The fallacies of MMT

Recently there’s been some spread in the influence of “Modern Monetary Theory”, a set of beliefs about monetary and fiscal policy championed by figures like Jamie Galbraith and Randall Wray. It’s been given (unfavorable) attention by Paul Krugman, lengthy blog posts by Steve Waldman and Nick Rowe, ample article space at new deal 2.0 and even a platform at the New York Times. Though MMT is certainly nonsense, I’m hesitant to spend much time refuting it—since it has roughly the same stature among economists that creationism commands among biologists, there’s a sense that I’m just tilting at windmills. Nevertheless, beliefs that are ridiculed by specialists can become powerful (and dangerous) outside the mainstream, and I think it’s best to confront them head-on.

MMT starts with lengthy historical musing about the origins of money. The claim is that we can’t think about “money” independent of the special status granted to a particular form of currency by the government. Specifically, the power to levy taxes in dollars is what gives dollars value in the first place.

All this may be true. But it’s also mostly irrelevant for the conduct of monetary policy. Perhaps government, using its power of taxation, nudged us into the equilibrium where green pieces of paper are used both as a medium of exchange and a unit of account. But that doesn’t mean that taxes, at the margin, have anything to do with the demand for currency.

In fact, this is easy to refute. Until the advent of QE, the vast majority of base money was held in the form of paper currency. The role of electronic bank reserves was incredibly marginal, fluctuating around $10 billion out of a monetary base of over $900 billion. Yet these electronic bank reserves were used to conduct the vast bulk of transactions in our economy. As James Hamilton points out, circa 2008 each dollar of electronic reserves was whizzing around on Fedwire an average of 350 times every day. Before the advent of interest on reserves, excess reserves were virtually nonexistent: banks were profit-maximizers, and there was no need to hold even a cent more than the 10% reserve requirement on checking accounts. The only question was whether to hold reserves in the form of “vault cash” (mainly in ATMs) or electronic accounts at the Fed.

In short: pre-2008 demand for base money—the asset that the government creates—consisted overwhelmingly of demand for cash.

But certainly no more than a trivial fraction of taxes are paid in cash. Employers don’t FedEx their withholdings to the IRS in envelopes filled with $100 bills; corporations don’t send a fleet of armored trucks when their bill comes due. Even Grandma probably sends a check. (In fact, if you’ll forgive my ignorance, I’m not positive that we’re even allowed to pay the feds in cash.)

Yet there has to be some reason why individuals are holding cash: it pays zero interest, which in normal times is well below even the safest alternative assets. And certainly there’s a demand for cash to conduct petty transactions: it’s how I pay at food trucks and Chinese restaurants too stingy to accept Visa. But this isn’t enough to justify the vast amount of cash in circulation, which at nearly $1 trillion amounts to over $3000 for every man, woman, and child in America. (That’s a lot of food trucks!)

In reality, of course, much of this cash is held abroad—estimates derived from formal cash shipments hover around 50%, but given the vast sums smuggled out of the country under questionable circumstances, the actual fraction is probably far higher. Cash is useful for tax evasion, illegal transactions, and storing value in a country without a reliable banking system. You’ll notice, however, that none of these explanations have anything to do with the MMT theory of money, in which the government creates demand for money by levying taxes (except insofar as taxes create demand for tax evasion!).

And even if (against all evidence) you accept the MMT view of money, it doesn’t become fiscally relevant. Suppose that you’re trying to patch a $20 trillion hole in the long-term federal budget. You raise taxes by $300 billion (in real terms) a year. At a real interest rate of 2%, this works out to $15 trillion. You’re most of the way there! Now consider the effect of your tax increase on demand for base money. As I’ve stressed, it’s abundantly clear that there is no significant effect, but let’s indulge MMT and make the ludicrous assumption that money demand will increase one-for-one with the rise in annual tax revenue. At a nominal interest rate of 4%, your annual seignorage income increases by $12 billion; the present value of the long-term increase is $600 billion. Not much compared to $15 trillion you raised from taxes, or the $20 trillion you owe—you’re still deep in the hole.

In fact, $600 billion is exactly 4% of $15 trillion. Even in the utterly implausible case where annual tax increases lead to a commensurate rise in the demand for money, the extra revenue from seignorage is just a fraction of the revenue you raised from taxes in the first place—this fraction being the nominal interest rate.

To sum up: MMT is wrong about money. Even supposing that it’s right, its impact is fiscally marginal. The government does have a budget constraint.

About these ads

112 Comments

Filed under macro

112 responses to “The fallacies of MMT

  1. casualreader

    Hey Matt,

    Nice to see you writing again. I’m not sure if I’m entirely right, but I think the MMTer use taxation as a way to monitor money supply – as in, they view increasing taxation as removing money supply, and decreasing it as increasing the money in the economy.

    For them, the story of the origins of money was just an illustration that the USD has value because we are paying our taxes in USD (how does this have to do with cash and the spiel about the that?). I think their question is: if the government states that taxes in the United States were now paid in another currency, say, the english pound, will the USD still have value? They don’t necessary state that taxes create the demand for money. Instead, they state that USD has a value above zero because of this ability to tax.

    Second, I don’t think MMTers believe that you would need to patch a long term federal budget, per se. To them, that is a non-issue, so you might be arguing past them in that example you make. The way they think about the deficit is this: in a currency issuing country, why does the deficit matter at all? The only way that it would matter is that it would creates inflation, but at that point 1) MMTers would argue that they would not opt for a budget deficit that large (only enough to reach some “optimal” level), and 2) if there was inflation, then they would manipulate money supply to stem this inflation.

    Cheers,

    Jack

  2. Hey Jack,

    Thanks for the comment. You say:

    “They don’t necessary state that taxes create the demand for money. Instead, they state that USD has a value above zero because of this ability to tax.”

    I think this would definitely be the right way of looking at the issue, and I’m rather agnostic about it. It’s hard to think about this kind of counterfactual, but if the government declared that it would stop accepting base money for payment of taxes, I can imagine the currency either completely losing value, or retaining it as the Fed maintained a commitment to a stable price level. (The problem is that the counterfactual world where the government stopped accepting dollars is such a strange world that it’s hard to “hold everything else equal” for the sake of an intellectual exercise.) Maybe MMT simply amounts to the claim that the former would happen. But then I don’t see how it gets off making claims about how the government can monetize the debt: it’s a long, long stretch from “the dollar wouldn’t have any value if the government didn’t accept it” to “the government can print enough dollars to cover the budget deficit and nothing too bad will happen”.

    I also agree with your characterization of MMTers’ belief about budget deficits, but I still don’t understand their position that they could avoid inflation. I addressed this more in my previous post. There are only two ways that the government can use monetary policy to ease its fiscal problems: seignorage and inflation. The problem is that the amount that can feasibly be raised by seignorage is very, very small compared to the federal budget–so as a matter of accounting, you’re left with inflation as the only route.

    Maybe MMT simply denies this arithmetic. (After all, the MMT is qualitatively coherent—it’s just that it gets the relevant magnitudes way off.) If so, the debate should be very easy to resolve. But I’m not quite so optimistic..

  3. casualreader

    Hey Matt,

    Thanks for the reply.

    As you stated, MMT simply denies this arithmetic. I think the way you characterized the US in your previous post looks at the US as a household instead of as a sovereign currency issuer.

    Think about the USD as a claim on US assets (be it a coke or real estate, whatever). What happens when you run a federal deficit is basically that you add claims on US assets (it’s kind of like debt is US equity, but not exactly). Well, to maintain a healthy economy, they need a specific level of private or public spending buying the assets that the economy produces. During a recession, you don’t have that spending.

    To MMT, debt is just increasing these claims on US assets. Is that a problem in and of itself? In a fiat currency state, it only seems to be a problem if everyone exercises those claims, thus increasing inflation. However, it seems everyone only wants to hold onto claims and not exercise them and buy assets.

    So, how do we buy assets to get out of the recession? MMTers favor running the deficit. They want to raise the number of claims on US assets to a point where people are exercising those claims (spending) at the “optimal” level. It doesn’t matter to them, per se, whether the deficit or the surplus stays at some level, increases, or decreases, or is huge, so long as the optimal level of spending is achieved. To them, “debt” and “surplus” are just numbers used to maintain this optimal level (note all of this is only true for a fiat currency state), and say nothing about the fiscal health of anything. They would not advocate even trying to cover the budget deficit, since in all probability it would change the “optimal level” achieved earlier. The only way they would run a surplus is if the “optimal” spending is reached, the government shouldn’t spend (note this is not to “cover” the deficit). In this sense MMT is more Keynesian than current mainstream economics.

    Note that in normal times when the “optimal” level of spending is achieved, if the government needed to take a proportion of total output by spending, they WOULD need to raise taxes in order to spend and take that output while avoiding inflation.

    Per your first point, I also agree that “ability to tax” is not a satisfying assumption. I think, however, they do all of this set up to say something else – it forms the basis for a fiat currency. If a government cannot tax in its native currency, then it cannot uphold its fiat currency, and since MMT is a theory on fiat currency, they must hold up that assumption.

  4. Thanks for the reply.

    I have mixed feelings about the Old Keynesian view that deficits are stimulative, which I’ll try to explore in a future post. I think it’s probably right in some ways, but not to the extent that Keynesian economists once believed—nor does it work through the same mechanisms.

    Narrowing in on the claims that distinguish MMT from standard monetary economics (or Keynesianism), I think that the critical paragraph is this:

    “To MMT, debt is just increasing these claims on US assets. Is that a problem in and of itself? In a fiat currency state, it only seems to be a problem if everyone exercises those claims, thus increasing inflation. However, it seems everyone only wants to hold onto claims and not exercise them and buy assets.”

    I think that this is a very apt description of how sovereign debt can work in a developed country—since it has benefits as a safe, liquid store of value, investors will be inclined to roll it over in perpetuity, and there may never be a day when everyone actually exercises their claims on the government.

    But as long as the government issues both bonds and money, the same logic doesn’t apply to money. In most ways, bonds and money are equivalent assets—they’re both ultimately claims on the government. The only differences are that (1) bonds pay interest and (2) money has additional liquidity properties that bonds lack. In equilibrium, these two have to precisely cancel out: the nominal interest rate on bonds must be such that the marginal investor, adjusting his portfolio between bonds and money, is exactly indifferent between the liquidity benefits from money and the higher returns from bonds. This constraint is impossible to escape, and thus the government can only choose nominal interest rates and the supply of base money jointly—there’s a one-to-one correspondence between the two, meaning that you can’t choose them both separately. (And this describes how monetary policy works in practice: when the Fed sets a target for the interest rate, it’s forced to engage in whatever open market operations are necessary to support that target. Like any monopolist, it can’t choose the price and quantity simultaneously—it can only select one, and let the market determine the other.)

    So what happens if the government decides to issue far more of its debt in the form of money, rather than bonds? After a point, the liquidity benefits of money are quite minimal, so that the cost of holding money relative to bonds (the nominal interest rate) has to be very low for the marginal investor to be indifferent between the two. We could safely say that interest rates would fall to zero.

    So the government would be paying nearly zero interest on its obligations. But this doesn’t mean that the real cost of financing the debt would decrease—after all, the real interest rate on government debt is pinned down by investors’ desire to transfer wealth over time. The real interest rate is almost always positive. In the long term, this policy of mostly monetized debt and zero nominal interest rates would simply mean sustained deflation and more or less the same real interest rates as before. In fact, the government would be slightly worse off from a fiscal perspective, because at zero interest rates it wouldn’t be making anything from seignorage.

    In a sense, therefore, I agree with MMT: it is possible for the government to issue most or even all of its debt in the form of money. But since at the margin, the liquidity benefits of such vast sums of money would be nil, this necessitates a nominal interest rate close to zero. And then the government would just be paying “interest” in a different way: rather than explicitly paying out to investors, it would see the real value of its obligations increase through deflation. Maybe those investors wouldn’t ever try to redeem their obligations en masse—but you could say exactly the same thing about debt issued in the form T-Bills. Nothing special happens when you monetize, except that you get long-term deflation.

  5. casualreader

    Okay, so given the 0% interest rates and the constant government spending that results in the buying up of output (and resulting in the creation of such money), doesn’t that at some point create inflation? Also, given that this continues, aren’t people going to start spending (there are more choices than just money, or bonds. People can spend it on real goods)? And as they do, won’t we reach optimal aggregate demand? At this point, this would create less of a necessity for the government to intervene.

    Could you expound further on your argument? Also, it would be really great if you did a post on old keynesianism!

    • casualreader

      I get the logic of the argument that leaves you at the 0%, but why is that the endgame?

      • It’s not necessarily the endgame. Whether or not 0% interest rates cause deflation or inflation is a very interesting and subtle question (this earned Minneapolis Fed president and former Minnesota professor Narayana Kocherlakota a lengthy online ribbing back in 2010.) I’d summarize it like this: if you’re going to have 0% interest rates into the indefinite future (which is necessary for holding much more debt in the form of money), mathematically the only possibility is a low, steady rate of deflation. Whether or not this equilibrium is actually implementable is another matter, and in reality we might just see hyperinflation because the Fed couldn’t convince anyone of its long-term commitment to deflation.

        Normally, of course, we associate low interest rates with inflation, and for good reason: conditional on following a long-term Taylor rule (one that pins down expectations in the distant future), a period of lower-than-expected interest rates will produce an inflationary boom. This is what we normally see in practice. The difference is that 0% interest rates are inflationary if everyone eventually expects to return to a policy regime with 4%, but a policy regime where 0% is the consistent interest rate target is actually deflationary. And since no one expects the Fed to implement long-term deflation, the former is what ends up actually happening. (These conclusions are partly bound up in the New Keynesian Phillips Curve, which I think is an insightful but flawed specification. A more realistic specification would probably have some backward-looking elements, in which case the question of whether 0% interest rates are inflationary would depend on both lagged and expected policy.)

        Now onto your question:

        “Also, given that this continues, aren’t people going to start spending (there are more choices than just money, or bonds. People can spend it on real goods)? And as they do, won’t we reach optimal aggregate demand? At this point, this would create less of a necessity for the government to intervene.”

        I partly agree, but I think this greatly limits the scope of MMT, much more than its advocates claim. In the government funds its stabilization policy by creating money, then (in normal times) we’ll see interest rates fall, inflation rise, and demand increase. Effectively, this would be wrapping fiscal and monetary policy into one (though, since it only takes a small amount of money creation to dramatically affect interest rates, monetary policy would be doing the vast majority of the work here—a fiscally irrelevant stimulus might still push the fed funds rate down from 5% to 0%). With enough stimulus, unless we were in a liquidity trap we’d reach optimal demand and no longer need to spend, like you say.

        But this doesn’t mean that the feds can finance any deficit using money—only the very small deficit that might arise from fiscal stimulus doubling as an open market operation. This strikes me as much less ambitious than the claims of most MMT economists, which usually involve saying that the government’s control of the money supply allows it to run pretty much any deficit it wants. Moreover, even in the special case of recessionary stimulus, the feds can’t permanently finance their spending by carrying it as money; after all, the resulting low interest rates are good during a recession, but not a sustainable long term policy. Eventually, unless they wanted permanently low interest rates (in which case the only long-term equilibrium is deflation, expectations of which would be contractionary during the recession itself), they’d have to move the debt out of money and back into bonds. Temporarily creating money has very little impact on the long-term budget constraint.

  6. I appreciate your attempt at an honest critique of MMT. There’s too much hot air on both the for and against sides; a more measured approach from either angle is appreciated.

    That said, your use of tax payment in reserve balances to critique MMT’s views of tax-driven money are off base, or at least a misinterpretation. Reserve balances ARE the “money” that settles taxes within MMT’s understanding of the modern monetary system. I explained all this back in 2004 within the context of (at that time) innovations in the payments system that were reducing the demand for reserve balances, FYI http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1723591

    Best,
    Scott

    • Hi Scott,

      Thanks for the comment. I’ve read through most of your paper, and I actually agree with almost all of it, though I don’t further infer that it’s always possible to finance deficits through money creation.

      I am in absolute agreement that the “puzzle” Friedman identified—how the Federal Reserve manages to have such a large effect on the economy with such a tiny change in reserve balances—is not really a puzzle at all. In fact, it reflects a somewhat embarrassing failure on Friedman’s part to understand why monetary policy matters: as I’ll write in an upcoming post, quantity equation monetary economics implicitly assumes that money matters directly because of its value as a medium of exchange (in which case there is a puzzle), while the modern view is that monetary policy matters due to money’s role as a unit of account (in which case the actual market for money can be arbitrarily tiny relative to the size of prices denominated in it). Add in short-term inelasticity of reserve demand, and the Fed’s day-to-day operations become easy to understand.

      I am not quite sure that tax liabilities alone maintain enough demand for reserve balances that the Fed can still effectively adjust the Fed Funds rate, but this isn’t a question that really keeps me up at night. In the worst case, we can simply change the rules such that banks have to hold (say) 2% of their 10% reserve requirement in the form of electronic reserves, and that will ensure consistent enough demand that the Fed will be able to do its job. (And if banks invent some substitute for a checking account to get around that rule, we can broaden the reserve requirement as needed.)

      My beef with MMT is that there isn’t enough demand for reserves to be fiscally relevant. Before interest on reserves, electronic reserves were at roughly $10 billion—this incorporated the demand from the need to pay tax liabilities and from other transactions in the economy. Debt held by the public, meanwhile, is approaching $10 trillion—literally a thousand times larger. How can we hold any sizable percentage of debt in the form of reserve balances without pushing interest rates down to zero (or whatever rate we’re paying on reserves)? I just don’t see the demand. This is the key question: how are you going to convince banks to hold even $1 trillion in reserve balances when they already only need $10 billion to conduct all their transactions? The disparity is enormous.

      And that’s why I focused on cash: since it’s at $1 trillion rather than $10 billion, it’s a form of holding debt that actually has nontrivial fiscal implications. But as we both acknowledge, cash is not used to pay taxes, so that this is peripheral to the MMT perspective on money.

      • Hi Matt

        You’re still misinterpreting a bit. There’s no requirement that the govt not issue bonds in MMT, though it is a policy proposal of several MMT’ers (not me, BTW). In fact, the govt CAN’T avoid issuing bonds or paying interest on reserve balances if the Fed wants a positive interest rate target. It’s not operationally possible (thus, the “no bonds” proposal of several MMT’ers assumes interest payment on reserves or a zero interest rate target–preferably the latter).

        Best,
        Scott

  7. All right, I suppose then that I don’t disagree with “MMT” per se, just with the very common claims deriving from it that the government can finance a substantial part of its debt through money. Indeed, it’s not operationally possible for the government to finance much of its debt through reserve balances unless it pays IOR (in which case it’s not really gaining anything) or it pushes nominal interest rates down to zero.

    My gripe, then, is that keeping nominal interest rates at zero isn’t fiscally beneficial in the long run at all. Unless you think that monetary policy can permanently affect the real rate of interest at which investors are willing to lend the government money, the real interest rate will still be positive, and zero nominal interest rates are only consistent in the long run with mild deflation. In this case, government debt is still costly—just that now, it’s costly because the real value of debt repayments is growing, not because the government is explicitly paying interest. There’s no fiscal benefit—and in fact, there’s a small cost, since at 0% the government earns nothing from seignorage.

    Since you don’t share this policy proposal, I suppose you’re the wrong person to ask, but I simply don’t see why holding nominal interest rates at zero forever would be fiscally beneficial.

  8. Try reading ‘The 7 Deadly Innocent Frauds of Economic Policy’ at

    http://www.moslereconomics.com/?p=8662/

    to get a full, non technical read on how it all works, and what our options are thanks.

    Warren Mosler
    http://www.moslereconomics.com
    MMT, First Generation

  9. Warren, I have looked at your list of 7 frauds and have the following comments:

    On Fraud 1: I agree that the government cannot formally default on nominal debts if it is willing to inflate away those debts. I have discussed at length, however, why this is not a practically relevant option; it has most of the drawbacks of a simple default and many additional disadvantages. Moreover, although the government can use inflation to eliminate existing debts, it can’t continue this game indefinitely: it ends when most of the current debts have matured, which (given an average maturity of 5 years) is very soon.

    On Fraud 2: You say that “Collectively, in real terms, there is no such burden possible. Debt or no debt, our children get to consume whatever they can produce.” I agree that claims about “borrowing from the future” tend to be overblown, but there are still several problems with this statement. First of all, in an open economy it isn’t true that “our children get to consume whatever they can produce”–you may need net exports to finance your external debts. Second, the government needs to raise money to finance its debt through distortionary taxation, which lowers the amount that it is possible to produce.

    On Fraud 3: I agree that the relationship between government deficits and savings is unclear, and so do most economists. In a simple neoclassical model where Ricardian equivalence holds, the net effect of government deficits on savings is zero. In an overlapping generations model, government deficits will generally decrease net savings, but not by as much as one might expect. There is also some ambiguity in the word “savings”; government debt provides an unusually liquid savings vehicle and can allow for more efficient store of wealth across generations (“savings”) while decreasing the fraction of savings held in the form of claims on capital investment. It is incorrect, however, to say that budget deficits necessarily “add to savings”, unless you don’t count the government’s asset position when you’re looking at savings.

    On Fraud 4: It is indeed important to make sure that the economy’s real capacity for production in the future is enough to support retirees. But public finance isn’t completely irrelevant: there still needs to be some mechanism to transfer value from workers to retirees, and that mechanism will generally involve taxes. (As I have stated repeatedly, seignorage is a form of taxation that gives us comparatively trivial amounts of revenue; other forms of taxation are both more efficient and necessary.)

    On Fraud 5: To the extent that trade deficits can be sustained indefinitely, they do indeed directly add to the standard of living. I think most economists would agree with this. Some are concerned that the US is building up “external imbalances” that will reverse in a sudden, cataclysmic financial crisis, as we’ve seen in many developing countries. I am actually not so concerned on this issue, since I think there are key qualities that distinguish the US from countries that have experienced these crises, in particular its overwhelming size and wealth. But I don’t think these concerns should be dismissed entirely.

    On Fraud 6: I find this absolutely incomprehensible. The “paradox of thrift” is relevant when the savings-investment market can’t equilibriate (i.e. interest rates have hit the zero lower bound), but it is not some kind of general barrier that prevents savings from ever being turned into productive investment. How do you think that business investments are financed? Someone is saving! (in a bank, the stock market, a venture capital fund, etc, etc)

    On Fraud 7: You are saying that running deficits today that will be financed by higher taxes tomorrow can be a good thing. I agree. In fact, at some level virtually all economists agree, even right-wing ones who do not believe in Keynesian stimulus. (They generally still believe that it is optimal for government to smooth its spending, as the marginal utility of government purchases doesn’t suddenly decline during a recession.) I think that monetary policy is a far more potent and direct tool than fiscal policy; maybe you do too, but since MMT conflates monetary and fiscal policy it seems impossible to talk coherently about the issue.

  10. Hi!

    paying back the debt- debiting securities accounts and crediting member bank reserve accounts- is not itself an inflationary event. it’s the ‘original’ deficit spending that is the inflationary bias, and not the mix of cash, reserves, and tsy secs it leaves in its wake as that spending takes place. All three are govt liabilities and net financial assets of the system. Note how even $3T of couldn’t put humpty together again and boost q1 to the Fed’s hoped for growth levels, though qe did apparently (successfully?) scare substantial numbers of portfolio managers out of the dollar, but that’s another story!

    There is nothing every stopping our children from producing what they consume. what is at risk is the ability to net import, which is consuming more than one produces. and paying the debt, including interest, is never more than debiting securities accounts at the fed and crediting reserve accounts at the fed. it doesn’t need to interfere with actual domestic consumption of goods and services, which is limited by real resources and people willing and able to work.

    Yes, I was writing about nominal savings, and not real savings. And if the world were totally ricardian, where cutting taxes did nothing, taxes could be cut to 0 without altering real outcomes, which, by the way, has been repeatedly pointed out by mainstream economists as well. So I don’t care how ricardian the world might be, with federal taxes functioning to regulate aggregate demand, unemployment is telling us we are grossly overtaxed for the size govt we have, and the credit conditions and ricardianess we have.

    The mechanism to transfer real goods and services to retirees is social security and medicare, with taxes functioning to regulate aggregate demand once the ‘right size’ govt is selected.

    we should at least stop the goose hunt where the next thing we’ll do is use our drones to kill it before we are buried in golden eggs…

    The causation is that savings is the accounting record of investment. Loans ‘create’ deposits, the issuing of financial liabilities, such as stocks, bonds, commercial paper, etc., ‘creates’ financial assets which is what those same things are called in the hands of investors.

    I’m saying higher deficits today will cause the economy to do what we are currently trying to get it to do through means that don’t achieve that goal.

    Hope this helps!

    • It’s only appropriate to put base money and government debt in the same box by saying “all three are govt liabilities and net financial assets of the system” if you’re prepared to accept zero nominal interest rates, or pay interest on reserves. Otherwise, there is a clear difference between money and bonds: one pays interest and the other does not. Investors need to be receive liquidity services from money over-and-above what they would get from bonds to be willing to pay this premium.

      I agree that once nominal interest rates are permanently at zero (or whatever rate is paid for IOR), there will be no difference at the margin between money and short-term government debt, but this is not the policy under which the government usually operates, and it is potentially problematic for other reasons.

      • even with any rate the fed might select the prices/yields of the tsy secs ‘gravitate’ towards ‘indifference levels’ and in that sense are always fungible for all practical purposes.

        In other words, the entire term structure of rates expresses indifference levels for the given investors and the given investments at any point in time.

  11. Historically, “MMT” is the blending of two different schools of thought – chartalism and endogenous money. What you have done in this post is critiqued chartalism. Personally, I’ve never been a big fan of chartalism (the idea that money has value because the state accepts it as payment for taxes). However, the endogenous money school of thought makes a number of great points. Throughout your post, you are assuming that “money” is something tangible, quantifiable, relatively fixed in quantity, and external to the economic system. In a modern capitalist economy with a developed financial system, it is not. As Warren mentioned above – money is created when banks make loans and banks make loans based upon their perception of real economic conditions. Thus money is inseparable from real economic activity.

    • Throughout your post, you are assuming that “money” is something tangible, quantifiable, relatively fixed in quantity, and external to the economic system.

      For the sake of convenience, I sometimes use “money” as shorthand for “base money”. “Base money” is both quantifiable and controlled by the central bank (though the amount of it is still endogenous to economic activity, since the central bank is trying to hit an interest rate target).

      I agree that “money” in the much broader sense of liquid bank deposits, or even liquid assets, is endogenous to real economic activity. In fact, I’ve made this point elsewhere, and will probably make it again in the future: the apparently stable relationship between monetary aggregates like M2 and nominal GDP is probably due more to nominal GDP affecting M2 (which consists mostly of liquid bank deposits that are claims on real investment) than the other way around.

    • see my paper:

      http://www.moslereconomics.com/mandatory-readings/a-general-analytical-framework-for-the-analysis-of-currencies-and-other-commodities/

      Here you can see the framework for analysis for any currency or commodity, which also ‘locates’ the endogenous money/circuitist schools and chartalists as well.

  12. JKH

    Interesting observations on the quantitative composition of the monetary base, but I think you’ve substantially misinterpreted the Chartalist element in MMT, and only glanced off the edge of what is a much wider MMT take on fiscal and monetary interaction.

    The Chartalist view is about getting the currency accepted from first principles. It’s not about explaining the full multiplicity of forms that the store of wealth can take.

    All taxes are ultimately paid in base money (and virtually all of that in reserves). But that doesn’t mean that a stock of base money per se must exist equal to some arbitrarily specified flow of taxation liabilities. Tax payments as they occur are almost entirely absorbed into the netting of flows in both directions at the level of interbank reserve settlement. This netting effect is most evident in those monetary regimes with a zero stock reserve requirement (e.g. Canada). So it is entirely unnecessary to map some function of bank reserve stocks to taxation flows over time.

    The Chartalist point is that it is the contingency of a tax liability that creates the demand for what starts out as a flow of reserves but ends up being the full constellation of money forms in a given fiat economy – bank reserves, CB notes, commercial bank deposits (that are convertible to CB notes on demand), etc. etc. – all of which are liquefiable to some degree such that the holder can sell them to raise cash for tax payments if necessary.

    More specifically, the tax driven demand for money has virtually nothing to do with the mechanism for controlling the Fed funds rate. In this context, paying taxes only requires that a bank reserve account be debited. It is the net effect – not the gross effect – of this type of transaction in the midst of all others that determines the momentary demand for reserves and its pricing impact and the CB’s response to that impact.

    • JKH

      “the tax driven demand for money has virtually nothing to do with the mechanism for controlling the Fed funds rate”

      that’s too strong – i mean there are other factors such as the contemporaneous expenditure driven (government) supply of reserves, and individual banks’ share of reserves due to non government transactions that affect reserve pricing – which the Fed must take into account as well

    • I actually agree with almost everything you’ve said here, especially the following: “Tax payments as they occur are almost entirely absorbed into the netting of flows in both directions at the level of interbank reserve settlement.” Absolutely. And when you say this…

      The Chartalist point is that it is the contingency of a tax liability that creates the demand for what starts out as a flow of reserves but ends up being the full constellation of money forms in a given fiat economy – bank reserves, CB notes, commercial bank deposits (that are convertible to CB notes on demand), etc. etc. – all of which are liquefiable to some degree such that the holder can sell them to raise cash for tax payments if necessary.

      …I think I made a similar point in my original post, arguing that it’s possible that taxes originally caused us to settle into the equilibrium where money the medium of exchange and unit of account, even if taxes don’t have much effect at the margin:

      All this may be true. But it’s also mostly irrelevant for the conduct of monetary policy. Perhaps government, using its power of taxation, nudged us into the equilibrium where green pieces of paper are used both as a medium of exchange and a unit of account. But that doesn’t mean that taxes, at the margin, have anything to do with the demand for currency.

      My point is simply that these deep historical considerations (what exactly led us to imbue money with value?) matter very little for the practical day-to-day operation of monetary policy, and whether or not the government can raise a substantial part of its income through seignorage. And that’s what bothers me about MMT: not the possibly valid insights about the historical and conceptual origins of money, but the audacious and plainly false claims about the government’s ability to finance itself through monetary policy.

      • JKH

        Fair enough.

        But then I don’t understand two things.

        First, I’m not sure MMT would disagree violently from what either of us is saying regarding the distinction between “strategic” and “operational” impacts of taxation on the demand for money. E.g. you say in your article, “let’s indulge MMT and make the ludicrous assumption that money demand will increase one-for-one with the rise in annual tax revenue”. I doubt that’s indulging MMT at all – from what I understand of MMT, I have no reason to predict they would make such an operational assumption about the relationship between taxation and “the demand” for base money. Indeed, I’ll predict with near 100 per cent confidence that they absolutely won’t. In terms of accommodating a deficit through base money, the fact is that the government has the option of creating interest paying bank reserves as an alternative to issuing bonds. This option is important both for its implicit value as an operational contingency plan in the context of the existing monetary architecture, and it’s more explicit value as a strategic plan for permanent change to that architecture, as enunciated for example by Warren Mosler in his suite of proposals for monetary system change.

        It is important to recognize that this option in either context really has nothing to do with the “demand for base money”. The government has two channels for the supply of base money to the system – banks and non-banks. The non-bank demand for base money is a legitimate demand, satisfied by issuing currency notes. The central bank does not operationally control that endogenous demand for base money. The bank demand for base money is more complex. It is up to a point (the option) entirely a function of interest rate control exercised by the central bank and the pricing of reserves that results in the context of their value for their core function as the interbank settlement medium of exchange. There is no context for the bank demand for base money that is separate from this operational interest rate control (except for inventorying a relatively small supply of notes to meet non-bank demand). The central bank supplies the level of reserves required to maintain that control. But it does have the option of supplying more than that. It still controls the interest rate (with a few minor exceptional inefficiencies) by paying interest at the floor rate. This option is supply side driven – the supply that the central bank wishes to inject as a function of broader asset pricing. It is being used in QE in the central bank’s attempt to have some influence on longer term rates. It was used earlier to accommodate swapping of less credit worthy assets to influence risk spreads on those assets. In all cases, it is the central bank that makes the market on where it is willing to swap interest paying reserves for such assets. It can make any market it wants in any of these assets, down to an interest rate of zero, presumably. And as noted above the government has the ultimate option of abandoning the new issue bond market altogether and simply allow net deficit spending flows to settle in as excess bank reserves in some combined treasury/central bank institution version of the monetary architecture. In all these situations, the banking system will be stuck with the level of interest paying reserves that is the result – as an aspect of the value of reserves in financial intermediation, pricing, and arbitrage – not as an aspect of the value of reserves for their core purpose of interbank settlement and the “demand” for those reserves in that context.

        Second, I’m not clear at all on path of your criticism of MMT beyond the Chartalist taxation perspective (perhaps I’ve missed previous discussion here). You say, “MMT starts with lengthy historical musing about the origins of money.” Exactly – but it’s only a start. I was disappointed precisely because your opening discussion about this part of it was so original and interesting – but was hoping for more beyond that. I’d encourage you to keep going.

      • I’m with JKH. There’s absolutely no suggestion in MMT that demand for base money increases 1-for-1 with tax revenues. My paper linked to above demonstrates this, and agrees with everything else JKH has said.

      • from the beginning:

        1. Govt. is desirous of provisioning itself
        2. It creates tax liabilities in its own currency of issue which
        results in real goods and services being offered in return for units of that currency
        3. The govt (and/or its designated agents) buys the real goods and services it desires in exchange for funds used to pay taxes and to fund savings desires.

        The US dollar is a simple, public monopoly.

        If the govt, the currency monopolist, restricts supply of its ‘product’- dollars in this case- by spending less than the need to pay taxes and the desire to net save dollar financial assets, the result is excess capacity we call unemployment- people looking for work paid in dollars who can’t find it.

        So if the tax creates more people for sale than the govt wants, it should either cut the tax or hire them.

        http://www.moslereconomics.com

      • TC

        I am barely an MMT person and understand it only fractionally as well as JHK and Scott. But this:
        “I doubt that’s indulging MMT at all – from what I understand of MMT, I have no reason to predict they would make such an operational assumption about the relationship between taxation and “the demand” for base money. Indeed, I’ll predict with near 100 per cent confidence that they absolutely won’t.”

        is spot on.

        Matt, this post is an honest attempt to further understanding, but I still think back to the original SRW comments about why MMTers get all pissy.

        People call them idiots, give the reason why MMTers are idiots, and it usually turns out that the MMTers don’t believe anything close to the thing that makes them idiots.

        Don’t group this crowd in with the crazies. If something seems immediately stupid about MMT, look into it more. I think you’ll find it to be richly rewarding.

        Even the Employer of last resort, even this most outlandish of proposals, look into why this particular policy proposal is such a big part of MMT thinking. You’ll run into effective demand vs. aggregate demand, and it rocks.

  13. Donald A. Coffin

    I’m fully aware that this comment is only tangentially (*very* tangentially) related to the real topic of this post. But I have always felt some skepticism about the hypothesis that vast sums of US currency is being held outside the US. I recognize that this may be true, but there’s some evidence that things may not be quite so simple.
    Using nominal values for (for example) currency and demand deposits, it appears quite clear that the volume of currency becan to expand fairly rapidly in about 1974, with average annual compound growth in the currency component of M1 from 1974 to 2010 being about 6.5%.
    And the nominal value of US currency in circulation has increased by a factor of about 10 during that period.
    And also during that period, the nominal volume of demand deposits grew at an average annual compound rate of only 1.5% per year, increasing by only about 70%.
    Where it gets interesting is when we do two things. First, look at money *per capita*, because, after all, the population of the US increased by about 45% between 1974 and 2010. Second, adjust that for rising prices, so we’re looking at *real* *per capita* holdings of currency and of demand deposits.
    When we do that, we find that real per capita holdings of currency rose by only 3% per year, while real per capita holdings of demand deposits actually fell by 1.5% per year (with almost all the decline coming between 1974 and 1983). And real per capita holdings of M1 grew by only 0.8% per year. (In 1974, demand deposits accounted for about 75% of M1; by 2010, only 34%.)
    Now, surely a part of the reason behind the delative decline in demand deposits was monetary innovation (and, early on, rising interest rates).
    But these changes, it seems to me, need some additional consideration. And I don’t seem to recall having seen much discussion of it anywhere. (Or, in all likelihood, I’m missing something blindingly obvious.)

    • I think this is an interesting observation, and my guess is that it’s due to a massive increase in US currency circulating abroad. One near-unbelievable fact about currency in circulation is that there are more $100 notes than $20 notes: http://www.federalreserve.gov/paymentsystems/coin_currcircvolume.htm (I carry around $20s all the time for small purchases, but $100s seem near-useless.) The vast majority of the value of currency is held in the form of $100 bills. And clearly, these cannot be serving any conventional transactional purpose: they’re either being used to store money in the underground economy or abroad.

  14. true, but i just stated it has nothing to do with the mechanism
    ;)

  15. First, great conversation, and welcome back Matt!

    A few comments:

    1) Matt, you write that

    I agree that once nominal interest rates are permanently at zero (or whatever rate is paid for IOR), there will be no difference at the margin between money and short-term government debt, but this is not the policy under which the government usually operates, and it is potentially problematic for other reasons.

    Do you view payment of IOR as a temporary change in US monetary institutions? My understanding is that it is a long-planned, permanent change, and we should revise the terms of our conversation accordingly. (See, e.g. the Fed’s consideration of a “floor” model of monetary policy, rather than traditional policy that expects reserve demand to result from reserve requirements and clearing needs. Even the widely adopted channel policy pays IOR, although with reserve supply managed so that quantity if interest-paying reserves is small.)

    Money and government debt are coherently distinct when money is non-interest-bearing, debt pays positive yields, and there are frictions that prevent the private sector from holding all-debt portfolios that are swapped into transactional money on-demand. When these conditions fail to hold — when there is no opportunity cost to holding money compared to holding an overnight bond, when bond markets are sufficiently liquid that mobilizing short-debt intro transactions is easy — the distinction falls away. Just as a descriptive, institutional matter, I suspect that the distinction is fading, due both to technological change and intentional decisions by the Federal Reserve. Do you disagree?

    2) You point out that, if reserve policy is set such that interest rates on bonds are zero, and if in equilibrium real yields are positive, then arithmetically, the zero-interest term structure that some MMT-ers advocate is consistent only with long-term deflation. As we saw in the fracas that followed when Kocherlakota made the same point, there are lots of subtleties and causal questions here — short term vs long term effects, would the purported equilibrium ever come to obtain under zero-interest conditions, etc. But lets put all of that aside. A more fundamental question is this: how do we know that the long-term equilibrium real interest rate is positive at all?

    That might sound like a dumb question. We’ve had centuries of experience with bonds and interest rates, and we’ve observed positive long-term real interest rates over the period. But that doesn’t eliminate the possibility of structural change. Over these centuries, we’ve grown collectively richer. In economies that tolerate substantial inequality, we’ve developed a sizable cohort of savers that are astonishingly wealthy, for whom the marginal utility of a dollar’s current consumption is near 0. Arguably the convenience yields and insurance value of holding money do not diminish as fast as consumption value, such that there is some level of wealth at which real interest rates would naturally turn negative. (Also, entities like the People’s Bank of China have also emerged, which derive different sorts of “convenience yields” from pursuit of policy objectives that, as a side effect, imply expanding portfolios of dollar-denominated securities. But this might not be a durable fact.) Anyway, the net effect of these changes is that, conceivably, the marginal buyer of government securities could well be willing to suffer negative yields for the trouble. If you consider this possibility, it has considerable explanatory value. After all, during the 2000 “savings glut” we had “oceans of liquidity” casting about for yield via increasingly exotic means. On the one hand, that suggests some dissatisfaction with very low yields on vanilla goverment debt. On the other hand, it suggests relentless downward pressure on yields. The response to these conditions by the financial sector was to expand the supply of “safe” debt securities trying to satisfy “traditional” expectations regarding achievable low-risk yields. If the financial sector had not done this, perhaps we would have observed real yields on part of the term-structure dipping negative (and we’d have had to deal with the disruption to the solvency of pension funds and insurance companies that would imply). Post-crisis, we have both seen substantial reductions of liquid wealth and substantial expansions of “vanilla” government debt, and still, real yields are low to negative across the curve. Again, the system seems to be bucking against a trend, that the marginal buyer of truly safe debt would be willing to pay a premium for the insurance and time-shifting service that debt offers, rather than requiring a yield to defer present consumption and liquidity loss.

    Do you think there might be anything to this? And if so, how might if affect your view of (some) MMT-ers suggestion that a zero-interest rate policy might be consistent with ordinary price stability, defined as an anticipated and moderate level of inflation?

    Sometimes we hold these debates about “monetary policy” and “fiscal policy” as if there are timeless truths. Maybe part of the reason why the MMT view is rising to unusual prominence now, more than 60 years post Abba Lerner’s famous essay (and there are much older antecedents I suspect), is because times and institutions are changing. Seigniorage is ultimately a function of the nonmonetary yields associated with holding government debt, relative to the value of current expenditures or alternative investments. Observed seigniorage is a small fraction of this, and is, as you’ve pointed out, a function of institutional arrangements that split these yields among governments, banks, borrowers, and savers in the economy. Maybe we are living through a shift in the relative value of these convenience yields and alternative expenditures. All of a sudden, theories that seemed at best arithmetically uninteresting really come to matter.

    • I think you’re right on target with both of these points, Steve. I do think, though, MMT’ers would argue that the theories that previously seemed “arithmetically uninteresting” don’t just start to matter but are the appropriate starting point for analysis of the monetary system. Not doing so makes it difficult to understand the sort of regime changes we’ve been going through for decades or even centuries.

      • Scott,

        Thanks for the pointer. (Your recommendations have always been great.) I’ve downloaded the Wray paper, but it will have to wait until tomorrow.

        I agree that the MMT perspective is an important perspective from which to understand any monetary system (even commodity or fixed exchanged rate systems are enriched by the compare and contrast). The question of how “arithmetically relevant” MMT is, in the very narrow sense of how much more latitude for fiscal policy it might offer than other perspectives will, of course, vary. MMT-ers are perfectly happy, I think, to acknowledge that under some circumstances, even a currency issuing government may have little fiscal latitude, it must trade off taxation or inflation. It’s just that those circumstances come up less frequently for currency issuers than for currency using governments (just as financial constraints bind corporate action less tightly for an all-equity firm than for a highly leveraged firm). The question of how “arithmetically important” MMT is will vary on a case-by-case basis, despite the universal applicability of the analytical tools.

        I do think MMT has become controversial and prominent now, though, precisely because it is “arithmetically relevant” now. MMT-ers claim that in the US we are significantly less fiscal constrained than would be predicted by other perspectives. Arithmetic relevance isn’t the same as analytical relevance, but it sure does raise the stakes!

    • Steve, thanks for the comments!

      Just as a descriptive, institutional matter, I suspect that the distinction is fading, due both to technological change and intentional decisions by the Federal Reserve. Do you disagree?

      I think that it’s certainly possible, though it all depends on whether the Fed decides to implement interest on reserves as a long term policy. But I don’t think that this matters much for my criticism of MMT’s claims that the government is immune to a budget constraint: if the Fed is resolving the issues I identify by paying interest on reserves, then there is a cost to financing debt, much the same as there is a cost through traditional financing via bonds. I don’t deny that the Treasury and the Fed could theoretically be merged, and that the entire debt could be held in the form of interest-paying reserves (though I don’t think it would be a good idea!). I do deny that this meaningfully relaxes the fiscal constraints facing the government.

      A more fundamental question is this: how do we know that the long-term equilibrium real interest rate is positive at all?

      This is a very interesting question. Before addressing it, however, I want to emphasize that this doesn’t make much difference for the plausibility of claims (often made by MMT advocates) that the government isn’t fiscally constrained because it has the power to issue money. If the equilibrium real interest rate is negative, then the government’s fiscal constraints are greatly relaxed (though they still exist!). This happens, however, regardless of how the government finances its debt—whether it be through bonds or reserves. Unless one believes that the long-term equilibrium real interest rate is somehow intimately connected to the share of government debt held in bonds vs. reserves (which seems doubtful), the ability to create money is irrelevant.

      I think that it’s possible that the long-term real interest rate on US government debt will turn out to be surprisingly low, even negative. I doubt that this will happen with many other kinds of securities, however, even seemingly safe ones: there’s something special about the (perceived) liquidity and safety characteristics of US debt that have allowed it to trade at a substantial premium. Will these premium continue in its present form? Possibly, though I suspect that through much of a last decade we’ve been experiencing a “perfect storm” for low real interest rates, as massive developing countries (China especially) with immature financial and monetary systems have stashed their savings in the one vehicle liquid and deep enough to accommodate them.

      In fact, I think this militates against the lackadaisical MMT attitude toward deficits. The US has the power to issue debt with unparalleled properties of liquidity and safety. There is great demand abroad for an asset with these characteristics, and hence the US is able to finance its debt at relatively low interest rates. But although the demand is high, it is not unlimited; in particular, it’s largely the result of a small set of institutions (central banks, sovereign wealth funds, etc.) pursuing a mandate that is idiosyncratically weighted toward “perfectly safe” assets. If the US issues these assets in too great a quantity, it runs a good chance of saturating the market, and the marginal investor will be someone who doesn’t care as much about the characteristics that make US debt truly valuable. Like any good monopolist, the US should try to capture the inframarginal surplus, and avoid creating so much of its debt that the yield rises and the surplus is transferred to the inframarginal investors instead.

      • Matt,

        Thanks back for the thoughtful response. I don’t think we’re disagreeing very much, either the two of us or the MMT-ers.

        MMT is sometimes used (foolishly IMHO) as an excuse for, as you put it, lackadaisical fiscal policy. That is, it is one thing to say (as the MMT as well as lots of other people do say) that right now there is little downside to expansionary fiscal policy, because the elasticity of the price level to deficit expenditure is low. (That’s a weird but not unheard of way of putting things in mainstream macro. There is a perfectly orthodox tradition — Scott Fullwiler first turned me onto it — called the fiscal theory of the price level. John Cochrane’s work on the subject frankly reads to me like full-throated MMT.)

        It’s another thing to say that there is never a downside to expansionary fiscal policy, because a currency issuer cannot default.

        In my reading, in nonpolitical conversations, MMT-ers claim the former, not the latter. They do claim that the currency issuer that borrows in its own scrip need never default. But they do acknowledge a fiscal constraint, some level and character of expenditure and borrowing at which the price level would shift. I don’t think they’d have a problem with your surplus-maximizing monopoly issuer claim that there are inframarginal deficit/yield combinations that might maximize the total real resources the public sector could recruit through issue of new paper. Fundamentally, everyone from Ron Paul to Warren Mosler on the outside right on through to Ken Rogoff on the inside acknowledges that there are real constraints on government expenditure so long as government wishes to manage the value and yield of its scrip in terms of real goods.

        The substantive dispute has to do with the character of the constraint, the degree of slack in the near term and the means of measuring the constraint over a longer term. The MMT-ers, in my view, have the more parsimonious but also perhaps the more dangerous position. They do not have a theory of the price level or budget constraint (although if I’m reading Scott’s work correctly, they are fairly certain that their policy prescriptions would leave long-term debt to GDP contained within some long-term bound that they do not characterize ex ante). The MMT-ers claim practical claim, I think, is two-fold:

        1) We don’t know precisely where the budget constraint is — we don’t have a good theory, and we have reason in fact to believe any boundary to be unstable boundary, fluctuating as a function of private and external-sector preferences;

        2) a currency-issuing state that borrows in its own currency and has effective taxing power has tools sufficient tools to play things by ear without cornering itself catastrophically.

        One way of putting it might be that the mainstream view sees us walking along a tightrope, balancing sometimes contradictory objectives. We must manage fiscal and monetary policy so that prices are stable and yields moderate on the one hand, but also so that the level of economic activity is sufficient to enable full employment. There are fraught tradeoffs between these goals.

        The MMT view is that rather than a tightrope, we are driving along a gently curving lane. Yes, we could still go off the road on either side, and we do have to turn the steering wheel with the road. But it is not so difficult, to follow the curvy road, to notice when we are drifting and correct course. So we shouldn’t sweat it so much. If resources are underutilized and it seems like activity is too low, we should deficit spend money into people’s pockets and expect that the price level won’t shift, and let a real balance effect encourage a return to utilization. If we are wrong, we will notice we are drifting. We have a steering wheel, we have a brake, we’ll be fine.

        Which of these two characterizations is more accurate strikes me as more an empirical than a theoretical question, once you concede that we don’t have a model of fiscal constraint that is sufficiently well-identified to make quantitative predictions. (Rogoff & Reinhart’s 90% of GDP rule of thumb doesn’t strike me as compelling.)

        Still, we are left in a difficult place. In practice, we have to decide. Shall we be lackadaisical, as you put it, about fiscal matters, or shall we be neurotic right now. It’s one thing to say we don’t know whether we are walking a tightrope or driving a country lane. It’s another thing to decide whether you are willing to test a curve that you’d never navigate on a while. One rule is to adopt a precautionary principle and assume hazard, because that seems safer. But then we have questions: “at what cost?” and “safer for whom?” There are real costs to caution, and they are not borne evenly. So it’s unsurprising that there is a political struggle.

        But in the economic sphere, I think that we are manufacturing controversy unnecessarily. We know we don’t have an adequate predictive theory of fiscal constraint. We can discuss factors that lend themselves to looseness (current unemployment, the apparent tendency over the past decade toward negative real interest rates, etc.). We can discuss factors that suggest caution: commodity price inflation [at least until today], US dollar weakness, 90% rules of thumb, etc.) We can place different degrees of weight on that evidence, and we undoubtedly will, undoubtedly based in part on our own circumstances and political commitments along with our considered technocratic judgments. Ultimately, I think the MMT-ers are simply highlighting perfectly coherent and orthodox factors that suggest looser constraint than is conventional. Their argument about money is wholly analogous to commonplace notions in corporate finance about how leverage creates financial risk and distress risks compared to equity financed enterprises of equal fundamental value. They could be wrong about the degree of non-constraint, and we each have our own little cost/benefit calculators on fiscal matters. But they are not actually as weird as they claim to be. It strikes me as sociological more than economic that both the “mainstream” and the “post-Keynesians” have separated themselves across frontiers of mutual petulance. I wish we’d police those borders less carefully. I’d like to invite you along for some transgressive panty raids, jumping from one side of the fence to the other.

        The interesting issues, I think, are the ones you highlight. Are the circumstances that are suppressing interest rates and providing large convenience yields to holders of US paper temporary or permanent? Is this a made-in-China “perfect storm” or a durable fact. The MMT-ers think they have a theory general to all currency issuers. You suggest that there’s something special about the US, its liquidity and safety. I think you’re both right, and we need a theory that explains the heterogeneity of fiscal constraint, both longitudinally and cross-sectionally. I think MMT-ish intuitions are a lot less relevant to, say, an Indonesia than to an America, even though Indonesia is a currency issuer. Small open economies have foreign currency denominated operating liabilities even when they do not have foreign currency dominated debt. There’s a whole landscape of circumstances, some of which look like the world MMT-ers describe, some of which look like hair-trigger constraint. Rather than fighting which place is more real, Chicago or Katmandu, it’d be better to concede the possibility of both and characterize the forces that would render them so different.

        A small point:

        If the equilibrium real interest rate is negative, then the government’s fiscal constraints are greatly relaxed (though they still exist!). This happens, however, regardless of how the government finances its debt—whether it be through bonds or reserves. Unless one believes that the long-term equilibrium real interest rate is somehow intimately connected to the share of government debt held in bonds vs. reserves (which seems doubtful), the ability to create money is irrelevant.

        I love this point, but I only half agree. We agree that it is not the fact that money is used as a medium of exchange, but the fact that it offers a convenience yield that renders it in-demand despite offering low or negative financial yields. Money or debt, as you say, does not matter at all, as long as we posit the tacit yield.

        However, one of the tacit yields that a security can offer is direct usability or easy conversion to the medium of exchange. So the two are not entirely unrelated.

        My view of finance has become increasingly obsessed with tacit yields, and liquidity — convertability to the medium of exchange — is only one source of those. There are several others. But I’m tempted to flip the classification. Rather than say it’s not money, but tacit yields, that renders the US less fiscally constrained than one might expect, I’d argue that we should think of tacit yields as the defining characteristic of money. Whatever it is that people want to hold for reasons other than consumption value or financial yield is money-like. Moneyness, defined this way, can be impermanent and fickle, but I think that’s all right. Wasn’t it Cochrane again who argued that the dot-com bubble resulted in part from the fact that tech-paper took on certain characteristics of money?

        I’ve abused your comment thread, and apologize for that. You’ve been admirably generous in responding to people, and I don’t want to abuse your time too. Though of course I welcome anything you have to say, please don’t feel obligated. It is delightful to have you in the economics blogosphere. You’ve got a sharp mind and a good voice. I look forward to reading everything you write.

  16. Max

    I’m not an MMT heavy but perhaps I can add something. The reason why some MMTers suggest not issuing bonds isn’t because they believe that increasing reserves would “finance” anything, but rather, because it vividly illustrates that the popular idea (popular even with economists) that the bond market is in control is a fallacy. The bond market has no power over a currency issuing government, so concerns that the U.S. could become the next Greece are baseless. This is true under current institutional arrangements; nothing needs to be changed.

    • beowulf

      The bond market has no power over a currency issuing government, so concerns that the U.S. could become the next Greece are baseless. This is true under current institutional arrangements; nothing needs to be changed.
      Correct, until the Tsy-Fed Accord of 1951, short term rates were capped at 0.375%, long term rates capped at 2.5% (the Fed buying up any Treasuries the market didn’t), lets say the Tsy-Fed Un-Accord of 2011 decides that we’re going back to the pre-1951 regime, except with Tsy issuing nothing longer than 3 month T-bills. yes they could do that under the current institutional arrangements. And with that, you just saved Matt having to raise $300 billion a year in new taxes (and no doubt getting hammered at re-election).

      So if $300 billion a year gets you 75% there ($15 trillion), then $$400 billion a year gets you 100% to paying off $20 trillion (paying off the national debt is ALWAYS a bad idea, but we’ll play along), Matt said in his hypo there are 4% nominal rates, swell that means our current debt service is $800 billion a year (ouch!). But wait, as old debt rolls over and are replaced with T-bills with 0.375% cap (and with Fed refunding its net earning to Tsy, the effective rate will be even lower), that net interest amount will keep dropping, until one fine day, we’ll reach 0.375% nominal interest rate and net interest will be less $80 billion a year. There will be no reason to ever pay off the debt, so long as economy’s average growth rate exceeds 0.375% (well again, could be less thanks to Fed refund).

      Our debt to GDP ratio will keep shrinking and we can spend that freed up $780 billion a year on human needs instead of wasting it on pointless rent seeking. Sounds crazy, but this is pretty much how we funded World War II (average interest rate on debt was less than 1% and that was before Fed had begun its practice of refunding earnings).

      • You are implicitly assuming that by using its power to change nominal interest rates (which we all agree that it possesses), the government can change the real interest rate that investors demand from it on its debt. Why do you believe this? The conventional view of the real interest rate states that it is determined (at least in the long run, in which the Fed cannot use price stickiness to manipulate it) by entirely real factors, like investors’ intertemporal preferences and the supply of investment projects. Why do you think that investors will forever accept a lower real return on wealth in equilibrium just because the Fed has manipulated nominal variables?

      • Matt,

        The way we do this is by rejecting the argument that there is a real rate set by real factors.

      • Furthermore, it appears that you are making the Kocherlakota assumption that a low nominal rate necessarily implies deflation. That particular view is absolutely not accepted by everyone (or even most?) in the profession, as the response to his speech demonstrated.

        I have, however, believed for a long time that the crux of the matter between MMT and others is this point about how nominal and real interest rates are set, and how much the latter “matter.” I think my MMT colleagues have been placing emphasis on the wrong things to a degree, at least in terms of debating other economists.

        Best,
        Scott

      • OK, last one. I should start waiting a few minutes before hitting enter, I guess.

        An attempt at clarification: If you’re referring to the equilibrium real rate vs. the actual real rate, then yes, your point makes sense, at least within the mainstream view (if you said this above and I missed it, my apologies). In other words, you’re suggesting that the equilibrium real rate must be positive (which would seem to be Kocherlakota’s point), not the prevailing real rate if the Fed keeps the target at roughly zero. Obviously, the mainstream view is that the latter would keep us from actually achieving the equilibrium real rate. Again, though, MMT rejects the notion of an equilibrium real rate, which–yet again–is probably the crux of the matter.

        Best,
        Scott

    • Max (I’m referring to the post several entries above), you’re correct than in a literal sense the “bond market” has no power over a currency issuing government, since the government could switch to issuing its debt in the form of money rather than bonds. But why should this matter? If the “bond market” is concerned about the government’s ability to repay its debt (as it is in Greece), then why shouldn’t the same be true for the “reserves market”? Why shouldn’t banks be just as reluctant to hold government debt in the form of reserves as bond traders are to hold it in the form of T-bills? If solvency is the underlying concern, then there should be no difference.

      • Max

        “(I’m referring to the post several entries above), you’re correct than in a literal sense the “bond market” has no power over a currency issuing government, since the government could switch to issuing its debt in the form of money rather than bonds. But why should this matter?”

        It’s directly relevant to the political debate, because people are convinced that the government has to run a pro-cyclical fiscal policy, otherwise the bond market will demand higher interest rates and a default becomes possible.

        If you understand that issuing bonds is a favor that the government does for investors – a welfare program if you will – then this worry becomes laughable.

        “If the “bond market” is concerned about the government’s ability to repay its debt (as it is in Greece), then why shouldn’t the same be true for the “reserves market”?”

        If money and bonds are issued by the same government, then the default risk is the same, i.e. none. There is no solvency concern. In the Euro zone, the bond market regards Germany as the de facto currency issuer. Greece is in a position analogous to a U.S. state.

  17. Detroit Dan

    Awesome comment, beowolf. Thanks…

  18. Hi Matt,

    Though MMT is certainly nonsense, I’m hesitant to spend much time refuting it—since it has roughly the same stature among economists that creationism commands among biologists, there’s a sense that I’m just tilting at windmills.

    In light of this very good discussion, and the various qualifications of your own position you’ve made above, do you still believe that “MMT is certainly nonsense. . . “? And do you still suggest, as you do above, indirectly, that MMT is analogous to “creationism?”

    • My opinion has not changed on matters of fact, though perhaps I’d define MMT somewhat differently. I still think the notion that control of the monetary base relieves the government of the need to satisfy a budget constraint is absurd, and akin to creationism. Since I viewed this belief (based on my previous interaction with MMT advocates, as well as an examination of their positions in public debate) as a central tenet of MMT, I felt that it was reasonable to call MMT “nonsense”. (I suppose we can have a meta-debate about whether it is reasonable to call a school of thought with some defensible beliefs but one central and ludicrous one “nonsense”… but that’s mainly semantic and doesn’t interest me very much.)

      Since many commenters here have reacted negatively to my portrayal of this “chartalist” belief as a core principle of MMT, I think that maybe I should reevaluate my terminology to make my criticism more explicit; I want to be clear that it’s the Randall Wray “don’t listen to S&P” editorials that are ridiculous, not every historical observation that MMT has made with regard to money.

      • TC

        A few simple questions: Is the constraint being satisfied now? How do you know? How can you tell the behavior of the future governments will keep the no Ponzi condition satisfied given what we know today?

      • Thanks for clarifying that, Matt. You said:

        I still think the notion that control of the monetary base relieves the government of the need to satisfy a budget constraint is absurd, and akin to creationism. Since I viewed this belief (based on my previous interaction with MMT advocates, as well as an examination of their positions in public debate) as a central tenet of MMT, I felt that it was reasonable to call MMT “nonsense”.

        I agree that it’s a central tenet of MMT that there is no Governmental Budgetary Constraint on nations with non-convertible fiat currencies having floating exchange rates and owing no debts in currencies other than their own, except for self-imposed budgetary constraints. But I don’t think your arguments earlier have shown that this view is nonsense. MMT economists say that such nations are always capable of spending by marking up private sector accounts since there are constitutional limits on their ability to spend. Now, I think that this is a fact. Can you really sensibly challenge this?

        I know you can say that the unlimited exercise of this authority by the Government can have terrible effects. And, I don’t think any MMT economists either disagree with you on that or that the Government ought to engage in unlimited spending. What MMT people do say, instead, is that all Government spending should be assessed from the viewpoint of its impact on the economy and society and should be evaluated in terms of whether it fulfills public purposes.

        Government deficit spending that created inflation would not fulfill purposes, so generally MMT is opposed to such spending unless there’s good reason for thinking that any ensuing inflation is counter-balanced by other positive benefits of that spending.

        Now one may look at deficit spending after the output gap is closed as a Governmental Budgetary Constraint, if one wants to. But that’s not what president Obama is talking about when he says we can’t afford this spending or that spending because we are running out of money. No, he’s saying that the Government can only fund its spending by taxing or borrowing, and this, not to put too fine a point on it, is either a lie, or a gross display of ignorance. Surely you would not endorse such a distortion?

        Above, in replying to Warren you said:

        “On Fraud 1: I agree that the government cannot formally default on nominal debts if it is willing to inflate away those debts.”

        You may think the above, but Warren has never said that, so he does not agree either with you or that this is something that MMT economists advocate. In fact, I dare say that he probably thinks that any such construal of what MMTers think is pure nonsense and a result of being unwilling to read MMT texts very carefully.

        What Warren says is that whatever the level of accumulated Government debt, the Government’s capacity to spend in the here and now is exactly the same as if it had accumulated no debt at all. Whether it’s spending in the here and now will inflate away its debts, is an entirely different issue. MMTers think that as long as an output gap exists in the economy demand-pull inflation will not occur, and so the currency will be not be inflated.

        They also think that if the output is gap is closed and such inflation does follow because of unwise continued deficit spending that it will occur because of that excessive deficit spending regardless of the level of previously accumulated debt.

        I’ll end now because it’s late. But I also advise that you reconsider each of the comments you made on Warren’s frauds. Each and every one is either demonstrably wrong-headed or terribly unclear or both. Anyway, goodnight for now. Hopefully, in the next few days I’ll have a chance to reply to your
        post with one of my own.

      • Blogs that don’t allow one to edit comments for typos after you’ve published are really annoying. Anyway this statement above should read:

        MMT economists say that such nations are always capable of spending by marking up private sector accounts since there are no constitutional limits on their ability to spend. Now, I think that this is a fact. Can you really sensibly challenge this?

  19. JKH, you say:

    I doubt that’s indulging MMT at all – from what I understand of MMT, I have no reason to predict they would make such an operational assumption about the relationship between taxation and “the demand” for base money. Indeed, I’ll predict with near 100 per cent confidence that they absolutely won’t.

    I agree that there is no basis to think that MMT supporters actually believe this. Here is the reason why I nevertheless discussed it as “indulging MMT”: for reasons that I cannot comprehend, MMT advocates seem to think that there is an important connection between their claims about the origins of money in the ability to tax and the ability of a government issuing its own currency to escape any budget constraint. The “higher taxes lead to a substantial increase in reserve demand” channel is the only way I can imagine these two points having much to do with each other. Thus I discussed it.

    And as noted above the government has the ultimate option of abandoning the new issue bond market altogether and simply allow net deficit spending flows to settle in as excess bank reserves in some combined treasury/central bank institution version of the monetary architecture. In all these situations, the banking system will be stuck with the level of interest paying reserves that is the result – as an aspect of the value of reserves in financial intermediation, pricing, and arbitrage – not as an aspect of the value of reserves for their core purpose of interbank settlement and the “demand” for those reserves in that context.

    Sure. I agree that the government could conceivably issue all its debt in the form of reserves, and then pay whatever nominal interest rate on these reserves is jointly consistent with (1) the real interest rate that investors demand to hold that amount of debt and (2) the long-term target inflation rate.

    I fail to see, however, what fiscal benefits this arrangement could conceivably bring. The government still needs to finance its debt at the market-determined real interest rate. Unless the composition of debt between reserves and bonds somehow influences this real interest rate (which seems implausible, and would demand substantial elaboration), its financing costs are essentially the same as before.

    If MMT is merely claiming that the government could hold its debt in an alternative form, then I’m fine. But this doesn’t seem to be what happens in practice: instead, we get a lot of rhetoric about how the ability to create reserves makes the government immune to fiscal concerns in general. And I don’t see how the decision to hold debt in the form of interest-paying reserves rather than T-Bills can justify this rhetoric at all.

  20. “A few simple questions: Is the constraint being satisfied now? How do you know? How can you tell the behavior of the future governments will keep the no Ponzi condition satisfied given what we know today?”

    I don’t think that the government necessarily needs to abide by the exact “no Ponzi condition” included in most macroeconomic models. It seems likely that the special characteristics of government debt are valuable enough that the government can perpetually maintain debt at a certain fraction of GDP.

    However (and this is where I see frequent misunderstandings), perpetual deficits do not necessarily even mean that the no-Ponzi condition is being violated. As long as the underlying real rate of interest in the economy is higher than the rate of GDP growth, a government that indefinitely maintains debt at (say) 100% of GDP will not violate any no-Ponzi condition: as time goes to infinity, the date-0 present discounted value of debt will approach 0. Of course, the real interest rates paid on government debt are usually below the rate of real GDP growth, but this isn’t a correct comparison—the implicit return on government debt also includes liquidity services, it’s quite possible that the true overall return is above the rate of real GDP growth. In that case, conventional macro models that assume a no-Ponzi condition are perfectly fine.

    No matter what, however, I think that governments probably can maintain some level of debt (expressed as a fraction of GDP) into perpetuity. So I’m not terribly wedded to no-Ponzi conditions as a way to understand public finance.

    But I have a hard time understanding how this is relevant to MMT. If the Treasury can violate a no-Ponzi condition, it can do so regardless of the form in which it is issuing its debt. Why should things magically become easier when it issues most of its debt in the form of money?

    In fact, it seems like they would become (slightly) harder, since the government would no longer extract the inframarginal surplus from the additional liquidity services provided by base money over T-bills. (The marginal value of base money for transactions purposes goes to zero pretty quickly once you issue an appreciable fraction of your debt in that form.)

  21. Let me get even more macro to make the point.

    Goods and services (including labor) offered for sale with dollar price tags on them are all the govt. of issue needs to make the purchase.

    The goods and services offered for sale at those prices are the ‘offered side’ of the market. That price is already there.

    And these willing sellers of goods and services in the market place know what they are getting in exchange for their goods and services- dollar balances in their bank accounts.

    The prices they are asking already incorporate information such as their cost of production, alternative goods and services currently offered for sale in exchange for dollars, outstanding dollar debts and obligations the seller may have, and possibly some knowledge of prior levels of their govt’s spending and and their govt’s financial obligations.

    But in any case, the govt. is always ‘operationally’ capable of buying whatever is offered for sale with dollar price tags on it. Any constraints on buying what’s already offered for sale are necessarily political and self imposed.

    And in the US today that buying/payment in the first instance creates a reserve balance in a Fed member bank account. And the govt. can offer securities accounts at the Fed as an alternative account for dollars at the Fed, and when someone elects to shift their dollars from their reserve account to their securities account for whatever reason, we call that ‘funding the govt’ and ‘the govt going into debt’. Whatever. That exercise of providing various alternative accounts at the Fed to clearing balances has no bearing on the govts ability to make payment the way it always does, via the crediting of member bank reserve accounts.

    Yes, govt. buying could cause the subsequent offered prices from willing sellers to be higher, and yes, if that turns into a continuous increase in prices we call that inflation.

  22. JKH

    “We get a lot of rhetoric about how the ability to create reserves makes the government immune to fiscal concerns in general. And I don’t see how the decision to hold debt in the form of interest-paying reserves rather than T-Bills can justify this rhetoric at all.

    Actual or potential reserve creation addresses a separate issue, which is the fallacy of hysteria about “solvency”, “bankruptcy”, and “we’ve run out of money”. The operational and/or policy option of not issuing bonds means these characterizations are false.

    Also, they distract from the more substantial issue of “fiscal sustainability”.

    I’m suspicious of a budget equation that depends rigorously on the present value of future surpluses, requiring that to achieve “fiscal sustainability”. It seems like a static, intellectually lazy approach to the issue. Dynamic modelling such as illustrated in Fullwiler’s paper suggests there are other paths for achieving sustainable outcomes, without requiring future surpluses comprehensively or even at all. History shows that debt has not been repaid yet. Why should it be repaid in the future? For that matter, why should it be repaid anymore than money should be “repaid”?

    Steve Waldman’s comments as usual are interesting, in this case partly because of the importance of a single word in these debates: “constraint”. You can see that Waldman has used it here in a different way than Mosler has, at least in their respective comments here. The most frequent MMT use of the word is in its application to institutional barriers that should be freed up or ignored because they can be freed up – e.g. bonds instead of reserves. These are the “constraints” that mislead on the basic issue of “solvency”.

    Waldman addresses the (different) constraint of an acceptable inflation boundary. MMT prefers to discover that boundary incrementally – not through present value calculations and pre-emptive fiscal decisions based on financial ratios and bad ideas about solvency – all of which impede discovery of the true inflation boundary over time.

    “For reasons that I cannot comprehend, MMT advocates seem to think that there is an important connection between their claims about the origins of money in the ability to tax and the ability of a government issuing its own currency to escape any budget constraint. The “higher taxes lead to a substantial increase in reserve demand” channel is the only way I can imagine these two points having much to do with each other. Thus I discussed it.”

    The connection is between the authority to tax and the ability to issue currency and have that currency accepted. The connection between taxation and reserve demand is a minor operational detail, usually emphasized only for the benefit of those who wish to understand the flow of taxation in the context of monetary operations. Otherwise, taxes reduce income, saving, equity (net financial assets), and aggregate demand – and are a key element in the MMT approach to discovery of the acceptable inflation boundary when fiscal measures such as reduced taxes are pursued to increase aggregate demand and employment.

  23. Peter D

    For a non-economist like me, MMT is really very simple: the govt deficit should accommodate the non-govt sector’s (incl. foreigners) desire to save in your currency + any growth in the economy. Saved money is not inflationary and growth “pays for itself”. If the govt wishes to spend in excess of these, then inflationary effect should be counteracted by taxes.

    Constructs like IGBC are almost meaningless. IGBC is an ex post condition. It is just like saying that you cannot drive your car and accelerate forever. Does it tell you that you cannot accelerate right now? No, in fact, if you’re going uphill you probably want to floor that gas pedal just so as not to decelerate (I am glad to see Steve using the car analogy as well!)

    IGBC is important only insofar as there is an observable that tells you that you might be violating it. And what that observable might be? Seems to me, it could only be demand-pull inflation, in which case MMT tells you’re either spending too much or undertaxing.

  24. Matt,

    You write:

    “Sure. I agree that the government could conceivably issue all its debt in the form of reserves, and then pay whatever nominal interest rate on these reserves is jointly consistent with (1) the real interest rate that investors demand to hold that amount of debt and (2) the long-term target inflation rate. ”

    This shows that you are conflating two separate issues within MMT. I think that’s what several have attempted to explain to you already.

    Issue 1: Your first part ” I agree that the government could conceivably issue all its debt in the form of reserves, and then pay whatever nominal interest rate on these reserves” is what we agree with and we should stop debating this point because everyone’s basically talking past each other. In other words, we agree that the central bank at the very worst can set the nominal interest rate on the national debt.

    Issue 2: Your second part–“then pay whatever nominal interest rate on these reserves is jointly consistent with (1) the real interest rate that investors demand to hold that amount of debt and (2) the long-term target inflation rate”–is where we disagree. MMT’ers, as I explained above, reject the idea of an equilibrium real rate. We see it, instead, within the context of a desire by the non-govt sector to net save out of current income relative to the govt’s deficit.

    I think that all the other issues here–for instance, whether fiscal constraints should be seen as a tightrope versus a winding road–are secondary to this issue (though I am not implying in the least that these issues aren’t important), at least as regards the debate between MMT and other economists.

    Best,
    Scott
    Best,
    Scott

  25. JKH

    Here is an unconventional comment on the interface between MMT’s view of rates and the mainstream notion of some “equilibrium real rate”:

    I fail to see the usefulness at all in the concept of an “equilibrium real rate”.

    First, since central banks set the short term nominal rate, they effectively set the short term real rate. This is because the central bank has a continuous option to change the nominal rate. Therefore, any change in the assumed inflation rate which changes the prevailing short term real rate is in effect a change in the real rate as decided by the central bank. The option of changing or not changing the nominal rate is equivalent to an option of changing or not changing the real rate, notwithstanding any change in the assumed inflation rate, at any time. The option and its continuous availability mean the central bank effectively sets the short term real rate. The mere fact that the assumed inflation rate changes while the central bank leaves its nominal rate unchanged doesn’t mean it does not set the real rate in real (continuous) time. So it becomes clear that the central bank controls both nominal and real short term rates (allowing for unconventional easing at the zero bound).

    The question then becomes – what is the nature of the relationship between short and long term rates, nominal or real?

    MMT is quite clear in arguing its case from the nominal perspective. The central bank sets the short term nominal rate, and the market forms expectations about the path of that central bank process over time. The fact that market expectations fluctuate means that bond prices are volatile – much more volatile than the short term policy rate. Some may argue that other factors come into play, but these are ancillary –the expected path of short rates is the most rational persistent force driving government bond market yields. And the actual path of short rates set by the central bank wins out in the end. The fact that the Fed can engineer inverted yield curves in more conventional tightening patterns is demonstration of this. In that case, the Fed wins out over the “bond vigilantes” – through deserved, merciless short term tightening that eventually drives bond yields lower. Or, in unusual times like now, it wins by deserved dismissal of concerns from those who imagine inflation more than they can demonstrate its threat.

    MMT takes this anchored view of rates in some of its policy options, such as a bondless deficit expenditure system – or, as Mosler sometimes suggests, a system that restricts Treasury debt to Treasury bills. Either of these options effectively dismisses the usefulness of term bond market instruments that are mostly a bet on the expected path of short term policy rates.

    So my question is – what is the usefulness of an “equilibrium real rate” concept when the Fed pulls the string on the only rate that really matters – nominal or real – and does so in response to a changing world where any imagined “equilibrium” rate must surely be as volatile as much as bond market prices themselves? Why does the MMM rationale for dismissing the usefulness of term government bonds not apply to the very concept of an “equilibrium” real rate as well?

  26. Dear Scott and JKH,

    I’m glad that we’ve crystallized the debate—it seems that the main dispute surrounds whether the Fed can meaningfully adjust the long-term real interest rate.

    So it becomes clear that the central bank controls both nominal and real short term rates (allowing for unconventional easing at the zero bound).

    Absolutely. So far we agree.

    MMT is quite clear in arguing its case from the nominal perspective. The central bank sets the short term nominal rate, and the market forms expectations about the path of that central bank process over time. The fact that market expectations fluctuate means that bond prices are volatile – much more volatile than the short term policy rate. Some may argue that other factors come into play, but these are ancillary –the expected path of short rates is the most rational persistent force driving government bond market yields. And the actual path of short rates set by the central bank wins out in the end.

    The long-term nominal interest rate can indeed be viewed as the market expectation of cumulative short-term rates set by the Fed. But this ability to set the long-term nominal interest rate through expectations of the policy rate doesn’t necessarily translate into an ability to set the long-term real interest rate.

    Consider the channel through which the Fed influences short-term real interest rates. At least as understood by most economists, it works like this:

    1. The Fed increases the short-term nominal rate through open market operations.

    2. In the long term, a lasting increase in the short-term nominal rate would result in an increase in the inflation rate. In reality, however, the inflation rate is determined through the adjustment of sticky prices (or maybe sticky information, if you like Mankiw and Reis’s model). Since an increase in the short-term nominal rate is generally not understood to mean a long-term shift to a higher inflation target, forward-looking price setters do not shift to an inflation trajectory that cancels the increase in the nominal rate.

    3. This results in an increase in the real interest rate. This increase is contractionary enough that it actually causes a short-term decrease in inflation (perhaps even deflation).

    Here there is a specific market friction accounting for why the Fed can defy supply and demand in the savings/investment market and change real interest rates in the short run: sticky prices. But sticky prices do not last forever, and the Fed’s ability to control real interest rates disappears over time.

    I’m curious: what market friction does MMT believe is sufficient to give monetary policy long term control of the real interest rate? You seem to need an entirely different model for understanding long-term savings and investment decisions, one that makes them mostly independent of the desire to hold real wealth over time, or the availability of investment projects paying real returns elsewhere in the economy. What is this model?

    Scott, you appear to be aiming at a partial articulation of this alternative model when you say:

    We see it, instead, within the context of a desire by the non-govt sector to net save out of current income relative to the govt’s deficit.

    But this isn’t really any different from mainstream macro. Where does the “desire of the non-govt sector to save out of current income” come from? To most economists, this desire is ultimately pinned down by consumers’ and businesses’ decisions to transfer their resources across time: maybe you’re saving from retirement, or maybe you’re borrowing to finance a new home. Why should these decisions be altered, in the long run, by purely nominal intervention from the Fed? If consumers are rational, it’s hard to see why this would happen. Maybe you think that consumers are irrational (and, of course, I’d agree to an extent), but there’s no reason why this irrationality would operate in a direction that you claim. In fact, I can just as easily see it going the other way: consumers suffering from money illusion might be pissed off at the low nominal rate and refuse to save, decreasing the supply of savings and causing an increase in the real interest rate.

    And the actual path of short rates set by the central bank wins out in the end. The fact that the Fed can engineer inverted yield curves in more conventional tightening patterns is demonstration of this. In that case, the Fed wins out over the “bond vigilantes” – through deserved, merciless short term tightening that eventually drives bond yields lower. Or, in unusual times like now, it wins by deserved dismissal of concerns from those who imagine inflation more than they can demonstrate its threat.

    Sure, the Fed can engineer an inverted yield curve over the span of a few years. (This is in line with the standard story about how the Fed’s ability to alter real rates is determined by price stickiness, which also operates on a span of somewhere between 6 months and a few years.) I don’t think that there is any empirical evidence, however, that the Fed can change the long-term real rate.

    When the Fed is expected to be up against the zero lower bound for an extended period, in a sense it has more power than usual to change medium-term real interest rates (though precisely because it’s facing the zero lower bound, its normal channel doesn’t work). This is because at the zero lower bound, the real interest rate is stuck above its equilibrium level—possibly for a while. If the Fed manages to raise the expected inflation rate, it can alleviate this problem and cause a meaningful decrease in medium-term real interest rates. But again, this is because the zero lower bound is an instance where equilibrium in the savings/investment market isn’t operational, and the forces of supply and demand want to push the real interest rate lower.

    Aside from the zero lower bound, however, a halfway competent central bank can prevent the savings/investment market from falling too far into disequilibrium, and this isn’t relevant.

    In sum, the Fed’s power to alter the short-term real interest rate does not extend to the long-term. Incidentally, one good way to see why this doesn’t follow is by looking at formal models where it provably does not. I would refer you to either Woodford’s book or Gali’s shorter and more approachable one. You might disagree with the models in these books, but they provide an excellent rejoinder to claims that the Fed’s short-term powers are also long-term ones—these are logically consistent environments in which the Fed has short-term influence but the long-term real interest rate is pinned down by other factors.

    • JKH

      I should confirm that I don’t speak directly on behalf of MMT – I just observe/opine on it, sometimes in the context of reactions to it. If I implied that MMT believes the existing short rate Fed policy mechanism is effective in accomplishing anything much useful in the long run, that is very probably not the case. So I doubt that MMT would say the central bank has “an ability to set the long-term real interest rate”, or at least express a related idea in quite those terms, but Scott is the better one to articulate that.

      “I don’t think that there is any empirical evidence, however, that the Fed can change the long-term real rate.”

      Tabula rasa – I’m deeply suspicious of the very idea of “the long term real rate”, possibly in epistemological terms, if I knew for sure what that meant. At the very least, I’m not sure what the relevance of “the” rate as such is.

      What is “the long term real rate” anyway? Is it some 10 year or 30 year or perpetual hypothetical Treasury bond real rate which the actual market continuously searches out like the truth? As an economic phenomenon, does it change in continuous time? If not, why not? And if it is important, what good is it as an ex ante rate in determining anything about “equilibrium” beyond the moment at hand? And doesn’t such an idea of equilibrium change continuously as well?

      The type of interest rate that I can see as more relevant is one that can be measured ex post, now in respect of the past period, or in the future in respect of the future past period. And this sort of interest rate hardly ever reflects a single rate that was born ex ante and stayed that way. It is generally some sort of accumulation of varying rates over time. The central bank policy rate would be a good elementary example of a single rate that changes over time (usually) before it arrives at a final accrued rate for the time period in question. A more complex generic example is a portfolio of rates that originated singularly ex ante, but collectively can only be measured ex post, due to the time lag of individual pricing contributions. Scott has done some interesting work looking at the cost of the US federal debt over history. That cost in fact reflects the historic accrual of a weighted average constellation of ex ante bond rates that happened to emerge at points in time in the past. The ex post nominal and real rates for that history are calculable. But how on earth do we relate that sort of empirical ex post rate effect to a concept of some ex ante “rate” that appears at a point in time right now? And why would an equilibrium real rate not be in a state of constant flux anyway?

      “In the long term, a lasting increase in the short-term nominal rate would result in an increase in the inflation rate.”

      I didn’t follow the related discussion very closely, but isn’t that from Kocherlakotaville?

      Agree that the zero bound case is interesting and somewhat separable.

  27. JKH, I think your epistemological concerns about the “long term real interest rate” are good ones. I don’t think that there is something clear called the “long term real interest rate” that economists can measure. At the same time, however, I think something like the long term real interest rate does exist, and that it’s an important price to consider.

    First of all, on a very meta level, there is always some ambiguity about what economists should do when there is no highly compelling empirical evidence—as is the case with long-term real interest rates, where for all the usual issues of endogeneity and confounding variables it is virtually impossible to econometrically single out the effect of (say) monetary policy.

    In these circumstances, I think it is important to have some baseline theoretical model of what is going on. This model does not need to be complicated, and it will certainly not be completely correct, but it is almost always better than throwing up one’s hands and falling back on either arbitrary intuition or sketchy empiricism.

    And in the case of real interest rates (and many other prices), there is a very simple framework to keep in mind. There are two sides: savers and debtors/entrepreneurs. Savers want to transfer wealth from today into the future (based on the trajectory of their real income and demand for consumption), while debtors want to do the opposite. Entrepreneurs (or, more accurately, all entities receiving funds from capital markets) want funding for projects that will yield real economic returns. There is no reason why any of these decisions, to a first approximation, should depend on purely nominal considerations like the trajectory of the price level. In the short run, yes, the Fed can make a difference (because monetary policy has real effects through price stickiness), but why should any of this matter in the long term? People have a certain level of real income today, a certain level of real consumption they’re targeting in retirement, and preferences about transforming the former into the latter.

    This is not a model that makes clear exactly what the real interest rate will be; it is, however, an intuitive framework for thinking about the issue that suggests that monetary policy should not have a large long-term impact.

    Moreover, it’s not as if the “real supply/demand” view of real interest rate determination is entirely lacking in empirical support. Countries with a long cultural tradition of saving, or a relatively high fraction of middle-aged workers saving for retirement, tend to have lower domestic real interest rates. Countries with some exogenous reason for a boom—e.g. a sudden increase in investment caused by newly discovered natural resources—tend to have higher real interest rates. Real interest rates in an open economy change according to capital inflows and outflows. Causal empirical observation suggests that a classic, “real supply and demand” view of the real interest rate has a lot going for it, even if monetary policy accounts for most of the short-term fluctuations in the rate.

    And even if you don’t accept the “long-term real interest rates are barely affected by monetary policy” baseline, I don’t see why the alternative is clearly that long-term real interest rates are pushed down by loose monetary policy. We can all agree that there are enormous differences between the short term and the long term, and as such it’s quite a leap from “lower nominal interest rates cause lower real rates in the short term” to “lower nominal interest rates cause lower real rates in the long term”. Maybe they cause higher real rates—without some kind of model, who can say?

    I didn’t follow the related discussion very closely, but isn’t that from Kocherlakotaville?

    I might do a post on this at some point, since the Kocherlakota episode was one of the most frustrating, everyone-talking-past-each-other online exchanges I’ve ever seen. To be completely honest, I think many of the parties involved (on both sides) revealed a depressing level of ignorance about how monetary policy works.

    If you believe that long term real interest rates are set by real factors (and that they are usually positive in the long run), Kocherlakota was right in a formal sense: a near-zero nominal interest rate is only consistent in the long run with deflation.

    But this did not make his comments relevant or insightful. First of all, in an operational sense lower long-term nominal interest rates are actually implemented by a higher nominal interest rate rule. (I know this sounds paradoxical, but stay with me.) Central banks around the world almost always use some kind of “Taylor rule”, in which interest rates rise more than one-for-one with inflation. Along the target inflation path, a 4% inflation target will result in nominal interest rates that are 2% higher than the nominal interest rates associated with a 2% inflation target. But conditional on a particular level of inflation—say, 4%—the higher inflation target will mean a lower nominal interest rate. For instance, if the Taylor rule coefficient on inflation is 1.5, and the underlying real rate is 2%, then conditional on 4% observed inflation, a 4% inflation target implies a nominal interest rate of 6%, while a 2% inflation target implies a nominal interest rate of 7%. This distinction is a mathematically simple but conceptually difficult subtlety of monetary policy, one that Kocherlakota seemed to miss.

    Second, and on a related note, in practice a decline in the fed funds rate almost always means an increase in inflation because because it not understood to mean a long-term shift in the target inflation rate. Thus price-setters do not lower their expected inflation rate, and in fact short-term inflation increases a bit from the demand pressure caused by lower real interest rates. This is made mathematically explicit by New Keynesian models like those in Gali’s book.

    Since the logic of this area is so twisted—there are all kinds of pitfalls like short term vs. long term, on-equilibrium-path vs. off-equilibrium-path, etc.—it’s easy to see why it gets such poor treatment in blogospheric debates.

    But the statement I made before…

    “In the long term, a lasting increase in the short-term nominal rate would result in an increase in the inflation rate.”

    … is still correct, in the sense that any equilibrium trajectory with a permanently higher nominal rate should (all else equal) lead to a permanently higher inflation rate.

  28. Ramanan

    SRW,

    Good points about nations such as Indonesia and … “So it’s unsurprising that there is a political struggle.”

    Also thanks for confirming the other day that you are a balance-of-payments worrywart. Its my favourite topic. The world used to be run in Keynesian principles in the “golden age” and suddenly in the 1970s, Keynesians found themselves cutoff with government policy. The usual constraints highlighted is what the Keynesians *already knew* about demand management but it was Nicholas Kaldor (who incidentally coined the phrased “convenience yield” and also argued that “the conventional and narrowly financial standards, as the true criterion of budgetary policy”) who later went to critique the Keynesians via the sectoral balances approach (though he was the face of Keynesianism in the UK). Even Nick Rowe observed the connection to Keynesianism in the UK in the 1970s recently in his blog.

    IMO, the balance of payments constraint is now hounding the US. It is not surprising to see US Treasury Secretary Tim Geithner continuously putting pressure on China and other Asian Tigers to do something about their currency.

    MMTers intuition doesn’t work for the US itself! Whatever rhetoric MMTers say about asset/liability, “money” is an asset without a liability in MMT.

    Harry Johnson once mentioned that Keynes did a huge disservice to Macreoeconomics by not starting with open economy macro and leaving it for end chapters in whatever he wrote.

  29. on the one hand giethner wants china’s currency to go up and the dollar down, on the other he’s worried they might dump their dollars and drive the dollar down.

    just one of many hopelessly confused mainstream discussions

    • Ramanan

      No its not confused. If the US’s net foreign assets situation keeps deteriorating, it increases the fragility of the dollar.

      On the other hand, asking China to revalue its currency, which amounts to depreciating the dollar is to reverse sectoral imbalances so that situation leading to a rapid decline of the dollar is avoided.

      Surprised this point is not appreciated.

      • so it’s ok to ‘work’ to get the dollar weaker for one reason and stronger for another…

      • Ramanan

        “so it’s ok to ‘work’ to get the dollar weaker for one reason and stronger for another…”

        Can’t figure what that exactly means. But “working” for a depreciation of the dollar puts the external imbalance of the United States on a more sustainable path. Reduces payments to foreigners and reflates the local economy because consumers buy more local products. More fiscal expansion can be carried out.

      • Tom Hickey

        What evidence is there that the USD getting more “fragile.” The dollar has been depreciating from about the time that the US was complaining it is overvalued, i.e., when the trade deficit ballooned. The dollar is not yet low enough as far the US is concerned. The “fragility” is net exporters freaking out that their US market might dry up.

    • Ramanan

      Let me offer a weak analogy.

      You think tax rate changes are sufficient to contain aggregate demand.

      For lets say, the government reduces taxes to boost demand. When the pressure on aggregate demand starts to build (and hence prices as well), it will have to increase tax rates. But tax rates increase prices! Isn’t that self-contradictory.

      Its not because the price increase due to tax increase is temporary.

      • Tom Hickey

        Rising prices and lower effective demand results in less sales, which leads to inventory build up, sending a signal to either cut back production or reduce prices and accept margin compression.

  30. all federal taxes is do is regulate aggregate demand.
    (apart from a few ‘special’ taxes targeted towards behavior modification)

    and how does raising personal tax rates (not that I’d propose doing that- I prefer a real estate tax) raise prices?

    • Ramanan

      Well there are all kinds of taxes and even firms pay taxes. If firms need to pay more taxes, they will increase prices.

      • Tom Hickey

        Only if they can. Right now some firms are suffering from margin compression because they cannot pass on rising commodity prices proportionally.

      • Ramanan

        “Only if they can. Right now some firms are suffering from margin compression because they cannot pass on rising commodity prices proportionally.”

        Talking of generic situations. At any rate, other nations have seen price rises of non-commodities due to rising commodity prices.

        To be more specific, increase in the VAT rate in the UK has seen price rises and affected headline inflation.

        To be clear on my point – a tax rate adjustment on personal income taxes to control aggregate demand cannot be achieved so easily. Consumers may demand that firms take up the burden of tax rises and in that case a higher tax on producers will lead to price rise.

  31. Tom Hickey

    Depends on what the US means by “strong dollar.” The US is continually saying that it wants a strong dollar while complaining that the dollar is overvalued. It’s pretty clear that “strong dollar” means low inflation in the US, and “overvalued” means that the US want to see further dollar deprecation relative to net exporters. The policy seems contradictory but it is not contradictory at all as I see it. This has been the US position for some time.

    • The US wants it both ways–they don’t want China or any other country to dump Tsy’s because that will depreciate the $, but they want China and any other country doing likewise to stop manipulating to allow the $ to depreciate and improve the trade balance. These two positions are inherently contradictory, obviously. That is Warren’s point.

      • Tom Hickey

        Scott, I think that there are two things going on. First the general condition of dollar overvaluation, which applies to countries other than China. Most of the net exporters are intervening to keep their currencies lower that would be otherwise.

        Secondly, the Chinese peg. The US wants China to either end the peg, which the US knows China will not do, or else drop restrictions on US exports and investment in China, which the US is pushing as realistic and China is resisting. I suspect that the US would gladly trade a slow pace of currency revaluation for a bigger share of the Chinese economy.

      • Ramanan

        These things can be phrased in different ways. Such as for a given trade balance, a stronger dollar is preferable or for a trade balance which is the United States’ favour, a stronger dollar is preferable. Thats more on the wishful thinking side.

        On the practical side of things, both the Federal Reserve and the US Treasury want the Asian Tigers to revalue their currencies – in particular China because China gains unfair trade advantage in trade by keeping its currency depreciated.

        Just because politicians are confused doesn’t mean asking China to revalue its currency is a confused strategy itself. Its hoped that the United States’ trade imbalance situation improves – consumers purchase more domestic products, entrepreneurs will produce more after seeing more demand for their products and more fiscal expansion can be carried out.

      • wh10

        If the end goal is currency devaluation, then I still don’t understand why the fear of tsy dumping is not contradictory. Thanks.

      • Ramanan

        “If the end goal is currency devaluation, then I still don’t understand why the fear of tsy dumping is not contradictory. Thanks.”

        Lets not make too much a baboon out of Tim Geithner.

        A controlled currency devaluation achieves a reversal of imbalances. At least it is hoped so. Smooth movements of the dollar alone can be helpful for the US but that is not under the control of the US Treasury.

        A sudden rapid movement of the dollar is unwanted.

        It is ridiculous to compare the two.

  32. Tom Hickey

    Matt R. I have a hard time figuring out how economist can seriously talk about the real interest rate, which presupposes knowing the inflation rate, when there is a huge pile of money bet on this in opposite direction and major figures on both sides arguing cogently for opposite conclusions. One of the points of contention among traders is whether the trend in the US is inflationary or deflationary, let alone with the present rate of inflation is and what futures rates are expected. Respected economists like Marc Faber are even warning about possible hyperinflation, while the Fed has clearly been fighting deflation.

    Here’s the point. If economists actually could know these things they would quit their day jobs, become traders, and retire early with a few billion dollars. I tend to believe what Warren Mosler says, since he as the money to show for making the right calls, against what economists who are making not very much given their years of education are saying.

    • I’m not saying that economists can predict the long-term equilibrium real interest rate—just that it exists.

      By the same token, any economist who told you that he could predict the path of oil spot market prices over the next 50 years would be crazy. That doesn’t mean that oil prices aren’t determined by supply and demand, or that government intervention into the market would have more than a transitory impact.

      • Max

        An alternative view is that stable equilibrium doesn’t exist without government intervention. Uncontrolled markets have multiple temporary equilibrium states and jump between them.

        In the case of a fiat money system, the government is in a powerful position to dictate returns since there is no physical supply constraint. So why should money earn a rent, like land does?

      • I’m not saying that economists can predict the long-term equilibrium real interest rate—just that it exists.

        Well, if you don’t have a coherent theory to explain past long-term “equilibrium real interest rates” and you also can’t predict the current “long-term interest equilibrium interest rate,” then what reason do you have to believe that is remotely scientific that an “equilibrium long-term interest rate” exists. You’ve accused MMT theorists of “creationism,” presumably because you believe that certain aspects of MMT theory are “metaphysical”, even “religious” in character, as is creationism. But what then of your belief that “equilibrium long-term interest rate” exists? isn’t that a “metaphysical,” even “quasi-religious” belief motivated by your seeming commitment to the neo-liberal economic paradigm with its unquestioning faith in the ultimately determinative influence over the international economic system of the bond markets? So, who’s the creationist now?

        More importantly, since your disagreement with MMT seems to be coming down to your belief in the existence of an “equilibrium real long-term interest rate,” shouldn’t you be trying to devise a crucial experiment or test of the theory that such a rate exists? And, if you can’t imagine how to do so, doesn’t your continued belief in its existence, come to down to a metaphysical belief without practical utility? And, if that’s so, if we were to reject MMT-based policies because of this belief of yours, wouldn’t we as Trader’s Crucible recently said in connection with his recent criticism of the IGBC, be:

        . . . little more than cave people sacrificing the lamb and burning it on a stone altar to hold back the wrath of the always angry gods for another year. . . . >

        You also said:

        By the same token, any economist who told you that he could predict the path of oil spot market prices over the next 50 years would be crazy. That doesn’t mean that oil prices aren’t determined by supply and demand, or that government intervention into the market would have more than a transitory impact.

        While this is true. it doesn’t mean that oil prices in the spot market are entirely determined by supply and demand either. While I’m no expert on the spot market, it appears to me that the Saudi monopolist sets the boundaries of prices on the spot market, and that the ling-term path of these prices will be determined very much by whether we choose to remain dependent on oil, and to the extent that we do where the Saudis decide to set the prices.

        Finally. as you can gather from my previous comments, I believe that your earlier comments that the MMT view asserting that ” control of the monetary base relieves the government of the need to satisfy a budget constraint is absurd, and akin to creationism,” was quite unreasonable and also irresponsible. I asked you earlier whether you still believed that after a lengthy exchange with MMT supporters. You replied by doubling down.

        Now there has been much more discussion of the differences between your views and MMT, and the result is that your contention that there is a GBC comes down to the truth of your belief that an “equilibrium real long-term interest rate” exists, and you have shown us that 1) you have no good theory accounting for the existence of such a rate, and 2) you cannot predict the long-term path of this so-called real interest rate constraint in order to test any theory you do have. So, since, at best, the validity of your own view that an IGBC derived from this constraint exists, is as much a hypothetical as the MMT theory that there is no such constraint, are you ready now to give up the “absurd” and “creationism” which were surely designed to marginalize MMT?

  33. letsgetitdone,

    I will be very busy for the next week or two will probably not be able to respond much more to comments in this space.

    First of all, when I talked about MMT as akin to “creationism”, I was making a statement about its status in the economics profession, not implying that it has metaphysical or religious elements. I think that this is amply clear from the context: “…since it has roughly the same stature among economists that creationism commands among biologists…”

    Second, let me observe that if MMT merely believes that the government has a great deal of power to alter the long-term real interest rate through monetary policy, then most of its statements about how the government can always meet its obligations by creating money are irrelevant. The Fed can keep nominal interest rates at a very low level while still holding a relatively small portion of federal debt in the form of money. Therefore, if the monetary authority’s ability to influence long-term interest rates is the only channel through which MMT addresses concerns about the government’s solvency, then all the discussion about how the government can simply pay its bills through crediting reserve accounts is unnecessary and even misleading: since money demand is quite elastic with respect to interest rates, maintaining nominal rates at a low level for the long term would necessitate only a small monetization of the debt, and would be entirely consistent with ordinary fiscal practice.

    Since MMT puts such an enormous emphasis on the government’s ability to pay by crediting reserve accounts, however, I am forced to conclude that interest rate policy is not the only channel through which MMT believes it can address fiscal issues. Instead, there is some kind of claim that investors are inherently far more likely to accept holding government liabilities in the form of money than in the form of bonds. I believe that this claim is extremely wrongheaded—and, at the very least, not given anything near the substantiating logic that it deserves.

    Third, I think that your approach to dealing with uncertainty in our understanding of the macroeconomy is at some level very strange. At best, you can argue that my working hypothesis of a “long term equilibrium real interest rate” is unsubstantiated. But you lack strong evidence for any alternative working model of long term real interest rates. If we follow your arguments to their logical conclusion, therefore, we should really throw up our hands and say that we don’t understand how financing costs vary with respect to monetary policy and debt issuance. Amid such uncertainty, however, we still have to make decisions about fiscal and monetary policy, which in the absence of one truly convincing theory must rely mainly on a mix of cautious incrementalism and empirical evidence.

    And as I survey it, the empirical evidence is not at all favorable to MMT. Many countries have tried to finance their debt in an appreciable way through money creation. To my knowledge, all such attempts have failed and led to severe inflation. More broadly, countries that have attempted to accumulate debt over a certain fraction of GDP (say, 100%-150%) have almost inevitably been forced to default and experienced some kind of economic crisis. The only exception is Japan—this against dozens (probably hundreds) of cases where governments with high debt burdens have either experienced severe inflation or default. If we are indeed uncertain about the theoretical mechanics of real interest rates and sovereign debt, any even remotely prudent approach to policy should take this empirical experience into account—and I think it is clear that this experience suggests the importance of fiscal caution. This is true even if my arguments about real interest rates are total bunk.

    Now, I know that MMTers would complain that all the countries that ran into hyperinflation and sovereign default simply weren’t doing it right (for whatever reason). That is legitimate: you may have a theory that suggests governments can use monetary policy to effectively negate their fiscal constraints, even if they haven’t successfully managed to do it so far. But given the history of economic disaster associated with such attempts, I think you should be very, very certain about your modeling framework before you can responsibly operationalize your theory and rely on money creation to address our budget concerns. You should have some chain of logic that makes you absolutely positive that your policy won’t run into the same disasters that have befallen countless nations throughout economic history.

    I see no such reason for confidence. Again, at best you have argued that my notion of an equilibrium real interest rate is theoretically and empirically questionable. You do not have a clear alternative model through which monetary policy can permanently lower real interest rates; instead, your claim seems to rely mostly on tenuous extrapolation from the observed short term effects of monetary stimulus. This is not, to say the least, a compelling basis for making policy decisions that have proven to be extraordinarily risky in the past.

    And there are strong arguments for why the Fed should not be able to change the long-term real rate of interest by an appreciable amount. Consider this: the Fed can induce massive changes in the nominal interest rate (on the order of, say, 5%) with a change in base money on the order of $100 billion. This is in a country where the balance sheet of the household sector alone has over $70 trillion in assets. Why should such a tiny intervention be able to move the real rate of interest at which such a massive pool of assets is priced? This cries out for some kind of explanation.

    And mainstream economics has an explanation. The fed funds rate is determined by the marginal value of transaction and liquidity services provided by base money, which must equal the yield spread between base money and otherwise equivalent assets that do not provide those services. At the margin, these transaction and liquidity services are not so important to the macroeconomy (though the inframarginal benefits are enormously important); thus a relatively small intervention relative to the macroeconomy can still have a large influence in this market.

    Since the market for the liquidity services of money is not very important at the margin, monetary policy cannot have much influence through this channel. However, money has another role that is much more important at the margin: it is used as the unit of account for virtually all transactions in the economy. A change in the expected path for the value of this unit of account has massive implications for all transactions and contracts whose nominal terms were set before the change. Thus a seemingly small intervention into the market for liquidity services provided by money can, due to money’s role as a unit of account, reshape the meaning of sticky nominal prices and contracts and have a significant economic influence.

    But the only reason that a shock to the unit of account has any influence is that prices and the terms of contracts do not immediately adjust. If they did instantly adjust (in some kind of frictionless economic nirvana), the unit of account would become an irrelevant sideshow to the more fundamental forces of supply and demand that determine economic outcomes. In the real world, prices and terms of contracts are usually set for a while, and until they are reset we get a real economic effect when the Fed effectively redefines the unit of account. But when we’re looking at timespans long enough to include many changes in nominal prices and rewrites of contracts, this effect is no longer operational: in the long run, when businesses will renegotiate their contracts to accommodate whatever change has occurred in the path of the unit of account, the Fed’s ability to influence nominal variables no longer has much of a real impact.

    This is a pretty convincing story. It’s not flawless, and there are some rough edges to work out, but the mainstream economic account of the real effects of monetary policy provides a surprisingly robust explanation for why tiny interventions by the Fed can have massive short-term impacts. Does MMT has some alternative story with comparable internal coherence and empirical support? Again, this a pretty tough hurdle for a theory to surmount: it is incredibly unintuitive that such a macroeconomically miniscule amount of base money creation can result in such impressive impacts. And if you’re going to be so confident in your theory that you’re willing to embark on a path that has crippled dozens of economies in the past, it better be good enough to put the mainstream account to shame.

    Finally, let me note that I do not believe that the government is entirely powerless to affect the real interest rate. Just as Saudi Arabia is a big influence in the oil market, the United States government is a very large presence in the market for investments, and it possesses enough market power to have some influence on real rates. But this does not give the US the power to borrow any amount it wants at any price it wants, any more than it gives Saudi Arabia the power to sell all its oil in the long term at the price of its choosing. (In fact, Saudi Arabia is quite constrained by the market: it hews so religiously to OPEC quantity targets because it knows that an massive spike in the price of oil will cause losses in the long term that outweigh the short-term benefits.) Both entities are powerful but ultimately subject to the even vaster forces of global supply and demand.

    • Max

      A mainstream economist discovers a flaw in the Euro system:

      http://www.voxeu.org/index.php?q=node/6484

      And here’s what Warren Mosler wrote about the Euro 10 years ago:

      http://www.epicoalition.org/docs/rites_of_passage.htm

      It’s difficult to square such prescience with the claim that MMTers are idiots who have been empirically refuted. It seems rather that the evidence is forcing the mainstream to rethink their theories.

      • This is very powerful evidence of your lack of knowledge about mainstream economics. In reality, many economists were extremely skeptical about the euro. There is an entire branch of economics called “optimal currency area” theory that has been around for at least half a century (there was even a Nobel Prize awarded for it), and it identifies several conditions that are important for an area to be suitable for a single currency — such as labor mobility and fiscal unification, which clearly did not apply in the case of the Euro.

        If you want to understand just how strong the mainstream skepticism of the Euro was, you should read this article from a few years ago, which was dedicated to mocking all the mainstream economists who questioned whether the Euro was a good idea.

      • Max

        Paul Krugman said it was “a paper I wish I had written (there is no higher praise).”

        I assumed he wouldn’t write that if it was a rehash of a well-worn mainstream Euro critique.

      • yes, many were skeptical from the start and remains so.
        and for a variety of reasons.
        and as the euro teeters, they all claim its for their reason.

        My point has been the optimal currency area thing is not the reason. In fact, Charles Goodhart, a central banker who more than understands monetary operations and that currencies are ‘tax driven’ (and who I respectfully do not agree with on many other issues) wrote about exactly that as well.

        Nor is the reason that the wrong initial exchange rates were selected.

        My reason, shared by some mainstream economists, is that the member nations entered the union and became ‘currency users’ with prior levels of debt incurred and ‘appropriate’ (in response to savings desires for their currencies) when they were ‘currency issuers.’ This put them in ponzi from day 1. Currency issuers are necessarily not (nominally) revenue constrained when servicing their debt as currency users are. It’s like the euro members turned themselves into US states, which are also currency users and revenue constrained.

        And note that California’s debt to gdp is only maybe 5%, and the other states relatively low as well, as compared to currency issuers. Markets just won’t let them get all that high. Luxembourg, for example, was never a currency issuer, and only had about 15% debt to gdp, while Belgium, a currency issuer, with a massive pension fund industry causing local incomes to go unspent, had a 125% debt to gdp ratio.

        Ponzi works just fine on the way up, it’s the way down where it tends to all comes apart.

    • Sorry to hear that you’ll be too busy to reply over the next couple of weeks, but thanks for your previous reply Matt. Here’s mine in dialog format.

      First of all, when I talked about MMT as akin to “creationism”, I was making a statement about its status in the economics profession, not implying that it has metaphysical or religious elements. I think that this is amply clear from the context: “…since it has roughly the same stature among economists that creationism commands among biologists…”

      Joe: Then it’s not your personal view that MMT is “akin to creationism”?

      Second, let me observe that if MMT merely believes that the government has a great deal of power to alter the long-term real interest rate through monetary policy, then most of its statements about how the government can always meet its obligations by creating money are irrelevant. . . .

      Joe: I really don’t know what you mean by this statement or why it might be irrelevant. Who cares what MMT believes? The issue is whether the MMT proposition that the observed long-term interest rate can be determined by Government action in economic systems with Governments sovereign in their own currencies is true or false? Also, who’s talking about the “long-term real interest rate?” You’re the one assuming that such an interest rate exists and that it is different from the observed interest rate. I’m not assuming the existence of any such hidden variable, or the existence on a long-term equilibrium in a fiat monetary system. Nor do I believe that you have any evidence that such a variable exists. That is one of the major points at issue between us. I’m saying that your belief in such a variable is religious in nature, and is a conjecture that you have admitted you cannot even test. So, if it’s not testable, then why bother us with it? Why should it make any difference in macro-economic outcomes? And, if it does make a difference, then why can’t you measure it?

      . . . The Fed can keep nominal interest rates at a very low level while still holding a relatively small portion of federal debt in the form of money. Therefore, if the monetary authority’s ability to influence long-term interest rates is the only channel through which MMT addresses concerns about the government’s solvency, then all the discussion about how the government can simply pay its bills through crediting reserve accounts is unnecessary and even misleading: since money demand is quite elastic with respect to interest rates, maintaining nominal rates at a low level for the long term would necessitate only a small monetization of the debt, and would be entirely consistent with ordinary fiscal practice.

      Joe: Who said anything about the Government’s ability to influence long-term interest rates being the only channel relevant to solvency? In writing the above critique, I think you’ve made the implicit claim that you’ve read enough MMT to have developed a pretty good understanding of what the MMT approach has to say. But this statement really makes me doubt that this is true. MMT’s claim about the Government’s basic economic solvency isn’t primarily based on its ability to control interest rates in the bond markets. While that ability is there, the primary influence channel relevant to solvency is the Government’s capability, if it so wills, to spend without issuing debt, thus ending the bond market in US currency entirely.

      In other words, it’s our currency, and our authority to appropriate spending. And anytime we like we can use our currency, and the Government’s authority to pay obligations incurred in that currency. MMT’s chief channel of influence isn’t monetary policy at all. It is fiscal policy. You would know that if you had read any significant part of the MMT literature.

      Since MMT puts such an enormous emphasis on the government’s ability to pay by crediting reserve accounts, however, I am forced to conclude that interest rate policy is not the only channel through which MMT believes it can address fiscal issues. Instead, there is some kind of claim that investors are inherently far more likely to accept holding government liabilities in the form of money than in the form of bonds. I believe that this claim is extremely wrongheaded—and, at the very least, not given anything near the substantiating logic that it deserves.

      Joe: There is no such claim. The only claim is that legal tender around here is US money. Since the US Government can spend/create money as needed to implement Congressional appropriations it can always but whatever goods and services are available for sale in US Dollars. Now people may prefer base money or they may prefer securities when they a choice between them. But when there are no securities because, for example the US Government has quit issuing them, then investors have little choice but to either buy real goods and services with their base money or lend it to others.

      Third, I think that your approach to dealing with uncertainty in our understanding of the macroeconomy is at some level very strange. At best, you can argue that my working hypothesis of a “long term equilibrium real interest rate” is unsubstantiated. . . .

      Joe: But that’s only part of what I argue, and it’s not the best part. More fundamentally, I’m arguing, and you’ve admitted, that you have no way of measuring a “long term equilibrium real interest rate,” and testing for the existence of such a construct. Now, I recognize that such a rate may exist, and that one day, someone may find a way to measure this construct and then test for its existence. Stranger things have frequently happened in the history of science. However, for the present, you are in the position of making a “metaphysical” posit, and so far, at least, don’t seem willing to grant that this posit may not exist. Its existence is a conjecture, and at this stage it is an unsubstantiated conjecture, in the same way that the existence of God as a causal force in the universe is an unsubstantiated conjecture. And, for my part, I think that this is the best that you can currently argue against MMT’s claim that a Government that is sovereign in its own currency can determine the interest rates it will pay on its securities, or even whether it will pay any interest at all.

      . . . But you lack strong evidence for any alternative working model of long term real interest rates. If we follow your arguments to their logical conclusion, therefore, we should really throw up our hands and say that we don’t understand how financing costs vary with respect to monetary policy and debt issuance. Amid such uncertainty, however, we still have to make decisions about fiscal and monetary policy, which in the absence of one truly convincing theory must rely mainly on a mix of cautious incrementalism and empirical evidence.

      Joe: Well first, I think you need to read Scott Fullwiler’s papers on how interest rates are determined in systems where the Government is sovereign in its own currency, as well some pieces by Forstater and Mosler and Bill Mitchell. Here are the links to Scott’s papers to start: http://www.cfeps.org/pubs/wp-pdf/WP53-Fullwiler.pdf and http://www.cfeps.org/pubs/wp-pdf/WP34-Fullwiler.pdf Mosler and Forstater is here: http://www.cfeps.org/pubs/wp-pdf/WP37-MoslerForstater.pdf Forstater and Mosler is here:
      http://moslereconomics.com/wp-content/graphs/2009/07/natural-rate-is-zero.PDF Bill Mitchell’s material is here: http://bilbo.economicoutlook.net/blog/ He has many posts explaining why the Government can control interest rates on securities, including material on “the natural rate of interest.” These references show that MMT does have models for explaining and conditionally predicting the long-run tendencies of interest rates associated with alternative Government policies.

      So, second, in light of MMT’s theoretical development and empirical evidence referred to in these papers, it’s just not the case that we have to rely on a policy of cautious incrementalism. Instead, I think we need to rely much more on crucial tests of MMT models and available empirical evidence that is relevant to MMT models.

      And as I survey it, the empirical evidence is not at all favorable to MMT. Many countries have tried to finance their debt in an appreciable way through money creation. To my knowledge, all such attempts have failed and led to severe inflation. More broadly, countries that have attempted to accumulate debt over a certain fraction of GDP (say, 100%-150%) have almost inevitably been forced to default and experienced some kind of economic crisis.

      Joe: I think that ‘evidence’ testing theoretical models has to be relevant to those models. You know if you attempt to test Galileo’s law of falling bodies using feathers in an atmosphere, and claim that the behavior of the feathers refutes Galileo’s law, you will be dead wrong. Because to perform that test on feathers you need a vacuum. Similarly, here, if you want to test MMT’s predictions about interest rates in economic systems having Governments sovereign in their own fiat currencies, then you have to use data taken from such systems, and exclude data from systems that don’t fit this condition. The data you’re referring to above applies to nations that aren’t sovereign in their own fiat currencies for the most part. When you look at economic systems since 1971, when the world went off the gold standard, and only at those that fit the sovereignty condition, then I think there is NO EVIDENCE anywhere refuting what MMT has to say about interest rates, or about demand-pull inflation. The empirical record on MMT is not mixed. You cannot find even one refutation when you take the currency sovereignty conditions into account.

      On the other hand, Japan, does fit the sovereignty condition, and it has done everything wrong from the neoliberal viewpoint of worry about runaway interest rates. With its debt-to-GDP ratio, it should be a hyper-inflating wreck right now according to neo-liberal theory. But obviously, the Japanese Central Bank is seeing to it that interest rates there are even lower than they are in the US. The Japanese case is a crucial test of MMT thinking and there’s no refutation of it at all.

      “The only exception is Japan—this against dozens (probably hundreds) of cases where governments with high debt burdens have either experienced severe inflation or default. . . . “

      Joe: That Japan is “an exception” is a very bad and careless argument in my view. As I say just above there’s not one case that fits the MMT definition of a Government sovereign in its own currency that has severe inflation or default due to economic necessity. Ironically, WW II Japan is a case where such a Government has gone into default, but this default, when Japan refused to repay its obligations to its enemies is clearly a case of a political rather than an economic default. Japan could have issued the currency necessary to repay its debts, but it chose not to for political reasons.

      . . . If we are indeed uncertain about the theoretical mechanics of real interest rates and sovereign debt, any even remotely prudent approach to policy should take this empirical experience into account—and I think it is clear that this experience suggests the importance of fiscal caution. This is true even if my arguments about real interest rates are total bunk.

      Joe: I have to say that I think this argument is a two-edged sword. First, you may be uncertain about the theoretical mechanics of real interest rates but we are not. We look at Japan and other nations sovereign in their own currencies, and we do not see a hint of a rise in core demand-pull inflation when there is an output gap. And we don’t see any reason to refrain from using expansive fiscal policy until we do see some evidence of demand-pull inflation. We don’t believe in the bogeyman under the bed that we can’t observe or measure. And we don’t believe that the suffering of poor and middle-class working people should be endured because the well-off, including professional economists have this unsubstantiated and theoretically incoherent fear that uncontrollable inflation might ensue.

      In fact, from our point of view, present economic policies in most nations are not only imprudent because they risk a severe double-dip recession, but are demonstrably and observably damaging to many millions of people who are not well-positioned to benefit from Government policies that find high unemployment tolerable because they lower wages and labor costs for businesses. So, in short, I deny that present fiscal policies are “cautious” and “prudent,” or consistent with empirical evidence. Instead, I think they are damaging, oppressive, cowardly, and ignorant, and that they misapply data and mis-characterize it as “evidence,” when it doesn’t meet the necessary criteria for such an interpretation.

      Now, I know that MMTers would complain that all the countries that ran into hyperinflation and sovereign default simply weren’t doing it right (for whatever reason). That is legitimate: you may have a theory that suggests governments can use monetary policy to effectively negate their fiscal constraints, even if they haven’t successfully managed to do it so far.

      Joe: Nope! We’re not saying they weren’t doing it right. We’re saying that the Governments that ran into hyper-inflation and had sovereign defaults weren’t sovereign in their own currencies, and that nations that are sovereign in this respect haven’t had hyper-inflation and defaults, even when they’ve had stupid economic policies. The MMT assessment of Japan, for example, is not all favorable. They may have shown that the Government can force low interest rates regardless of what the bond markets have to say about one’s debt-to-GDP ratio, or one’s “credit rating;” but they have also shown that if that Government won’t use fiscal policy to create full employment, it can easily condemn itself to lost decades and even lost generations, even if it is sovereign in its own currency.

      Further, MMT theory is not about using monetary policy to negate fiscal constraints. It is about freedom to use fiscal policy and evaluate the consequences of Government spending, without having to worry about solvency considerations.

      But given the history of economic disaster associated with such attempts, I think you should be very, very certain about your modeling framework before you can responsibly operationalize your theory and rely on money creation to address our budget concerns. You should have some chain of logic that makes you absolutely positive that your policy won’t run into the same disasters that have befallen countless nations throughout economic history.
      I see no such reason for confidence. . . .

      Joe: Matt, I don’t understand how you could undertake to write this post without doing more careful research into MMT. What did you do, read SRW’s post and Krugman’s critiques and conclude that you understood MMT? What else did you read? Look, this is a serious school of thought with a very large body of literature behind it, especially in recent years. It is facing a competing paradigm that has failed abjectly to predict the Great Recession, and which is now failing to create prosperity during the recovery period because it is so paralyzed with fear over inflation that it is repeating the policies of Herbert Hoover.

      Other than among professional economists steeped in the neo-liberal paradigm, no confidence exists among people in the predictions and projections of conventional economics. When you say be cautious, you are saying let’s do what we were doing before, be more careful and prudent about it, but use the same underlying theory. We are saying that this is not prudent, because the neo-liberal theory with its undue faith in the market’s self-regulating characteristics has been very decisively refuted again, as it was in the late 1920s and early 30s. It is just wrong, and it is not prudent, bt risky and irresponsible, to keep using it as you are doing.

      Instead, it is much more prudent to use theory that is unrefuted to devise our economic strategies. I submit to you that MMT is as yet unrefuted; and that the the data you’ve pointed to as refuting it is not evidence against MMT, because it’s not about Governments sovereign in their own currencies. Before you say that there’s not much evidence corroborating MMT, I think you need to be much more well read in the MMT literature than you are.

      There’s all manner of data that should not have turned out in the way that it did if MMT were not true. Japan’s interest rates are only one case. Argentina’s recovery after its default using MMT-like policies is another. The occurrence of the US recession at the beginning of the Bush Administration, after years of surplus, is another. The history of depressions or recessions following on periods of surplus in this country is yet another. Even events like the hyper-inflations in Weimar and Zimbabwe, make sense from the viewpoint of MMT. MMT could have predicted both if it were used in those situations.

      Recently, there’s all kinds of evidence that the sector financial balance model of MMT, an accounting identity, is useful for describing transaction flows among the Government, private and external sectors of the economy. If you read the MMT blogs you will soon find that daily events occur which make a great deal of sense from the viewpoint of the MMT model, but no sense from a neo-liberal point of view at all.

      Again, at best you have argued that my notion of an equilibrium real interest rate is theoretically and empirically questionable. You do not have a clear alternative model through which monetary policy can permanently lower real interest rates; instead, your claim seems to rely mostly on tenuous extrapolation from the observed short term effects of monetary stimulus. This is not, to say the least, a compelling basis for making policy decisions that have proven to be extraordinarily risky in the past.

      Joe: Again, this is just not right. We do have coherent theories about this. Please read the references I suggested, and my previous comments. Based on our models there’s every reason to believe that the Government can control long-term interest rates, and can keep them at or near zero if it wants to. On the other hand, there is absolutely no reason to believe that interest costs projected by CBO, OMB, and other organization over the next 10, 20 or 50 years are even remotely correct. These are simply pie in the sky projections, without any basis, that are being used to scare people into austerity programs that are creating suffering in the here and now.

      And there are strong arguments for why the Fed should not be able to change the long-term real rate of interest by an appreciable amount. Consider this: the Fed can induce massive changes in the nominal interest rate (on the order of, say, 5%) with a change in base money on the order of $100 billion. This is in a country where the balance sheet of the household sector alone has over $70 trillion in assets. Why should such a tiny intervention be able to move the real rate of interest at which such a massive pool of assets is priced? This cries out for some kind of explanation.

      And mainstream economics has an explanation. The fed funds rate is determined by the marginal value of transaction and liquidity services provided by base money, which must equal the yield spread between base money and otherwise equivalent assets that do not provide those services. At the margin, these transaction and liquidity services are not so important to the macroeconomy (though the inframarginal benefits are enormously important); thus a relatively small intervention relative to the macroeconomy can still have a large influence in this market.

      Since the market for the liquidity services of money is not very important at the margin, monetary policy cannot have much influence through this channel. However, money has another role that is much more important at the margin: it is used as the unit of account for virtually all transactions in the economy. A change in the expected path for the value of this unit of account has massive implications for all transactions and contracts whose nominal terms were set before the change. Thus a seemingly small intervention into the market for liquidity services provided by money can, due to money’s role as a unit of account, reshape the meaning of sticky nominal prices and contracts and have a significant economic influence.

      But the only reason that a shock to the unit of account has any influence is that prices and the terms of contracts do not immediately adjust. If they did instantly adjust (in some kind of frictionless economic nirvana), the unit of account would become an irrelevant sideshow to the more fundamental forces of supply and demand that determine economic outcomes. In the real world, prices and terms of contracts are usually set for a while, and until they are reset we get a real economic effect when the Fed effectively redefines the unit of account. But when we’re looking at timespans long enough to include many changes in nominal prices and rewrites of contracts, this effect is no longer operational: in the long run, when businesses will renegotiate their contracts to accommodate whatever change has occurred in the path of the unit of account, the Fed’s ability to influence nominal variables no longer has much of a real impact.

      This is a pretty convincing story. It’s not flawless, and there are some rough edges to work out, but the mainstream economic account of the real effects of monetary policy provides a surprisingly robust explanation for why tiny interventions by the Fed can have massive short-term impacts. Does MMT has some alternative story with comparable internal coherence and empirical support? Again, this a pretty tough hurdle for a theory to surmount: it is incredibly unintuitive that such a macroeconomically miniscule amount of base money creation can result in such impressive impacts. And if you’re going to be so confident in your theory that you’re willing to embark on a path that has crippled dozens of economies in the past, it better be good enough to put the mainstream account to shame.

      Joe: I’m sorry Matt, but I’m not clear on why this example is relevant to our discussion. I agree that the Fed can have an enormous influence on interest rates. In fact, MMT research has shown that the Government including the Fed and Treasury can determine the FFR, and, in fact, can drive this rate down to zero and also drive the rates on securities down to very close to zero. Neo-liberal economics cannot explain this. It cannot explain Japan’s interest rates except by introducing ad hoc paradigm-conserving hypotheses. It cannot explain why long-term interest rates in the United States are as low as they are.

      I don’t think the MMT story in answer to the question you posed above would be very different from the narrative you gave. But I think that MMTers would also say, “who cares.” This isn’t about monetary policy. It’s about fiscal policy, realizing the full potential of the economy, and getting full employment. As we do that, increased Federal spending will have a deflationary effect on interest rates with a natural tendency toward zero. At that point, the Government will have to decide whether it wants rates at or near zero. If it doesn’t, it will have to pay higher rates of interest on reserves in order to hit a higher FFR target. If the effect of fiscal policy creates demand-pull inflation, then the Government will start destroying base money through increased taxation and cease its deficit spending. It will trade off the negative aspects of inflation against other more positive consequences of its spending.

      Finally, let me note that I do not believe that the government is entirely powerless to affect the real interest rate. Just as Saudi Arabia is a big influence in the oil market, the United States government is a very large presence in the market for investments, and it possesses enough market power to have some influence on real rates. But this does not give the US the power to borrow any amount it wants at any price it wants, any more than it gives Saudi Arabia the power to sell all its oil in the long term at the price of its choosing. (In fact, Saudi Arabia is quite constrained by the market: it hews so religiously to OPEC quantity targets because it knows that an massive spike in the price of oil will cause losses in the long term that outweigh the short-term benefits.) Both entities are powerful but ultimately subject to the even vaster forces of global supply and demand.

      Joe: I’m glad you don’t believe that the Government of the US is entirely powerless to affect the real interest rate, but that is still a far cry from saying that the Government of the United States can set the interest rate it desires. So, I think we still disagree. I’ve indicated above why MMT thinks the Government can control the interest rate so I think that any further discussion should be on the details of the work from Scott, Warren, and Bill that I cited earlier. I think the reference to the Saudis as a case is interesting because by looking at the Saudi case we’re beginning to get at the idea that the United States is a monopolist when it comes to its currency, and also when it comes to its own securities. As a monopolist it can set interest rates on its own securities. Some of the issues connected to this capability are discussed in the references I gave but Warren Mosler’s paper here: http://moslereconomics.com/mandatory-readings/soft-currency-economics/ is also very relevant as is Pavlina Tcherneva’s http://www.epicoalition.org/docs/Pavlina_2007.pdf

      I urge that you read these pieces and learn more about the implications of the Government’s position as the currency monopolist. That’s a very important aspect of MMT.

      • Deus-DJ

        “Joe: I have to say that I think this argument is a two-edged sword. First, you may be uncertain about the theoretical mechanics of real interest rates but we are not. We look at Japan and other nations sovereign in their own currencies, and we do not see a hint of a rise in core demand-pull inflation when there is an output gap. And we don’t see any reason to refrain from using expansive fiscal policy until we do see some evidence of demand-pull inflation. We don’t believe in the bogeyman under the bed that we can’t observe or measure. And we don’t believe that the suffering of poor and middle-class working people should be endured because the well-off, including professional economists have this unsubstantiated and theoretically incoherent fear that uncontrollable inflation might ensue.

        In fact, from our point of view, present economic policies in most nations are not only imprudent because they risk a severe double-dip recession, but are demonstrably and observably damaging to many millions of people who are not well-positioned to benefit from Government policies that find high unemployment tolerable because they lower wages and labor costs for businesses. So, in short, I deny that present fiscal policies are “cautious” and “prudent,” or consistent with empirical evidence. Instead, I think they are damaging, oppressive, cowardly, and ignorant, and that they misapply data and mis-characterize it as “evidence,” when it doesn’t meet the necessary criteria for such an interpretation.”

        This is precise, and correct. The neoclassical view of the economy is dangerous not only for this reason, but for the associative reason that it promotes financial instability through deregulatory policies. So they shoot us in the hand and in the foot…and they deny that they’ve done either one. Of course both of these contribute to the large inequality thats propped up, the largest since the great depression. So really they shoot the economy in the hand, foot, and the chest.

        Most neoclassical economists know better than to actually argue publicly against a viewpoint that has consistently attacked them since the days of Thorstein Veblen, but not this guy, apparently.

  34. Adam

    Too bad this reasonable argument started out with a invidious taunt.

    Also, don’t biologists usually *win* their arguments against Creationists?

    • They do. But the question is: who is the real creationist here. If my argument is correct that Matt’s commitment to market determination of long-term interest rates is in no way empirically based, then it seems to me that the show is on the other foot.

  35. beowulf

    Second, let me observe that if MMT merely believes that the government has a great deal of power to alter the long-term real interest rate through monetary policy, then most of its statements about how the government can always meet its obligations by creating money are irrelevant… [then downthread]… And there are strong arguments for why the Fed should not be able to change the long-term real rate of interest by an appreciable amount.

    I’m reminded of the joke where the preacher asks a man if he believed in baptism by submersion, “Believe in it? Hell, I’ve seen it done!”
    To reiterate what I wrote above “until the Tsy-Fed Accord of 1951, short term rates were capped at 0.375%, long term rates capped at 2.5% (the Fed buying up any Treasuries the market didn’t)”. And for 10 years, from 1942 to 1951, long term rates did not rise above 2.5% thanks to Fed market tap operations.

    So you have it backwards, its only because the govt can create money that it has the absolute power to alter the long term interest rate… and a whole lot more besides (e.g. prime rate, required reserves, margin or down payment requirements, transaction fees on anything that moves through the FRS).

    I do take Warren’s point, if the FFR falls to zero, why even bother issuing long bonds? it’d makes life simpler if Tsy only sold T-bills– which are currently yielding 0.03%. At that rate, on Matt’s $20 trillion public debt, instead of $800 billion a year (at 4%), debt service would be around $6 billion a year. We’ll need to dramatically cut taxes and/or raise spending to make up for deflationary effects of net interest payments dropping by 99%.
    And that too would require a long thread to hash out. :o)

    • beowulf

      Sorry, first paragraph above should be in quotes.

      • Thank you beowulf, a cogent point indeed. BTW, another post on the debt ceiling and coin seigniorage with a link to your post at Correntewire, FDL, ourfuture.org, and scheduled for tomorrow at DailyKos’s Money and Public purpose MMT group.

  36. Deus-DJ

    I find it funny that you, Dr. Rognlie, have been debunked on every claim you’ve made against MMT, and you refuse to admit that you were wrong. Rather than apologizing you instead say “I’m glad WE’VE crystallized the debate…” Seriously? Not to mention that you’ve ignored every post that has debunked any certain claim you’ve made about MMT.

    Now that you finally realize where the real point of emphasis is, on the long term real interest rate, you still refuse to let go of any of your polemics against MMT(and you’ve said as to why: because its not popular it must be wrong, hence your comparison of it to creationism). There are two distinct policy paths that occur as a result of your perspective and that of MMT. What yours suggest is that deregulation and free and open markets are best. MMT suggests that, as long as we’re not at full capacity that it is ok for the government to spend and that the deficit should have no bearing on any discussion of whether or not we should have full employment. It’s obvious the consequences: your ideology created modern finance, and you continue to support the structure that exists(not necessarily the big banks, but policies that led to the big banks). I would recommend you read a formal rebuttal to your claims here:

    http://nakedkeynesianism.blogspot.com/2011/05/monetization-of-debt-what-does-it-do.html

    • Thanks for your previous reinforcement of my long comment Deus-DJ. I swear if I read another word about “equilibrium” I’ll scream. Neo-classical economists seem to think that social systems (of which economic systems are but a sub-type) are like Newtonian chemical and physical systems. Most of them never seem to have heard about complexity science, or chaos theory, or any of the newer outlooks people use to try to understand human interactions.

      Generally, I’ve appreciated Matt’s civility during this long discussion. But as someone else remarked, the initial attempt to marginalize MMT by characterizing it as “absurd” and “akin to creationism,” wasn’t civil discourse. You and I have both tried to get Matt to back off his labeling of MMT in this way, in light of the way this discussion has proceeded, since, at a minimum, it amply demonstrates that MMT supporters have very strong arguments to offer against Matt’s own theories and formulations.

      These arguments, founded in the MMT perspective, have clearly given Matt enough trouble in exchanges that it should be apparent to any neutral observer that whether MMT is true or not, it is certainly not either “absurd” or “creationist.” Yet Matt will not admit to his error in this characterization.

      I’m one of those who think that serious people interested in the science of economics should be willing to admit error when it is plain that they have made one. I’m not saying that Matt is plainly in error in positing a long-term real equilibrium interest rate as a constraint on Government spending; though I have argued that this belief is plainly highly conjectural. But I do think that most of the commenters here have clearly shown that MMT is neither “absurd” nor “akin to creationism.” These characterizations are not only uncivil, but they are a plain error, and I, like you, am still awaiting Matt’s owning up to this error.

  37. Pingback: Our Story so far… « The Traders Crucible

  38. Pingback: It’s my fault Matt Rognlie misunderstands MMT « The Traders Crucible

  39. I think we all agree that MMT is only partly characterized by a claim that the government can bring down its long-term costs of financing by using monetary policy to lower interest rates. Let me first address MMT’s more central claim, which is that since the government can always create money to pay for its expenditures, there is no binding “government budget constraint”.

    Here’s the problem: why should individuals be more willing to accept holding government debt in the form of money rather than bonds? They are both government liabilities. If you think that the government can fund itself by printing massive amounts of money, you have to assume that the public will be okay with holding that money as a store of value—which might be true to a point, but is similarly true with bonds.

    Admittedly, there are some important differences between bonds and money. If someone doesn’t want to hold a government bond, she can sell it in exchange for money; it is easy to see how a mismatch in supply and demand of government bonds might have severe consequences. If someone doesn’t want to hold so much money, the same option doesn’t exist; instead, she has to spend that money on goods and services. But this doesn’t mean that excess money printing isn’t problematic; the effect isn’t as immediate as a run on government bonds, but if everyone wants to adjust their money holdings downward and does so by buying goods and services, there will be inflationary pressure—possibly very severe inflationary pressure, depending on the mismatch between the amount of money in circulation and the willingness of individuals to hold it.

    At some level, you all seem to acknowledge this: you say that there will not be inflation when there is an output gap. I agree. Since excess real money balances cause inflation through the channel of excess demand, it is impossible for a demand-driven output gap to coexist with inflationary pressure from debt monetization. If this is the only claim that MMT is trying to make, then I’m fine with it.

    But you seem to be making the much stronger claim that the government’s budget constraint is not an issue, regardless of whether there is an output gap. After all, the government spends for many reasons entirely unrelated to cyclical stabilization. What happens if the government runs up too much debt when there is not an output gap, and the private sector becomes unwilling to hold that debt in any form, money or bonds? There will be massive inflation. Even if the government runs up debt during a recession, it will need to keep that debt in some form after the recession. I see no way that MMT gets around the fact that the private sector has to be willing to hold government debt, regardless of what form it comes in.

    Honestly, you even seem to admit this:

    If the effect of fiscal policy creates demand-pull inflation, then the Government will start destroying base money through increased taxation and cease its deficit spending. It will trade off the negative aspects of inflation against other more positive consequences of its spending.

    Okay! So there is an effective constraint on government spending: if issuing debt in the form of money leads to inflation, you’ll need to pull money out of the system, and to do so you need taxation. My claim is that this is a situation that will inevitably occur if the government runs up too much debt; thus the ability to tax is a constraint on government spending.

    • TC

      Hi Matt,

      I am not one of the big shots in MMT. But I don’t think MMT would disagree with much of what you say here.

      This is a post on the Solvency vs. Debasement that I wrote, and there is a bunch more at the blog.

      Of course we can have inflation. And the level of inflation is THE soft constraint on spending. It isn’t a hard constraint in that we cannot spend more – we can – but rather a soft constraint in “is this extra inflation worth it?” The answer to this question can be “No.”

      http://traderscrucible.com/2011/03/29/solvency-and-value-insolvency-and-debasement/

      Before you say “no Ponzi”:

      http://traderscrucible.com/2011/04/29/chapter-2-in-which-the-traders-crucible-slays-the-intertemporal-government-budget-constraint-and-mr-rowe-demonstrates-his-worth/

      I know that this isn’t published in a journal. But I think it is an important way to think about some of the premises of this section of the MMT vs. most economics debate. Can we observe this stuff in real time? If not, how can we be expected to make rational economic calculations about it?

    • wh10

      Matt, thank you for the continued discussion and interest.

      “But you seem to be making the much stronger claim that the government’s budget constraint is not an issue, regardless of whether there is an output gap.”

      Ok, let’s align on the definition of “issue.”

      MMT says it’s not “an issue” from the perspective of *ability to spend.* From an operational perspective, there is nothing to stop the government from spending itself into oblivion. Yes, the consequences may be inflationary or hyperinflationary, but there is no literal “constraint” on its *ability to do so.*

      MMT says it *is* an issue from the perspective of *inflation.* As you have already recognized above, MMT firmly states government spending becomes an inflationary issue once the output gap is closed. It’s MMT basics that inflation *is* the bogey.

      “Honestly, you even seem to admit this:
      If the effect of fiscal policy creates demand-pull inflation, then the Government will start destroying base money through increased taxation and cease its deficit spending. It will trade off the negative aspects of inflation against other more positive consequences of its spending.”

      Matt, I think you’re trying to catch MMT in an “aha!” moment of weakness, but MMT openly and transparently recognizes everything in this quote. This is not an admission of theoretical flaw. It’s MMT fundamentals.

      “Okay! So there is an effective constraint on government spending: if issuing debt in the form of money leads to inflation, you’ll need to pull money out of the system, and to do so you need taxation. My claim is that this is a situation that will inevitably occur if the government runs up too much debt; thus the ability to tax is a constraint on government spending.”

      We’re back to the definition of “issue” and “constraint” from above. There is no literal constraint, but there are potential consequences. Your claim is MMT’s claim if we define “runs up too much” as beyond the output gap, which you seem to be doing as well. And again, the whole taxation to reduce AD thing is MMT basics.

      Does this make sense?

    • I say it this way.

      The govt levies a tax
      The private sector offers real goods and services in exchange for the sum of the funds needed to pay the tax, plus any funds it desires to net save.

      If the non govt sectors don’t want to net save anything at all, the govt will not be able to deficit spend without simply driving up prices without getting more real goods and services.

      So then the question is whether offering tsy securities/offering a higher term structure of rates/alters savings desires. And in what direction.

      I see two channels- the interest income channel and the cost channel

      First, for every dollar borrowed there is a dollar saved (loans create deposits, etc) so apart from govt. securities altering rates simply shifts income between savers and borrowers. And which might have the higher propensity to consume has been nearly impossible to discover.

      Second, the govt is a net payer of interest to the non govt sectors.

      So the Fed hiking rates/tsy driving up longer rates adds income and net financial assets to the non govt sectors. Seems to me that’s like a bumper crop, and makes dollars less valuable, but that’s just my take. And so it’s no surprise to me that nations with CB’s setting higher rates tend to have more aggregate demand and higher rates of inflation (however defined).
      But I could be wrong on this. The propensities could indeed be very different to the extent higher rates are ‘deflationary’- just that I don’t ever recall seeing it happen that way.

      Third, interest rates are part of the cost of holding inventory and the cost of production, and in today’s world no small part of pricing happens to be cost push (yes, lots of imperfect competition, firms driving for market share, etc), so that makes rate hikes inflationary as well.

      To your point, it’s not impossible that higher term structures of rates actually add to ‘inflation’ and lower rates help keep it down.

    • Matt, thanks for returning. You say:

      Here’s the problem: why should individuals be more willing to accept holding government debt in the form of money rather than bonds? They are both government liabilities. If you think that the government can fund itself by printing massive amounts of money, you have to assume that the public will be okay with holding that money as a store of value—which might be true to a point, but is similarly true with bonds.

      I think you’re still missing the MMT point. A Government sovereign in its own currency never “funds” its spending. It just spends. The public always needs some money, because people can only pay their tax liability with dollars. Bonds won’t do. Gold won’t do. Oil won’t do. Only dollars will do.

      If the public finds dollars less attractive because the Government has spent too much, the effect of that is to drive prices up. It is not to make the Government run out of money, or to impose a budget constraint, unless Congress loses its nerve and won’t appropriate funds needed by Government. If this occurs we are talking about a political constraint and not one of financial capability. Remember the MMT claim is that a Government sovereign in its own currency can’t become insolvent. It says and claims nothing about political error or stupidity.

      In you latest formulation, you’ve pointed out that too much spending can lead to inflation and that the desire to avoid or contain inflation represents a budgetary constraint on the Government. and you say:

      Okay! So there is an effective constraint on government spending: if issuing debt in the form of money leads to inflation, you’ll need to pull money out of the system, and to do so you need taxation. My claim is that this is a situation that will inevitably occur if the government runs up too much debt; thus the ability to tax is a constraint on government spending.

      I think you’re really fuzzing up the issues here. MMT economists agree with what you’ve said just above including the idea that money is debt, and that too much spending will cause inflation. For example: Randy Wray’s piece here:

      http://www.newdeal20.org/2010/07/20/deficits-do-matter-but-not-the-way-you-think-15355/

      is very clear about this.

      However, in making the above argument, you’re shifting the ground of what people are arguing about. The politicians, Pete Peterson, the Hooverians, the deficit hawks, and even the deficit doves, are all making the claim that the Government can run out of money and be forced into insolvency.

      The argument is that deficits that are too large, and debt-to-GDP ratios that pass a particular level, can cause the bond markets to raise interest rates or even decide not to buy Treasury securities. So, they claim, the Government won’t be able to borrow money to meet its obligations, and the Government will literally run out of money for lack of lenders.

      It is this argument that MMT claims is false for Governments sovereign in their own currency. Your reply above doesn’t touch the issue of whether Government faces this kind of insolvency constraint. MMT’s position is that it doesn’t.

      There is no level of the national debt, or the deficit, or the debt-to-GDP ratio that can render the Government insolvent or destroy its capacity to spend the money it needs to spend to buy goods and services for sale in Dollars. So, the President is wrong when he says we can run out of money, as is Pete Peterson, or anyone else who says so.

      Now, what you’ve just said above is what MMT people are continuously saying: specifically that Government spending has to be evaluated in light of its anticipated consequences. One of these possible consequences is inflation, and, of course, when Government spending is likely to create demand-pull inflation it is to be avoided unless avoiding it results in consequences that are even worse than inflation. So yes, the Government has spending constraints; but these are not related to solvency or an incapacity to spend. They are related to the effects of contemplated spending and whether these effects fulfill the public purpose or have negative consequences we need to avoid.

      In short, I don’t know any MMTers who don’t agree that good economic policy will sometimes cause us to constrain spending. But these constraints should be self-imposed based on our estimates of what the effects of spending are likely to be. They will not be constraints based on magic levels of national debt numbers beyond which we must not go, or magic debt-to-GDP number limits compiled from comparing large numbers of nations with grossly different currency regimes on all fours in an attempt to induce specious generalizations about all economic systems.

  40. Tom Hickey

    If this is the only claim that MMT is trying to make, then I’m fine with it.”

    A chief claim, but not the only claim.

    “But you seem to be making the much stronger claim that the government’s budget constraint is not an issue, regardless of whether there is an output gap.”

    See Scott Fullwiler’s paper on the IGBC, “Interest rates and fiscal sustainability,” for a definitive answer.

    http://www.cfeps.org/pubs/wp-pdf/WP53-Fullwiler.pdf

    From the concluding remarks: “The sustainability of fiscal policy as determined via the orthodox IGBC framework is irrelevant for understanding the workings of a modern money economy.”

  41. the fed could announce tomorrow that it had, say, a 1% bid for tsy secs of any maturity, for any quantity, and the tsy curve wouldn’t trade any higher than 1%.

    the consequences would be those of the lower interest rates, and, if Japan is any indication, those consequences aren’t the hyper inflationary nightmare some imagine. But I agree it remains to be seen.

    however, there are no consequences per se for whatever quantity of tsy secs the fed actually purchased, again, beyond the consequences of the lower interest rates.

    it’s about price, and not quantity, when it comes to the fed

    • Again, I am arguing that the government cannot use monetary policy to manipulate real interest rates in the long term. I fully acknowledge that the government can set nominal rates at any level it likes; my point is that very low nominal interest rates are only consistent in the long term with disinflation or deflation, because the real interest rate is determined by non-monetary considerations.

      By citing Japan, you are doing a very good job of making my case. Japan has experienced long-term zero nominal interest rates and persistent deflation, which is exactly what I said is necessary for a policy of keeping nominal rates at zero to be possible for a sustained period of time. And deflation doesn’t seem to be working out very well for them…

      • Tom Hickey

        MMT’s point with respect to Japan, and Richard Koo’s point, too, is that the way to combat deflation resulting from a “balance sheet recession” (Koo) in which the desire to save/delever results in demand leakage is by running deficits to accomodate the domestic private sector’s saving desire taking into consideration the external position, too. Japan is a net exporter, which redues demand leakage to saving, but not enough to accomodate the need. Larger deficits are required, and Japan will remain under deflationary pressure until balance sheets are repaired. Japan’s large amount of outstanding tsys has neither resulted in inflation nor raised the interest rate, even with rating agency downgrades. There is no IGBC problem either. All of which MMT explains.

        The same is true of the US. The US will experience what Japan has until balance sheet repair is complete. That will take some time at the current rate unless government accommodates this with deficits.

      • So seems you are saying ‘inflation’ (however defined) is not a function of nominal interest rates.

        I tend to agree, though I do think given current institutional structure low nominal rates are ‘deflationary’ via income interest channels on the demand side, and costs of holding inventory and costs of production on the supply side. So it’s no surprise to me that 0 interest rates in Japan and now the US haven’t ‘worked’ as anticipated.

        However, for me the reduced demand and positive supply side effects are a good thing, as they mean taxes can be that much lower for a given sized govt. and given credit conditions.

        Also, seems to me there is no ‘long term’ rate of inflation, but only current rates of inflation (again, however defined).

        What we do have are longer term ‘inflation’ indexed securities that imply longer term ‘real’ rates based on whatever they are indexed to. However those markets are subject to disproportionate ‘technicals’ rather than actual inflation expectations, again, whatever that means and implies.

        Last, on a gold standard, for example, the long term rate of inflation would imply the price of gold changing accordingly. That is, if gold were 35/oz today and 35/oz in all the fwd markets that would imply 0 inflation as defined.

        With fiat currency/floating exchange rates the notion of ‘inflation’ is a lot more problematic at best.

        But again, I suspect we all agree on these things, and that Austrians should have no issues with ‘mmt’/no issues with anything I’ve written.

        Not to say that they won’t have issues with what many ‘mmt’ers’ might/do state, just as I do.

      • beowulf

        “Again, I am arguing that the government cannot use monetary policy to manipulate real interest rates in the long term. I fully acknowledge that the government can set nominal rates at any level it likes”
        Real Interest Rate = Nominal Interest Rate – Inflation, so there’s no disagreement here. The question then is how to control inflation? In a word (or two), fiscal policy. MMT agrees with Wynne Godley and Marc Lavoie that, “if the fiscal stance is not set in the appropriate fashion that is, at a well-defined level and growth rate—then full employment and low inflation will not be achieved in a sustainable way. We also show that fiscal policy on its own could achieve both full employment and a target rate of inflation.”

        http://tinyurl.com/3jlb385

        “Japan has experienced long-term zero nominal interest rates and persistent deflation, which is exactly what I said is necessary for a policy of keeping nominal rates at zero to be possible for a sustained period of time.”
        Remember the Tsy-Fed Un-Accord of 2011 I postulated above? Eliminating your need to pay $800 billion in annual debt service would rather dramatically expand the output gap unless filled with spending hikes or tax cuts. If its not filled, then near-0 interest rates would be deflationary and your economy would look like… Japan. Japan’s problem (if there is a problem, more on that in a sec) isn’t their monetary policy, its their inadequate fiscal policy.

        Of course I’m oversimplifying. Up until their recent tragedies, Japan’s economy was doing just fine. Toyko-based journalist Eamonn Fingleton has long made the point that Japan intentionally poor-mouths its economy to the West:
        “For Japanese officials this is a matter of national security: Japan’s previous image as the juggernaut of world trade had proved dangerously counter-productive by the late 1980s. Since then the myth of an often absurdly dysfunctional Japan has been assiduously projected into the Western press. The result is that Western policymakers who once feared Japanese economic expansionism switched to pitying the “basket case”.”

        http://www.fingleton.net/?p=919

        At the height of Japan’s worst recession since the 1970s, their unemployment rate peaked at 5.2% (in this country, the Fed would consider that dangerously close to “full employment”) As of early 2011 (per this St. Louis Fed chart), Japan had it down to 4.3%. Deflation is working out for them just fine.

  42. Pingback: Quora

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s