Rebuttal to Bob Murphy’s Critique of Modern Money Theory

I find it very laudable that heterodox economists are debating this topic so carefully and always with reference to facts. Here are my thoughts on Bob Murphy’s post The Upside-Down World of MMT:

In it he writes:

According to many proponents of MMT, “deficits don’t matter” when a sovereign government can issue its own fiat currency, and all the hand wringing over the government’s solvency is absurd.

Until I see what statement or MMT hypothesis specifically he is referencing here I have to call this a strawman argument. I don’t think any MMTer I’ve read has ever said “deficits don’t matter”. According to MMT, deficits are needed to provide the private sector with sufficient claims upon the public sector (= net private saving), but if deficits go beyond net private saving demanded by the private sector they can cause demand pull inflation, which most argue is indeed undesirable.

In fact, the MMTers claim that given the reality of a US trade deficit, a sharp drop in the government’s budget deficit would hamper the private sector’s ability to save. Thus, the Austrians are unwittingly calling for a collapse in private saving when they foolishly demand government austerity.

To be precise, MMT doesn’t say the private sector’s ability to save is hampered, but its ability to NET save. The reason net private saving is important is not arbitrary, it is because we observe that drops in net private saving tend to cause large indebtedness of the private sector to itself (household debt to income ratio rises), culminating in a depression or a severe recession. In fact, 7 of the 7 times the US federal government ran a budget surplus were followed by a depression (6 times) or a severe recession (1 time).

Bob Murphy then spends some time accurately figuring out that in this context MMTers are not talking about all private sector saving, but only net private saving, not exactly a new revelation, but it’s good to see he went through the mechanics of figuring it out.

Now Nick Rowe and the MMTers are certainly correct when they observe that “private saving net of private investment” can’t grow without a government budget deficit (again if we disregard foreign trade). But so what? The whole benefit of private saving is that it allows for more private investment.

He’s missing the benefit of a net private savings buffer that makes actors in the private sector more comfortable to spend and invest.

He also indirectly touches upon a common mistake made by orthodox economics and also Austrian economists: They think that banks need people to deposit savings before the bank makes loans allowing borrowers to invest. This is empirically not the case, and it turns the causality of saving vs investing upside down. Banks in a fiat money system don’t wait for people to deposit money with them before they can make loans. They make loans, which create deposits. Those bank checking deposits created out of loans are the savings generated in the private sector after a loan has been made. Now, in order for the loan to result in actual investment spending, the borrower needs to use his checking account deposit to purchase machinery, computers, build a factory, etc.

Banks make loans regardless of how much reserves they have on deposit. They obtain the reserves needed as per reserve requirement after the loan has been made, with a couple of weeks of time lag, and the central bank then accommodates the private banks’ demand for such reserves at all times, if needed. If this wasn’t the case the interbank lending rate would move outside of the Fed’s declared policy range which doesn’t happen in a floating fiat money system.

It should be pointed out that any savings generated by private bank lending is always matched by a corresponding private bank loan. This is why private sector savings generated out of private bank loans do not change the private sector’s net position against the public sector. Insufficient claims upon the public sector and excessive private debt in turn are precisely the economic conditions MMT warns about.

The fans of MMT should therefore stop pointing to those identities as if they prove the futility of government austerity during an economic downturn. Those tautologies, and the cherished equations of the three sectors, are consistent with post-Keynesian and tea party economics.

I don’t think any MMTer has ever claimed that those accounting identities prove anything. They are a helpful starter, but Murphy doesn’t even begin to talk about the private sector’s empirical demand for the net savings buffer I’ve explained above.

As a final way to illustrate the non sequitur of the equations involving government budget deficits, note that we could do the same thing with, say, Google. Go back through all the equations above, and redefine G to mean “total spending by Google.” Then C would be “total consumption spending by the-world-except-Google,” and so on.

After doing this, we would be able to prove — with mathematical certainty — that unless Google were willing to go deeper into debt next year, the world-except-Google would be unable to accumulate net financial assets, in the way MMTers define that term. The proper response to this (perfectly valid) observation is, Who cares?’

This is an invalid comparison because Google is not the issuer of the currency and doesn’t have the power to unilaterally impose arbitrary tax debts upon others. This is the whole point of the sectoral balance identity: we define the different actors in the fiat money economy based on their relation to that fiat money: the issuer of the money (public sector), the user of the money (private sector), and issuers/users of other fiat monies (foreign sector). Murphy thinks these definitions are arbitrary, when they are not at all.

Finally, in the section “Not All Spending and Income Are Created Equal” Murphy explains to us that purely looking at numbers and GDP figures doesn’t tell us much about the work actually performed and the value actually created inside the economy. This is true, and fortunately nobody competent in MMT has ever made any claim to the contrary as far as I’m aware.

MMT doesn’t claim that government spending can create infinite wealth, just that if you start a monetary economy by imposing a tax on people, there are reactions on the part of the taxed people, and those have to be taken into consideration in all your pertinent models and theories.

I have made the case before that for the past 5000 years we’re aware of ALL monetary economies have been started in that manner. Bitcoin may be a very recent exception to that rule, and I’ve written about how Bitcoin imitates the fiat money system in some ways.

Austrian economists by and large ignore this causality because they believe in the barter theory of money, where you can indeed hallucinate into existence an economy that functions without taxation and government spending, such as the one Murphy elaborates on in this article, where he neatly imagines how Robinson Crusoe could easily save up coconuts without any government deficit spending. This model is helpful if you’re ever stranded on a lonely island with coconut trees. It isn’t if you’re trying to understand the workings of a complex monetary production economy.

I have been guilty of this myself so I’m not speaking from a pulpit by any means. I’m just pointing out where empirical evidence has helped me correct my thinking on some aspects of fiat money economics.

By the way, here’s where you can read my full post about Modern Monetary Theory.

Related Posts:

Podcast: What They Don’t Tell You About The Fiat Money System

I had a great chat with Stefan Molyneux at his call in show, starting at 1m 50s:

We discussed the origins of the monetary system and its implications on government deficits, private sector saving, debt, and many other things.

Check out Stefan’s amazing one of a kind show, which I’ve been following for over 7 years at this point, at

Related Posts:

Tweetstorm: Important & Largely Unknowns Facts About Fiscal Policy

This tweet storm by MMT economist Stephanie Kelton nicely sums up empirical facts about fiscal policy in a sovereign floating fiat money system that conventional wisdom tends to be nescient or ignorant about:

Related Posts:

Modern Monetary Theory

With permission from Being Libertarian, I am reposting an important exclusive piece I wrote that lays out some fundamental changes in how I’ve come to understand the functioning of a fiat money economy, and its economic implications on debt, deficits, and other hugely important concepts:

For quite a while now I’ve been pretty unsatisfied with mainstream as well as Austrian economics-based takes on the global economic situation, in particular phenomena such as record low to negative interest rates in countries with record debts (such as Japan), massive excess reserves, and QE 1 through infinity without much consumer price inflation, etc. No economic school I had learned about offered fully coherent answers regarding those.

So I thought it may be worthwhile to throw another heterodox economic school of thought commonly labeled “Modern Money Theory” (MMT) into the mix.

In particular, I’ve been reading L. Randall Wray’s book Modern Money Theory: A Primer on Macroeconomics for Sovereign Monetary Systems which I highly recommend. It’s also available for free in online only format.

I believe that a lot of what is being revealed in this book is of relevance to all of us, regardless where we stand politically. MMT offers prescriptions that can work within a libertarian context just as much as they can in the context of a more interventionist state. Most of all: MMT offers a plain and unbiased description of how fiat money systems actually work.

In this post I will provide a rough overview.

Spending and Taxation in a Sovereign Money System

MMT primarily describes a monetary system in which a country’s currency cannot be redeemed against anything else, such as a commodity or some other foreign currency, and in which the currency exchange rate floats freely on the foreign exchange market. Those are characteristics that apply to today’s US dollar for example. (MMT then also helps understand redeemable and pegged currencies, but first we need to understand the workings of an unconstrained currency system to get there.)

In such a monetary system, the government imposes a tax on the population it governs, that is, it claims the right to forcefully appropriate peaceful and innocent individuals’ property on a recurring basis.

It decrees, for example, that everyone shall owe an income tax in US dollars by year’s end. It then proceeds to create and spend those dollars in the form of paper bills and coins with numbers on them (or in the form of electronic bank entries called bank reserves that are convertible into bills and coins on demand, but more on that later) representing their tax redemption value in order to obtain products and services. It moves resources from the private sector into the so-called public sector.

Individuals who get paid in those dollars will then find ready acceptance of this money in the private sector, given that most individuals will need it to pay the imposed tax. Money prices and voluntary exchange in the private sector emerge as a result.

On a side note: There is indeed lots of historical evidence suggesting that money has been issued in this manner for millennia. In earlier days some of the tokens issued and accepted in taxes and fines for punishment were precious or base metal coins with the sovereign’s head minted on them (which would circulate at a significant premium on the metal value), or in some regions so-called tally sticks were used as money tokens by the king to obtain goods and services. Anyone possessing such a tally stick could return it to the crown in place of tithes which others would have to settle in real produce. The book I mentioned earlier also delves into commodity money (such as gold and silver coins) in that historical context, including redeemable token currency systems like the gold standard.

The important point that MMT makes here is this: Taxes drive money. The sovereign’s arbitrarily-imposed tax liability has to be discharged using the money tokens issued by that same sovereign. Government spending is a means of introducing money into the economy. Taxes then destroy that money. Spending and taxation are thus not fiscal, but rather monetary policy. The government doesn’t first raise taxes in order to obtain money to spend, it rather creates the money first by spending it into existence and then removes it from the economy via taxation.

The Central Banking System

The government’s treasury has a checking account bank, the so called central bank. In the US this is the Federal Reserve Bank (Fed). The US Treasury’s balance in this bank account is replenished to $5 billion dollars at the end of every day, but is generally not counted as part of the country’s money supply. The reasons behind this setup are of purely legal nature, but ultimately the processes of spending and taxing to create and destroy money respectively work exactly as outlined above.

(In order to understand this system better, initially it really helps to view the central bank and the treasury as one and the same institution: the government. Even if the legal nature and partial ownership structure of certain central banks, such as the US Fed, may lead one to believe that it operates “independent” from government, the end result is for the most part the same. For details I highly recommend the chapters 3.6 and 3.7 in the primer I mentioned above.)

Then there are the country’s private banks where people can hand in their money tokens for storage and safekeeping and write checks against their balances for convenience’s sake. As per the rules of the system, these banks also get to maintain their own checking accounts at the Fed. The money in these checking accounts is commonly referred to as “bank reserves”.

Say I present a $100 bill that was introduced via government spending into the economy to such a bank’s teller. The bank takes the bill, marks up my bank deposit by $100, hands the bill to the Fed, and the Fed then marks up the bank’s checking account at the Fed (its “bank reserves”) by $100 as well.

If I want to withdraw $25 in cash and the bank doesn’t have the cash on hand, it needs to ask the Fed for $25 in cash. The Fed will deliver the cash and mark down the bank’s reserves by $25. The bank hands me the $25 and marks down my bank deposit account by $25.

As you can see, cash and bank reserves are interchangeable. Together they form what is called “high powered money”. They are the only means of payment that the treasury will accept in tax payments.

Let’s say I owe $75 in taxes to the government and make payment via check. The treasury will ask the Fed to mark down my bank’s reserves by $75 and the bank will mark down my demand deposit by that same amount. The money has disappeared from circulation. I have been made poorer in comparison to the sovereign currency issuer but the money didn’t really “go anywhere” other than the treasury’s account at the Fed which gets replenished anyway, no matter what taxes get collected or not. (By the way: If you were to withdraw your money and make tax payment in old paper bills, the money would simply get shredded!)

Or let’s say I bank at bank A and I make a check payment of $75 to Bob who has a checking account at bank B. Bob presents my check to bank B, bank B presents it to the Fed, and the Fed will mark down bank A’s reserves by $75 and mark up bank B’s reserves by that same amount. Bank A marks down my deposit account and bank B marks up Bob’s deposit account in turn.

So you can see that banks need bank reserves in order to (a) facilitate tax payments, (b) meet interbank settlement requirements, and (c) facilitate cash withdrawals. If a bank doesn’t have the necessary reserves it’ll incur an overdraft on its reserve account at the Fed and needs to obtain the reserves after the fact to cover that deficit. This is what’s commonly referred to as the Fed’s “lender of last resort” function.

The bank obtains such reserves by offering savings accounts and CDs to individuals looking to earn interest instead of holding cash. Furthermore it can also borrow reserves for a short period from other banks who don’t need them at the moment on the interbank lending market. Finally, it can also borrow reserves from the Fed directly, and generally only against safe collateral, such as treasury securities, or sell those securities to the Fed outright.

Government Bonds and Interest Rates

Cash currency in circulation and the aforementioned bank reserves at the Fed generally pay low to no interest. Thus banks like to move some of those reserves into interest bearing securities. Private individuals have that same option. This is why the government sells bonds of varying maturity. Ultimately government bonds are nothing but savings accounts or CDs at the Fed. All that happens when a bank or an individual buys a government bond is that bank reserves are debited and a securities account is credited with said treasury security.

When an interest payment on a bond is made the Fed marks up the bond holding entity’s bank reserves electronically. When the bond matures, the bond is removed from the securities account, and reserves are credited with the bond’s face value amount via keystrokes. There is no tax money that the Treasury needs to raise anywhere to make this interest or principal payment.

Since the private banks can always move their reserves into bonds of any maturity and since there is no default risk, the interest paid on those bonds constitutes a bottom level for interest rates. There would be no point in loaning out money to private borrowers at lower rates. Thus the Treasury and the Fed can manipulate the overall term structure of private loans by setting a floor below which rates won’t go. This is most commonly the case for very short term instruments, such as Treasury Bills (1 year or less), which the Fed sells to private banks to move bank reserves away from the private banks if the overnight interest rate on the aforementioned interbank lending market drops below the desired policy rate.

It should be noted that a few years ago the Fed started paying 0.25% on reserves, effectively making the sale of short term credit instruments obsolete. But the economic effect is ultimately the same: The interbank lending rate won’t drop below this level.

Why manipulate these interest rates in the context of this overall framework you may ask? That’s a good question, and indeed, MMT suggests that in a sovereign money system there’s really no reason to do so. If the Fed stayed out of it, the short term interest rate on the interbank lending market would probably tend towards zero, since banks will always want to get rid of their excess reserves to maximize interest earnings.

Why sell longer term government bonds like the Treasury does, effectively setting a risk free rate and thus a floor for longer term loans? Again a good question! In fact, MMT ultimately suggests that beyond very short term Treasury Bills at most there’s really no reason for the government to be floating long term bonds.

Within the confines of today’s fiat money system, MMT actually offers the most libertarian alternatives regarding interest rate management and government bonds: let the overnight rate go wherever market conditions amongst private banks let it go, and don’t issue any long term government debt at all!

Private Bank Lending

As outlined above, private banks make it more convenient for people to store and use their money. When I deposit cash in my bank account, then the bank essentially writes me an IOU, saying that it owes me cash on demand, in the form of a checking account. The government, which can always create new bank reserves via spending, guarantees those IOUs via deposit insurance in case the bank runs out of reserves due to mismanagement, which is why they are widely accepted and generally considered as good as cash or “liquid”. It is thus not a coincidence that checking accounts are generally considered part of the money supply. You can say that private banks are in reality public/private partnerships where the bank’s deposit liabilities are publicly guaranteed and the rest (liabilities to bondholders, shareholders, etc.) is private. This explains why when a bank goes under its stock price tends towards zero (the shareholders lose), yet deposit holders are made whole by government’s deposit insurance.

But a much larger function of banks is lending. Lending occurs when a bank identifies a creditworthy borrower who is offering his rather illiquid IOU (in the form of the promise of future payments) to the bank, while the bank in turn offers its own very liquid IOU (a government insured promise to convert into cash on demand) in exchange. In return for offering a more liquid IOU the bank charges interest. The differential between the interest it pays to depositors and other creditors and the interest it charges on its loans determines its profit.

In a very simplified yet relevant model of an economy with only one bank, you can think of a bank extending a loan to an entrepreneur, marking up his checking account by say $100,000. The entrepreneur then uses that money to pay sellers of labor (employees) and of commodities (vendors), marking down his own checking account while marking up the others’. (You can see here by the way how for every loan, there is a matching bank deposit. Not because the bank had to obtain deposits first, but rather because every loan generates a matching bank deposit.) He then produces widgets that he sells to those same people, so that ultimately his bank checking account is marked up again while the buyers’ checking accounts are marked down. He can now repay the bank loan which ends the loan cycle and withdraws the previously issued money from circulation. We can see here that much like taxes do for high powered money, bank loans create a drain on and thus a demand for checking account money.

However, if the entrepreneur is unable to sell the widgets (while the recipients of his payments hold on to their checking account money) he defaults on his loan. The bank needs to write it off and lose out on expected interest revenues, yet it still owes cash to the deposit holders on demand. If some deposit holders were to demand cash (or write checks if we add more banks to the model) then the bank would have to sell assets to the Fed to obtain the reserves needed. If it has no more assets to sell it would have to obtain a reserve loan from the Fed that it needs to repay at some point or face insolvency. So even if a bank can create new checking account money into existence without any bank reserves at all, it is only interested in extending loans to creditworthy borrowers or else it may face insolvency. This model can be easily extended by adding more banks and borrowers with the same net flow of funds over time.

We draw two important conclusions: Banks don’t wait to obtain bank reserves and then lend those reserves out later. They rather identify creditworthy borrowers first, create loans, and then obtain bank reserves from the Fed later if and when they’re needed. Furthermore, in spite of their ability to create money out of thin air, banks have no incentive to create an indefinite amount of loans. Rather loans are made when willing and able individuals stand ready to make successful investments. If loans go sour, the bank is on the hook (as it should be).


To summarize, the most important things I took away from this exercise are:

  • In a free floating irredeemable fiat currency system the government cannot be forced to default on its debts (it could of course choose to suspend interest and/or principal payments, but that would be a policy choice, not one of financial necessity)
  • In such a fiat money system the government creates money by spending and destroys it by taxing
  • Taxes drive money
  • Government bonds are basically just like savings accounts at the central bank
  • Private banks lend first, then obtain reserves
  • Fractional reserve lending is naturally constrained by the creditworthiness of borrowers

In my opinion MMT helps us explain and understand the current fiat money system much better than conventional or Austrian economics. This is coming from an Anarcho-Capitalist who has read books like Mises’ Human Action, Theory of Money and Credit, Socialism, and many others, so don’t think I say this lightly!

Even if we find moral or economic flaws in a system, it is better for us to understand how it actually “works” right now in order to make policy recommendations down the road. I believe MMT offers a lot of insight on that front and I will provide some contemporary and relevant examples in subsequent posts.

Related Posts: