Krugman on (or maybe off) Keen

Flattr this!

Paul Krugman has just commented (twice) on my most recent blog about my paper for INET. In one sense, I’m delighted. The Neoclassical Establishment (yes Paul, you’re part of the Establishment) has ignored non-Neoclassical researchers like me for decades, so it’s good to see engagement rather than wilful (or more probably blind) ignorance of alternative approaches.

Click here for this post in PDF

Figure 1: Krugman’s first piece

There is a bizarre asymmetry in economics: critics of Neoclassical economics like myself read Neoclassical literature avidly, no because we agree with it—far from it—but because we feel obliged to understand why they hold to their counterfactual views on the economy.

Most Neoclassical economists, on the other hand, don’t even bother to consider critics within their own ranks—let alone critics from without. So to have a paper referred to is definitely a plus.

In another sense, I’m appalled, because Krugman’s comments put on display that very ignorance of Neoclassical literature—let alone of alternative economic thought.

For instance, Paul refers to many of the propositions in my blog (it’s clear that he hadn’t read the paper on which it is based) as “assertions about what is crucial, without much explanation of why these things are crucial.”

One of these “assertions” is the key role of the change in debt—rather than saving out of current income—in financing investment.

Well Paul, in that paper you will find references to the extensive theoretical and empirical literature from which that assertion was derived. I could start with non-Neoclassical authors like Schumpeter, but let’s lead with someone from within The Citadel (as Alan Kirman once called the Neoclassical orthodoxy: Alan Kirman, 1989, p. 126): Eugene Fama. The “assertion” that the change in debt was the main source of funding for investment was confirmed by Fama and French in a pair of empirical papers:

The source of financing most correlated with investment is long term debt. The correlation between I and dLTD is 0.79… These correlations confirm the impression … that debt plays a key role in accommodating year-by-year variation in investment.” (Eugene F. Fama and Kenneth R. French, 1999, p. 1954)

“Debt seems to be the residual variable in financing decisions. Investment increases debt, and higher earnings tend to reduce debt.” (in an unpublished draft of the same paper).

Or consider Alan Holmes’s crucial paper in 1969, in which he fought an unsuccessful campaign against the later experiment in Monetarism (far from being a “strict Monetarist”, as Paul jibes at one point, I and my Post-Keynesian colleagues and forebears take money seriously while simultaneously being trenchant critics of Friedman’s simplistic Monetarism—see for example Nicholas Kaldor, 1982). Holmes, then Senior Vice-President of the New York Federal Reserve, noted that the key Monetarist policy prescription of regulating the economy by “a regular injection of reserves” was based on “a naïve assumption” about the nature of the money creation process:

The idea of a regular injection of reserves—in some approaches at least—also suffers from a naïve assumption that the banking system only expands loans after the System (or market factors) have put reserves in the banking system. In the real world, banks extend credit, creating deposits in the process, and look for the reserves later. (Alan R. Holmes, 1969, p. 73)

Holmes would turn in his grave at Krugman’s naïve assertion, half a century later, that banks need deposits before they can lend:

If I decide to cut back on my spending and stash the funds in a bank, which lends them out to someone else, this doesn’t have to represent a net increase in demand. (Paul Krugman, 2012)

As Randy Wray observed, that is “the description of a loan shark, not a bank”—or of a hypothetical world in which banks need deposits before they can lend. In the real world, as Holmes points out above, bank lending creates deposits. That’s why banks matter in macroeconomics, and it’s not “Banking Mysticism” to point this out: it is “Banking Armchair Theorism” to ignore them in macroeconomics.

Neoclassical economists have ignored this point for decades, which is why you have to look to the non-Neoclassical literature to truly understand money creation and the crucial role of banks. Schumpeter put it clearly during the last Depression: he described the view that Krugman puts today, that investment (which is what the most important class of borrowers do) is financed by savings, as “not obviously absurd”, but clearly secondary to the main way that investment was financed, by the “creation of purchasing power by banks … out of nothing“. This is not “Banking Mysticism”: this is double-entry bookkeeping:

Even though the conventional answer to our question is not obviously absurd, yet there is another method of obtaining money for this purpose, which … does not presuppose the existence of accumulated results of previous development, and hence may be considered as the only one which is available in strict logic. This method of obtaining money is the creation of purchasing power by banks… It is always a question, not of transforming purchasing power which already exists in someone’s possession, but of the creation of new purchasing power out of nothing. (Joseph Alois Schumpeter, 1934, p. 73)

Figure 2: Krugman’s second piece

Why does it matter that “once you include banks, lending increases the money supply”? Simply, because the endogenous increase in the stock of money caused by the banking sector creating new money is a far larger determinant of changes in aggregate demand than changes in the velocity of an unchanging stock of money. And in reverse, the reduction in demand caused by borrowers repaying debt rather than spending is the cause of the downturn we are now in—and of the Great Depression too.

Figure 3 shows the ratios of private and public debt to GDP in America from 1920 till now. Non-neoclassical economists like myself, Michael Hudson, Ann Pettifor, the late Wynne Godley, Randy Wray and many others (see Dirk J Bezemer, 2009, and Edward Fullbrook, 2010 for fuller lists of those who warned of this crisis before it happened–including of course Nouriel Roubini, Dean Baker, Robert Shiller, and Peter Schiff) were shouting that the post-1993 explosion in private debt was unsustainable, and would necessarily lead to a crisis when its rate of growth slowed (let alone turned negative), for years before the crisis began (my first academic warning of the dangers of rising private debt is shown as SK1, and my first public warning that a crisis was imminent is shown as SK2 on Figure 3). We were ignored, in large part because only Neoclassical economists like Krugman, Bernanke and Greenspan had the ear of the public and politicians.

Now the crisis is the defining economic event of our times, and years after it began, the only period to which the recent boom and bust in the private debt to GDP ratio can be compared is the Great Depression.

Figure 3: Aggregate Private and Public Debt

Yet Neoclassical economists like Krugman continue to assert that the aggregate level of private debt, and changes in that level, are macroeconomically irrelevant, when even casual empiricism implies that changes in the aggregate level of private debt are associated with Depressions.

So while I welcome any Neoclassical economist at the forthcoming INET conference taking up Krugman’s call (“I hope someone in Berlin presses Keen on all this”), in reality Paul, empirically oriented non-Neoclassical economists like myself are the ones challenging the unsupported assertions of Neoclassical economics—not the other way round.

Bezemer, Dirk J. 2009. “”No One Saw This Coming”: Understanding Financial Crisis through Accounting Models,” Groningen, The Netherlands: Faculty of Economics University of Groningen,

Fama, Eugene F. and Kenneth R. French. 1999. “The Corporate Cost of Capital and the Return on Corporate Investment.” Journal of Finance, 54(6), 1939-67.

Fullbrook, Edward. 2010. “Keen, Roubini and Baker Win Revere Award for Economics,” E. Fullbrook, Real World Economics Review Blog. New York: Real World Economics Review,

Holmes, Alan R. 1969. “Operational Constraints on the Stabilization of Money Supply Growth,” F. E. Morris, Controlling Monetary Aggregates. Nantucket Island: The Federal Reserve Bank of Boston, 65-77.

Kaldor, Nicholas. 1982. The Scourge of Monetarism. Oxford: Oxford University Press.

Kirman, Alan. 1989. “The Intrinsic Limits of Modern Economic Theory: The Emperor Has No Clothes.” Economic Journal, 99(395), 126-39.

Krugman, Paul. 2012. “Minsky and Methodology (Wonkish),” The Conscience of a Liberal. New York: New York Times,

Schumpeter, Joseph Alois. 1934. The Theory of Economic Development : An Inquiry into Profits, Capital, Credit, Interest and the Business Cycle. Cambridge, Massachusetts: Harvard University Press.


About Steve Keen

I am Professor of Economics and Head of Economics, History and Politics at Kingston University London, and a long time critic of conventional economic thought. As well as attacking mainstream thought in Debunking Economics, I am also developing an alternative dynamic approach to economic modelling. The key issue I am tackling here is the prospect for a debt-deflation on the back of the enormous private debts accumulated globally, and our very low rate of inflation.
Bookmark the permalink.

174 Responses to Krugman on (or maybe off) Keen

  1. RJ says:

    “The true backing of money is the faith or trust people place in it.”

    Sorry but this is complete nonsense. Nonsense that for some reason you want hold onto.

    (Would you deposit money with someone or a bank that held no assets to back your deposit – the banks liability- whatsoever.)

  2. Dannyb2b says:

    “They have value because if you don’t pay your taxes with them you go to jail, and the state can enforce that. Believing the ‘medium of exchange’ myth helps oil the works, but it is not necessary and often violated. You can be forced to work for whatever the state determines in our state system.”

    Note and coins have value in part because people trust the these to be a stable unit of account, store of value and medium of exchange. Yes you have to pay taxes in this currency which creates a demand for it. Agreed. Still no need to create money as debt though.

    “Money is always the liability of a third party as a matter of accounting – and therefore technically a debt. Trying to suggest that it can be anything else means that you haven’t understood the accounting.”

    That is only how we choose to account for money in our current unstable system. We can improve upon this and treat money similar to a normal asset because it would represent a stabilizing benefit.

    “The problem is not who issues the liabilities. It is the mechanism that controls the quantity of lending that matters. Centralised quantity management is too rigid for a dynamic economy.”

    Both who issues money and the mechanism are important. You can centralize decision making which increases accountability and at the same you can make money enter the system in a balanced decentralized manner.

  3. NeilW says:

    “We can improve upon this and treat money similar to a normal asset ”

    Even a ‘normal asset’ has a liability. That is the accounting.

  4. alainton says:

    I posted on the latest NYT thread

    Paul I think many of us can see who is the true banking fantasist here – and the empirical data proves it. Your view of saving not only belongs to the old school loanable funds days but even back to the mid 19th century wages fund days where banks as seen as being limited by the wages of their depositors.

    If it were the case the bank were limited by deliveries from the ‘armoured car’ of the fed then how come there is no correlation between physical money base and M3? Why doesn’t the fed control inflation by limiting printing of physical money? Of course if it did the ATMs would won dry in a couple of days and we would have a revolution. These days the causation runs the other way, monetary base of is set by issues of regulation and financial instability so that funds deposited with the fed are driven by the endogenous growth of M3.
    Your ideas are full of stock flow confusions. It is the flow of future deposits that limit the rate of future expansions of of the power to lend not the initial wages or profits stock. Go back and read some Taussig.

    Steve Keen being a central figure on the emerging Stock Flow Consistency school doesn’t make that mistake. Indeed he is working on a model which integrates horizontal and vertical money and has interbank interaction – initially to prove that the MMT’r were only 80% right but im sure will now have a much bigger and much more easily disproven target.

    Build your own model of financial intermediation and we will make our own minds up about empirical correlation. History will judge.

  5. Dannyb2b says:

    “Even a ‘normal asset’ has a liability. That is the accounting.”

    What liability does an asset like my bicycle have?

  6. NeilW says:

    “Put simply armoured cars from the Fed control the money supply!! Has anybody really argues a cash constraint since the Gold Standard went, or even in a 100 years.”

    No, and for the simple reason that for there to be a constraint there has to be a mechanism for enforcing that constraint.

    There is no law without enforcement.

    What Krugman misses is that reserves and cash manage the *payment* system, not the lending system. And therefore if the central bank tries to enforce a constraint there, the payment system starts to break down.

    If you have a cash constraint then people can’t get cash, or ATMs start to run out. And you would likely cause panic in that situation – as the level of cash is at or near the minimum required to clear the cash payment system anyway. So it is the same problem as reserves – if you restrict those reserves (a quantity restriction) you have to expect your interest rate to blow out and for one or other of the banks to go bust due to cashflow issues.

    And we know that none of the central banks would allow that during the GFC and that there has been no regulation or resolution procedure put in place to deal with banks that can’t clear their payment system.

    So the central bank is controlled on the payment system by the simply political fact that the public will not allow them to enforce their constraints on the payment system even if they wanted to.

    The only enforcement the central bank can do is on capital ratios, and even then they are reluctant to put a bank into administration for rule violations.

    Banks cheat, because it is in their interests to do so – as is holding as much of the economy to hostage as they can get away with.

  7. NeilW says:

    “What liability does an asset like my bicycle have?”

    A valuation ‘reserve’ in the unit of account.

    That’s why bank reserves are called bank reserves even though from the bank’s point of view they are assets. What they really mean is ‘central bank reserve account’.

  8. Dannyb2b says:

    “A valuation ‘reserve’ in the unit of account.

    Just in case the value of my bike changes in dollars?

  9. NeilW says:

    “Just in case the value of my bike changes in dollars?”

    Yep, which it will because like most assets it depreciates.

    Eventually it is useless and you move the asset to the liability account to eliminate it.

    Think matter and anti-matter – equal and opposite that annihilate each other when brought together.

  10. Dannyb2b says:

    “Just in case the value of my bike changes in dollars?”
    Yep, which it will because like most assets it depreciates.

    Then its not relevant to count money as a liability because it always has the same nominal value. Unlike a bike which might depreciate in monetary terms. Right?

  11. alainton says:

    I should say that the evidence is that Krugman did read Steve’s INET paper if only at a skim.

    There is no evidence that he reads this blog, but he might have skimmed through the Berlin conference papers after pulling out and Steves would have jumped out as it was written as a repost to Krugman and Eggerston. That the blog post is an attack on this paper is less obvious.

    Krougmans original blog reads as hey I cant be there, don’t have time to read up and riposte, so someone do it for me to maintain my reputation.

  12. centerline says:

    “Banks cheat, because it is in their interests to do so – as is holding as much of the economy to hostage as they can get away with.”

    This was the basis of my proposition a few days ago (thereabouts) in questioning the structure of central banks – and the significant potential for a conflict of interest therein. Without addressing a clear and socially responsible mechanism/regulation for the the creation of money we are unfortunately going to see any given system corrupted.

    I know that I am off on a different tangent than the modeling and theory you guys regularly flog so well. But, I have to insist how important it is to recognize what lies outside the equations that is influencing them. The compliment to this of course is more accurate modeling of the how system functions, reducing the potential for exploitation.

    I have a suspicion that work like Steve’s is going to gain much more press in the coming years.

  13. alainton says:

    Steves paper – page 5 (Harvard numbering please 🙂

    ‘[bank lending] increases the aggregate amount of money in circulation, increasing aggregate demand in the process—and predominantly financing investment or speculation rather than consumption’

    Of course if the rising price of an asset is used to leverage a further loan then it does finance consumption as people are spending money they assume they will get in the future from rising asset prices – the classic example being second mortgages before a property crash. The increased consumption has a temporary income effect on those in the consumer goods and personal service sectors – as well as the above wealth effect – who then try and ‘get in on the act’ of diverting their increased savings to asset speculation.

  14. Dannyb2b says:


    In other words every asset in our system of accounting has a liability in the sense thats its valuation may change. But that doesnt make it liability. Therefore there is no reason to issue the currency as debt either. If people that hold cash wish to adjust for its change in purchasing power in their accounts because the value of their money may change thats fine. But when money is created its not useful to recognize it as debt.

  15. alainton says:

    And a follow up to the above – it is not the consumers who are behaving irrationally – (although historically unwisely) – contra Bernanke.

    Rather it is the misevaluation of risk by banks in valuing the counterparty to the loan – such as property assets. If we have a steady ratio of purchases to sales then the valuations are low risk, but the uncertain possibility of a property crash makes this less and less likely. Of course if everyone tries to sell the same type of asset at the same time to pay off a loan then the market crashes. Now banks seems to have priced in this uncertainty as a ‘risk’ of a further property price downturn, but it becomes self fulfilling as it restricts expansion of the money supply and restricts expansion of the asset (housebuilding) setting up the potential for a further house price boom. Each bank behaving ‘rationally’ increases volatility. There can be no rational expectations when asset prices are radically uncertain.

    There is a solution to this, insurance, underwriting an asset price floor. And states can enforce this through providing such insurance or requiring it through regulation. However this insurance can itself become a source of financial instability if it is used an asset to secure further loans. This is the main reason why the European CDR market is like a pack of cards. Ultimately the only way a firewall can be created is through a state backed debt write off with monetary expansion being used to secure the balance sheets of banks that would otherwise go to the wall. Observers will note that despite violent protestations to the opposite that is just what the eurozone has been forced to do in the last year – and the one to watch at the moment is a Dutch bank looking likely to go belly up from a green enforced haircut. Of course so far state monetary measures have been used as bank welfare rather than to write off householder debt, and steves conclusion is called ‘cranky’!

  16. RJ says:

    Double entry book keeping means that asset = liabilities + SHF

    So a bike (an asset) as part of total assets WITHIN ONE COMPANY will be supported by total liabilities + SHF’s

    This is different to a FINANCIAL ASSET that is always backed by a financial liability HELD BY ANOTHER PARTY.

  17. Dannyb2b says:

    “This is different to a FINANCIAL ASSET that is always backed by a financial liability HELD BY ANOTHER PARTY.”

    A financial asset such as note or coin does not need to be backed by a financial liability to any party. There is no need for this primal idea it is counterproductive to issue money as debt. Its just how the system is structured at present but this needs to change.

  18. Steve Hummel says:

    So make an accounting of the national assets. Then, because the individuals of the nation are the inheritors not only of their cultural heritage of productive capacity, but are also the individuals who do the work which enables the REAL credit that results, you’ll have a nation of 330 million capitalists. Distribute the dividend and compensate the businesses who discount their prices to consumers and money will become what it most essentially is, a ticket for the distribution of goods and services. And of course those who work and create things will have their additional remuneration, and Banks will no longer have their monopoly on credit, in all likelihood their “tail wagging the dog” control of the government, their ability to extort the individual by their usurpation of his/her REAL credit and democracy will have finally actually come to economics.

  19. NeilW says:

    “Then its not relevant to count money as a liability because it always has the same nominal value. Unlike a bike which might depreciate in monetary terms. Right?”


    Every asset has an equal and opposite liability in accounting.

    You want to call something an asset, then there is a liability somewhere in the same accounting system to balance it.

    All the time, every time. Without fail.

  20. NeilW says:

    “A financial asset such as note or coin does not need to be backed by a financial liability to any party. ”

    Yes it does. The accounting demands it. When a central bank note is issued the central bank note reserve account goes up by the same amount.

    When that note is banked at a private bank, then the note goes back to the central bank to be swapped for ‘bank reserves’. Then the central bank pays interest on those reserves to control monetary policy.

    Notes and coins are just like Government Bonds to a private bank.

  21. Steve Hummel says:

    Money, other than for one’s legitimate inheritance, needs to be issued for production and then extinguished as a result of consumption. Any speculative financial outlay that cannot be bound back to actual production of SOMETHING OTHER THAN SIMPLY MORE MONEY needs to be either VERY tightly regulated or declared illegal. No more Roving Cavaliers of Credit.

  22. RJ says:


    Debt backs

    -Money and it also back
    -Our savings

    So money (a financial asset)= debt (a financial liability)

    but we need money to save. This money can then be converted to Govt bonds. So

    bonds (a financial asset) needed for savings (as we age) = debt (a financial liability)

    This is why Govt debt is so important. Without it it is almost impossible for everyone to save for their retirement. Because the burden of debt for the non govt sector becomes too large. Whereas for Govt it is almost totally irrelevant esp at current low levels.

    But until people take the time to understand double entry book keeping this will not be understood.

  23. Steve Hummel says:


    Okay, but why not create an entirely separate accounting and distributing entity which is firewalled off from the government by all due separation of powers etc. and consequently less manipulable by the financial and producing powers that be? Then mandate it to distribute the monies for the dividend and the retail business compensation as well as to the Banks, all at 0%, and with the proviso that they only lend it within very strict regulated guidelines. Any already existing monies can be lent as Bank and borrowers see fit. We’ll call it The National Credit Office.

  24. RJ says:


    The problem with many (including monetary reform sites) who want to change to another banking system is

    They no not understand the current system
    They do not know what money is and moneys relationship to debt
    They do not understand double entry book keeping or financial accounting
    When pressed on exactly how the new system would work (including the JEs) they hide behind meaningless statements

    But just one point. Interest is an expense to the Govt but REVENUE to the non Govt sector.

    So 0% interest would smash the non Govt sector. Interest expense is meaningless to the Govt (its just a journal entry) but critically important to savers and pension funds.

  25. alainton says:

    Demockracy on the NYT blog

    As Will Rogers used to say: “A remark generally hurts in proportion to its truth.”

    After all, Keen is, in effect, saying what Krugman has believed, taught, and written about is fundamentally wrong. He’s saying Krugman has believed, in effect, in the geocentric solar system. …
    The real question is whether a Nobel Laureate / Princeton Professor / respected pundit can say something like “Holy Mackeral! All that stuff I believed was not accurate.”

Leave a Reply