Wednesday, April 10, 2019

More "Interesting" Thoughts about Socialist Monetary Policy

Interest income is forbidden in a socialist society due to socialist ideology.   Someone is earning income on their money, when the only legitimate form of income is from labor.

In a previous post, I described a socialist monetary policy that was based upon the issue of money to pay wages to government workers (which are all workers) and then money is absorbed as it is spent in government stores and shops (which are all of them.)

If the workers are paid by crediting deposit accounts and their accounts are debited when they purchase consumer goods and services, then the payment of interest would be technically simple.   Further, it would seem that interest rates could play something like the role they do in a market economy--including one with a monetary policy using interest rate targeting.   Paying higher interest would reward workers for saving.   The reduction in spending on consumer goods and services would allow for a reduction in the production of consumer goods and services without creating shortages.   That would free up resources to produce capital goods, which would allow for the production of additional consumer goods and services in the future.   Those added consumer goods would allow workers to both spend the amount they saved as well as any accumulated interest income.   Those workers who saved would be especially rewarded for allowing an expansion in total future consumption. 

I was more concerned, however, with the use of interest rates as a tool of monetary policy.   That is, the possibility of responding to a shortage of consumer goods and services by increasing the interest rate.  As above, this should motivate additional saving and so reduced spending on consumer goods and services now.   This seems to make an increase in the interest rate an alternative to increasing the prices of the consumer goods and services.   Higher interest rates would reduce inflation--more or less.

Interestingly, these interest payments would seem to increase the quantity of money.   In the socialist economy without interest payments, all money is created by the payment of wages to workers.   Paying interest on saving would be a new source of income and a new source of money creation.    An increase in interest rates would entail an increase in income and money creation.   This would seem to add to the demand for consumer goods and services and so tend to create more pressure to increase their prices to avoid shortages. 

Of course, this can be analyzed as an income and substitution effect.  The income effect would be to increase the demand for consumer goods and services and exacerbate inflationary pressure.   The substitution effect would be to reduce current consumption spending and expand future consumer spending as described above.

In a market economic system, an increase in interest rates also increase interest income but directly reduces profits (at least accounting profits.)   There is no particular reason to believe that total income rises.  The use of monetary policy to raise interest rates typically involves a contraction in the quantity of money rather than an expansion.   Even leaving aside the impact on private sector investment (which does not exist in the socialist economy described here,) an increase in interest rates reduces expenditure.  (These "decreases" frequently involve a dampening or possible reversal of increases of consumption or investment due to other causes.)   The situation is much more ambiguous in the sort of socialist economy described here.

Can these considerations about the effects of interest payments on the quantity of money and income in a socialist economy be applied to a market economy?   Or more especially, to a market economy in transition to socialism?

Monday, April 8, 2019

Socialist Monetary Policy

Consider a command economy that doesn't fully dispense with money but rather pays wages to the workers in all of the nationalized industries and charges them prices for the consumer goods they buy in the various government owned and operated shops and stores.   In other words, they are not quite ready to have people show up for work according to their perceived abilities and take away consumer goods and services as they need them.   Those who fail to show up to work don't get paid, and those who don't get paid cannot buy consumer goods and services in the shops.

In capitalist economies, monetary policy is typically managed by adjusting policy interest rates.  For there to be interest rates there must be lenders and borrowers.   The lenders earn interest income from the money they have accumulated and lent.   They are earning income from property and not labor.   This is exploitation and inconsistent with socialist principles.

However, in a socialist economy money can simply be printed up and paid out as wages and then retired when it is spent on consumer goods and services.  If money is credited and debited to deposit accounts of workers as they are paid and spend their incomes, then the creation and destruction of money would be literal.   If wages are paid in currency and then paid over to government shops, it would be silly to shred currency when received and print it anew.   Currency would need to be collected from the retailers and then distributed to the various agencies and departments employing workers.    Worn bills would be shredded and new currency printed as needed.

Would this system result in inflation?   It would only create inflation if the government chose to raise the prices it sets for consumer goods and services in its shops.   Inflation would be a choice by the government.

What could happen is shortages of consumer goods and services at the prices set by the government.  This depends on how many consumer goods and services are produced but also by the workers as they choose what and how much they buy.

It would be possible to correct a shortage by raising prices resulting in inflation.   However, those shortages could also be corrected by reducing the wages workers are paid.   And finally, it would be possible to expand the production of consumer goods and services, perhaps by shifting resources from the production of additional capital goods, consumer services provided without charge like medical care or education or still other goods like domestic security or military preparedness. 

A socialist system would have no role for a "monetary policy" based upon the manipulation of interest rates.   And while there is a quantity of money outstanding at any time and a demand to hold it by workers, it is doubtful that the quantity theory really applies.   There is noting like gold mining or even trade surpluses resulting in additions to a stock of gold.   Further, the usual vision of "velocity" of money being spent, received, and then spent again would play little role in a system where for the most part money is created to pay workers' wages and then received by government shops and extinguished.

Now, the socialist system described above does allow workers to save.   They can simply accumulate currency or else balances in deposit accounts.   To the degree there is saving, the money created by the payment of wages will exceed the money absorbed by the sale of consumer goods and services.   Saving would involve lower demand for consumer good and services than if there were no saving.  That would allow resources to be devoted to the production of capital goods, expansion of products distributed at no charge like healthcare or education, or the military or security services.

If deposit balances are used, it would be possible to pay interest on those balances.   This would mean, of course, that workers who saved would earn income on their property, which is contrary to socialist principles.   Leaving that aside, such interest might encourage saving and so free up resources from the production of consumer goods and services to allow the production of additional capital goods or perhaps expand "free" social services or even military or policing.    Interestingly, if additional capital goods are produced, this could allow for an increase in the production of consumer goods in the future, providing additional goods and services for the workers who saved and earned interest to buy. 

Further, if a shortage of consumer goods and services were to develop, paying higher interest on deposits would encourage saving and correct the shortage without the need to raise prices or lower wages.    Inflation would be controlled by higher interest rates to the degree it encourages saving.

Collecting taxes on wages, of course, is just another way of describing a decrease in wages.   Collecting some kind of sales tax on purchases is simply another way of increasing prices.  In fact, aren't the prices paid for consumer goods and services really just a tax for taking the various consumer goods and services?

Perhaps these principles can be applied to understanding the monetary system of a capitalist system?   Or, more to the point, the transition of a capitalist system to the new socialist order! 

Sunday, April 7, 2019

Government Default and MMT

Advocates of Modern Monetary Theory (MMT) have claimed that most economists have recently come around to their position that a government that borrows in terms of its own currency cannot default on its debt.   It can simply create new money out of thin air and pay off its debt as it comes due.

This is hardly a novel idea. While a government could default on debt denominated in a currency that it can issue rather than create money to pay it off, the conventional wisdom among economists has always been that it is much more likely that a government that has no other way of paying debt as it comes due would likely issue money to pay it off.   That is the principle behind the fiscal theory of the price level.    

It is at least possible that section four of the 14th Amendment  to the U.S. Constitution would be interpreted as mandating the issue of legal tender currency in payment of debt as a last resort.   The provision was aimed at keeping southern representatives in Congress from blocking the payment of Civil War debt or trying to get Confederate debt paid with U.S. funds.   But the plain language of the first sentence, "The validity of the public debt of the United States, authorized by law, including debts incurred for payment of pensions and bounties for services in suppressing insurrection or rebellion, shall not be questioned,"  might be interpreted as prohibiting explicit default.  

The creation of new money to pay off part of the U.S. national debt would be illegal.   However, it would be legal for the Federal Reserve (the Fed) to purchase government bonds on the open market.   As long as the Fed purchases bonds from investors that the Treasury is selling to them to refinance the national debt, there should be no problem in selling enough new bonds to pay off bonds that are coming due.   The problem is that the Fed's legal mandate is to provide price stability and maximum sustainable employment.   This puts significant limits on the amount of government bonds that the Fed can purchase.   Of course, Congress could amend the Federal Reserve Act in any number of ways, including just requiring that the Fed purchase bonds directly from the Treasury if necessary to raise money to pay off existing bonds as they come due.   

If budget deficits are so high that the national debt grows faster than the economy, then eventually interest on the national debt will outstrip the total income generated by the economy and necessarily outstrip total amount of tax revenue that can be raised, even leaving no provision of current public goods or any personal income to allow for taxpayer survival.   Well before that point is reached, some kind of default is inevitable.   To avoid such a default, before the point at which the interest burden becomes unbearable, the budget deficit must be reduced so that the national debt grows no faster than the economy. Such a budget deficit is sustainable.

On a positive note, slowing the growth of government spending to less than the growth of the economy and keeping the tax system unchanged will result in a shrinking budget deficit and eventually make the burden of interest on the national debt proportionately less severe and further eventually lead  to a budget surplus that will finally result in there being no national debt at all.   

On a negative note, if investors believe that the budget deficit will never be reduced enough to become sustainable, then they may immediately refuse to refinance the existing national debt by purchasing new bonds to pay off the old bonds as they come due.   This is because when the inevitable default occurs at some future time, no investor wants to be caught holding government bonds.   Immediately before that, no investor will buy government bonds.   But if no investor buys government bonds, then the government cannot refinance its existing national debt and so the crisis occurs immediately.   The logic of this situation is that a crisis could occur at any time due to a loss in confidence.

It certainly will make some difference if the government expressly defaults on the national debt, simply paying limited principal and interest as opposed to issuing new money to pay them off as they come due.   The second option results in hyperinflation.   The money used to pay off the bonds will be worth much less.   However, this hardly matters.   If the government is going to create a hyperinflation to pay off its national debt at some future time, no bond investor wants to be caught with the bonds when their real value is greatly reduced.   This will make it impossible to refinance the national debt, requiring the hyperinflation in the nearer future and so on, such that it could happen at any time.

The primary difference between the two scenarios is that express default on the national debt would directly impact government bonds, while inflationary default would effect all contracts denominated in units of the government currency.   While this more catastrophic result might motivate a reduction in the budget deficit to sustainable levels, my own view is that explicit default of the national debt is superior to hyperinflationary disaster. 

The Virginia School approach would require a monetary constitution that mandates some sort of price level stability.  Inflationary default would simply be unconstitutional.   And, of course the better known Virginia School approach would be that constitutional rules should limit the budget deficit so that it is easily sustainable.   Basically, government spending should be funded by current taxation.    Still, I also favor running a slight budget surplus when the economy is growing on trend, shifting to balance when the economy grows less than trend and budget deficits when the economy shrinks.  I think raising taxes or cutting expenditures in a recession is undesirable.  In my view, a constitutional debt limit, with the limit set so that the interest on the national debt is easily sustainable, is probably the least bad approach.

Anyway, the notion that excessive budget deficits will eventually lead to hyperinflation (without some reform of the monetary and fiscal constitution) is my view.   I have not lost all hope that budget deficits will be reduced to become sustainable and I am not predicting imminent financial catastrophe. Still, that hyperinflation is a more likely scenario than express default is not something that I discovered after learning about MMT.   I would also note that this purported insight from MMT hardly provides assurance that proposing massive new government spending programs to be paid for by money creation is responsible.

Saturday, April 6, 2019

Modern Monetary Theory

I believe it was Robertson who claimed that what was true in Keynes was not new and what was new in Keynes was not true.   Much the same can be said of Modern Monetary Theory (MMT.)

It is almost certain that a monetary system can be developed where the government creates money to fund its spending and uses its tax system to generally absorb sufficient money so that the price level and inflation vary only a limited amount.   All that is necessary is that what is conventionally described as the budget deficit--government spending less tax revenues--be planned such that only a modest growth rate in the quantity of money occurs. Of course, tax revenues actually depend on economic activity and so the actual growth rate of the quantity of money would vary, but the result would be only modest variations in the growth path of the price level and so the inflation rate. Targeting the growth rate of the quantity of money, much less the growth path of the price level or inflation, by adjusting the tax system on the fly would be nearly impossible and adjusting government spending would also be difficult and likely politically impossible. 

However, the sale of interest bearing bonds to absorb money would allow for what amounts to a very conventional monetary policy.   Excessive monetary growth could be offset by what amounts to open market sales--the sale of what would be newly-issued government bonds under this monetary framework.   If it were necessary to contract the quantity of money to target the price level or inflation (or nominal GDP,) then this would simply require a sale of bonds of greater value than the current budget deficit.   Some advocates of MMT apparently seek to avoid bond finance.   This appears to be a matter of anti-capitalist ideology.   Interest bearing bonds allow investors to earn interest income and investment income entails exploitation of the working class.

None of this is new and simply reflects the "standard" view of the government budget constraint.   Government spending must be funded by either taxes, borrowing, or money creation.    Now, to the degree any advocate of MMT emphasized that government first spends newly created money and then absorbs it by taxes or bond sales, then it a bit silly.  In reality, the flow of government spending over time is matched by the flow of tax revenues, bond sales, and money creation.   Before or after is not really economically significant. 

The "Virginia School" approach to monetary theory almost entirely focuses on the need for a "monetary constitution" governing the creation of money.   However, there have always been hints of "old Chicago" due to James M. Buchanan's influence.    Fundamentally, government spending should be funded by taxes, so that voter/taxpayers can make a sensible decision regarding the allocation of resources between the production of private and public goods.   However, the Buchanan approach to monetary institutions implies that there is usually some budget deficit funded by money creation.   Some part of government spending is funded exactly as proposed by MMT.   While this might make up a substantial portion of  spending for a very frugal government, the usual assumption is that the appropriate rate of government funding by money creation is a small portion of government spending.

For many years, following the lead of "new Chicago," or monetarism, Virginia School thinking assumed a very predictable nominal budget deficit reflecting a fixed growth rate for the quantity of money.   However, the "old Chicago" approach assumed at least the possibility of a variable growth rate of the quantity of money.   These changes would be necessary to offset changes in velocity, (or accommodate changes in the demand to hold money) and so maintain a stable price level.   Interestingly, the Virginia School has little more interest in "bond finance" of budget deficits than MMT.   However, the problem isn't any purported exploitation of workers due to interest income but rather concern that the current generation of voter/taxpayers would be using debt finance to fund current benefits at the expense of future generations of voter/taxpayers.

Of course, the Virginia School advocates changes in public finance and monetary institutions.   MMT claims that not only should the government fund all of its expenditures by money creation, but it also insists that is what really happens. MMT claims that under the monetary and fiscal institutions of the United States today, government spending is funded by newly created money and when taxes are paid, money is destroyed. Their distaste for government bonds may sometimes leave it unsaid, but they also claim that the sale of bonds destroys money too.

There is more than an element of truth to their claim, but in the final analysis, government spending in the U.S. is funded by taxes and borrowing.   So what is the element of truth?    The U.S. Treasury has a deposit account at the Federal Reserve (Fed.)   When the Federal government makes payments, the funds are drawn from its account at the Fed.   Those receiving the payments generally deposit them at a bank (depository institution.)   Funds are transferred from the Treasury's deposit account at the Fed to the bank's depository account at the Fed.    It is apparent that this transaction simply transfers money from the Treasury's account to the bank's account.   However, the Fed does not count the Treasury's deposit account as part of reserves and it does count the deposit accounts the banks keep at the Fed, so whenever the Treasury spends money, "reserves" increase.

The opposite occurs when the Treasury receives receipts of either tax payments or bond sales.   The Treasury deposits the checks or electronic payments it receives in its deposit account at the Fed.   The Fed shifts funds from the deposit accounts of the banks used by the taxpayers or bond investors to the deposit account of the Treasury.   Because the deposit account of the Treasury is not counted,  "reserves" as reported by the Fed decrease.

Reserves makes up a portion of the monetary base, which also includes currency held by the nonbanking public.  The logic above suggests that when the government spends money, the monetary base increases and when taxes are paid or government bonds are sold, the monetary base decreases.   This argument supposedly shows that MMT isn't proposing anything new but rather is simply a recognition of what the U.S. government already does--fund its spending with newly created money.

If the Federal Reserve were to target reserves or the monetary base, then it would have to make offsetting transactions.   Each time the government spent money, the Federal Reserve would need to make an open market sale.   Each time the government received a tax payment or receipts from the sale of bonds, it would need to make an open market purchase.

However, this is really not the most useful way to think about it.   The government receives tax payments frequently and it receives proceeds from the sale of bonds more frequently still.   It also makes payments frequently.   These represent the flow of funds through the Treasury's account at the Fed.   When the Treasury's balance at the Fed changes due to a mismatch between its receipts and expenditures, the Fed's measured reserves change in the opposite direction.    The limit to the increase in the quantity of reserves reported by the Fed due to an increase in government spending not matched by tax revenue or bond sales is the Treasury's existing balance at the Fed.   In other words, unlike other entities, if the Treasury spends the money it holds, this shows up as an increase in the quantity of money reported by the Fed rather than a transfer of money balances.   Economically, the effect would be the same, the result being measured as an increase in the quantity of money rather than a reduction in the demand to hold money by the U.S. Treasury.

As explained above, if the Fed targeted the quantity of reserves or base money, and it continued to measure it without including the Treasury's balance at the Fed, then changes in that balance would result in changes in the quantity of base money.   The Fed would need to offset those changes by open market operations.   If the Treasury reduced its balance at the Fed, the Fed would need to sell securities on the open market.   If the Treasury increased its balance at the Fed, the Fed would need to purchase securities on the open market to leave the quantity of money unchanged. 

While the Fed could invest in private securities, if it follows its traditional policy (pre-2008, anyway,) then the Fed would be trading government bonds in place of the Treasury.   If the Treasury reduced its balance at the Fed and the quantity of money increased and the Fed sells government bonds to reverse that increase, then it is the Fed selling government bonds on the market rather than the Treasury.   Because the Fed is an independent agency of the Federal government, if the balance sheet of the Fed and the Treasury are consolidated, then government spending is being funded by bonds.

It seems to me that it is unnecessary for the Fed to control the quantity of reserves and base money so tightly.   If slight transitory fluctuations in the Treasury's balance at the Fed result in slight transitory variation in the quantity of reserves and base money, it would likely have little or no effect on spending on output, the price level, or inflation.   Interestingly, if the Fed targets the federal funds rate, as it has for many years, then if the changes in reserves due to changes in the Treasury's balance have some impact on interbank lending, then keeping the Federal Funds rate on target would require that the Fed make the offsetting open market operations described above.

When thinking about the flow of government spending, tax receipts, and the sale of government bonds, the impact of fluctuations in the quantity of reserves or base money due to variation in the Treasury's balance at the Fed are trivial.    More importantly, the laws of the United States set up a tax system that generates revenue and the government creates a budget based upon spending that tax money.   Any budget deficit must be funded by bond sales by the Treasury.   There is a limit on the total amount of bonds that the Treasury may sell, a limit that is periodically increased by Congress.    While Congress could change the laws, the fiscal institution existing today is that the Treasury must obtain funds from taxes or sale of bonds before it spends it.   However, it would hardly matter if it were not necessary to first accumulate money in the Treasury account as long as bond sales matched the budget deficit over the budget period.

Does the U.S. government fund any of its expenditure by money creation?    The conventional wisdom among economists is that it does.   The government generally does run a budget deficit, which it funds by selling bonds.   However, some of those bonds are purchased by the government's own central bank, with newly created money.   While it is illegal for the Fed to purchase government bonds directly from the U.S. Treasury, it hardly matters that the bonds must pass through some middleman.   The Treasury sells bonds on the market and the Fed purchases some of them on the market.   Since the Fed is an independent agency of the U.S. government, looking at their consolidated balance sheet shows that when the Fed owns U.S. government bonds, it is the government owing money to itself.   From this perspective, the U.S. government funds itself largely through tax receipts, but also through the sale of bonds to private investors (and other governments) and finally through money creation with a sort of legal fiction that that it is borrowing money from its own central bank.

While there is an important element of truth to this perspective and it reflects the Virginia School approach to the role of money in public finance, it is probably not wise to dismiss all the "legal fictions."   If the Fed was bound to a money supply rule, especially a rule that implied a constant growth rate of the quantity of money and if the government runs a substantial budget deficit all the time, then this approach may be the best way to make sense of the situation.   However, the Fed has never operated according to such a rule. In recent years, it has targeted the inflation rate and its legal mandate is to seek price stability along with maximum sustainable employment.  This mandate has generally resulted in a growing quantity of base money, but a policy of offsetting changes in velocity (or in other words, accommodating changes in the demand to hold money,) to stabilize prices implies that sometimes the quantity of money must decrease.   As a practical matter, while the Fed may typically accumulate more and more government bonds, it may sometime accumulate an unusually large amount to offset a transitory decrease in velocity.   But it would then need to be able to sell those bonds to decrease the quantity of money when velocity rises again to a more normal level.   

Of course, the Great Inflation occurred during a period in which the Fed had the same legal mandate as today.   And 2% inflation is hardly a literal interpretation of stable prices.  Further, there has been serious discussion of raising the target inflation rate without any consideration of changing the Fed's legal mandate.  However, consider a central bank that takes its legal mandate seriously--something more than a suggestion.   Then in a very real sense what is supposedly "fiat money" amounts to type of borrowing. 

This would especially apply to the Virginia School approach of a monetary constitution mandating a stable price level.  Under such a rule, a temporary decrease in velocity would require an increase in the quantity of money.   However, there would be a legal requirement that such money be later withdrawn from circulation when velocity recovers.  The best way to frame this situation is that the extra issue of money represents increased borrowing and when it is later withdrawn, it would be paid back.   If there were an existing national debt made up of outstanding government bonds, then a decrease in velocity would require open market purchases, replacing some of those bonds with money.   The best way to understand that is that interest-bearing debt that is not suitable for use as money is replaced by debt that can be used as money.   When velocity rises again, then the quantity of money must be reduced by open market sales.   But what that means is that interest bearing debt that is not suitable as money replaces debt that is suitable for use as money. 

Debt that can be used as money can be issued in some amount at a zero interest rate.   Interest is hardly ever paid on hand-to-hand currency.  Since "fiat money" is frequently identified with such paper notes, lack of a nominal interest return is sometimes taken to be a defining characteristic of money.   Of course, most money takes deposit form which can pay interest.   That includes the reserve deposits banks hold at the Fed.  They can, and since 2008 have, paid interest.

If the price index to be stabilized was simply the price of gold, then it would be obvious that this would be a gold standard.   The monetary constitution would require changes in the quantity of "fiat money" so that the price of gold is stable.   The issuer must stand ready to withdraw it from circulation if the demand to hold it falls so that its value relative to gold would remain fixed.   If the rule was enforced by requiring that paper money be redeemed with gold, then paper money would plainly be a type of debt instrument.

From this perspective, even with only a weak commitment to an inflation target, the money issued by the Fed can be seen as a type of debt.  Abandoning its inflation target would be a kind of default.   Currency bears no interest but reserves do.   The Fed has a balance sheet that includes both government debt and private debt, particularly mortgage backed securities.  The Fed earns more than than it pays and generates revenue.   It uses some of the revenue to cover its expenses, but transfers most of it to the U.S. Treasury.  From this perspective the U.S. government is not funding part of its deficit by monetary creation.   It is rather that one source of government revenue is profit from its central bank.

Suppose the U.S. balanced its budget and so spent no more than it collects in tax revenue.   The Fed would still earn profits and transfer them to the U.S. Treasury.   If the demand for base money increased, then the Fed could issue more.  It must do so to keep inflation on target.   The Fed would typically earn more profit on its larger balance sheet.   If the demand for base money were to fall, for example, due to financial innovation, the Fed must stand read to shrink its balance sheet.   It would pay part of the money it issued back by selling off some of its assets.  As a result, it would typically make less money for the Treasury.

So, there is an element of truth in the claims of MMT.   It is certainly possible to have a monetary institution where the government creates money when it spends it and destroys money when it collects taxes or sells bonds.   A relatively conventional monetary policy would be possible by selling bonds as needed to keep the quantity of money and the price level and inflation under control.   However, it is not the way the U.S. fiscal and monetary institutions operate now.   If the Fed is truly committed to keeping inflation on target, then it is probably best to see government spending as being funded by taxes or else borrowing.   In other words, the "legal fictions" requiring bond finance of budget deficits reflect a more fundamental reality.