Winner of the New Statesman SPERI Prize in Political Economy 2016

Sunday, 15 February 2015

The size of the recent macro policy failure

In my Vox piece, I did a simple exercise to show how important fiscal austerity has been in the US, UK and Eurozone. If government consumption and investment had grown by 2% from 2010 onwards, and assuming a multiplier of 1.5, GDP could be around 4% higher in all three ‘countries’. [1]

It cannot be emphasised enough what a huge waste of resources this represents. If 1% growth was lost each year, then by 2013 that gives a cumulative loss of 10% of GDP. That approximation works well for the US. It also roughly fits with Eurozone estimates based on simulations of the NIGEM and QUEST models described here, but the Rannenberg et al study that I have discussed generates cumulative GDP losses up to twice as large. The UK is different from the US because austerity was concentrated in the early years. Using the same methodology (i.e. a multiplier of 1.5) you get a cumulated loss of around 14% of GDP.

For the UK I’ve often quoted a smaller figure of a 5% loss, but based on an analysis which I have always been careful to describe as conservative. It takes OBR estimates of the impact of austerity, which uses lower multipliers (although it does include the impact of higher taxes, which I ignore), and then assumes that all this lost GDP was recouped in 2013. Both differences are equally important in going from 14% to 5%.

Why, for the UK, do I tend to quote the conservative estimate? Four reasons. First, the government is fond of using OBR analysis when it suits them, because their work has some authority. Second, the OBR analysis is more detailed and comprehensive, and it implicitly allows for some monetary policy offset, which may be reasonable given how high inflation was in 2011. (I discuss this issue in much more detail here.) Third, I thought there was some poetic justice in assuming that all of the GDP growth in 2013 was simply a bounce back from earlier austerity, given that many people argue that 2013 growth vindicated this policy. [2] Fourth, losing 5% of GDP is bad enough, so there seemed no gain in using a higher figure, particularly when most of mediamacro act as if the number is zero. But if you asked me what my best guess is, it is nearer 14% than 5%. [3]

As I show in the Vox piece, if US GDP was 4% higher in 2013 it would be above the CBO’s current estimate of potential. The same is true for OECD estimates of potential for the UK and the Eurozone. But all three estimates assume that ‘trend’ or ‘potential’ GDP, or whatever you want to call it, has slowed substantially following the Great Recession. In a subsequent post I want to consider how reasonable it is to assume potential GDP is independent of actual GDP, and why even my 10% (14% for the UK) figure could be an underestimate.

Whether it is 5% of GDP, or 10%, or more, it is numbers like this that I had in mind when I wrote these two posts. They illustrate all too clearly the asymmetric risks that I talked about there. If these numbers are right, but monetary policy makers are nevertheless broadly content with their performance over the last five years, they either have a completely distorted view of the costs of inflation [4], or they have become fooled by a belief in the divine coincidence: that they only need to look at inflation to judge performance. (They could believe that they did not have the tools for the job, or that they had the wrong target, but if that is what they thought that is what they should have said.)

Going from the past to the present, Paul Krugman recently talked about the difference between insiders and outsiders on policy. This also reflects my own experience in the UK. How do we outsiders change this? I think the best place to start is by getting the insiders to think about the costs of fiscal austerity. Once you do that, you realise how large the recent failure of macro policy (but not macro theory) has been, and therefore how important it is not to carry on making the same mistake.  


[1] I write could, because any extra demand growth might - or might not - have been offset by a tighter monetary policy. If it had been offset, in this counterfactual world I would then be writing posts about the foolishness of monetary policy.

[2] We would then have a perfect example of my ‘closing a part of the economy down to boost future growth’ repost, which I used when people wanted to describe 2013 UK growth as vindicating austerity. Paul Krugman prefers being hit by a baseball bat.

[3] So rather than my conservative estimate that austerity lost every adult and child in the UK £1500, my best guess is nearer £4,000. (That is £10,000 per average UK household.) The equivalent number for the US (10% of GDP per capita) is just over $5,000, and for the Eurozone  E3,000.
 
[4] A weak recovery probably shaved the odd percentage point off inflation between 2011 and 2014, but if you ask most people how much they would have been prepared to pay for this, I doubt if the answer would be in the thousands of dollars, euros or pounds.


47 comments:

  1. I saw Ben Broadbent, ex-senior economist at Goldman Sachs and now MPC member at the BoE, giving a pretty smug interview to Sky News today.

    Given that Krugman likes to use information put out by the economic team at Goldman Sachs - always with no link - I wonder how their model tallies with Broadbent's optimism?

    ReplyDelete
  2. If you accept that long term trend GDP would put the UK at circa £2100 billion, allowing for a global deflater; and, GDP at current market price is about £1720 billion, then the UK is about 18% short on the long term trend line.

    Alistair Darling was running an 11% deficit through Q1 2010, the UK economy was pulling out of the dive and then Osborne arrived and killed the recovery flat with his slash and burn the public sector austerity plan. If Darling had a second chance, he would probably have run nearer 14% deficit through 2010.

    The bottom line is Darling would have the UK economy where it is now, three years back! the higher deficit would not matter a toss. No bond vigilante can take on a central bank in its own sovereign currency; of which, it is the monopoly supplier. The UK Treasury and its Central Bank can never run out of Pounds Stirling, there is no Bill, presented in Pounds Stirling, it can't pay. It never ever has to borrow Pounds Stirling from anybody!!! (Acorn)

    ReplyDelete
    Replies
    1. Sterling. It's Sterling.

      Delete
  3. Lost GDP could have another cause entirely. Poor monetary policy. A policy partly induced by a false belief that monetary policy doesn't work effectively at the ZLB. Something for which you bear partial responsibility. You should examine your conscience rather than campaigning for greater fiscal policy that will always be snookered by an IT-targeting central bank: ie the theory of monetary offset.

    ReplyDelete
    Replies
    1. That has to be the most far-fetched attempt to smear someone I've seen in a very long time, James. You have finally jumped the shark.

      Taa taa.

      Delete
    2. Wren-Lewis has been asking for both so he nothing to apologize for. See helicopter drops. It's legitimate to ask which one is more effective or about the differences in how they work. One is more direct than the other.

      Different people want to do different things so why not compromise and do all of the above?

      Delete
    3. James in London,

      Monetary offset (MO) is nonsense. It’s far from clear what it actually consists of, but if it consists, as per your suggestion of the idea that central banks “always snooker” fiscal stimulus, that’s obvious nonsense given that during the recent recession, central banks far from “snookering” fiscal stimulus, actually AUGMENTED it with interest rate cuts.

      In contrast, there’s the definition of MO given by the high priest of MO himself, i.e. Scott Sumner. At the link below he defines MO as the idea that if the CB thinks fiscal stimulus will push inflation above the 2% target, the CB will partially or wholly negate that fiscal stimulus. Well of course. Who ever said otherwise? MO is one huge non-theory, or if you like, statement of the obvious.

      At the link below see in particular the paragraph just before and just after the supply / demand diagram.

      http://mercatus.org/publication/why-fiscal-multiplier-roughly-zero-0

      Coincidentally I had a “robust” exchange of views with Scott Sumner and Kenneth Duda on Scott’s blog yesterday on this subject. Word search for “Ralph” at the link below if you’re interested.

      http://www.themoneyillusion.com/?p=28674


      Delete
    4. Ralph Musgrave, both of your interpretations are idiosyncratic. The Scott Sumner site consistently assumes that the Fed controls nominal output via anticipated long run printing. You might challenge the claim that expected permanent money expansion results in inflation, but it is not limited to "he defines MO as the idea that if the CB thinks fiscal stimulus will push inflation above the 2% target, the CB will partially or wholly negate that fiscal stimulus." If you hold that inflation follows expected base levels, then the Fed is implicated whenever nominal metrics permanently diverge from a "preferred" pathway. Monetary offset is the hypothesis that the Fed controls nominal metrics, like a preferred inflation pathway, such that they are implicated in any given realized inflation path where fiscal policy is also occuring.

      Delete
  4. This comment has been removed by the author.

    ReplyDelete
  5. Well, if you want to make mediamacro pay attention you have to give them something they can hang their coats on. Take your figure of 10% GDP loss. That works out at roughly £2,500 per man, woman and child, or £10,000 per family. Now those are numbers the average man gets emotionally involved with, and that's just what mediamacro is looking for - something to unsettle its readership.

    You might think this a silly question "Which is bigger, 5% or 10,000?" and few people will ask for more details - they'll ignore the small number and fear the larger. "There's now't so queer as folk!" as they say in Yorkshire...

    ReplyDelete
    Replies
    1. Part of that gain would be borrowed (as in x amount of debt per man, woman and child) and another part would go to the very rich. There's nothing wrong with being dramatic, as long as you're correct about the basics.

      Delete
  6. A multiplier of 1.5 would not be enough to be sustainable, not even in France. You'd need a multiplier of ~2.5 for that in the UK and Eurozone and ~4.5 in the US. Any multiplier above 1 could function as temporary economic stimulus, but that's a long way from convincing the people who matter that such stimulus should be implemented.

    ReplyDelete
    Replies
    1. I just do not understand this. If you have flu you take some form of medication. You do not refuse that medication because it does nothing for you when you are well.

      Delete
    2. "If you have flu you take some form of medication". I actually do not. healing without medicine gives your body the opportunity to build up an immunity, besides most of the time medicine only masks symptoms instead of dealing with the underlying causes (stress and lack of sleep, for most people). Similarly you're going to have to convince the people who matter that a) a fiscal medicine will work as advertised, b) it won't interfere with dealing with underlying causes and c) that the patient is actually sick, not just getting older or returning a more sustainable state after a period of unsustainable stress. Don't get me wrong, I favor economic stimulus through infrastructure projects when it's clear what kind of crisis is going on, but at the same time I recognize there have to be limits (when you have a deficit of 7% of GDP and an already high debt you can't just spend lots more on economic stimulus) and that there are legitimate arguments against the idea that the crisis is really a simple one that can be fixed with stimulus money (after all it has to be paid back and you don't want it to end in the pockets of China and the world's billionaires).

      Delete
    3. I have done (a), (b) and (c), so you need to produce arguments about why I am wrong. If you think there are limits, you need to justify them.

      Delete
    4. You have done so for the UK, but you keep telling us that the same recipe should work for the European continent.

      It is not my job to convince governments and central banks against economic stimulus so I don't need to produce arguments why you would be wrong (I do not even think you are wrong in the case of the UK, as long as your proposed stimulus is accompanied by long term reforms).

      As to there being limits. The market determines those limits, you have to borrow from someone and for that they have to trust you, they might not (or only against extreme interest rates) when you borrow a lot and already have a high debt.

      Delete
    5. P.S. I think it would be very effective (at least in countries that have moderate elements on the rightist side of the fiscal spectrum) for proponents of stimulus to put an emphasis on forms of stimulus that are clearly temporary, such as infrastructure projects. Too many people are using stimulus-related arguments to hold up reforms or to give away social program goodies that are politically very hard to retract once in place. Economists may not consider such things but politicians sure do.

      Delete
    6. Indeed: http://mainlymacro.blogspot.co.uk/2013/11/something-we-can-all-agree-on.html

      Delete
  7. CPI in the UK has been above 3% in both 2010 and 2011... It was significantly above 2% target in 2012 and 2013. When you assume a multiplier for additional fiscal stimulus, what do you assume for monetary policy? If monetary policy is unchanged, what would have been the inflation outcome?

    I also find it peculiar that you are not mentioning the fact that the BOE has just decided they are not at the ZLB any more. Or that we have now seen several banks adopting negative rates. Even if you assume there is a lower bound (maybe -0.5% or -1%), we were still significantly way from that number it now seems (with potentially 1%-1.5% worth of easing still available). I think you can argue that this mistaken belief is as big an error.

    Also arguing that there are no distributional effects from fiscal policy is not very reasonable. You can argue that UK tax policy doesn't do enough redistribution or that the government shouldn't have hit the poor as much, but I think you have to admit that bigger deficits are most likely going to be eventually covered by higher taxes on the better off...

    As a better off myself, I am still okay with the current taxation level, but significant increases from here will start to feel pretty confiscatory. I already pay well over half my income... The cut in top income tax was likely unnecessary, but pushing the rate much higher will likely have significant effects. If you include the employer side on NI, you are already looking at almost 60% tax (depending how much you get as a salary), plus VAT and other taxes... You will force people to incorporate somehow if you push it higher...

    ReplyDelete
    Replies
    1. Your para 1: see my footnotes (1) and (4).
      Your para 2: It was clear in 2010 that the BoE regarded 0.5% as a lower bound. Fiscal policy was decided in that context
      Your para 3: It is interesting that you associate higher deficits now with higher future taxation. Yet deficits are now being reduced without much impact on taxation. So why not in the future?

      Delete
    2. So are you arguing the policy target is wrong and should be changed? I don't see how you can square your view otherwise. There must be a point at which it becomes foolish to try to stimulate the economy further. Let's assume the government had followed your suggestions (with no monetary policy change), and inflation run at 5% in 2010. I don't think the inflation itself is a big problem, but I do worry that it is a signal that gov spending is already pushing on supply side constraints, which would make it a bad bet in terms of multiplier. I guess the distributional effect could be still valuable, but it becomes a straight transfer at that point.

      I think it's reasonable to assume with gov spending will fall on top half of tax-payers. The tax/benefit system is progressive and deficit spending is likely more effective if targeted towards the poor (since you don't want the money to be saved)... Given this progressiveness, even if gov spending has a positive return, the return is likely negative for the top half. By definition non-budget constrained households could hire the unemployed if they wanted to. The government forcing them to do it must be a net cost on them... it's more complicated than that, but I think it's a reasonable approximation.

      The only situation I could see that there is no distributional effect is if there is still long run fiscal capacity (since you are just running down an asset owned by the state effectively). But that's a limited resource, so it can be used only once, so it cannot be a tool to be used all the time... Maybe it was available this time, but I have my doubts to be honest.

      Delete
    3. The MPC did not raise interest rates in 2010/11 because they saw that the increase in inflation was temporary, and not the result of a mismatch between supply and demand. How could they think otherwise, given unemployment at the time. Core inflation might have been a little higher without austerity, but this basic point would still have been true. (Actual inflation would probably have been lower because there would have been no increase in VAT.)

      In the UK austerity has largely hit the bottom and top of the income distribution. So without austerity both groups would have been better off in the short term. In the longer term the additional debt would have had to be paid off, but how that is done is a choice for any government. It could raise taxes, or reduce government spending.

      Delete
  8. Simon, I am curious as to why the sign and (roughly) the size of the multiplier seems to be relatively uncontroversial empirically in the UK, but not in the US. All of Barro's estimates are of multipliers <1. I cannot help but conclude that empirical economics is not "objective": if I know the authors name, I can predict whether his/her empirical work will reveal a multiplier > or < 1. For example, if I gave identical datasets to C Romer and R Barro, I am willing to bet which will estimate the higher multiplier ... what's going on?

    ReplyDelete
    Replies
    1. I fear your conclusion is partly correct - you could say the same about studies that look at the impact of the minimum wage. However what tends to happen is that those with a political view tend to focus on the empirical results that suits them, which can make the impression of bias greater than it actually is. Here is a reference to a meta-study that collects together different studies:

      http://www.boeckler.de/pdf/p_imk_wp_117_2013

      However most of these studies are problematic not so much because bias may have led to data mining, but because they fail to take into account the possibility or otherwise of monetary offset. I tend to use numbers that come out of thinking about simple models which are widely accepted, and which allow for the impact of monetary policy. Luckily this way of doing things is consistent with the broad mass of empirical evidence.

      Delete
  9. This is a terrifically important rhetorical direction to take the argument. Thanks.

    It would seem part of the effort should be to face down the "secular stagnation" crowd, who seem to determined to shift the economic target to meet the errant arrow of policy.

    ReplyDelete
  10. "So rather than my conservative estimate that austerity lost every adult and child in the UK £1500, my best guess is nearer £4,000. (That is £10,000 per average UK household.)"

    This is where faith becomes indistinguishable from demagoguery

    If the government had spent all that money on infrastructure as you propose, it would have raised salaries and employment of construction workers. Some of that would have spilled over to the rest of the population, but very much less than the amounts you calculate.

    So, essentially you are deceiving the public. Isn'tthat propaganda?

    ReplyDelete
    Replies
    1. Who do you think uses the infrastructure?!

      Delete
    2. The Japanese and the Spaniards spent like mad on infrastructure - and what did it get them?

      Delete
    3. Better infrastructure?

      Delete
    4. Bridges to nowhere; airports where no airplane ever lands, higher debt!

      Delete
    5. My reply wasn't meant to be flippant. If you look at Spain more generally, the EU development money that went into the country in the initial years transformed the country. Remote areas where travel anywhere took the best part of a day now have a comprehensive road system. Some was wasted of course (I always forget that private companies never ever mal-invest) but in general the country's infrastructure is vastly improved. Cynics might say it makes it much easier now to sell German cars and white goods to a new class of consumers ...

      Delete
    6. Perhaps I was to flippant. Spanish roads are better than German ones.I haven't been to Japan.

      The real question is: Would high public spending on infrastructure cur the recession? That is what SLW says. Of course it might help some peopl, at the price of raising public debt. But do we really have the tools to stop a recessions and bring about the same GDP we had before the recession?

      That is where SWL loses himself in fantasyland.

      Delete
  11. And what do those construction workers do with their new income? For all but the very rich, income~=expenditures, so they spend it on food, clothing and recreation. So the business of the local sandwich shop, department store, and mini-golf all increases, they hire more staff, who also spend their money- hence the multiplier effect, which is always larger when people are out of work.

    ReplyDelete
    Replies
    1. If it was as simple as that...

      What if they pay off their mortgages? Isn't that the definition of a balance sheet recession?

      Anyhow, multipliers are so controversial that they are largely articles of faith. What is the historical evidence?

      Delete
    2. @Anon 13:36

      If they pay off their mortgages that may make banks a little more comfortable providing new mortgages and loans to small businesses. Even if none of this happens the government has bought infrastructure at a discount, infrastructure that would have to be built eventually anyway and which can make a region more attractive for development by corporations which have saved up capital.

      I do agree that multipliers hardly have scientific rigor behind them. I'd say economic stimulus should be done adaptively, meaning it can easily be scaled up or down every say, six months.

      Delete
    3. Banks are awash with cash.

      Of course, better infrastructure is good.

      But what we are discussing is the theory that public expenditure can fully cure recessions and produce sufficient growth even if a large part of the population has to pay off debts.

      SWL claims that that would have brought the average household - not just the construction workers - 1o,ooo pounds.

      There are people who studied economics and then thought macroeconomics was a fairy-tale,

      http://www.enlightenmenteconomics.com/blog/index.php/2012/10/unicons-higgs-bosons-macroeconomics/

      SWL is doing his best to confirm that.

      And to plug a political party.

      Delete
    4. "SWL claims that that would have brought the average household - not just the construction workers - 1o,ooo pounds."

      He's just trying to illustrate what the GDP numbers mean. He's obviously not saying we'd all have been given a check for ten grand. He means the money could have been used for public benefit rather than wasted. You can think of that as propaganda if you want

      Delete
    5. Sam17 February 2015 at 16:12

      If you are right then it is a clear case of propaganda.

      But you are wrong: Nowhere does he say that those 10,000 pounds would have been used for public benefit.

      Delete
  12. Well argued, as usual, Professor. Keep up the great work!

    ReplyDelete
  13. Dear Professor Wren-Lewis, I am a big fan of your blog.

    I actually have just started a blog of my own and, inspired by the likes of yourself and Michael Pettis, have come up with a suggestion to counteract the flow of excess savings into the UK by using QE and writing off government debt to the same value as the current account deficit.

    If you had time to read it I would be most honoured.

    http://andricopoulos.blogspot.ch/2015/02/why-bank-of-england-should-write-off.html

    ReplyDelete
  14. This is a very nice and informative post i am very helpful from your post. Thanks for share it.

    ReplyDelete
  15. We shouldn't waste ANY resources.
    Job guarantee funded by printing.

    ReplyDelete
  16. "In a subsequent post I want to consider how reasonable it is to assume potential GDP is
    independent of actual GDP"

    I look forward to reading this upcoming post. It seems to me that there could be a link through hysteresis, at least partially in employment markets. Rod Cross at Univ. Strathclyde has long spoken of "strong" hysteresis in the unemployment rate, so that the equilibrium unemployment rate depends on historical performance. I did some work with Rod on this (as a tame mathematician) and he was certainly convinced and convincing. It would certainly provide a link between potential GDP and actual GDP.

    A recent discussion paper by Rod is here:
    http://repo.sire.ac.uk/bitstream/10943/547/1/14-002.pdf
    and some of the work we did together is here:
    http://www.tandfonline.com/doi/abs/10.2753/PKE0160-3477340302#.VONrmPmsW6k

    ReplyDelete
  17. Applying the current, and possibly, the next government's front loaded "austerity" model to public funding of higher education, [...] in the social sciences, I think the last seven years have shown that the "economics" subject, for one, has not proved to be value for money.

    ReplyDelete
  18. The judgement on the UK's austerity programme must revolve around three questions:

    1. In the summer of 2010 were the Tories correct in their analysis of the problem and was austerity the first best option?
    2. In 2011-12 should the Tories have changed to plan B when there was a danger of double dip recession?
    3. Was the recovery in 2013 unnecessarily delayed due austerity or was our recovery (earlier that another major economies) at least partial the result of government policy?

    I am not sure that answers are as obvious as some would think!

    ReplyDelete
  19. Recognising that government 'investment' is really shifting value from the productive part of the economy to a less productive part, what if the 'multiplier' is less than 1? Where would that put us?

    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.