Winner of the New Statesman SPERI Prize in Political Economy 2016


Saturday 24 August 2013

Missing the point at the IMF

The IMF have just published a working paper entitled: ‘Assessing the Impact and Phasing of Multi-year
Fiscal Adjustment: A General Framework’. Or to put it more simply: should austerity be front loaded or delayed? A really important topic and one where the views of the IMF are of some importance.

I guess if you call anything a ‘General Framework’ you are taking a risk. But honestly, if you also write this

“our framework does not explicitly model the monetary policy response, which could have an important impact on output”

then you have no business using the word ‘Framework’, let alone ‘General’. [1]

We need to go through the logic one more time. When monetary policy is not constrained (we are not at the Zero Lower Bound), monetary policy can (and to a first approximation should) completely offset the impact of any fiscal consolidation. The multiplier in that case will be approximately zero. [2] However if we are at the ZLB, then within the current monetary policy framework (essentially inflation targeting), and unless you are really optimistic about unconventional policy, the ability of monetary policy to stimulate aggregate demand is severely compromised. As a result, any fiscal multiplier will be substantially greater than zero.

Now consider two periods. In the first, we are at the ZLB. In the following period, we are not. Consider two fiscal consolidation programmes. In the first, everything is front loaded into the first period. In the second, nothing happens in the first period, and all fiscal consolidation takes place in the second. Design the two programmes so that we end up with the same debt to GDP ratio by the end of the second period, so they are neutral in this respect.

What is the overall impact on output of the two programmes? Frontloading hits output in the ZLB period, with possible hysteresis effects in the second. Delaying consolidation until the second period has no impact on output whatsoever, because any impact on output is offset by monetary policy. Simple. So the choice is a no-brainer - you delay fiscal adjustment until the ZLB period has ended.

You would think that with these very dramatic implications for the optimal path for fiscal consolidation, allowing for monetary policy would have to be part of any ‘general framework’. Not the whole of any such framework, of course. You would want to consider the particular situation of countries without their own monetary policy. You would also want to consider countries where credibility was so low that delay would raise interest rates on debt (which the paper does do). And of course many countries are not at the ZLB. However some rather large ones are (like the US or the Eurozone as a whole), so ignoring monetary policy in any ‘general framework’ is just crazy.

What the paper does do is allow the size of the multiplier to vary with the output gap. Now you might think that this does something similar to allowing for a monetary policy response and the ZLB. However the way the paper sets things up it does not, because it fails to allow for the fact that outside the ZLB, monetary policy can offset the impact of fiscal policy. In their simulations the gradual (not front loaded) consolidation paths still involve large output losses, because they assume that without fiscal adjustment the output gap would be zero, so delaying fiscal adjustment creates a large negative output gap, which leads to a large multiplier. So this completely misses the idea that monetary policy could and should offset the impact of fiscal consolidation once we are well clear of the ZLB.

This is by now such an obvious and basic point I can only wonder why it is not incorporated into the analysis. By ignoring this point, what has been done is just inapplicable to some major economies. I do not like being so critical and blunt, but this is no academic debating point. And I would hate to think that this reasoning has been ignored precisely because its implications about the timing of fiscal consolidation are so clear.

[1] The paper says, immediately after the quote I give in the main text: "This is reflected implicitly in the fiscal multiplier assumptions in our framework (which can be allowed to vary with the output gap), but future work would be needed to explicitly integrate the monetary policy response into the picture." I discuss below why the 'reflected implicitly' is misleading, and also how the impact of monetary policy could have been captured quite simply.

[2] Approximately, because policy may be targeting inflation instead of or as well as output, and the inflation/output implications of monetary and fiscal policy may differ. It would be wrong in this case to say that multipliers would still be positive because monetary policy is not perfect (and to use something like a Taylor rule to reflect that). Here we are looking at planned fiscal consolidations, which the monetary authorities will know about well in advance. Of course uncertainty means that monetary policy makers will not exactly offset the impact of expected shocks, but they may over compensate (negative multiplier) as well as under compensate (positive multiplier). 

12 comments:

  1. Your observations, while quite true, ignore the now obvious desire by powerful people to manipulate the implications of income distribution (including tax payments). The whole austerity argument is an attempt to limit government spending and its redistributive effects.

    Economics has "known" for decades the harm caused by pro-cyclical policies and decisions: how government spending should offset the herd's market decisions. This is what should be the focus of current economic debate - not some made-up notion of "austerity".

    Austerity is merely the excuse used by powerful (and obscenely wealthy) people to do what is in their selfish, minority interest: restrict government from counteracting the "efficient" market, in which they can apply their power to extract ridiculous economic rents, or commit outright fraud in the knowledge that they are "too big to prosecute".

    History will hopefully be unkind to today's elites. They have violated principles from econ 101, and the whole "austerity" debate (and monetary policy confusion at ZLB) is a smoke screen to make it seem as if there is uncertainty about the best policies. It's all fiction, though.

    ReplyDelete
    Replies
    1. and their henchmen, the economists who sow the seeds of obfuscation

      Delete
  2. There’s a fundamental logical flaw in the “fiscal consolidation” idea, as follows. It’s based on the notion that national debts are currently higher than the average post WWII level, ergo they must be reduced. (Though incidentally the debt is currently lower, in the case of the UK, than in the ten years or so following WWII).

    Anyway… the above bit of logic is no different to saying “the temperature in my house is higher than two hours ago, ergo the temperature is currently too high”. In fact the above over-simple idea is spelled out in the first sentence of the introduction to that IMF paper.

    The IMPORTANT QUESTION is: what’s the optimum size for a country’s debt (and likewise, what’s the optimum temperature in my house).

    MMTers’ answer to the latter “optimum” question is thus. National debt (along with monetary base) is a net asset for the private sector, and the larger the stock of that asset held by the private sector, the more the private sector will spend all else equal. Therefor the size of the debt needs to be whatever brings full employment.

    And that in turn is much the same as Keynes’s dictum: “Look after unemployment, and the budget will look after itself”.

    ReplyDelete
  3. Thanks for an interesting post.
    However, just to be fair to the IMF, you said that this is "A really important topic and one where the views of the IMF are of some importance.", but with regard to the above working paper, this claim is quite misleading. It is said very clear (in bold) in the box above the abstract of the paper that "This Working Paper should not be reported as representing the views of the IMF." So, I don't think you should think about this paper as the official "views of the IMF".

    ReplyDelete
    Replies
    1. Most IMF papers are published with this disclaimer.

      Delete
  4. Anon:
    You can not possibly believe that the IMF publishes papers that have no point relative to current economic circumstance or proposal for future dicta just for the hell of it. No one believes that. No one.

    ReplyDelete
  5. This comment has been removed by the author.

    ReplyDelete
  6. What's missing from this paper is what happens if a country does fiscal stimulus followed by contraction (what they call "adjustment"). They could model this easily but the answer will show that there is a drop in the ratio of debt to GDP that is faster than all others scenarios - remember the law of motion of government debt shows that it's all about nominal GDP growth. Plus, you'd have a much higher GDP per capita.
    Finally, don't you mean the multiplier will be between zero and one, not zero? If it was zero, then the purpose of monetary policy would have to be to completely offset government spending (irrespective of all other factors) rather than grow employment and prevent inflation as the economy approaches full employment.

    ReplyDelete
  7. The trick with the Austerity 'argument' is that it is always sold as the only alternative. No argument is given to confirm that no alternatives exist, it is just stated, many times, all over the news. And since austerity is indeed the only alternative for a private person, a family, or a business, it is easy to convince most people that the same is true for countries, governments and monetary systems. Basically, Austerity is being sold by not arguing for it, ever. It is the TINA argument of Margeret Thatcher.

    ReplyDelete
  8. That is a working paper, so it is the opinion of the authors, and that is from the Strategy and policy Review department, which is known to churn around garbage.

    ReplyDelete
  9. Have a look at peil.nl the 25 August polls.
    From the further questions asked you can easily derive that the failing policies of the present government caused a lot of disappointed people to move to other parties. Economy usually being the (by far) no1 issue on which people determine on whom they will vote.

    FYI.
    Standard move is VVD to CDA/D66/Wilders. Seen the fact that Wilders gained the most (by far), very likely he got most of the deserters.
    Same PvdA. Standard move is to SP (Socialists). Other alternatives won hardly anything anyway.

    Also clear that there is no electoral platform for further austerity. Which is a related issue of course. Imho very unrealistic longer term. Present welfarestate looks unaffordable longer term with aging hitting in and much cheaper competition fromn the East. Will give a nice/interesting collision lateron. But for now a nice Catch22 no platform for cuts and the EU demanding them (and the economy not starting up). A nice collision in its own right.

    Clearly looks good proof that this is putting populists in the saddle. Like with immigration earlier and EU/Euro populists are the only ones against. From a democratic pov in Holland not an issue people can vote for several Euro-sceptic parties. But in Germany it is half the population against transfers no party available to vote for. Or before in the UK (hole now closed by IP). Anyway it is clearly creating a climate in which populist parties prosper.
    And not without reason it is against the ideas of large groups and if things not work or work out (like now on the economy on earlier on immigration) you look a bunch of idiots as traditional parties.

    NB You made the Dutch press with the earlier post on this subject btw.


    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.