Thursday, 21 February 2013

The 'cello approach to monetary policy

I've mentioned in a previous post the idea of using fiscal tools to support monetary policy. In this post I want to explain what I mean.

The prevailing view of fiscal policy is that it concerns the financing of government and the behaviour of the population. It has nothing to do with monetary policy and indeed can be antagonistic to it. In this era of monetary dominance, emphasis has been on the effect of fiscal policy over the long run, and in particular the use of "automatic stabilisers" to dampen the effects of the business cycle. We can say that from the monetary dominance perspective, the best fiscal policy is one that is set up to act as a counter-cyclical buffer and then left alone.

And yet.....I have been discussing in recent posts the use of government debt as a money substitute in financial markets. That implies that fiscal policy must have a monetary effect, since government debt issuance is a key part of fiscal policy. The other key part of fiscal policy is taxation - and this too can have a monetary effect. The mistakes of the past that led to fiscal and monetary policy being in opposition to each other were due to the fact that fiscal policy was NOT seen as having a monetary effect. Sargent & Wallace's paper "Some Unpleasant Monetarist Arithmetic" presupposes that governments set their budgets without considering the likely impact on bond yields. In today's world this seems highly unlikely. If anything now it is central banks that seem less concerned about this: governments around the world are imposing suffering on their populations in order to earn the trust of international investors and prevent bond yields rising. So although Scott Sumner is (probably) correct to note that a determined central bank can always choke off expansionary fiscal policy, at present all the signs are that central banks (with the possible exception of the ECB) would do nothing of the kind. On the contrary, there are thinly-veiled cries for help from central banks to governments that appear deaf. Recently, the Bank of England came very close to telling the UK government to loosen fiscal policy.

Sadly, a generally looser fiscal stance may prove to be politically unacceptable. But I think there is a role for targeted use of fiscal tools to support a generally loose monetary policy. We can think of these targeted fiscal tools as fine-tuning. A large stringed instrument such as a 'cello usually has two sets of tuning keys - the pegs just below the scroll, which set the general pitch level: and the tailpiece keys below the bridge, which make small adjustments to the pitch that are too subtle for the pegs.
I see monetary policy and targeted fiscal tools in much the same way. The general policy stance is established by the central bank and conducted principally through interest rate policy and open market operations. But small inconsistencies, and tailoring policy to the circumstances of particular agents, can often be dealt with more easily with targeted taxation.

For example, in a twitter conversation with Miles Kimball recently about the problem that physical cash poses for negative interest rate policy, I suggested that it would be easier to impose a tax on vaulted cash and/or on large or frequent cash withdrawals than it would be to introduce a negative interest rate on physical cash or eliminate cash completely. After all, as Miles points out, a tax is the fiscal equivalent of a negative interest rate. Conversely, a tax credit (or a benefit) is a positive interest rate. Targeted taxes or tax credits can therefore be used to create proxies for interest rates in areas where monetary policy transmission is weak.

Another example might be imposing tiered taxes on bank asset expansion to limit the amount of credit that can be created. These taxes would not be like the credit controls of the 1960s, where banks could lend to a limit but then had to stop: banks would still be able to lend any amount they wished, but at an individual bank level, the more they lent the more expensive it would become. Taxation might be a more effective brake on excessive bank lending than either monetary or regulatory policy: monetary policy is too much of a blunt instrument to act at the individual bank level, and regulatory limits remove responsibility for lending policy from bank management.

Taxes like these would be Pigouvian taxation, of course - taxation designed to influence behaviour, rather than raise money. And as with all Pigouvian taxation, any money raised should not be used to finance government spending, because the whole point of the tax is that as people's behaviour changed it would eventually reduce to zero. Government cannot rely on a source of funding that is designed to disappear.*

Fiscal fine-tuning of monetary policy could also include debt issuance. Debt issuance has the same effect as reverse QE - it reduces bank reserves and removes money from circulation. Debt issuance itself is therefore by definition monetary tightening and cannot possibly be inflationary. The inflationary effects of debt issuance come from spending the money received, not from the debt itself. When debt is used to "fund" government spending, the government has already entered into the spending commitments by the time it issues the debt, so the path of future inflation is already set: it is irrelevant whether the deficit arising from those spending commitments is funded by currency issuance or debt. Furthermore, as the government's budget is a matter of public record, investors know what the spending plans are well before the debt or the currency to fund them are issued, and will respond accordingly. A very loose fiscal stance is therefore likely to result in exchange rate falls and/or bond yield rises long before the actual funding for that spending is needed. This is why a monetary authority attempting to choke off the inflationary effects of government profligacy "after the event" is doomed to fail: the intervention has to be when the spending commitments are entered into, not when the debt is issued. (This, by the way, is the reason for my caveat on the Sumner Critique that the sharp-eyed among you will have noticed in the third paragraph of this post.)

There has been quite a bit of discussion recently about the role of government debt in the monetary system. It  seems to me that the near-money nature of short-term government debt, particularly, enables it to be used as a monetary policy tool. I could envisage a role for short-term government debt issuance as a gentler form of monetary tightening in a fragile economy with a large private debt overhang where raising interest rates could be very damaging: obviously the money raised through this debt issuance could not be used for government funding, as that would counteract the tightening effect. Daniela Gabor notes that most sovereign debt management offices (DMOs) already intervene routinely in the repo markets, providing liquidity to shadow banks in much the same way that central banks do for regulated banks. What is this if not using a fiscal tool (government debt) for a monetary purpose?

There may also be monetary tools that could be used for fine-tuning, but in my view fiscal tools should be part of the toolkit. Why hamper the monetary authorities by denying them access to fiscal tools that would make their job easier?

Freeing fiscal tools from the objective of government funding would allow them to be used to control the money supply for the economy as a whole, in conjunction with monetary policy. In fact because they act directly on broad money, they can have a more immediate and direct effect on the "real economy" than monetary tools. Fiscal tools are powerful, which I suspect is one of the reasons why people are reluctant to use them. Properly applied, they can be extremely effective. Improperly used, they are disastrous. Which is why I would rather they were limited to a fine-tuning function within the "envelope" of a general monetary policy stance set by the central bank. Without that discipline the result could be very high inflation, collapsing bond yields and a currency in free-fall.

I am certainly not proposing an undisciplined fiscal policy. We learned that lesson in the 1970s. Nor am I suggesting that monetary policy should no longer be dominant. What I am suggesting is that fiscal tools have a role within the monetary framework, particularly now that government debt is widely used in financial circles as a money substitute. We should not refuse to use them because of past mistakes.

Related links:
Floors and ceilings - Coppola Comment
Some Unpleasant Monetarist Arithmetic - Sargent & Wallace
Sumner Critique - Market Monetarist
Bank of England Governor outvoted on QE - Guardian
Central banking and bubbles - The Economist
What monetary policy can and can't do - The Money Illusion
Storify: Kimball/Coppola/Sargeant
Getting Leeway on the Lower Bound for Interest Rates... - Confessions of a Supply-Side Liberal
Pigouvian taxation - Wikipedia
Sovereign debt managers and repo markets - Helicopter Money (Gabor)
When Governments Become Banks - Coppola Comment
Government debt isn't what you think it is - Coppola Comment
Central banks, safe assets and that independence question - Coppola Comment

...and the rest of my posts on safe assets, too

* The weakness in this idea is of course that government is bound to want to use the money. We've seen this already with the levy imposed on banks' wholesale funding, which was designed to encourage them to rely less on short-term wholesale funding and more on longer-term sources such as retail deposits. This has been an amazingly successful intervention: banks have improved their loan/deposit ratios from an average of 137% at the time of the financial crisis to about 105% now. Consequently, the amount of money raised from the bank levy has fallen considerably. But far from celebrating this, the Government complained about the lack of revenue, and in the Autumn Statement increased the levy to compensate. This completely misses the point!


  1. A more appropriate response is to ditch monetary policy as a fine tuning response and enhance the automatic fiscal stabilisers.

    The central bank should not be in charge. They are not elected to be in charge. The central bank should advise. Politicians should decide - because politicians can be sacked by the people.

    The problem here is the obsession with twiddling with interest rates and trying to do things indirectly.

    And I suspect that is to do with the United States fundamental problem with government that lay at the heart of why it was set up in the first place.

    Here in the UK we don't have that problem. The central bank is owned by HM Treasury - de facto and de jure.

    Central bank 'independence' from government is just a political choice - like public healthcare. It is not a fundamental constraint.

    The current operational control structure using a central bank department and a treasury department doesn't work properly. We need a different design.

    1. Neil, I don't agree with you about any of this, I'm afraid. Automatic stabilisers are not a "fine tuning" response and neither is monetary policy as presently operated. I am suggesting use of fiscal tools as an enhancement to existing approaches to monetary policy, not to replace it. This requires cooperation between central bank and government, of course, since the central bank does not, and should not, have authority to tax. But to ditch monetary policy completely? Absolutely not. You only need to look at the mistakes of the 1960s and 1970s to see why fiscal policy alone is insufficient. Or read Sargent and Wallace's paper.

    2. Frances, I might have to side with Neil a bit in this instance. I completely agree that fiscal tools should be used as a form of monetary policy but don't see why the central bank should control those tools.

      As you point out, fiscal policy is inherently monetary and is more capable of constricting excessive lending than merely raising interest rates (unless the CB is willing to go to extraordinary levels). Fiscal policy is therefore more capable of moderating business cycles (success is a different matter). However, fiscal policy generally involves picking winners and losers much more directly. For that reason I think Neil is correct that unelected officials should not be given that (extra) power.

  2. I think I would have more faith in a Pigouvian tax effect if the tax were directly visible to the borrower rather than a few people in a small number of banks. Why not introduce a Stamp Duty like tax on borrowing. The proceeds could be used to lever the Pigouvian effect by reducing other taxes such as VAT, Income tax etc., so maintaining economic growth while dampening credit expansion (aka bank assets). The overall effect being neutral in the long run.

    1. Well, of course there is already Stamp Duty on house purchases. As the majority of lending is mortgages, you would be hitting the same people twice.

      More importantly from my point of view, it wouldn't be an "expansion" tax because it would have to remain at the same level irrespective of the amount of borrowing a bank had done. Unless you think the "stamp duty" tax would automatically increase as a bank's lending book increased, which would discourage people from going to it for loans? But how on earth would you police that? We don't have real-time financial reporting.

    2. Stamp Duty on UK property is massively less than typical levels sur le continent where bubbles, except in Spain (another story) were less bubbly :-) Also all properties typically attract CGT (even owner occupied principal homes, but with an annual amortisation allowance)

      Realtime payments of Stamp Duty to HMRC as the approved loan is entered into the bank's IT system is more complex than an online PayPal transaction? The nature of the loan data could also be captured for analysis. The amount of the loan granted to the borrower would net of the Stamp Duty.

  3. A more interesting way for paying the Stamp Duty might be Mobinos ( Maybe a special type of Mobinos, such as Tobinos, could be used? The amount of Tobinos issued could be within the control of the CB. See: www.

  4. Frances says, “Why hamper the monetary authorities by denying them access to fiscal tools that would make their job easier?” Why indeed.

    In fact why not take it a stage further and just merge monetary and fiscal policy: i.e. when stimulus is required, just have the government / central bank machine print money and spend it (and/or cut taxes).

    That’s the policy advocated by Milton Friedman, Keynes, Abba Lerner, Positive Money, Prof. Richard Werner, the New Economics Foundation, and me. Advocates of Modern Monetary Theory also tend to go for a merged system.

    I suspect important motive for keeping monetary and fiscal policy separate is that it gives a host of windbags something to witter on about: e.g. economics commentators working for newspapers. And then academics just hate simplicity: it puts them out of work. They prefer the complexity of separate monetary and fiscal systems, rather than simply unifying them.

    1. I will write about "why not", Ralph. The reasons are not to do with economics and much to do with politics - and human nature.

    2. Before doing so, you might like to read an inspiring(??) article by me on the subject: