Thinking the unthinkable – a proposal for demand management

In this post I am going to propose a counter-cyclical policy that I think should in principle find support from economists of many different schools, including post-Keynesian economists as well as market monetarists.

I believe that current monetary policy is convoluted and indirect, especially when interest rates are near zero. It operates through the banking sector, and the behaviour of banks is a highly nebulous and complex subject in which there is a lot of disagreement. It is also not understood well by the general public and some even find the current primary dealer system alarming. The transmission mechanism from unconventional policies like quantitative easing is contested, especially by those from the endogenous-money camp. While market monetarists have some highly complex inter-temporal hot potato, portfolio adjustment and signalling transmission mechanism, many have difficulty finding plausibility or potency in this story, and it goes far beyond conventional money quantity effects, hence my description of it as indirect.

What I propose is extremely simple and not especially novel (although almost as an unspoken rule very rarely discussed): monetary policy should bypass the banks. Instead, a fully independent and non-partisan committee (like the FOMC, perhaps even the exact same people) should decide on a specific quantity of government spending that is to be financed by brand new money rather than by debt or by taxes (before you tell me how insane, immoral, counterfeit or idiotic this is, please read my response to some of the criticisms at the bottom) in accordance with some rule or target. This would be done by adding new money to the treasury’s account at the central bank (for you ‘spend first’ or ‘keystrokes’ MMTists, see the criticism section below). This in principle should not find objection, I believe, from market monetarists; after-all, monetarist models do not depend on or have any specific role for bank behaviour (as people like Steve Keen like to point out), it is still at its core just adjusting the quantity of money.

On the other hand it is far more direct, there would no longer be endless debates about the transmission mechanism: government spending is widely dispersed and pays the salaries of people across the country and of all classes. This doesn’t inherently require or encourage statist or anti-market policies either, this new money could simply finance a reduction in sales tax for instance, which would have the added benefit of lowering supply side inflation to counter-act any demand side inflation created by quantity effects. It also does not relinquish traditional monetary policy, an increase in the quantity of new money supplied to the government would inherently decrease the amount of new bonds supplied, which in turn would lower their yield and vice versa for a reduction in new money: hence this new ‘expansionary’ or ‘contractionary’ monetary policy has the same effects as before on interest rates. And even if it doesn’t, I see no reason traditional open market operations could not be conducted in addition to this policy.

Another benefit of this policy is that an increase in the money supply isn’t inherently caused by an increase in credit, thus discussion about unsustainable credit booms or fiat ponzi schemes where credit must be extended forever to pay-off the interest of the old credit would be greatly diminished. This would also be enormously helpful to those who worry about high household debt, this new money could be used to finance principal reductions and all kinds of other policies aimed at reducing indebtedness.

This is also entirely consistent with the rules based proposals advocated by Scott Sumner; the aspect of market monetarism that has much more bipartisan support (Michael Sankowski of the Monetary Realist camp for instance has said he is in favour of NGDP targeting). The quantity of new money, as before, could be set in accordance with a path for the inflation level or nominal gross domestic product growth. This time however there would be no doubt as to whether the target could be met or not, which is why I really think its something market monetarists should support.

Of course, one criticism I would expect to see, especially from monetarists, is that this policy is too radical and unnecessary, ‘we don’t need to change the demand management system since the current system is fine for that’. On the contrary, as I explained before I think monetary policy is highly convoluted and I would argue not optimal; Occam’s razor, I believe, would favour my proposal instead, which is far more simple in not relying on bank behaviour or credit, in essence changing policy structure such that it actually behaves more closely to monetarist models.

Now let me explain what this isn’t. This is not ‘printing our way to prosperity’, this is purely increasing demand, it is not a supply side policy. I agree with people like Cullen Roche and really most economists in that long run growth relies entirely on productivity and innovation (I would add institutional quality to that as well), not the quantity of money. This policy is not aimed at making us all more productive, instead it is merely a policy aimed at stabilising fluctuations in demand.

Of course there are serious criticisms that could be levied at this proposal, it is not perfect, I will analyse a few here:

People have non-linear, unpredictable or volatile demand for money (hyperinflation): This argument, favoured by Austrians, is that a policy like this would remove any ambiguity about what modern money really is (i.e. ‘nothing’, ‘toilet paper’, ‘not backed by anything’), and thus people, rationally or irrationally, will reject the currency, leading to hyperinflation which cannot be controlled merely by adjusting the quantity of money. While I can sympathise with this argument, it’s really not clear that this is actually the case. According to Bill Mitchell, the demand for money is not as simple as conventional wisdom suggests, fiat money still has utility for people because it can, for instance, extinguish tax liabilities. Others, like John Kay or Cullen Roche emphasize money as a social convention, useful in as much as other people also find it useful. Furthermore, I don’t think it is unclear to people what modern money really is regardless. Any statistics of M2 or the monetary base will show you that money has increased massively without being backed by anything and without causing hyperinflation. Suffice to say, I find the idea of people spontaneously rejecting the dollar, or the pound, not very plausible.

Moral hazard: perhaps a stronger argument, also favoured by Austrians. The government could easily get addicted to this free money, and this soft budget constraint (of which there is a large and interesting literature, useful for analysing centrally planned economies and their failures) would cause the government to become progressively more inefficient due to the incentive to cut down on wasteful costs becoming more and more diminished. This argument is a persuasive and important one, however I’m not sure if my policy would actually change anything, as Cullen Roche shows: the way treasury auctions are set up with primary dealers in countries which are sovereign in their currency (as opposed to say Euro-zone countries) makes the idea that the government would ever have much trouble finding funds or would have to suffer high interest rates due to investor risk premiums nearly impossible. In this case, there is already moral hazard and I don’t see how my policy has changed anything. You could argue that high quantities of debt in themselves are simply scary, which might put off government spending too much money, but it seems pretty clear to me in the face of ever increasing public debt that governments, despite their rhetoric, don’t really care that much.

Unnecessary, government spending is just key-strokes, they don’t have to worry about how much money is in their account: this would be from MMTists. First, I agree that the only thing preventing the government behaving in this way is self-imposed legal constraints. On the other hand, procuring funds is how governments today actually work, as Cullen Roche, former MMTer who started his own school after finding MMT too detached from the real world (ironically) explains:

“More recently, MMTers have started using two conflicting descriptions of monetary options.  They now use a “general case” and a “specific” case to explain why the government “spends first” in their general case and actually procures funds first in the specific case.  This is an evolution in MMT ideas that obscures and hides the fact that MMT’s long-held positions on “operational realities” are in fact shifting and/or misleading….

…To argue that the government does not procure funds to spend is a semantic debate because the government must always procure funds to ensure that the private sector allows the government to deal in the social construct (money) of the people (government of course being a construct of the people inherently making money a creature of the people and not just the state as MMT says)…”

It is also in our interests to make sure new outside money is created by an impartial and non partisan or perhaps rules based process, rather than entirely subject to the whims of short term thinking politicians, I hope I can find some agreement with MMT there.

Demand is not deficient, we are always at full employment: used by some RBCists and Austrians. While I disagree, obviously, this simply suggests that we should never need to manage demand, it doesn’t suggest  that this would be a bad policy at doing so should we need to.

In sum, I believe this policy to be simple with very few complications (far less compared to current demand management policy), no doubts about its effectiveness at what it does (increasing/decreasing demand, with no pretense about it increasing productivity or long run potential growth), and that suffers from criticisms which I believe are not strong enough to shoot this policy down.

Update: one more issue I forgot to address

Infrequency of budget decisions: this is something market monetarists are also likely to raise as a criticism, pointing to the advantages central banks have in being able to react more immediately to changing conditions. The remedy is simple, every quarter the committee can announce a quantity of money it projects will need to be provided for public expenditure during the next fiscal year, until private sector forecasts of NGDP or inflation (for instance) for that year are on track, which in turn is likely to affect growth in the current year by altering expectations. As market monetarists say, much of monetary policy is essentially signalling and expectations.

Even then, it is possible that some tax decisions could be decided quarterly instead of annually, although depending on the tax this could create some uncertainty. You could also set up independently financed government institutions, such as public investment banks, that could receive government support decided on a quarterly basis. I’m sure we could find plenty of schemes with quarterly or semi-annual expenditure decisions should expectations not be enough, so I don’t find that criticism too worrisome.

Update 2:
Steve Randy Waldman, in an email, expressing support for the idea points out that this policy shares some similarities with his own and Haito Zhang’s policy proposals, in that  “they share with your proposal a preference for fiscal spending or “helicopter money” rather than bank-mediated leverage as a macro policy instrument.” Steve’s proposal is even more radical than mine, not only bypassing banks but the government too, directly depositing money to individual bank accounts (with perhaps some lottery mechanism to prevent moral hazard), while Haito’s is an interesting democratised fiscal expenditure/investment program combined with an NGDP target although is not too specific on the financing. Getting political support for a proposal like mine would be extremely hard, but I think Steve’s proposal would probably be even harder. I’d argue that Hatio’s & Steve’ s solution are unnecessary while my proposal is simpler and logistically easier to administer, but they are interesting and innovative approaches none the less.

  1. anon said:

    Your proposal is not wrong per se, but there is no real difference between “giving brand-new money to the government” and “buying government bonds via QE operations and rolling them over indefinitely”. The latter also allows for more flexibility, especially when contractionary policy is needed to rein in excess NGDP. Under your proposal, the government would have to sell new government bonds and destroy just enough money to meet the target, which is weird and not something it (or the Fed) is set up to do.

    • Britonomist said:

      There is a difference; first of all in general, financing through the bond market creates a liability for the government, even if the fed buys the bonds back from the market, there is still a liability to the fed. The Fed could allow a default on that debt, but there is much uncertainty about whether the fed will even do that, and on top of that there is definitely uncertainty about whether additional QE will happen in the future at all, I find it difficult to believe that the uncertain prospect of future QE creates relief for the financing of government expenditure. The government cannot make spending decisions on the basis of possible future QE as the decision to do QE is made after expenditure has been decided, under my scheme the government can be certain that some of its financing will come from new money, there is no uncertainty and no liability, which relieves pressure for deficit spending. Further still there is uncertainty regarding whether these money injections currently are permanent or whether it will be unwound in the future, with many in the market essentially expecting and unwinding, thus somewhat ‘sterilizing’ the monetary injections. This would not happen under my proposal.

      Secondly, the current system just sucks money out of the banks for financing followed by an issuing of new money by the fed that just goes back to the banks with a tiny bit of interest, it’s an asset swap and means the creation of ‘net financial assets’ as MMTers might call it is much smaller; under normal times this means that monetary expansion is inherently debt expansion (which can be problematic), under the ZLB it is unclear if this new money goes anywhere at all. Under my scheme all new money just goes straight to the government and is spent into the economy in a way that does not necessitate more debt (although of course more debt will likely be expanded as more money enters the system, I’m not denying that), further there is most certainly more ‘net financial assets’ in the economy since new money has been spent into the economy without liability and without removing assets from the economy.

      As for contractionary policy, I’m not sure if it would ever get to the stage where the Fed would have to ‘destroy money’ rather than just simply stop issuing new money, but even then I see no reason why this couldn’t be a simple task. Obviously the current structure of the government/central bank is not set up for this, none of what I suggest is, that doesn’t mean the system can’t be changed. There are also other conventional ways that monetary policy could be contracted, for instance there could be a hike in reserve requirements or an increase in the rate the central bank lends to other banks.

      • anon said:

        If the QE is rolled over indefinitely, then that’s the a “permanent money injection”, and it creates no liability to the government. Yes, there is uncertainty about the extent of current or future QE. Some of that uncertainty can indeed be removed by committing to a policy regime such as NGDPLT, but there will always be some residual uncertainty depending on future money demand. The central bank cannot remedy this residual uncertainty, nor should it. If it were large enough to matter (it probably isn’t: consider the irrelevance of seigniorage revenue for modern governments) governments could hedge it by issuing securities linked to the amount of monetary base.

  2. W. Peden said:

    I could endorse this, if only to build a broader coalition for NGDP targeting. However, if I understand you correctly, you overstate the degree to which discretionary action is necessary, even from an MMT perspective-

    “This doesn’t inherently require or encourage statist or anti-market policies either, this new money could simply finance a reduction in sales tax for instance, which would have the added benefit of lowering supply side inflation to counter-act any demand side inflation created by quantity effects.”

    – that’s not bad, as far as it goes, but it goes too far. Instead, the core of your idea-

    “Instead, a fully independent and non-partisan committee (like the FOMC, perhaps even the exact same people) should decide on a specific quantity of government spending that is to be financed by brand new money rather than by debt or by taxes [snip] in accordance with some rule or target.”

    – can be accomodated simply through a funding policy: when NGDP is above a target range, expenditure is overfunded; when NGDP is below a target range, expenditure is underfunded. Any additional tethering of demand to keep NGDP within the prescribed range could be accomodated through OMOs (with or without a central bank, in the sense of an institution acting as lender-of-last-resort).

    Overfunding the PSBR was used as an instrument in the 1980s in Britain with significant success, although its distortionary effects on financial markets apparentely made it unsustainable as a permanent strategy (the same is true perhaps of QE). Underfunding operates on the exact same identity-

    The broad money supply = PSBR – debt sales to the non-bank public + bank lending to the private sector + net external financing

    No increased expenditure or reduced taxation is therefore necessary under almost any conceivable scenario. The Treasury can simply borrow more from banks, creating more new broad money, or spend “printed money”. That eliminates any ZLB liquidity trap scenarios, leaving only Keynes’s scenario where an increase in broad money does not lower interest rates, which only works in models where the only substitutes for money are bonds i.e. not the real world. Discretionary action, beyond funding policy, would be necessary only if underfunding the entire PSBR was insufficient, which seems fairly implausible, especially given the existence of automatic stabilisers which boost the PSBR in a depression.

    So this would imply policy that is liquidity trap proof (so the MMT/PK crowd should like it) and involves a very bare minimum of discretionary action (so those of us in the classical liberal/monetarist crowd can like it).

    • W. Peden, I agree, this doesn’t require any discretionary action at all. I mentioned the breathing room for discretion to show that a policy like this would easily help with AD even if you have an extremely pessimistic view about the potency of new money being issued, for all others though discretion is unnecessary. Scott Sumner said the same thing to me, this was my reply:

      “Scott, this isn’t fiscal policy, it doesn’t imply a change in government expenditure, it is merely increasing the money supply, governments may not change their expenditure just because there is more money, they may instead keep their expenditure constant but borrow less (probably the most likely outcome). The result: more money directly injected into the economy with no change in expenditure, the decision to do fiscal policy (changing expenditure or taxation) is still a decision by governments, not central banks. It is true that my policy would allow more breathing room for increased expenditure, but as a commenter points out, QE does too.”

      • W. Peden said:

        I see; so we’re in agreement.

        The point about QE making room for extra expenditure is a very good one.

  3. Hepion said:

    I can give you some commentary from MMT viewpoint

    1) All money is debt. There is no debt-free money. When government issues money it goes into debt.

    2) In a present system, when government is sovereign in it’s money, its debt is unlike any other debt in the economy. The method of going into debt is by issuing money. Bond sales are done afterwards to conduct monetary policy. Not as a funding mechanism. Bonds are sold as much as is needed, not less, not more, than what is needed to allow the CB to hit the overnight interest rate target. See professor Wray’s answer here:

    3) Do we need more unelected technocrats undermining democracy? No, we don’t. We need more powerful automatic stabilizers to act countercyclically stabilizing economic output.

  4. Hepion said:

    Cullen Roche has received money from Pete G. Peterson and has become a policy hack. I would not listen to him. Besides, hes just a blogger that learnt everything from Warren Mosler &co. It makes more sense to trust folks who have done actual research to these issues, no?

    There is this argument that some law supposedly prevent US government from having an overdraft in it’s accout at the Fed, but nobody cannot find the passage in law that prevents it. It seems more like long run practise that is often told as if it were a law that limits governments ability to spend. US. constituon says that even questioning the validity of US debts is unconstitutional. Fed refusing to pay debts on time as intructed by treasury would break the constitution. Besides it would plunge whole monetary system and therefore whole economy into chaos. If there is one purpose for which Fed exist it is to provide stability. It is what is strifes for in every action. It would take really lunatic leadership from the Fed to refuse to make these payments when instructed by treasury, if they even have that operational capability in practise. Massice scale economic crisis would get politicians quicky change the arrangements anyway. Imaginary constraint leading to imaginary problems leading to imaginary solution anyway.

    This is a fools argument.

  5. Britonomist said:

    @Hepion: To be honest, I don’t quite think you’re representing the MMT position quite accurately, or are presenting a misleading account of it from my understanding, but thank you for your input.

    I don’t care where Cullen received money from, as far as I’m aware it is (or was) his job to look after peoples’ money, so that is uninteresting and uncontroversial. I don’t agree with 100% of what he says, but most of his claims seem to be well researched and evidence based.

    • That’s a wetoolh-ught-lut answer to a challenging question

  6. Nathanael said:

    Steve Randy Waldmann’s proposal would actually be trivial to administer: cut a check to every person registered with Social Security.

    As for *political* difficulty, that is in the realm of political science. The primary political obstructions to sane policy right now are (1) plutocrats with psychopathic levels of short-term thinking who want to hoard all the money, (2) lunatics with theological beliefs which mean that they want to “punish” poor people, (3) theocrats who are actually trying to eliminate democratically elected government as we know it, (4) people who just want to ‘make the black President fail’. The four groups are not exclusive, and form the three major voting blocs composing the Republican Party. I don’t see that any plan which materially benefits the general population will get the support of groups (1) or (2). While (3) and (4) would happily support such a plan *after* their preferred ruler is installed into power (look at the checks cut by G W Bush), they won’t support the *current* government doing anything helpful.

    So the political problem is unsolvable until the Republican Party shrinks into irrelevance. After it’s done so, Waldmann’s proposal may be as politically viable as yours. (After all, Bush pretty much implemented it.)

Leave a Reply

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

You are commenting using your 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

%d bloggers like this: