Winner of the New Statesman SPERI Prize in Political Economy 2016


Tuesday 14 January 2020

Monetary and fiscal cooperation: the case for a state dependent assignment


In December last year Mark Carney said
“In a global liquidity trap, central banks cannot be the only policy makers who do “whatever it takes.” There are clear gains from coordination, with other policies – particularly fiscal policy”

I of course agree, as would most academic macroeconomists. So would any sensible informed fiscal policy maker. But of course this didn’t happen in the Global Financial Crisis from 2010 onwards in some key major economies, including the UK.

Carney’s statement, which follows similar statements by the central bank governors of the Fed and ECB, goes against what I have called the ‘consensus assignment’. The consensus assignment has monetary policy looking after the stability of aggregate demand and inflation, while the fiscal authority looks after government debt. In the UK at least this consensus assignment is deeply embedded in the way the media thinks about policy.

In 2009 George Osborne gave a speech in which he said
“[New Keynesian] Models of this kind underpin our whole macroeconomic policy framework – in particular the idea that by using monetary policy to manage demand and control inflation you can keep unemployment low and stable. And they underpinned the argument David Cameron and I advanced last autumn – that monetary policy should bear the strain of stimulating demand….”

This is a statement of the consensus assignment. The irony of it was that shortly before the speech was given UK interest rates hit their lower bound.

Is the consensus assignment still the best way to run policy after short interest rates hit their lower bound? In 2010, in Europe and the UK at least, central banks acted as if it was. Indeed they went as far as to advise fiscal policymakers to embark on austerity. Carney’s statement is an implicit acknowledgement that central banks had been wrong to do that.

The trouble with unconventional monetary policy is not that it does not work, but it does not work reliably. The scandal in 2010 was that while the Bank of England was suggesting in public that unconventional monetary policy could replace conventional interest rate policy, in reality they had little clue how much effect any change in unconventional monetary policy would have. An unpredictable and unreliable instrument is not a good basis for a policy regime when a better instrument is available, and that better instrument is fiscal policy. I think negative interest rates fit into this category of unreliable instruments, simply because we cannot for obvious reasons assume linearity.

So how do we ensure as far as we can that fiscal policy makers will not repeat the mistakes of 2010 in the next recession? The first best would be to have better fiscal policy makers, but alas that is not always possible. There are three widely discussed possibilities.

  1. The first is MMT. This in effect reverses the conventional assignment, with fiscal policy doing the demand and inflation stabilisation in all states of the world. If that happens debt looks after itself. I am not in favour of MMT, because I think independent central banks have been very successful at controlling inflation, and a government using fiscal policy would be less successful.

  2. The second is Helicopter Money. If you are prepared to call Helicopter Money (HM) monetary policy, this preserves the consensus assignment by giving the central bank a new tool. HM is more reliable than unconventional monetary policy, because HM is just like a tax cut, and we have a lot of data on the impact of tax cuts on consumers. Like tax cuts, HM will not work if all consumers are Ricardian, but they are not.

HM is only possible with the agreement of the government, preferably well before it is actually needed. Two key things have to be agreed. The first is the distribution mechanism, where I suspect some governments would prefer something other than a reverse poll tax. The second is an agreement to back the central bank, by which I mean supply it with the assets it requires to claw back at least some of the HM when the economy recovers. The only difference between HM and a bond financed fiscal expansion is that probably some or all of the bond issuance is delayed until after the economy recovers.

Central banks worry that governments will renege on their commitment to back the central bank. My response is that any government that would not back their central bank so it can fight inflation is also a government that would be prepared to abolish its independent central bank, so the concern is of no interest. I suspect also central banks think HM looks like fiscal policy to most people, and they shouldn’t be doing fiscal policy.

I would add a further point on HM. It will not stop a government using what I call ‘deficit deceit’ in a recession: pretending the deficit is too high and requires spending cuts, because the government wants to scare people into accepting a smaller state than would be popular otherwise. HM would avoid this fiscal consolidation influencing output because its demand effects would be offset by the central bank. But a shrinking of the state beyond anything that is popular in normal times is also almost certainly sub-optimal, and can have devastating political as well as economic implications, like those we have seen in the UK, and you could argue HM encourages this.

  1. The third, and most likely, is central bank advice. If the central bank thinks that a recession is coming where rates will hit the lower bound, it advises the fiscal authority that some fiscal expansion is required. This is fine if we are trying to combat a fiscal authority that is just ignorant on these matters. The central bank could also convince a fiscal authority that was worried about financing its debt, by for example agreeing to monetise the expansion needed by doing the corresponding amount of QE, or more simply to neutralise any failure by private agents to buy debt.

My concern here is with a government that said thanks for the advice, but we prefer to focus on reducing the deficit using spending cuts. Would the central bank be prepared to make its advice public? It might not do so if it was concerned that the government would reciprocate by starting to tell the central bank what do so. You could therefore argue that this strategy could be either ineffective, or may threaten central bank independence.

So how can you stop a government that is determined to use the rising deficit in a recession to shrink the state? Of course you cannot, but you can try and create the conditions that will put maximum political pressure on it not to. I suggest above that HM fails to do this, and central bank advice is unlikely to either.

What would be more effective is for macroeconomists and central banks to start being honest about the consensus assignment. As a near optimal policy regime that assignment is dead. Instead macroeconomics suggests what could be called a ‘state dependent assignment’. In most states of the world, central banks stabilise the macroeconomy just as they do now. However in an economic downturn of sufficient size (where ‘sufficient’ is to be defined) the assignment flips, and fiscal policy makers are in charge of stabilisation. In non-technical language, fighting recessions becomes the government’s job.

I think this is something that most academic macroeconomists and some central bankers have accepted implicitly but not explicitly. One reason is I think pedantic. Of course in the state dependent assignment the central bank does not stop trying to stimulate demand in a recession by at least keeping rates low, but there are compelling political economy reasons to highlight the responsibility of governments in this respect.

Those familiar with Jonathan and my paper on fiscal rules will recognise our knockout when rates hit their lower bound as one operationalisation of a state dependent consensus assignment. But it is not an ideal mechanism because switching the assignment should depend on forecast events. Others, like the IPPR and Resolution Foundation, have suggested alternative schemes that come under the umbrella of a state dependent assignment. There is a great deal of work required to figure out the best mechanisms, and also to think about who has control over when switches (both on and off) happen, and whether there is a role for the central bank and/or fiscal council in advising the government on effective stimulus packages.

To conclude, central banks are now recognising that fiscal stimulus is required in significant economic downturns. This is in contrast to the GFC, when many fiscal policy makers enacted austerity. One of the reasons they were able to enact austerity was the dominance of the traditional consensus assignment in the mind of the public. Our most effective way of preventing this happening again is to make a state dependent assignment the new consensus assignment.






4 comments:

  1. Quoting you: "So how do we ensure as far as we can that fiscal policy makers will not repeat the mistakes of 2010 in the next recession? The first best would be to have better fiscal policy makers, but alas that is not always possible. There are three widely discussed possibilities.

    The first is MMT. This in effect reverses the conventional assignment, with fiscal policy doing the demand and inflation stabilisation in all states of the world. If that happens debt looks after itself. I am not in favour of MMT, because I think independent central banks have been very successful at controlling inflation, and a government using fiscal policy would be less successful."

    In response,
    (1) Regarding the need for fiscal policy in a recession, I guess I agree.
    (2) Regarding what you call “MMT,” this term may be a reasonable shorthand for MMT’s policy proposals regarding fiscal and monetary policies (assuming a sovereign currency situation, as the U.K. of course already enjoys). I would prefer the term “MMT’s proposed policies.” Of course, MMT proposes policies related to fiscal and monetary policy that would be in effect at all times, not just during recessions. Moreover, MMT is largely descriptive in nature, making it more than its policy proposals.
    (3) You are thinking independent central banks are needed to control inflation, the standard New Keynesian position. If independent central banks are needed (which you presume will not decide to hold nominal rates constant, becoming more pro-MMT than the government--not a given in my opinion!) to control inflation, the cost is then unemployment and the need to generate environment-destroying private-sector growth, and especially investment. That becomes hard to sustain during recessions. At the moment, I’ll not attempt to challenge the whole New Keynesian story of central bank credibility, etc.

    ReplyDelete
  2. I think the easiest option for the central bank would be to announce the buy of government bonds by QE, coupled with a public explanation that the goal is to allow government to do what they think best to ensure a quick recovery without worrying about public debt during the recession.

    Phrased carefully, it avoids the central bank telling the government what to do while defusing most of the political arguments in favor of austerity.

    ReplyDelete
  3. Where have all comments gone, doesn't anyone read this blog anymore?

    ReplyDelete
  4. Simon:

    Here is my attempt to formulate a robust operating framework for the Fed so that it can respond to the challenges you outline: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3494567

    ReplyDelete

Unfortunately because of spam with embedded links (which then flag up warnings about the whole site on some browsers), I have to personally moderate all comments. As a result, your comment may not appear for some time. In addition, I cannot publish comments with links to websites because it takes too much time to check whether these sites are legitimate.