A Few Questions for Exogenous Money Proponents

The Krugman/Keen controversy seems like as good a time as any to post my views on endogenous versus exogenous money.

Naturally, I support the endogenous money theory, as it has all the evidence behind it. Economists even seem to acknowledge that the standard textbook story is somewhat behind the times, although they then cling to the money multiplier model, choosing to add epicycles instead of abandoning their core theory (they have a habit of doing this).

As far as I’m concerned, the evidence is so overwhelmingly opposed to exogenous money the burden of proof is on them. So here are a couple of questions:

(1) Why do expansions of the broader measures of money generally precede rather than succeed expansions of the base? Surely in the money multiplier model banks would require reserves before expanding lending? Evidence:

There is no evidence that either the monetary base or M1 leads the cycle, although some economists still believe this monetary myth. Both the monetary base and M1 series are generally procyclical and, if anything, the monetary base lags the cycle slightly.

(2) A corollary of (1): why do fiat expansions not necessarily result in M1+ expansions? For example 2008:

and similar results can be found during the onset of the Great Depression:

(3) Surely, if banks can create deposits with loans (which they simply can in the real world), by definition a loan adds new purchasing power and increases nominal demand? So by extension, the level of private debt in the economy is crucial to understanding it? How else do you explain the robust empirical link between private debt growth and indicators such as unemployment, housing prices and GDP?

Attempts to answer this framework should also avoid the standard circular tactic of assuming the money multiplier, then describing everything in terms of the money multiplier. But even when doing that, I just don’t see how you can square all the evidence with the exogenous money model. Prove me wrong.


, , , ,

  1. #1 by dkuehn on March 30, 2012 - 12:58 pm

    So I’ve always thought of the money multiplier as an upper bound on how the money supply could expand – certainly not a determinant of the money supply. Reserve requirements and the demand for reserves are two quite different things.

    I think you can say “demand for reserves swings independently of M1 movements, so money supply is endogenously determined” and still think that given the existing structure of demand for reserves, manipulation of M1 can still impact the money supply relative to the counterfactual. I would not call that “exogenous money” – I would call that a recognition that exogenous policy still impacts a system of endogenous money creation.

    • #2 by Unlearningecon on March 30, 2012 - 2:51 pm

      I’m not sure of the nature of the evidence I’d require for this, but I’m pretty sure that the banks are not constrained by reserves – they simply lend out what they want and find the reserves somehow, often via the CB accommodating this demand. The CB could theoretically choose not to, but then the economy would grind to a halt.

      As such, I’m not sure of the ‘demand and supply for reserves’ approach, at least in the short term. I appreciate the CB has impact over the long term.

      Banks are, of course, limited by regulation, their own risk perceptions and capital requirements.

  2. #3 by JW Mason on March 30, 2012 - 2:29 pm

    if banks can create deposits with loans (which they simply can in the real world), by definition a loan adds new purchasing power and increases nominal demand?

    You are right about endogenous money, but this doesn’t follow. It is perfectly logical to believe that money is fully endogenous, but also that expenditure is fully determined by current income (plus expectations, exogenous interest rates, etc.) with banks simply accommodating desired expenditure.

    Or in other words, when you say “adds new” and “increases”, you mean relative to what baseline?

    • #4 by Unlearningecon on March 30, 2012 - 2:59 pm

      Relative to the existing money supply, that is.

      Are you saying that if private debt is demand driven, we need to look at what drives demand instead of private debt?

  3. #5 by Blue Aurora on March 30, 2012 - 5:03 pm

    Isn’t the money supply in practice, partially endogenous and partially exogenous?

    Out of curiosity Unlearningecon, do you consider yourself more of a Kaleckian or a Kaldorian in terms of economic growth?

    • #6 by Unlearningecon on March 30, 2012 - 10:52 pm

      Yes, this is true. But I would argue it’s ‘endogenous with exogenous constraints’ rather than ‘exogenous with endogenous constraints’ – I feel neoclassicals argue the former.

      I’m not well versed in Kaldor’s work yet, but I do plan to be as I’m aware he was the most devastating critic of Friedman’s monetarism, predicting its failure years in advance.

      As for Kalecki, I favour his emphasis on investment and treatment of some marxist concepts. There’s also the obvious observation that a lot of ‘Keynesian’ work could be considered Kaleckian.

      • #7 by Blue Aurora on March 31, 2012 - 6:08 am

        I see. Will you read more of Kaldor’s work on growth theory, though? He seems to have anticipated Romer’s work on endogenous growth a few decades earlier. Here’s a book on Kaldorian macrodynamics that you might want to read. It’s by a Professor of Economics that I met at Trinity College.


      • #8 by Unlearningecon on April 2, 2012 - 2:33 pm

        Thanks for that.

  4. #9 by Frances Coppola on March 30, 2012 - 7:45 pm

    I think there is a causation problem here. Which comes first – the demand for debt, or the availability of debt? Does the availability of cheap credit make people more willing to borrow to spend? If it does, then debt must drive nominal demand and Keen is correct. Alternatively, does the demand for debt drive its availability? Do banks simply respond to existing customer demand for loans, or do they actively entice customers to borrow more by cutting rates and reducing standards?

    In neoclassical economics banks simply respond to customer demand for loans, and demand can be controlled by raising the cost – hence the emphasis on interest rates as a means of controlling the growth of spending. The assumption is that people will forego spending if the cost of borrowing rises too much, but that at lower interest rates people will still only borrow to spend on things they would have bought anyway.

    I think the evidence does not support the neoclassical view of banks as passive intermediaries. There is no doubt in my mind that banks actively inflated the credit bubble through aggressive lending policies, and that customers did borrow far more than they would have done if banks had not behaved like this.

    • #10 by Unlearningecon on March 30, 2012 - 11:00 pm

      Under a Keynesian framework, the level of consumption is roughly stable, and the rate of interest is a large factor in determining the rate of investment – the lower it is, the more there will be, and the more it will be spent on sustainable investments. The rate of interest is set by the Central Bank.

      This also requires regulations to limit consumer lending, capital requirements and capital controls. But it appears to be a good way to control the both growth of private debt and what that private debt is used for.

  5. #11 by Will on March 30, 2012 - 8:02 pm

    “choosing to add epicycles instead of abandoning their core theory”

    I know that this is common usage, but it’s a historically inaccurate metaphor. The story that Ptolemeic astronomers kept adding new epicycles is a Whig-history myth. The epicycles were one of the original elements of Ptolemy’s system in the Almagest, and the modifications that subsequent Ptolemaic astronomers made did not add any new ones.


    Sorry, I know I’m being pedantic, this just happens to be a bee that I have in my bonnet.

    • #12 by Unlearningecon on March 30, 2012 - 10:55 pm

      I don’t consider it pedantic, I consider it incredibly interesting – thanks!

      The internet has taught me that things most people consider to be true, generally aren’t.

  6. #13 by Louis-Philippe Rochon on March 31, 2012 - 2:09 am

    Your post was quite interesting and a validation of much of what post-Keynesians, like Basil Moore, Marc Lavoie, Nicholas Kaldor, John Smithin and myself, have been arguing for many years. We would claim, however, that money is fully endogenous and there are no exogenous constraints, except the banks’ own willingness not to lend, based on their expectations, in an uncertain world, of the future levels of aggregate demand. but as long as banks judge borrowers are creditworthy, there is no reason not to supply credit.

    • #14 by Unlearningecon on March 31, 2012 - 5:39 pm

      Surely the base rate counts as an exogenous constraint?

  7. #15 by JMRJ on March 31, 2012 - 1:52 pm

    I think before you start debating “exogenous v. endogenous money” you would be well advised to figure out what “money” is in the first place. You can chase your tail all you like, of course. Just don’t expect to arrive at any new destinations.

    • #16 by Unlearningecon on March 31, 2012 - 5:40 pm

      Economists have various measures of money, some of which I alluded to in this post. See here.

      • #17 by JMRJ on March 31, 2012 - 5:47 pm

        “Measuring” money and defining it are not the same thing.

      • #18 by Unlearningecon on April 1, 2012 - 9:31 pm

        Measuring it is sufficient for the purposes of this post, though.

  8. #19 by Unlearningecon on April 1, 2012 - 8:20 pm

    Louis you obviously don’t like to be disagreed with, but I’m not particularly interested in random abusive comments or splitting hairs over terminology. Also if you want unsubscribe do it yourself, I can’t do it for you.

  9. #20 by AFG on April 8, 2012 - 1:39 am

    I tend to agree with what I think is you’re view (I like to say, money is endogenously created, but exogenous factors can shift the Marginal Revenue and Marginal Cost curves from which banks choose how much money to create), but I don’t think you’re questions are particular difficult to answer from the classical view.

    1. Multiple Answers
    a. Expectations – Banks increase lending because they predict the Fed will increase reserves
    b. Short term v. Long Term – They create money in the short term, but in the long term, their lending will converge to some multiple of reserve levels

    2. Once again expectations, people don’t believe that the expansion in reserves will be permanent (Fed won’t accept temporarily higher inflation to get back to the right path). Krugman has written some fairly famous papers on this point. See Here: http://web.mit.edu/krugman/www/trioshrt.html

    Others like Scott Sumner would agree with this view, but also contest the empirics of this. Even small changes like QE have produced a large response, because markets predict that the small changes signal larger changes to come. They would say that looking at measures like M1 doesn’t capure the full impact, and look towards things correlated better with NGDP (like stock market prices).

    3. There are too many possibly ways to answer this one different views. I think the general theme would be that private debt might matter, but it depends what type of debt, how risky, how it is distributed, etc. There is a correlation between private debt and those indicators, because all things being equal, it CAN cause them, but doesn’t in all circumstances.

    Some traditional conservatives or those of the Austrian persuasion would say that it only worth looking at when government policies cause it to deviate from the equilibrium rate (Lower interest rates move consumption forward and Fannie/Fred encourage mortgage debt).

    Traditional liberals would agree that it matters only when it deviates from the equilibrium rate, but explain it with compensation, behavioral explanations that make finance inherently unstable or sectoral imbalances (induced by Chinese monetary policy, for example).

    Finally, those that believe in EMH (like Sumner) would not disagree that private debt can matter. But no one can usefully identify when and why it does. Therefore, we shouldn’t waste our time with it.

    PS – I will also say that what I said what I think your view is (stated above) is not all that different than people like Krugman, Rowe, or Sumner. I, for the life of me, cannot figure out why MMTers and the rest of the endogenous money crowd spends so much time distinguishing “operational constraints” and incentives.

    No one really thinks anymore that the Fed precisely determines the money supply at every moment in time. It affects incentives for economic activity (then it’s a definitional question of what M1 or M whatever is “money supply”, with the Fed having varying degrees of control over each one), and different schools disagree about the transmission mechanism, political constraints, and effectiveness.

    I also don’t think your crowd recognizes that, in theory, as long as the money supply/inflation/NGDP is not *Perfectly* inelastic with respect the reserves or some other thing the Fed controls and there is at least partial Ricardian equivalence (people react to expectations), then monetary policy is theoretically very powerful. They could set any inflation target and then increase reserves by 4 quadrillion percent until the economy responded. As long as the Fed has no constraint on how far it can go and credibly promises to go as far as necessary to reach its target, then it can theoretically hit any target.

  1. An FAQ for Mainstream/Neoclassical Economists « Unlearning Economics