Simon Wren-Lewis: Why Is It Taboo?
Professor Simon Wren-Lewis asks,
If Quantitative Easing (QE), why not helicopter money? We know helicopter money is much more effective at stimulating demand. Helicopter money is a form of what economists call money financed fiscal stimulus (MFFS). In their current formulation independent central banks (ICB) rule out MFFS, because the institution that can do the stimulus (the government) is not allowed to cooperate on this with the institution that creates money (the ICB). In a world where governments - through ignorance or design - obsess about deficits when they should not, it turns out that MFFS or helicopter money is all we have left to prevent large negative demand shocks leading to deep and prolonged recessions. So why is it taboo?(Also: please see the comments on this article at Mike Norman economics to see Tom Hickey's take on this.)
I would argue that this subject is "taboo" because legally the central bank has no fiscal powers, and the central bankers probably have no desire to get involved with it. The flow of research dollars and prestige for mainstream monetary economics is largely under the control of central banks, and so the research community is not going to rock the boat. You can add to this the extreme distaste of discretionary fiscal policy that was pushed by free market economists, which intellectually captured the mainstream.
But there are some good reasons to pay little attention to "helicopter money", even if you are sympathetic to discretionary fiscal policy.
- There is no good way of implementing the programme.
- The concept of "money-financed fiscal stimulus" makes no sense, either according to modern mainstream economics, or Modern Monetary Theory (MMT).
I cover these points in turn.
No Good Way To Implement The Programme
The important thing to remember is that all spending has to be authorised by the legislative body (at least in political systems that are descended from the British parliamentary system). The central bank is allowed to undertake certain types of transactions independently (such as repos and open market operations), but it cannot just spend money on random goods and services just because it is "independent". But we could imagine that the legislature could delegate limited spending powers for emergency counter-cyclical purposes.
Even if this could be passed through the opposition of fiscal conservatives, there would still be huge implementation issues. The central bank would need to have a "neutral" or "fair" means of sending money to households. The problem is that there is no mechanism currently in place that would allow them to do so - unless an income guarantee was implemented.
Let us say that we want the central bank to send to $100 to "all citizens". How does it do that? The government has voter lists, but those are not continuously up-to-date. If the payment is done via the tax system, such as by changing tax withheld, it will only affect those who are paying taxes in this manner.
Creating a new bureaucracy to administer a programme that might make a payment once a decade would be a spectacular waste of resources. It would have to be piggybacked on top of an existing system that deals financially with the bulk of households:
- tax withholding (which misses many households);
- a new programme, such as an income guarantee.
Given that the central bank has no good way of calibrating such a programme, I would prefer that this counter-cyclical fiscal policy be done via the automatic stabilisers. It could either be beefed up unemployment insurance, or a Job Guarantee. The first advantage is that the spending is going to where it most needed: households without jobs. The second advantage would be that spending is automatic, and so would not depend upon the forecasting accuracy of the central bank.
(Update: added this paragraph as a clarification.) Of course, this still leaves discretionary fiscal policy under the control of politicians. This will always be with us (as long as we have properly functioning representative governments, which is unlike the situation in Greece right now). And discretionary policy of this type was successfully applied during the peak of the financial crisis, even by nominally "anti-Keynesian" conservative governments. The mistaken policy tightening occurred later, and it was discretionary. As long you have realistic expectations (the business cycle will not be abolished), discretionary policy of this form is likely to be somewhat helpful. But we cannot generalise about the effectiveness of this policy, as it depends upon the quality and attitude of the elected government. To put it into economist jargon, there is no well defined "reaction function" for policy of this type.
(Update: added this paragraph as a clarification.) Of course, this still leaves discretionary fiscal policy under the control of politicians. This will always be with us (as long as we have properly functioning representative governments, which is unlike the situation in Greece right now). And discretionary policy of this type was successfully applied during the peak of the financial crisis, even by nominally "anti-Keynesian" conservative governments. The mistaken policy tightening occurred later, and it was discretionary. As long you have realistic expectations (the business cycle will not be abolished), discretionary policy of this form is likely to be somewhat helpful. But we cannot generalise about the effectiveness of this policy, as it depends upon the quality and attitude of the elected government. To put it into economist jargon, there is no well defined "reaction function" for policy of this type.
Money-Financed Fiscal Stimulus Makes No Sense
Modern Monetary Theory economists have launched a linguistic war on the word "financed" with respect to central governments. And they have a point, considering how its use leads to nonsensical ideas like "money-financed fiscal stimulus". I will explain why the term makes little theoretical sense, even if we accept the semantics.
All modern models of government finance will include an accounting identity for one period that looks like:
(Government Deficit) = (Increase in the monetary base) + (Increase in government debt).
(Technically, this is an approximation. This is a deficit defined in terms of cash accounting, and does not include non-cash items. Additionally, since Treasury Bills are issued at a discount, the second term should read "money raised through debt sales".) This is the non-controversial component of the inter-temporal governmental budget constraint.
This gives rise to a great deal of mythology about how deficits are "financed". How much is "financed" by debt, and how much is "financed" by money issuance?
This whole concept is nonsense. Government liabilities consist of money, and bonds, which are forward money. The private sector will arrange its holdings of government liabilities based upon things like the demand for currency, the need for required reserves in the banking system, and interest rates. The central bank has to buy/sell government bonds in the required quantities to hit its interest rate target.
The true (one period) accounting identity is:
(Government deficit) = (Increase of government liabilities).
Afterwards, the private sector adjusts its portfolio weightings of money and bonds based upon monetary policy settings and their preferences. Changes in bond and money holdings are entirely the result of those preferences.
Once again, this description holds for both Modern Monetary Theory and mainstream Dynamic Stochastic General Equilibrium models. This may not apply to older heuristics like the IS/LM model, but that just reflects their poor modelling of stock-flow relationships within a dynamic economy.
The only way of guaranteeing some spending will be "money financed" is if it is allocated to entities that hold the bulk of their financial assets in the form of currency (notes and bills). In the real world, this is either foreigners or people in the underground economy, such as mobsters. My suspicion is that the optics of creating a programme to allow the central bank to send suitcases of cash to criminals would be pretty bad.
What about the zero bound?
If rates are at zero, and stay there forever, money and government debt are indistinguishable. This is what Warren Mosler has been advocating. But in such a case, as long as people do not believe that interest rates will go negative, there is no reason to issue bonds. It is immaterial how deficits are "financed" in such a world.
But if there is any chance that interest rates will become positive, the central bank may be forced to sell government bonds from its balance sheet in order to keep interest rates at its target. Therefore, even if the deficit is initially "money financed", the government will still end up with extra debt outstanding once interest rates rise.
It's About Fooling The Gullible
The only reason to discuss "money-financed fiscal easing" is to fool people who do not understand fiscal policy. By making it look "interest free", it is supposedly better and more sustainable. These sorts of semantic games were useful for things like the Social Security (government pension) system; by pretending it is not a "pay-as-you-go" transfer system, it became politically untouchable.
But I doubt that linguistic games would work in this case. The average person on the street would reasonably ask why certain expenditures are "money financed" when others are not? People know that money is fungible (although they might not use the word "fungible"). And fiscal conservatives hate discretionary fiscal policy. They will oppose the programme regardless of how it is semantically packaged.
(c) Brian Romanchuk 2015
“There is no good way of implementing the programme.” Yes there is. See bottom of p.10 to 12 here:
ReplyDeletehttp://www.positivemoney.org.uk/wp-content/uploads/2010/11/NEF-Southampton-Positive-Money-ICB-Submission.pdf
“The only way of guaranteeing some spending will be "money financed" is if it is allocated to entities that hold the bulk of their financial assets in the form of currency (notes and bills). In the real world, this is either foreigners or people in the underground economy, such as mobsters.”
Complete nonsense: new money gets channelled into every household pocket if the new money is spent on the usual public sector stuff (roads, education, etc). The alternative (for those who are politically right of centre) is to cut taxes, in which case (again) money is channelled into ordinary household pockets (i.e. Main Street).
You mean normal fiscal policy. Or are you still of the mind that the Lords of Finance should have a veto over the elected house on quantity because they are so much better and wiser than mere politicians.
DeleteThe House of Lords in the UK lost its veto over budgets in 1910 as a constitutional matter. That has never been reversed. I'm not up on Canadian budget etiquette, but my understanding is that the Senate will not reject finance bills and that finance bills can only be proposed in the Commons.
There is no good reason why those constitutional arrangement should be subverted - other than by those who believe they are cleverer than the rest of us.
Neil,
DeleteI’m not specifically advocating the Positive Money system there. I’m simply making the point, which Brian Romanchuck doesn’t seem to get, that it’s perfectly feasible to have a system in which the state creates money and spends it (and/or cuts taxes) when stimulus is needed. Indeed, most MMTers agree with that, far as I can see. Certainly there’s plenty of articles on Mike Norman’s site which claim that the state should create “fiat” and spend it in a recession.
As to who decides (1) the aggregate increase in spending under such a system (politicians or technocrats) and (2) who decides how much public spending to allocate to education, roads or whatever, those are separate questions.
Ralph,
DeleteI realise I should have added "under the current institutional framework". I do not know enough about positive money to say whether it would match, but I did give an example which would allow it to work - an income guarantee.
As for currency holding, all the statistics say that "normal" households carry little. The bulk is held overseas and criminals. The bulk of US currency issued is $100 bills, but these are rarely used in legal transactions. Even if there was a safe way to distribute currency, it would be returned immediately to the banking system, where it would be effectively exchanged for government debt.
Most Keynesians prefer discretionary fiscal policy, and that applies to post-Keynesians (which includes MMT). I do not think it is possible to completely eliminate recessions with such policy. My view is that the forecasting accuracy of central banks is weak, and so there would be little advantage to give them added fiscal powers. This makes me somewhat of an outlier as a post-Keynesian.
DeleteDuring the crisis, the developed countries successfully applied small counter-cyclical stimulus. This was not enough to eliminate the recession, but it was all about you could expect. The error in the euro area is that they implemented austerity to turn off the automatic stabilisers, which was a discretionary fiscal error. In the US, fiscal settings are too tight, but that should be implemented by a structural tax cut or spending increase.
"The only reason to discuss "money-financed fiscal easing" is to fool people who do not understand fiscal policy. "
ReplyDeleteI think I agree. I probably have less objection to to helicoptor money than you, but in my view it achieves little that debt financed fiscal deficits wouldn't, other than to placate those that worry about debt.
Brian: "Government liabilities consist of money, and bonds, which are forward money."
ReplyDeleteThere's your mistake right there.
Bonds may be one of three things:
1. Forward money
2. Forward tax increases
3. Forward expenditure cuts.
I stand behind my statement. You are essentially adding a qualification that government liabilities are backed by future primary surpluses. This either follows from the Fiscal Theory of the Price Level (which is the reductio ad absurdem of DSGE macro), or neo-Chartalism. Of course, there is the usual issue of interest rates being below nominal GDP growth, which means that "primary surpluses" turn into a "upper limit of primary deficits".
DeleteBut even so, as a bond holder, it is an instrument that delivers money to me at a certain forward date (plus coupons, not that they really exist any more). Other people may be the ones paying the taxes that "back" the currency. It's forward money for me, and a forward primary surplus for the citizenry at large.
Additionally, in a world where interest is paid on deposits (reserves) at the central bank, there is no good distinction between bonds and the monetary base in the Fiscal Theory of the Price Level (or neo-Chartalism).
It is interesting how strong the urge to make the simple complicated is. Of course bonds are forward money. But a better way to put it, a more "MMT" way is that money is a current bond.
DeleteAs Warren puts it, government bonds are part of the economy's 'base money' spent by government and not yet used to pay taxes.
DeleteThey form a financial assets of the rest of the currency area.
The mistake always made by people outside MMT is to equate 'not yet used to pay taxes' with 'must end up as taxes'. That is not something that can be determined in the current period, or something that should be of concern. Whether they will be withdrawn as taxes in the future depends upon the net saving desires of the population in the future. In other words how much they wish to spend then.
The evidence of history is that they just roll up as the population grows and people get wealthier.
Other than that the payments bonds attract are exactly the same as welfare benefits or 'income guarantees'. They are government payments made to a section of the population in the current period and spent or saved by the recipients based upon their marginal propensity to consume.
Unemployment benefits are paid to those who hold the status of 'jobless'. Bond interest is paid to those who hold the status of 'bond holder'. There is no functional difference between those classes of people. It is a political choice to pay either.
This comment has been removed by the author.
DeleteDo bond payments (interest) and welfare benefit payments represent shares in the economy?
DeleteI think the answer is 'yes'.
If the answer is 'yes', then the source of interest or welfare benefit is very important. If the source of payment is an exchange with (or tax on) other workers or property owners, then property is being traded for property. The total amount of property in the economy remains unchanged.
If government borrows money, it is creating money and thereby dividing the economy into an increased number of shares. By borrowing either from the Central Bank or private economy, government is increasing the amount of property in the economy.
The choice of borrowing from the private economy or borrowing from the Central Bank, or borrowing vs taxing, is a political choice.
I read with amazement but all the nuances still fail to coordinate. I think I see the missing coordinating factor. Please consider this concept:
ReplyDeleteRemember that fiat money is easily created. A representative from the Independent Central Bank (ICB) can sit down with a representative from Government and trade products. ICB has money, Government has bonds. Make a trade and ICB has bonds and Government has money to pay everyone.
It is that simple.
Now to the points from Nick Rowe, The bond is both an accounting of how much money has been issued, AND, a simple promise to give back the same money.
Is the bond forward money? It is but to no matter. Just as the ICB sat down with government to exchange products, the bond holder can sit with the ICB and exchange bonds and money. So who cares? This is the weakness of Quantitative Easing.
To Nick Rowe's points 2 and 3, the same ICB-Government exchange concept applies. His points are correct but to no matter. Forward tax increases (to pay the debt) are a misunderstanding of what the bond really is. Just give the bond holder money and as he spends, sales taxes and income taxes will increase which allows government to recover the money it initially spent. An automatic stabilizer.
I will close and sum up what I think is the missing coordinating link: Government bonds are the accounting record of how much money government has spent, AND, how much money has been promised to be returned to bond holder.