Winner of the New Statesman SPERI Prize in Political Economy 2016

Thursday 9 April 2015

Cyclically adjusted deficits and instability

Jean Pisani-Ferry, currently advising the French government and former director of Bruegel (the Brussels-based economic think tank) has written a heartfelt plea for more stability in the Commission’s estimates of potential output. The reason is straightforward. The Eurozone’s fiscal rules require meeting targets for cyclically adjusted deficits within the next year or two. Every time estimates of potential output change, the target for the actual deficit also changes, and policy often has to respond immediately to meet the new targets.

Pisani-Ferry of course understands why this happens, and why cyclical adjustment makes sense in principle. He is not advocating returning to the days when targets ignored the economic cycle. However the problem he clearly has is in communicating to policy makers who are not economists why they need to change actual policy simply because someone in Brussels has revised their estimate of potential GDP. He writes

“Members of parliament – who are not technicians – are understandably disturbed when they are asked to pass a revised budget in response to an updated estimate. Not knowing the whys and wherefores, they end up perceiving such revisions as a source of artificial instability.”

However an obvious objection to his proposal might involve the following scenario. To meet its target the government is embarking on austerity which is also reducing GDP. The Commission gets new information which leads it to revise up its estimate of potential GDP, implying the need for less austerity. Should the Commission ignore this information for the sake of stability?

I would suggest that the ‘non- technical’ instincts of policymakers are right in this case, but the reason is more basic. Short term fixed date targets for the deficit are a source of instability. There is no economic reason to have such short term deficit targets, and there are plenty of very sound economic reasons not to have targets of this kind. In essence this is because the deficit should be a shock absorber, but by targeting it you make taxes or spending the shock absorber. Hence the perceived, and actual, instability.

It would be much better, as Jonathan Portes and I argue, to have deficit targets for 5 years ahead. Whether these should be fixed date or rolling targets would depend on how trustworthy the government was, and whether there was an independent and robust fiscal council that could provide an ‘implementation incentive’ for the government. (We would argue in addition that fiscal policy in the Eurozone needs to play a countercyclical role, both at the individual country level to correct imbalances within the zone, and at the aggregate level if interest rates are at the Zero Lower Bound.)

Would such targets need to be cyclically adjusted? In the context of an economy with its own monetary policy we would argue not, because within five years the central bank should have eliminated any output gap (in expectation). Whether cyclical correction via competitiveness effects works so quickly is less clear, and if it does not cyclical correction would still be required. However I suspect that if targets were for five years ahead, revisions caused by new estimates of potential would be less problematic, because the need for an immediate policy change would be reduced. (As an example, when the OBR first revised its estimate of potential in the UK, Osborne’s response was to extend austerity into the next parliament, rather than intensify current austerity.)

The key point is that targets for the deficit just one or two years ahead are foolish things to have, and cyclically correcting the target only makes them slightly less foolish. Indeed, I would go so far as to say they are primitive in macroeconomic terms. It is like telling consumers that they shouldn’t smooth their consumption, but instead vary their spending or income to keep their wealth at some fixed target level. You would only want to do that to a child, and policymakers should not be treated as children.


  1. "In essence this is because the deficit should be a shock absorber, but by targeting it you make taxes or spending the shock absorber. Hence the perceived, and actual, instability"

    If that is how deficit works in short term, then it is how it works in medium term.

    So then why this?::
    "It would be much better, as Jonathan Portes and I argue, to have deficit targets for 5 years ahead."

    This is correct:
    "The key point is that targets for the deficit .................... are foolish things to have,"
    I removed 'one or two years ahead' to make it fully correct.

    1. This comment has been removed by a blog administrator.

  2. Budget balance should be defined in goals, not an end in itself.

  3. 'A rolling target depending on how trustworthy the gov't was'. This is the heart of the problem, the same problem the ECB faces with Greece, does it defend it or chastise it?
    In the UK the embedded gov't financial architecture protects us against a financially irresponsible gov't who use austerity as to drive an un-democratic agenda, as intra-day overdrafts at the CB can be bundled into a buffer (the DMA) overnight. The cyclical (endogenously driven) deficit is allowed to stabilise the gov't sector relative to the non gov't. Until the Euro gets this type of arrangement they'll be chasing their tails on cyclical adjustment till the ship goes down.

  4. SW-L advocates “5 year ahead” targets. I object to ANY deficit target. I.e. I agree with Keynes who said “Look after unemployment, and the budget will look after itself”. And the word budget there is synonymous with “deficit”.

    To illustrate, it’s always possible that the population of any country “does a Japan” and insists on holding a much larger stock of government debt and/or base money. In that case, the state has absolutely no option but to run a relatively large deficit to enable to private sector to have that higher stock of savings, if the state wants to maintain full employment (aka zero output gap). In MMT parlance, the state has to meet the private sector’s “savings desires”.

    And if the interest demanded for holding debt rises significantly above zero (at full employment) that’s a sign that there’s too much debt. In that case some QE is in order combined with a rise in tax.

    In short, I suggest 3 targets: full employment, 2% or so inflation and interest on the debt equal or near to zero.

  5. Ah yes, France.
    If I remember correctly, the incoming French government was going to end 'austerity' and lead the way forward.
    How are they getting on with that?

  6. I think it is a nonsense to change potential of an economic continuously. The problem first arises with respect to how to measure this potential. Who can meet your targets, even when policy makers are themselves Economists. We must see the practical what is happening, not just make day dreams.

  7. A bit of a digression but Bruegel (particularly under Jean Pisani-Ferry) is a curious beast.

    It was an attempt to help build an intellectually credible framework for the, lets be kind here, strong priors of European Union institutional ordoliberalism. There were no real ordoliberal think tanks before because it is is a far more romantic than scientific philosophy and I think Bruegal fulfills a similar roll to conservative US thin tanks like the AEI, giving a dispassionate analytical veneer to a viscerally right wing position.

    Under Guntram Wolff there is a fair bit more diversity at Bruegel (that dreaded "Anglo-Saxon economics", and they do not mean new classical type), for instance the IMF's Mody, but the overall impression is still of a mixture of Bocconi boys and orodoliberal true believers trying to sound reasonable.

    People underestimate how much more influence ideological conviction has the edge over not just mainstream economics but evidence in EZ policy circles. We do political cults well in Europe, and this one has cost us a lot more than just growth and youth opportunities.


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.