's, if's, or but's.

With respect to this discussion, MMT'ers deny that the MMT approach can be reduced to ISLM and recast as a sort of monetarism, as we have already gone thorugh previously. MMT is fiscalism for good reason: The transmission from government expenditure to an increase in nongovernment NFA to increased NAD, and the opposite in the case of taxes, is clear. The transmission of interest rate management is not clear for a variety of reasons that MMT'ers explain. MMT'ers reject the assumption of that interest rates affect I & S in the way presumed, so that the ISLM model is not useful as a policy tool. For MMT the sectoral balance approach and functional finance are policy tools, using fiscal policy.

You also assume that taxation (revenue) funds expenditure. MMT denies this, in that in a fiat system, a monetarily sovereign government that is the monopoly issuer of a nonconvertible floating rate currency funds itself directly with expenditure. No taxes needed for funding. Not now. Not ever.

Expenditure and taxation are separate fiscal operations that do not depend on each other. Expenditure transfers real resources from private to public and increases nongovernment NFA in the exchange. Taxes withdraw net financial assets generated previously by the expenditure that created these assets ex nihilo. That is what "fiat" means. The connection between expenditure and taxation is that the currency that government creates ex nihilo is in effect a tax credit. That's it. Nongovernment must obtain these credits to satisfy its obligations mandated by government in the form of taxes, fees and fines.

The federal government in the US spends first and then taxes to reduce reserves as needed to control inflationary pressure and issues tsys to drain reserves from the interbank market. Taxation is a fiscal op. Tsy issuance is monetary op. Expenditure adds reserves, taxes subtract reserves, and tsy issuance converts reserves into another asset form without altering NFA. Monetary policy only alters NFA through the amount of interest paid, since interest payments increase nongovernment NFA. So lower rates are deflationary in this respect, and higher rates inflationary, according to MMT.

To understand MMT. it is necessary to recognize that MMT is a "new paradigm," in that it rejects many assumptions (norms) on which the current economic paradigm rests because, e.g., they do not reflect how the Treasury, cb, and financial sector actually interact — as Scott F. pointed out in another thread, as I recall.

I am happy that you are trying to understand MMT, but I think that without approaching it on its own terms, you are getting confused about what MMT is saying and more importantly at this critical juncture, what MMT is prescribing. ISLM is never going to capture what MMT is saying becuase its transmission is monetary (interest rates) and MMT is fiscal (expenditure and taxation). MMT is shouting out to dump monetary policy and switching to fiscal policy based on SFC macro models, the sectoral balance approach, and functional finance (and some other things, too). MMT sees the issue as lagging demand due to demand leakage stemming from increased saving desire and net imports, which government needs to offset fiscally with a corresponding deficit.

BTW, retiring the debt by increasing reserves to buy up tsys simply puts the reserves drained from the interbank market back into the market. This would mean that the overnight rate would go to zero unless the cb paid interest on excess reserves equal to its target rate, or else set the target rate equal to zero (the MMT recommendation). What you are proposing by transferring tsys to the Fed's book is the no-bonds recommendation of MMT'ers, which the Fed is now pursuing to some degree. You say that is might be better to buy back bonds now because there might be inflation in the future. The fact is that QE is depriving nongovernment of the added NFA of the interest to be paid and transferring it to the Tsy to reduce the deficit. This is just adding to the demand leakage that is crippling the economy now. If the interest were used to increase the deficit by that amount, this extraction would not lead to demand leakage. Moreover, we can deal with inflation should it arise fiscally, in a targeted fashion, instead using the blunt instrument of interest rates to contract the entire economy and increase the buffer of unemployed as we have been doing under NAIRU.

[RSJ]

Nick, the interest on MZM is the MZM Own Rate, which is negative in real terms, and about 2-3% less than FedFunds. IIRC, MZM Own rate has been about -1%. Perhaps -1.6% when adjusted for CPI.

And this does not take into account fees.

And yet households are willing to hold 7 Trillion of it -- yet government must pay the bond rate on 7 Trillion of public debt otherwise there would be an explosion of inflation? Really?

[RSJ]

Here's the data

[Andy Harless]

RSJ,

Part of the spread between the Treasury rate and the MZM Own Rate represents the banking (and quasi-banking) industry's costs, including the cost of capital to the banking industry. The rest (if there is any) must be monopoly profits. If the government starts taking retail deposits, it will also have associated costs, so it certainly won't be able to reap the entire spread. If you want to argue that it would be efficient to have a government money monopoly instead of inside money, you need to say either (1) why the government's costs would be significantly lower or (2) why there is monopoly power in the banking industry. I can think of arguments for both of these, but they aren't trivial questions.

[RSJ]

Notes and coins are insignificant.

Just because some small amount of interest is paid does not mean that there is no seignorage.

A more robust definition of seignorage income should be defined as differences in funding costs.

Banks can borrow at negative real rates. There is a limit, due to the limited demand for the public to hold zero maturity assets. Nevertheless that demand is not decreasing as a share of GDP.

Very few people use notes and coins as we move to a cashless society, but this has nothing to say about seignorage income. The seignorage income is still there. It should be measured in terms of spreads on funding costs, not by counting "coins".

[vimothy]

Andy, I think the problems associated with RSJ's proposal are if anything more serious than that. What advantage does the govt have in terms of the actual process of intermediation, i.e., why should the govt be "better" than the banking system at transforming illiquid assets into liquid liabilities?

[RSJ]

Andy,

The banking industry is an oligopoly, and has been from the very beginning, when governments tried to fix the price of deposit rates.

The Law of Large numbers is the best possible example of an increasing returns to scale industry.

I don't think that the government should necessarily offer retail banking services, but they should tax banks to ensure that their funding costs are at least the fed funds rate. A simple idea would be to tax all bank assets that aren't CB liabilities at the policy rate, and then give banks tax credits for any interest payments made to creditors that are not supplying bank capital, as defined by bank regulators.

In terms of evidence, you can point to more funds being paid out in the form of bank bonuses to employees than were paid out to shareholders. That, plus compensation in the industry more generally, is prima facie evidence (to me) that their profits are greater than their overall cost of capital.

[vimothy]

RSJ, in the UK (as in Canada), the BoE pays interest on reserve balances, so the govt only earns seigniorage on notes and coin.

[Clonal Antibody]

Nick,

Abba Lerner never wrote a book called "Functional Finance" The Federal Debt paper is the only time in his publications that "Functional Finance" is highlighted. I guess, that the term stuck to his writings.

To get at what the current MMT economists are about, you should probably start at some of the Levy Institute's working papers, and work backwards to get at the original sources, particularly Working Paper No. 272. Functional Finance and Full Employment: Lessons from Lerner for Today? by. Mathew Forstater http://www.levyinstitute.org/pubs/wp272.pdf The references from that paper should lead you to the relevant Lerner works.

There is a good website on Abba Lerner's work at http://www.economyprofessor.com/theorists/abbalerner.php and http://www.newschool.edu/nssr/het/profiles/lerner.htm

However, that will only lead you to his life's work, and not necessarily to the body of knowledge that developed from his ideas.

[JW Mason]

Functional finance means following this budget procedure if you are a government:

1. Decide the desired level of government expenditure, i.e. the share of society's real potential output thatis better allocated publicly rather than privately. (G)

2. Given G and the existing level of taxes net of interest and transfer payments, then:
A. If the output gap is too big and inflation is below target, reduce taxes or raise transfer payments.
B. If the output gap is too small and inflation is above target, raise taxes or reduce transfer payments.

3. Repeat in the next period.

The thing about this procedure is that you can follow it without knowing anything about the existing stock of government debt. In particular, it is never the case that you have to set G below the level that would otherwise be desirable because the existing stock of debt is too large.

If you can follow this rule, you are in a functional finance world. If you can't, you are not.

The argument that Lerner was specifically challenging was that there is an additional constraint on public borrowing beyond the economy-wide inflation constraint, such that if we act too aggressively to reduce the output gap (or raise the inflation rate) in this period, we will be forced to accept a larger output gap (or lower inflation rate) in some future period.

Of course you can redefine long-run budget constraint to mean inflation constraint. But at that point you are not using the phrase the way other people do. And you are not showing that the functional finance and mainstream views are equivalent, you are accepting the functional finance view and rejecting the mainstream view.

Evidence? Just look at the enormous effort - not just academic but in very important policy contexts like the European Growth and Stability Pact - to define acceptable limits to government deficits and debt-GDP ratios. If you agree with that - if you think that the optimal level of today's deficit is lower when the existing stock of debt is higher - then you are taking the mainstream view. If you think that the optimal level of today's deficit is the same regardless of the existing debt stock, then you are taking the functional finance view. They are not the same.

[vimothy]

JW Mason, What does this rule tell you to do when inflation or the stock of pubic debt becomes very large AND there is an output gap?

[Derridative]

"If r is less that g (forever), there's a free lunch from issuing debt"

I think this free lunch is a lot smaller than a lot of MMTers assume.

True, the government can run perpetual deficits while keeping total debt constant as a proportion of GDP -- but only by running smaller and smaller deficits (relative to GDP) over time.

If deficits are constant as a proportion of GDP over time, then the debt is growing faster than the economy b/c it is an accumulation of deficits that are growing as fast as the economy and then additionally paying interest.

This seems like a more significant constraint on deficit spending than MMTers realize, even if their understanding of the economy is accurate.

[vimothy]

Derridative has the best pseudonym since Kate Mossad.

[Nick Rowe]

JW: suppose there's a major emergency, so G has to be big this period, but you know it will be smaller next period. If private demand is roughly the same in the 2 periods, your version of FF would mean big taxes this period and small taxes next period. That is generally not optimal. The deadweight costs of distorting taxes will be bigger than if taxes were smoothed. Generally, it's much better to use bond-finance to handle fluctuations in the amount of G you want. That's Barro's tax-smoothing hypothesis. (Not the same as Ricardian Equivalence).

Clonal: Damn! My memory must be going. I'm sure I read a big beige book by Abba Lerner. And I thought the title was Functional Finance!

[JW Mason]

What does this rule tell you to do when inflation or the stock of pubic debt becomes very large AND there is an output gap?

Those are two separate questions. If the stock of debt gets large, you ignore it and keep doing what you were doing. You always set today's deficit the same way based in today's inflation and output, regardless of the existing stock of debt. Lerner was very explicit about this.

If you're getting conflicting signals from inflation and the output gap, that is a problem. Much (most) of Lerner's later work was addressed to exactly this question. His view was that this situation would arise if you had excessive levels of monoPoly power in the economy, which isn't a problem you can solve with macroeconomic tools. In any case, we can say that if this divergence is a persistent problem, we are not in a functional finance world. But that's still different from there being a long-run government budget constraint, in the way it's normally understood.

[JW Mason]

oops, italics were supposed to be the first paragraph only.
[Fixed NR]

[Nick Rowe]

Derridative: "If deficits are constant as a proportion of GDP over time, then the debt is growing faster than the economy b/c it is an accumulation of deficits that are growing as fast as the economy and then additionally paying interest."

That is a very helpful and new (to me) insight. By "deficits" you mean "primary deficits", yes?

I still can't quite wrap by head around the math, to get a back of the envelope calculation of how big it is. Care to explain a little more fully, with just a bit more math? (What percentage of GDP can you run a permanent primary deficit as a function of r and g?) Thanks.

[Clonal Antibody]

I am hoping that this takes care of the italics issue

[Nick Rowe]

Damn! Realised that my above comment in response to Derridative doesn't make sense. Still trying to get my head around the math.

[Clonal Antibody]

Nick,

There is always a problem when a society (public or private) uses interest bearing debt instruments to finance its spending. See Michael Hudson's The Mathematical Economics of Compound Rates of Interest: A Four-Thousand Year Overview - http://michael-hudson.com/2004/01/the-mathematical-economics-of-compound-rates-of-interest-a-four-thousand-year-overview-part-i/ and http://michael-hudson.com/2001/04/the-mathematical-economics-of-compound-rates-of-interest-a-four-thousand-year-overview-part-ii/

This is one of the reasons that in the current MMT thinking, the only way to make money work is to have a Zero or negative natural rate of interest. Dis-savings have to be forced by taxation, if savings are held in interest bearing instruments.

[anon]

"tax banks to ensure that their funding costs are at least the fed funds rate"

there are branch infrastructure, salary, and admin costs that require those spreads

its called retail banking

[Max]

JW, that is a brilliant concise summation. Thank you.

I would add that popular discussion of the danger of debt assumes an unstable multiple equilibrium situation where if the bond market "predicts" default, then default will occur. We see this playing out in the Euro zone due to the politically fragmented design of the system.

[vimothy]

JW Mason, They are related questions in that then govt can either finance its deficit by printing money or by issuing bonds. For "functional finance", the can opener is the assumption that the govt's budget constraint does not matter. So it's no surprise that it doesn't have any useful suggestions for states of the world in which it does.

[Determinant]

Why are getting angry at banks in Canada when they are effectively nationalized anyway? All bank stocks are broadly held by law. No individual may hold more than 10% of a bank's shares and foreigners may hold no more than 25% in total.

Bank stocks are a mainstay in every mutual fund, RRSP and pension plan in Canada, including the Canada Pension Plan and the Quebec Pension Plan.

Sure, most Canadians don't care about the slice of the banks they own, but the fact remains it's there. Bank profits come from Canadians and go right back to Canadians, to a large extent. They aren't Crown Corporations but the financial effect of their ownership is almost as broad.

[JW Mason]

suppose there's a major emergency, so G has to be big this period, but you know it will be smaller next period. If private demand is roughly the same in the 2 periods, your version of FF would mean big taxes this period and small taxes next period. That is generally not optimal. The deadweight costs of distorting taxes will be bigger than if taxes were smoothed. Generally, it's much better to use bond-finance to handle fluctuations in the amount of G you want.

Sure, no argument. We have two macroeconomic instruments, the tax level and the choice of bonds vs money. When we need to make big adjustments in G and are worried about the distortionary effect of offsetting tax adjustments, we'll want to turn the bonds-vs-money knob. But that's different from saying that there is an independent constraint on the size of the public debt or deficit. The desirability of smoothing taxes over time is completely separate from the question of whether the position of the public balance sheet is an independent consideration in setting tax levels.

The question is, given our decision about the optimal share of public expenditures in total output, and the observed inflation rate and deviation of total from potential output, do we choose a lower level of spending, or a higher level of taxation, if the government deficit and/or stock of debt is large, than if they are small? If you say No, you believe in functional finance. If you say Yes, you don't. It's really that simple.

[Nick Rowe]

JW: good thought-experiment. Let me re-state it. Suppose you have decided on the optimal paths of G and T and M over time. Then suddenly, just before you implement that policy, you find the government owes an additional $100 debt you had missed. How do you revise your optimal policy?

I say you increase T and/or cut G by (r-g)$100. (r is real interest rate, g is growth rate of real GDP.) If you don't do this, the debt/GDP ratio will grow without limit, which is not sustainable.

[vimothy]

Nick,

Re Derridative's comment. Set seigniorage to zero for simplicity; write the primary deficit as,

(G – T)/Y = ∆(B/Y) – (r – g)B/Y

Say that r – g = -0.01 and that ∆(B/Y) = 0

Then,

(G – T)/Y = (0.01)B/Y

As B/Y becomes small (Y is increasing over time), (G – T)/Y must become small as well.

[JW Mason]

Suppose you have decided on the optimal paths of G and T and M over time. Then suddenly, just before you implement that policy, you find the government owes an additional $100 debt you had missed. How do you revise your optimal policy? I say you increase T and/or cut G by (r-g)$100.

And I say you don't change T and G at all. Unlimited growth in the debt/GDP ratio is only unsustainable if you can identify a specific economic mechanism that brings it to a halt, within an economically-relevant horizon.

More concretely, if something in economics is sustainable for more than a few decades, it's sustainable, period. None of the relevant relationships are stable enough for models with longer horizons to matter. (Of course this is not the case for e.g. climate models!) Sure, "If something can't go one forever, it will stop." But it's equally true that if something *can* go on forever, it still will stop. And something that can't go on forever, can still go on for an arbitrarily long time.

[vimothy]

No, that's not right is it.

[JW Mason]

Anyway, we've gotten tot he crux of the matter here. To say the debt/GDP ratio cannot rise without limit, is to say that there's some limit to that ratio which cannot be crossed, independent of the macroeconomic effects of additional borrowing. Lerner's whole point is, No there's not.

[Nick Rowe]

vimothy: I'm not sure. Just can't get my head straight on it.

JW: Yes, we are getting to the crux of it. It's exactly the same sort of arguments that explain why a chain letter can't keep growing for ever (if the rate of interest on the chain letter exceeds the growth rate of the economy). Take an overlapping generations model for example. The old sell the bonds to the young, who grow old, and sell them to the next generation. If the debt/GDP ratio gets too big, the next generation of young simply won't be able to afford to buy the bonds off the old.

[vimothy]

Someone help me out here.

Say r – g = α < 0 and that ∆(B/Y) = 0. Then,

(G – T)/Y = αB/Y

Govt can run a constant deficit as a proportion of GDP equal to αB/Y while keeping ∆(B/Y) = 0...?

[Nick Rowe]

JW: which makes me note something ironic. The only model I can think of where the debt/GDP ratio could grow without limit is Barro's model of Ricardian Equivalence. For there, the old generation *bequeath* the bonds to the young, rather than selling them to the young. But in Barro's model, the choice between bond- and tax-finance is irrelevant anyway. Bond-financed tax cuts do not increase consumption demand or shift the IS curve.

vimothy: it sounds right when r exceeds g. Don't see why it can't also be right when r-g is negative. But my brain is getting old and tired.

Anyway, assume you are right! Suppose g=3% and r=2% and B/Y=30%. Ballpark for Canada. Then you can run a permanent primary deficit of 0.3% of GDP. About the same as seigniorage. A second nice little sideline, but not enough to pay the (T)bills ;-)

[Nick Rowe]

vimothy: I think you are right. If so, then you and Derridative have really cleared up something that my head wasn't clear on last night. Flashes of light going off!

Compare your formula to my formula for seigniorage. Properly interpreted, they are exactly the same!! Currency pays a real interest rate of minus 2% (in Canada, with 2% inflation target). So seigniorage on currency is exactly the same as seigniorage on any government liability that pays a real rate of interest less than the real growth rate.

Bingo!

[vimothy]

It looks right to me. Govts can run continuous primary deficits as a constant proportion of GDP and stabilise public debt as a proportion of GDP if r < g because the growth rate of Y is greater than the growth rate of the debt. If you want the debt and the economy to grow at the same rate, the govt primary surplus / deficit must make up the difference between r and g.

Impossible to think straight in this heat though. And like you say, ultimately pretty negligible.

[vimothy]

Nick, Cool! I love economics for those light-bulb moments. Which formula for seigniorage, sorry?

[Nick Rowe]

vimothy: seigniorage is dM/P, = (dM/M)*(M/P). If inflation is 2%, and real growth is 3%, then dM/M is 2%+3% for M to stay a fixed ratio to Nominal GDP.

assume M/P is 5% of Nominal GDP, and nominal GDP is growing at 3% real plus 2% inflation =5%. So seigniorage = (dM/M)*(M/NGDP)=5%*5% = 0.25% of GDP.

The only difference is that the real rate of interest on currency is negative, being minus 2% inflation.

[Clonal Antibody]

Nick,

Here is a more legible copy of the Lerner "Functional Finance and Federal Debt" paper

http://www.scribd.com/doc/29435145/Functional-Finance

[beowulf]

Nick, you're defining seigniorage as govt's profit on inside money, RSJ is saying govt should use taxing power to claim (or rather reclaim) seigniorage on outside money as well.

As for Abba Lerner, his Functional Finance ideas were further developed in his book The Economics of Control, I'll let Lord Keynes explain which parts to look at:
I have marked with particular satisfaction and profit three pairs of chapters-chap. 20 [Production and Time] and 21 [Interest, Investment and Employment I], chap. 24 [Interest, Investment and Employment III (functional finance)] and 25 [Capital, Investment and Interest], chap. 28 [Foreign Trade III (in a capitalist economy)] and 29 [Foreign Trade IV (in the controlled economy)].
Here is the kernel of yourself. It is very original and grand stuff. I shall have to try when I get back to hold a seminar for the heads of the Treasury on Functional Finance.
http://socialdemocracy21stcentury.blogspot.com/2010/09/would-keynes-have-endorsed-modern.html

Here's a pdf of Lerner's book
http://203.200.22.249:8080/jspui/bitstream/123456789/1052/1/The_Economics_of_control.pdf

[Ramanan]

Vimothy,

"Govt can run a constant deficit as a proportion of GDP equal to αB/Y while keeping ∆(B/Y) = 0...?"

The government has very less control over its deficit and is endogenously determined by the private sector. The question then is if that is the case, whats the worry ? The answer is that the government controls its expenditures and the tax rate. The deficit and the debt are determined by G, t (tax rate), and the private sector (such as via the propensity to consume). The fiscal stance of governments is such that it does not bring full employment. The private sector can grow by itself if its expenditure is greater than its income which is an unsustainable process because it leads to increased indebtedness and can cause a big recession.

[Nick Rowe]

Thanks Clonal and beowulf! Links updated.

beowulf: "Nick, you're defining seigniorage as govt's profit on inside money, RSJ is saying govt should use taxing power to claim (or rather reclaim) seigniorage on outside money as well."

Yep. (But both you and RSJ are using "inside money" and "outside money" in the exact opposite senses to which they are normally used in monetary theory.)

[beowulf]

YW Nick. I don't think RSJ used the term "inside money" so reversing the terms is just on me.

Instead of further mangling what RSJ's proposing, I'll just link to his post on the topic.
http://windyanabasis.wordpress.com/2011/03/28/leaving-modern-money-theory-on-the-table/

[JW Mason]

Beowulf,

Hey, thanks for that. I'm ashamed to say, I've never read The Economics of Control. From your link to my Kindle.

[RSJ]

"Anyway, assume you are right! Suppose g=3% and r=2% and B/Y=30%. Ballpark for Canada. Then you can run a permanent primary deficit of 0.3% of GDP. About the same as seigniorage."

IMO, that's not the right way to think about this. What r < g means is that given *any* primary deficit/GDP bound, there is a (finite) Debt/GDP bound that is a function of the deficit to GDP bound.

And more importantly, if you exceed your primary deficit/GDP bound for a while, it doesn't matter, because over time, your debt to GDP will fall back to target if you only lower future deficit spending to the target.

Therefore there is never any future tax obligation. At best there is an obligation that future deficit spending be reduced back to the target. The economy "forgets" past violations of the deficit/GDP bound, as time whittles them away without the need to run surpluses.

The best example was reductions in debt after WW2. There were no surpluses -- well, there were rare negligible surpluses, but the debt/GDP ratio fell rapidly just with time and economic growth.

[vimothy]

Nick, Understood.

[Min]

"So you are not worried about inflation. Instead you are worried about deflation. And you were a government that could print your own money. What would you do?
"You would print money and spend it. Or print money and use it to finance tax cuts. And you would keep on doing it, more and more, until you got to the point where you did start to worry about inflation. You first eat all the free lunches.

"That's the underlying kernel of truth in Abba Lerner's Functional Finance (pdf) [see also his book The Economics of Control (pdf)]. And I don't know of any mainstream macroeconomist who would disagree."

Gee, Nick, where are those guys hiding? Why aren't they advising gov'ts in the U.S. or Europe? If they aren't advising gov'ts, why aren't they writing op-ed pieces in newspapers or appearing on TV news shows to point out the sea of nonsense that we are awash in?

[RSJ]

Or perhaps a better way of looking at this is that Ricardian equivalence is a self-consistent logical loop:

Because r > g, future deficit spending implies future tax increases, and therefore the household sector is indifferent between holding government bonds (with deficit spending) and not holding government bonds (with a balanced budget), so the sale and purchase of bonds has no effect on rates.

The seignorage view is that households _do_ demand government bonds, and are willing to pay a premium for them over the return on capital.Households value the secure payment over the variable payment because they are credit constrained and therefore cannot consume according to PIH.

Because they pay a premium, r < g (where r is the risk-free rate, not the rate charged to the private sector). Because r < g, then there will not be future tax increases, and therefore they value them. Another loop.

But the second loop is limited, in that the public's insurance demands are limited, and so increasing the stock of debt beyond this insurance demand will result in higher rates.

The government can only enjoy the benefits of r < g to the extent that the stock of debt does not exceed the insurance demand. Beyond that, additional debt sales will cause r to increase.

Both "loops" are internally consistent, but only one -- the non-Ricardian loop -- is consistent with the data. It makes no sense for households to pay a premium for bonds over capital if they do not consider bonds to be wealth.

Whether or not r < g tells you which loop you are on.

[vimothy]

Ramanan, Primary deficit is G - T, and (you've already agreed that) the govt controls both those variables.

[Clonal Antibody]

Nick,

You may also find Keynes thoughts on Lerner's Functional Finance in this 2002 article by David Colander “Functional Finance, New Classical Economics and Great Great Grandsons” http://community.middlebury.edu/~colander/articles/Functional Finance, New Classical Economics and Great Great Grandsons.pdf

Quote:
It is a grand book worthy of one’s hopes of you. A most powerful piece of well organized analysis with high aesthetic qualities, though written more perhaps than you see yourself for the cognoscenti in the temple and not for those at the gate. Anyhow I prefer it for intellectual enjoyment to any recent attempts in this vein.

In the second of the two books which you have placed within one cover, I have marked with particular satisfaction and profit three pairs of chapters—chap 20 and 21, chap 24 and 25, and chap 28 and 29. Here is the kernel of yourself. It is very original and grand stuff. I shall have to try when I get back to hold a seminar for the heads of the Treasury on Functional Finance. It will be very hard going — I think I shall ask them to let me hold a seminar of their sons instead, agreeing beforehand that, if I can convince the boys, they will take it from me that it is so!1

The above letter from John Maynard Keynes to Abba Lerner celebrating his book The Economics of Control contains the essence of my view of functional finance
End Quote

[beowulf]

JW Mason, You're welcome. Its a pretty interesting read. I noticed Lerner thanked Friedrich Hayek (among others) in the preface. He'd been a student of Hayek's at LSE but this book is definitely the anti-Road to Serfdom (The Road from Serfdom?). Its a marvel Hayek didn't stroke out by Chapter 1, "The three principle problems to be faced in a controlled economy are employment, monopoly and the distribution of income... unless these primary tasks are accomplished the economy is uncontrolled." (p. 3).

[Determinant]

Gee, Nick, where are those guys hiding? Why aren't they advising gov'ts in the U.S. or Europe? If they aren't advising gov'ts, why aren't they writing op-ed pieces in newspapers or appearing on TV news shows to point out the sea of nonsense that we are awash in?

Because they are deeply unfashionable. The "lower taxes, free-market" Hayekian/Friedman crowd won the day. We spent three decades driving the Demand Managers from office and replacing them with inflation-hawks. Such is our reward for this decision.

The public understands taxes. They don't understand, or have forgotten the plague of chronic unemployment. Curiously the newspapers of the 1930's were also rife with the delusion of seeing inflation at every turn, just like today. Eerily like today. We've come full circle.

We need a celebrity. Somebody to go on PBS like Milton Friedman did and explain all this. Actually Milton Friedman's greatest economic contribution to both macro and micro was entirely practical: designing and implementing automatic payroll income tax deductions to finance WWII. It's what makes modern government possible. Much as he later hated the beast, he played a key role in creating it.

Aggregate Demand Management isn't something to be feared; it isn't innately anti-capitalist, even though the Road to Serfdom fans might think it so. ADM's flip side is Aggregate Supply. Under an ADM regime the free market, that is price-determined consumer-driven micro choices are what controls the distribution of aggregate supply. Prices determine how we slice the pie and provide the signals on what consumers want. What's wrong with that?

All ADM says is that Aggregate Demand can be played with to ensure that an economy can consume Aggregate Supply most of the time so that chronic unemployment does not result. You can sugar-coat it with some social programs but you don't have to. This appeals to me on humanitarian and compassionate grounds but it isn't strictly necessary.

It forms the basis of the post-war social contract but we've spent 30 years demolishing that. I think we were wrong, collectively.

[RSJ]

"Primary deficit is G - T, and (you've already agreed that) the govt controls both those variables."

Taxes are paid as a proportion of (nominal) income, which the government does not control. Government spending includes automatic stabilizers, which are also not controlled.

So the private sector can force the government into a budget deficit, merely by becoming poorer (earning less income) and requiring more benefit payments.

[Clonal Antibody]

Also from Mathew Forstater's "LERNER, Abba Ptachya (1903-1982)" http://www.cfeps.org/pubs/wp/wp52.htm

Quote:

Decisions concerning taxation are to be made only with regard to the economic effects in terms of the promotion of full employment, price stability, or other economic goals, and not ever because ‘the government needs to make money payments’ (Lerner, 1943, p. 354). Likewise, ‘the government should borrow only if... the effects’ of borrowing are desired, for example ‘if otherwise the rate of interest would be too low’ (Lerner, 1943, p. 355).

These points of view were repeated and elaborated by Lerner in his 1951 book, The Economics of Employment:

[T]axes should never be imposed for the sake of the tax revenues. It is true that taxation makes money available to the government, but this is not an effect of any importance because money can be made available to the government so much more easily by having some created by the Treasury. (1951, p. 131, original emphasis).

Likewise, ‘borrowing’ is also not a funding operation for Lerner.

What are the purposes of taxation and borrowing, if not to fund government spending? The purpose of taxation for Lerner is its ‘effect on the public of influencing their economic behavior’ (Lerner, 1951, p. 131, original emphasis) including, as we will see below, creating a demand for government fiat currency. Like taxation, borrowing is not a funding operation; rather, it is a means of managing reserves and controlling the overnight interest rate when the government runs a budget deficit:

[T]he spending of money...out of deficits keeps on increasing the stock of money [and bank reserves] and this keeps on pushing down the rate of interest. Somehow the government must prevent the rate of interest from being pushed down by the additions to the stock of money coming from its own expenditures.... There is an obvious way of doing this. The government can borrow back the money it is spending (Lerner, 1951, p. 10-11, original emphases).

Two extraordinary results follow from Lerner’s analysis here. First, the implication is that ‘borrowing’ logically follows, rather than precedes, government spending. In fact, this analysis questions the accuracy and relevance of the term ‘borrowing’ itself for discussing government bond sales. Second, note that the budget deficit is causing interest rates to fall, the exact opposite from what traditional theory has long predicted (i.e., ‘deficits cause high interest rates’).

The role of taxation and borrowing, reserve management and interest rate maintenance will become clearer upon examination of two, less well-known, Lerner contributions, ‘Money as a Creature of the State’ (1947) and his Encyclopaedia Britannica entry on ‘Money’ (1946), which place him squarely in the Keynes-Knapp Chartalist school, and which is key to fully understanding the possibility and effectiveness of Functional Finance. The ability of the government to conduct fiscal and monetary policy according to the principles of Functional Finance is made possible by the fact that, as the title of Lerner’s paper states, ‘money [i]s a creature of the state’:

The government — which is what the state means in practice — by virtue of its power to create or destroy money by fiat and its power to take money away from people by taxation, is in a position to keep the rate of spending of the economy at the level required to fill its two great responsibilities, the prevention of depression, and the maintenance of the value of money. (Lerner, 1947, p. 314)

In adopting this view Lerner followed Keynes ....

End Quote

[anon]

"Gee, Nick, where are those guys hiding? Why aren't they advising gov'ts in the U.S. or Europe? If they aren't advising gov'ts, why aren't they writing op-ed pieces in newspapers or appearing on TV news shows to point out the sea of nonsense that we are awash in?"

Perhaps I'm missing something, but wouldn't most economists agree in practice with what Nick Rowe is saying here? Disregarding the inflation hawks sitting on Fed committees (they are mostly playing reputational games; besides, nothing says you couldn't have inflation hawks in a MMT world), most economists expressly make Nick's point: below-target expected inflation is effectively a license to print money, and not doing this in 2008 was a clear failure of monetary policy (or "aggregate demand management", to use Determinant's term).

[Ramanan]

vimothy | April 23, 2011 at 11:12 PM,

No I was referring to the tax rate as opposed to total taxes.

Total taxes is dependent on the level of activity in the private sector.

Of course, there is nothing preventing the government from targeting a level of deficit. For example, if the deficit is more than the target, tighten policy and if less, relax ... but that policy leads to sustained unemployment and has to give in sometime.

[himaginary]

"When you get to the point that Aggregate Demand is high enough, so you start to worry about inflation, you use taxes to finance past, present, or future government expenditure precisely because you don't want to print more money and make inflation higher."

Some people say that it's too late when you start to worry about inflation. They say that the change from deflation to inflation is not continuous, and by the time you start to worry about inflation, two-digit inflation -- if not hyperinflation -- is inevitable. This claim is sometimes called ketchup theory.

Many people wonder why BOJ doesn't move more aggressively. The answer lies in this theory. Believe it or not, this theory is very prevalent among Japanese macroeconomists. (Although it isn't known by the name of "ketchup theory" in Japan -- maybe because we mainly use tube instead of bottle as ketchup container. :)

[Nick Rowe]

anon: "Perhaps I'm missing something, but wouldn't most economists agree in practice with what Nick Rowe is saying here? Disregarding the inflation hawks sitting on Fed committees (they are mostly playing reputational games; besides, nothing says you couldn't have inflation hawks in a MMT world), most economists expressly make Nick's point: below-target expected inflation is effectively a license to print money,...."

You are not missing anything. I don't think there's anything I have written here (with the possible exception of the bit I wrote right at the end on dynamic inconsistency when r is less than g) that any mainstream economist would disagree with. They might not have thought about it in terms of "Functional Finance". They might have expressed the same basic ideas in a different way. But that's all. (Many don't like the phrase "printing money" that I have used.)

In fact, what I have written here (except for the bit right at the end) is effectively institutionalised in Canada (and most countries). The job of the Bank of Canada is to find the free lunches, grab them, and hand them over to the government so the government can eat them. And the job of the government is to obey the long run government budget constraint, including the free lunches that the Bank of Canada gives it once a year when it hands its profits over to the government.

But not everyone is getting it, the way you are.

"....and not doing this [printing money] in 2008 was a clear failure of monetary policy (or "aggregate demand management", to use Determinant's term)."

Though most mainstream macroeconomists would disagree with you on this part. Scott Sumner has been arguing for this, but many would say that monetary policy can't do much at the ZLB, and/or that events happened too big and quick for us to blame central banks.

"Aggregate demand management" by the way is a fairly standard term, though perhaps it has a slightly old-fashioned Keynesian flavour.

[Nick Rowe]

himaginary: I think the "ketchup theory" is at the back of a lot of economists' minds. I'm surprised that the only explicit mention of it is in that Economist article.

Thank you by the way for translating so many of my posts into Japanese! I often read them, with Google translate, to see your added commentary.

Determinant: "It forms the basis of the post-war social contract but we've spent 30 years demolishing that. I think we were wrong, collectively."

I don't really see it that way. Most central banks target inflation. Rightly or wrongly, they see inflation targeting as a good way in practice to implement aggregate demand management. By "flexible" inflation targeting, they believe they can keep the unemployment rate about as close as is practically feasible to what they call the NAIRU, or the "natural rate", or what we used to call, oxymoronically, "full-employment unemployment".

In the last couple of years only, some however did seem to lose faith that aggregate demand management was possible. See my old post http://worthwhile.typepad.com/worthwhile_canadian_initi/2010/05/the-orthodox-loss-of-faith.html

And the Eurozone is a mess.

Clonal: thanks. But I would say that Lerner and Keynes have basically won their war on that point, at least to the extent that they had a valid point. What I have been trying to do in this post is elucidate the extent to which their point is valid.

RSJ: I like your 11.04 "loops" comment. I think you are onto something.

Except just for this bit, where you go badly wrong, in confusing what is wealth for the individual and what is wealth in aggregate:
"Both "loops" are internally consistent, but only one -- the non-Ricardian loop -- is consistent with the data. It makes no sense for households to pay a premium for bonds over capital if they do not consider bonds to be wealth."

In a Barro-Ricardo world, a government bond is wealth for the individual who holds it, just not net wealth in aggregate.

[Ramanan]

Beowulf,

Thanks for Abba Lerner's book.

It seems Lerner's attitude toward describing imports are benefits is totally different from what is argued in the Chartalists' blogs. Lerner seems to be describing situations in which protectionists measures are being taken by nations and in that scenario, imports would be benefits. That attitude is drastically different from saying something like "we can purchase anything foreigners want to sell us" and narrative around that.

[vimothy]

And if the govt didn't run a deficit, nobody could earn any money and nobody could save, amirite?!

Anyway, empirically you have something like govt expenditure constant over the cycle, and its take from net taxes moving pro-cyclically, falling in the bust as output contracts and transfer payments increase.

But the idea that the govt cannot control its deficit and so should not worry about its budget constraint is very wrong. Even in Godley and Lavoie's model, which Ramanan is referring to, where there is no economic activity whatsoever without govt spending, steady state equilibria are defined by constant ratios of financial stocks and flows, i.e. the steady state is characterised by precisely the same sort of rule for the national debt that Nick and I were just discussing!

Which brings me to a more general observation about the MMT blogregore: inconsistency. Because MMT is basically an ideology for many of its proponents, at times positions are advanced for political reasons, and done so without regard for the fact that they contradict other positions. (To be fair/clear, this doesn’t apply to Ramanan or RSJ). The endogeneity of the budget deficit is a classic Mitchell cliché, but it’s never put next to that popular deus ex machina, gubbermint sovereignty, and examined. Warren Mosler has said to me numerous times that the govt can basically take whatever it likes and do whatever it wants and the private sector has no choice but to take its lumps. (Or see this recent Winterspeak statement: "A government's capacity to impose taxes has nothing to do with the quality of the real economy it oversees and everything to do with its sovereign power.") So how do we get from there to it being totally unable to control its budget?

[Ramanan]

Vimothy,

Thanks for saying it doesn't apply to me :-) I am quite an MMT criticizer, though quite a fan of Post Keynesianism. Yes, I agree there is a lot of inconsistency in MMT definitions and it starts from the definition of "sovereignty".

At any rate, what you have written above is not an argument. For example, you argued that some definition of wealth is all wrong without checking that national accountants use the same definition.

It is crucial to get what is endogenous and what is exogenous right.

In the G&L models you are refering to, the long-run is just a state which is reached due to decisions taken by various sectors. The fun is in the states that lie between two steady states... how economies move from state A to B .. and how policy can affect output, what happens if consumers' propensity to consume decreases and those things.

Back to the argument, the government sets its expenditures and the tax rate not taxes. High activity leads to high inflow of receipts - for example during the dot-com boom, there were a lot of capital gains and high taxes received by the US government. The US govt didn't decide the AMOUNT.

For let us suppose, that the government tries to achieve a deficit of 2%. This can be done by increasing the tax rate. Highly likely that the increase in tax rates causes a deterioration of output leading to deterioration of the budget balance.

The insistence of controlling the budget is equivalent to Milton Friedman's insistence that the money supply be controlled.

[Determinant]

Well, I would disagree on the effectiveness of ADM coordination between the Bank of Canada and the Government, but it's really a government problem, not the Bank's. The Bank has to do what it's told to do which is target inflation.

Robert Skiddelsky has a good book on Keynes for today. For somebody like myself or Min there is a good table in there which summarizes the institutional role changes over the last 50 years as we moved from worrying about unemployment to worrying about inflation.

In an MMT world the Ministry of Finance would take a much more proactive stance in macro issues than it does at present. Rarely do we see income taxes adjusted for macro management anymore.

Though this thread has shown the difference between a sustainable budget deficit and a nominal surplus is so small as to be negligible given Canada's present interest rate, monetary and growth conditions. Lucky for us a nominal balanced budget isn't a bad thing that way and it helps make the politics easier to sell.

[Ramanan]

Vimothy,

Winterspeak is just an anonymous blogger and don't think should be included in MMTers.

Btw, it should be clearly identified what MMT is. For example Nick posted on Delong's website about the name "MMT" ;-)

"MMTers" are basically a few Neochartalists and can be categorized in Post Keynesian Economics thought you will see "MMT is not Post Keynesian" in Bill Mitchell's blog. Don't think PKEists themselves identify with Neochartalists. Here is a list of Neochartalists - Randy Wray, Bill Mitchell, Warren Mosler, Scott Fullwiler, Palvina Tcherneva, Stephanie Bell Kelton, Mathew Forstater, Eric Tymoigne ...

So there is some overlap but very little. What is PKE itself is debatable but I believe one can give some proper answers.

[beowulf]

"So how do we get from there to it being totally unable to control its budget"

Ha ha, you've independently come up with the Heisenberg uncertainty principle* (and since Heisenberg didn't publish in peer-reviewed economic journals, you'll get the credit). The govt can target budget deficit or aggregate demand but not both simultaneously.

*principle implies that it is impossible to determine simultaneously both the position and the momentum of an electron or any other particle with any great degree of accuracy or certainty. This is not a statement about researchers' ability to measure the quantities. Rather, it is a statement about the system itself.
http://en.wikipedia.org/wiki/Uncertainty_principle

[dilletaunted]

"Which brings me to a more general observation about the MMT blogregore: inconsistency. Because MMT is basically an ideology for many of its proponents, at times positions are advanced for political reasons, and done so without regard for the fact that they contradict other positions."

are you talking about actual mmt *economists*? or just random people you find on the internet? (btw, what you described there is not a contradiction, as mosler is always discussing how monopolists will determine prices or quantities, but not both) because i can find random neoclassical economics proponents dicking around on the blogosphere and present you with tons, tons of contradictions

[dilletaunted]

"And if the govt didn't run a deficit, nobody could earn any money and nobody could save, amirite?!"

so, so lazy

[beowulf]

"That attitude is drastically different from saying something like "we can purchase anything foreigners want to sell us" and narrative around that. "

Yes, I noticed that (granted we were still under the gold standard in the 40's). But like Donald Trump, it appears that Lerner was in-Ramananigm. :o)
http://moslereconomics.com/2011/04/21/robert-reichs-no-so-innocent-fraud/comment-page-1/#comment-50105

[vimothy]

Ramanan,

Yes, I respect you for that. I think disagreement is a defining characteristic of honest debate. Where do MMTers criticise other MMTers?

I am defining wealth as a stock with the present value of future income flows. This way of characterising wealth doesn't make sense in Godley and Lavoie's model—e.g. in the model SIM which we discussed on the last thread, wealth is just the stock of money the govt has spent into existence. There are no individuals in the model anyway, so why would they care about their future income?

I have all the EViews files though and agree that it's fun to play about with them.

I think I've been clear about what I believe to be endogenous and what I believe to be within the govt's power to control, so I won't argue with you about that.

More interesting, to me, is the way you associate a surplus with a contraction in output. A common MMT argument about the necessity of a deficit for growth or macro stability goes like this. Every time the US govt has run a surplus is has caused a massive contraction in output. Therefore, don’t run any surpluses and you won’t have any massive contractions in output. But if you put this next to the idea of budget endogeneity, you realise that the reason surpluses sometimes lead a crash might be because they naturally coincide with the peak of the cycle—because the govt’s take from net taxes is higher when output is higher, for the reasons you give.

Also, you’re not being fair to Winterspeak. All the non-pseudonymous academic bloggers respect him/her, so it seems reasonable to use one of his/her statements as characteristic of a particular line of MMT thought. In any case, as I wrote above, it’s a statement that is totally consistent with something that Warren has been pretty clear about.

[dilletaunted]

"But if you put this next to the idea of budget endogeneity, you realise that the reason surpluses sometimes lead a crash might be because they naturally coincide with the peak of the cycle—because the govt’s take from net taxes is higher when output is higher, for the reasons you give."

yes, because none of the times that the gov't has run a surplus has been because of the purposeful efforts of fiscal 'conservatives'

[Max]

This paper, "Implications of a Budget Surplus at Mid-Year 2000", was quite prescient:

http://www.cfeps.org/pubs/pn/pn0001.html

The author's main point is that economic growth based on private debt is dangerous. But nowhere does the author say that loose fiscal policy prevents bubbles!

[Min]

"The Long Run Government Budget Constraint is undefined."

Don't some mainstream economists believe that it is, that the long run limit of G-T = 0? Not only that, don't they believe that that belief is rational?

[Clonal Antibody]

Determinant,

You said:
"Because they are deeply unfashionable. The "lower taxes, free-market" Hayekian/Friedman crowd won the day. We spent three decades driving the Demand Managers from office and replacing them with inflation-hawks. Such is our reward for this decision."

From Lerner's biography at the New School http://homepage.newschool.edu/~het/profiles/lerner.htm

Quote:
As can be gathered from this (partial) list of contributions, it is shameful that an economist who contributed so much to the arsenal of economics should nonetheless have been condemned to roam in the professional underground. Any one of his contributions should easily have earned him a Nobel Prize, and the sum total of them earned him the recognition as one of the most remarkable economists of the century. But to the staid professionals of academia, Lerner was simply not "one of us"
End Quote

[Clonal Antibody]

On the Lerner comment above, I will say this much: I had two mentors (one was my thesis adviser, and the other was a philosophical mentor) who were nominated -- both were very deserving -- one did not get the award because "He was not one of us!"