Winner of the New Statesman SPERI Prize in Political Economy 2016

Thursday, 11 June 2015

Should we aim for budget surpluses?

Does it make sense to target a budget surplus in normal times within five years, as George Osborne suggested at the Mansion House last night? I’m afraid any answer to that has to first respond: define ‘target’ and ‘normal’. We do not have those details at the moment, so I’ll try and finesse them by asking whether it makes sense for the budget to be on average in balance within five years: more surpluses than deficits, but the occasional (abnormal) large deficit. [1] In this post I’ll ignore problems associated with the Zero Lower Bound for interest rates, which is a very good (irrefutable?) reason why we should not be seeing any fiscal tightening right now. Here I’ll focus on the longer term.

This question is really the same as asking what the long run target for government debt should be. I recently discussed an IMF paper which suggested that, as long as the market was happy buying the debt, there was no need for the government to reduce the level of debt from current levels (around 80% of GDP). That policy would imply running deficits of around 3% of GDP, which is a long way from a surplus. I also said that might be an extreme position. In this post I gave various paths for deficits and debt, where the other extreme was balancing the budget. A balanced budget could involve debt falling rapidly to around 40% of GDP by 2035, and by 2080 the debt to GDP ratio would be close to zero. I also gave various paths in between these two extremes.  

So which should it be: keep the debt to GDP ratio at around 80% as the IMF suggest, or get it to fall rapidly as George Osborne suggests, or something in between? Consider some popular arguments for going with George Osborne.

1) It provides scope to respond to another Great Recession without running out of what the IMF call fiscal space.

This is right in principle, but the numbers do not imply we need to get debt down that fast, unless we are expecting the equivalent of Great Recessions to happen in the future much more often than in the past. The IMF paper has some calculations on this (pages 12 and 13), and I looked at a particular experiment here.

2) We need to reduce the debt burden for future generations.

Under the assumptions in the IMF paper, the costs of getting debt down now exceed the future benefits. Again, that might be too extreme, but it would be very hard to justify a quick Osborne like reduction in debt on distributional grounds. That would mean that the costs of reducing debt would largely fall on the same generation that suffered as a result of the Great Recession, which would seem perverse.

3) Any individual would always want to pay back their debts quickly

Bad analogy. Here a country is more like a firm. Firms typically plan to live with permanent debt, because it has paid for its capital. The state has plenty of productive capital. To put the point in distributional terms, if we paid back most government debt within a generation, we would be giving that capital to later generations without them making any contribution towards it.

So it is hard to justify aiming for budget surpluses within the next five years. But I want to make one final point. How quickly you should reduce debt involves difficult technical issues. While I’m reasonably sure that the extremes of keeping debt at 80% of GDP or going for surpluses within the next five years are not optimal, that leaves a wide range of possibilities in between, and neither theory nor evidence gives us much guidance at the moment. This really is an area where more research is needed [2], and it would be good if the Treasury - the main interested party - was promoting that research. What we get instead are jokes about reactivating the Commissioners for the Reduction of the National Debt. (It was a joke, surely?) Sign of the times, I’m afraid. 
[1] It makes no sense to target any deficit/surplus number on an annual basis. The budget deficit should be a shock absorber, to prevent volatility in things that matter, like tax rates and spending decisions. (Shocks can be cyclical, but they can arise from other sources, so cyclical correction - even if it could be done well - does not negate this point; see Portes and Wren-Lewis.) That is why the coalition originally had a target for the deficit in five years time, which makes sense because it allows the deficit to be a shock absorber.

[2] Yes I know this is what academics always say, but on this issue it is absolutely true. Compared to the oceans of work on monetary policy, work on optimal government debt amounts to a puddle. One reason may be that central banks are good at encouraging and utilising academic research, whereas finance ministries are less so. 


  1. "1) It provides scope to respond to another Great Recession without running out of what the IMF call fiscal space."

    People, especially on the right, say this all the time. They don't mean what you think this means. They obviously don't mean create a low debt to allow for fiscal stimulus, because they don't believe in fiscal stimulus.

    And I don't think sitting on low debt, refusing to increase government spending, during a financial crisis is much help. Is it better than sitting on high debt, refusing to increase government spending, during a financial crisis? Hard to say, I suspect probably not a large difference.

    I absolutely agree there is little academic research on optimal government debt level (what about optimal private debt levels?), given that it is by far, the biggest *OBSESSION* of politicians and pundits. Why don't academics work in this area?

    1. "People, especially on the right, say this all the time. They don't mean what you think this means. They obviously don't mean create a low debt to allow for fiscal stimulus, because they don't believe in fiscal stimulus."

      Maybe - but most economists "on the right" do think that automatic stabilisers should be allowed to work (and if they are skeptical of discretionary fiscal stimulus it's that they worry about the efficacy of spending money quickly for the sake of spending money). See Lucas and Stokey (1983).

      In any case, the mere fact that Osborne is allowing the balanced budget rule to be relaxed outside of normal times is evidence against your concern.

    2. Obviously, "automatic stabilisers should be allowed to work". After all, they are automatic. No one is proposing to remove jobless benefits. The point is, if one's view is that fiscal stimulus is bad, producing a surplus because it "provides scope to respond to another Great Recession without running out fiscal space" is a nonsense argument.

    3. I think you may be missing something. You may need plenty of fiscal space for the next recession not to conduct countercyclical fiscal policy, but to bail out the financial sector. The implicit subsidy to our large banks has not gone away, but it relies on the government having the fiscal space to rescue the financial sector again.

    4. "No one is proposing to remove jobless benefits."

      Really? Is the squeeze on the jobless illusory?

    5. In order to answer the question “What’s the optimum size of the national debt?” one first has to answer the question, “should public investments be funded by borrowing or by tax?”. I just can’t get my head round that one. A Swiss academic Kersten Kellerman argued for the tax option.

      Second, I like SW-L’s reference to “implicit subsidies” paid to banks. The phrase “actual subsidies” would be better. The Fed loaned THIRTEEN TRILLION to sundry banks during the crisis. That’s best part of the US GDP!!. And the rate of interest? Well it certainly wasn’t Bagehot’s “penalty rate”. Much of it was loaned at a zero or very near zero rate. Some subsidy that.


  3. Are you not flattering Osborne by engaging in economic based arguments with regard to his plans for deficit reduction and ultimately bringing down the national debt when it is really it an ideological pursuit of having a much smaller state sector?
    (This doesn't mean you shouldn't continue the outstanding work in highlighting this government's economically illiterate policies by the way!)

    1. It does appear more of a stunt than anything else. An opportunity to wave his, err, something, in parliament and win big a vote. Oh and endlessly berate any opposition for voting against 'common sense' policies - like running a government like a household.

      Obviously I think it's nuts to try and run a normal times surplus within 5 years - just because of how far from that position we are at present. But it's not like there's a law against it...

      However, if you want to lock something down in a law, you should have the strongest evidence or theory basis for doing so - and that just ain't the case here. Passing laws willy-nilly is... well, it's not small 'c' conservative.

    2. Jason,

      It’s not so much a “stunt” as a trap for Labour. If Labour says what the intelligentsia knows perfectly well, namely that household analogies are nonsense, then Tory tricksters will accuse Labour of being irresponsible with the housekeeping money, and most voters will believe that. BTW, the intelligentsia consists me and a few other people..:-)

    3. Labour tried the tactic of ignoring the "Labour Spending Crashed the Economy" lie for the last five years. It was not successful, and gave the Tory tricksters a free run on the economy. Virtually everybody now, including Labour voters, think that it was excess spending by Labour in the years 1997-2007 that caused the crash and the huge increase in the deficit in the years 2008-10.

      If Labour ever wants to be in government again it has to counter Tory lies with the truth - and keep on telling it every chance they can.

  4. Putting these Keynsian theories into Law is like passing a law to say that the Sun orbits the Earth.

    1. Very ironic, coming from a man who treats quantity theory of money as if it's some amazing new insight hitherto undiscovered by the economics profession.

    2. I'm not saying it's new.
      Just that it is the first thing forgotten (ignored) when there is a desire to 'do something'.
      Now, if an economist came up with a suggestion that respected QTM then I'd be all ears.

  5. "We would be giving that capital to later generations without them making any contribution towards it."

    What a fantastic way to put it, I'd never conceptualized it like that.

    1. Maybe; if it were true. Don't forget that the unobservable public capital is not just funded by marketable debt, but also by the much larger (twice the size of the "national debt") non-marketable liabilities in the form of unfunded pension promises.

  6. Surely there's no possibility that it's going to happen anyway. If you look at the stock flows in the OBR projection, in the absence of any private sector investment, the future surplus relies on heroic levels of private indebtedness, and a huge drop in the trade deficit. Are either of those really likely?

  7. SW-L says in connection with Osbo’s debt reduction proposal: “It provides scope to respond to another Great Recession without running out of what the IMF call fiscal space. This is right in principle…”. Nope. Wrong in principle and for the following reasons.

    The fact that the debt is on the high side (e.g. 80%) in no way hinders stimulus should the latter be needed and for the simple reason that simulus DOES NOT require an increase in the debt: stimulus can be funded by new base money. Keynes pointed that out 70 years ago, though today’s so called “professional economists” don’t seem to be acquainted with Keynes.

    A quite separate and important question is: what rate of interest is being paid on the debt? If it’s anything much above inflation (i.e. if the REAL or inflation adjusted rate is positive) then it needs reducing. But that’s child’s play: print money and buy back the debt and as to any excessive inflationary consequences of that, deal with that by raised taxes (and/or public spending cuts) and “unprint” the money collected.

    Put another way, the “fiscal space” concept is complete rubbish, as I explained three years ago and as Bill Mitchell explained recently. See respectively:

    1. Let's assume that the government controls the central bank, so can finance any fiscal stimulus by printing money. That is probably fine in the short term, because inflation is not an issue. But once the recession is over, it will be. In that case the government has to reduce base money. It would therefore seem as if printing money does not avoid the fiscal space problem, but just postpones it.

    2. Tax. That's what it is for, no? To allow a government to cool an expanding economy so preventing inflation. I agree with Ralph on this occasion, there is no fiscal space problem, if there is a retrenched recession, zero lower bound, under utilised workforce, near zero inflation, pay settlements quiet, record low interest have as much fiscal space as you are ever going to have.

      Further, which government does not control it's central bank, surely that is prior? The UK most certainly does, it appoints it's chair, issues it's remit and can make it disappear with a wave of the Westminsterial hand. I'd call that control. A government running a fiat, non-convertible, free floating currency system can always "afford" fiscal stimulus. Yes inflation but only when we are running at full capacity...when was that last?

    3. The issue here is not whether the government can stabilise the public finances, it is whether it will. At some point debt becomes so high that default becomes politically more attractive than increasing tax or cutting spending. As long as that point exists, and some point markets will - quite rationally - start demanding a default premium.

    4. Japan demonstrates how far you can go without reaching that point.The fact that the IMF give it zero fiscal space (less than Ireland and Portugal!) shows how credible and useful the whole concept is.

    5. Simon, read this post:

    6. "some point markets will - quite rationally - start demanding a default premium."
      And yet JGB bears have been humbled.
      What's to stop this from happening at a lower level? What is the "some point"? Are there estimates? What if tax rises and spending cuts depress the economy and make the deficit worse? How much power do these " bond vigilantes " have?
      And let's say the government borrows from the central bank/prints money. So we swap cash with bonds. If the bondholders are saving, where is the inflation risk? And if they spend, the money gets spent and respent and as it goes round it generates tax, closing the budget deficit and no need to issue so many bonds. The money is either spent or becomes savings.
      Govt "debt" with a currency issuing government is actually the national savings.
      When the govt spends £100, it creates an amount of tax and amount of saving, as the money goes round and round. Tax is like cash back on the govt's credit card. If there is no saving, the government always gets all of its money back. Let's suppose there are savings. Then the government prints some bonds borrows these back and spends the money again. Bonds are money!
      Let's suppose the government just "prints money" and spends it, then the savings are just reserves/cash. Where is the inflation risk? Nowhere!
      The mainstream suppose due to the "money multiplier" this leads to inflation. But we know this is bullshit!
      Simon, please do a post on debt free money.

    7. "But once the recession is over, it will be."
      As it is if the spending is "funded" by borrowing.
      Printing money does not cause any additional inflation risk. This is seen with QE and Japan.
      Our government should abolish its 'borrowing' and just spend by 'printing money.' Rates at 0% forever! The 'borrowing' operation is completely voluntary and actually slightly increases inflation risk due to interest payments. It is voluntary spending that is corporate welfare. It is also used by conservative to frighten us.
      The limit to spending is real resources. This is the MMT view. So you agree with the MMT view? I will assume you do until you reply.
      Please endorse platinum coin to pay off the debt in the US and spending from the 'ways and means account' in the UK.


    9. The only other thing I would add is that no government anywhere spends by "printing money." There is a certain amount of money printed due to demands in the bsnking system. It is a hysterical term that conjures up images of Zimbabwe and hyperinflation.
      Government always spends money by crediting bank accounts. It can drain them with bond sales after but it does not reduce inflation!
      Will you be willing to have a discussion about your views and MMT? On what points precisely do you disagree? It frustrates me sometimes that hetredox papers are ignored and there is no open debates. I think we need debates going back to first principles.
      Anyway, thank you for the excellent blog and sorry for my many comments.

    10. "Let's assume that the government controls the central bank"
      It should be noted - central bank "independence" is complete BS! (Except in the eurozone where the governments submit to the unelected ECB for some reason.)
      Nobody can bounce a govt check.

    11. "At some point debt becomes so high that default becomes politically more attractive than increasing tax or cutting spending. "
      A government could go mad and default at any time. This has nothing to do with high public debt.
      Deficits are caused by nongovernment sector saving. So its not clear tax rises or spending cuts will necessarily lead to lower deficits.
      In reality, bondholders prefer to buy bonds rather than cash because they want a risk-free investment.
      And of course we can stop this bullshit by just not issuing debt at all.
      We are not on a gold standard anymore. No need to borrow back the savings

  8. Comment from Engin Kara via email:

    Keeping the debt-to-GDP ratio low during normal times makes perfect
    sense, I think. This is because when the ratio is low, fiscal policy
    may be more effective. I attach a figure to show this point:

    The figure comes from my paper, which I am currently revising for a
    journal. In the paper we use a New Keynesian model with Kiyotaki and
    Moore type financial frictions. As the figure shows, fiscal policy
    becomes less effective as the debt-to-GDP ratio increases. An obvious
    policy conclusion from this figure is that governments should try to
    keep the ratio low during normal times. This would allow them to use
    fiscal policy more effectively in times of crisis, when it is most

    1. Which might be perfectly relevant and valid if this government and chancellor ever had the intention of using discretionary fiscal policy in times of crisis; which of course does not mean bailing out to the tune of billions of £s those who actually caused the crisis.

  9. shouldn't the question be:

    "should we aim for private sector deficits?"

    1. No - Britain is not a closed economy.

    2. oh, so you think the question should be:

      "should other countries aim for private or public sector deficits?"

  10. Simon, This is a reply to your 11 June 2015 at 06:13 comment: the standard “reply” system doesn’t seem to work.

    The fiscal space problem (as I understand it) is that if the debt rises above some level, creditors demand a big increase in interest. If government goes for “print and spend” instead of “borrow and spend” then there’s no increase in the debt, so the fiscal space problem doesn’t arise.

    I agree the need to withdraw base money is a potential problem. Whether it’s insurmountable depends on how volatile private sector spending is. If it’s not too volatile, then no problem: an interest rate rise and/or budget surplus solves the problem.

    But suppose it proves impossible to implement deflationary measures fast enough after £X of “print and spend”. I suggest that in those circumstances much the same would occur under “borrow and spend”. Certainly private sector paper assets rise in both cases, which is the cause of any excess demand. But under the print option, the assets the private sector gets are more liquid (i.e. base money) as opposed to under “borrow” where the private sector gets bonds.

    Thus to get the same amount of stimulus in each case, the total sum involved under “print” needs to be less than under “borrow”. So I’m claiming (not with total confidence) that the “need to withdraw” problem is the same in each case.

  11. I Just wanted to say how refreshing it is to see an intelligent debate about the issues and technicalities rather than the usual myth-fest we get from the Tories and M-S-M.

    It is interesting to note that the Prof is arguing that we ought to be careful about total dept-to-GDP ratios to ensure effective responses to crises. I probably hadn't considered that enough. I do think it noteworthy how much lower the UK debt is compared with its historic highs. However I think that targeting the deficit is completely bonkers. Targeting sustainable growth is surely what we should be doing? And let the debt take care of itself in the long-term.

    I also think it's worth pointing out how Labour left the UK debt lower in 2008 than it was in 1997. But we all know they didn't fix the roof while the sun was shining coz Osborne keeps telling us this it must be true!

  12. Your economy still seems weak; a stimulus would be better than a balanced budget. But once you are fully utilizing the factors of production, then balancing the budget would be a wise idea.

  13. By the way, the future generations are likely to be richer than the current one. So why the fuss?

  14. SWL means well but ...

    A budget surplus or deficit is completely the the wrong target. To steer the ship of the UK economy the true compass is full employment. That means being able to walk out of a job in the morning and start a new one in the afternoon. By the picture on this blog SWL is just about old enough to remember when this was last possible in the UK. George Osborne should have been spit roasted and served as the main course at the Mansion House last night.

    Thats it fartig

    1. "The true compass is full employment"? Who ever said otherwise? Put another way, the ultimate objective behind the surplus / deficit argument is: how do we maximise employment without exacerbating inflation too much.

  15. Here I go again with the usual (often I even bore myself).

    I think the key issue related to the optimal public debt is Ricardian non-equivalence. I am quite confident that people mistakenly perceive government bonds to be net wealth when making consumption savings decisions -- that public debt creates the illusion of wealth. I think this illusion which strangely cohabits in most minds with the similar illusion that government bonds are net poverty -- a burden on future generations. I think that people forget about their (and their heirs) share of the liability when acting as consumers considering their wealth and optimal consumption and forget the owners of the bonds (and their heirs) when acting as citizens and deciding how to vote. A substantial part of the debt will pile on future generations twice, first when they inherit it, and second when they pay taxes to repay it.

    I note that you have written that the evidence is more against than for Ricardian equivalence. Yet in your previous post on debt seems to assume that there is Ricardian equivalence (as ytou did in your post with the guess that the government spending multiplier is roughly 1). I guess that this is the result of bending over backwards trying to be more than fair to anti-Keynesians. Certainly it is an odd to insist that government debt causes higher aggregate demand when (partially and reluctantly) defending Osborne. But it does or it doesn't whatever Osborne says.

    My guess on the effect of debt is that it causes higher consumption and higher aggregate demand for given real interest rates. To achieve target inflation, the monetary authority will cause higher real interest rates. This will cause lower investment (mostly in housing). I'd tend to guess that this should reduce welfare (illusions cause mistakes which are on average costly, although they may cancel out other mistakes). I would also guess that the effect is more important than the increase in deadweight losses do to the variance over time of tax rates (which should be second order in the standard deviation of tax rates).

    I can imagine favoring public debt if one is very worried about secular stagnation. But "In this post [you] ignore problems associated with the Zero Lower Bound for interest rates," and so, it seems to me, do the IMF analysts.

    I guess I am asking whether you believe in Ricardian equivalence or not.

    1. I'm not sure why you thought that in my previous post on debt (I assume you mean the one discussing the IMF paper) I was assuming REq held. Both there and here I'm talking about long run issues, not business cycle macro. It is the case that I think issues about long run debt are best analysed in OLG type models rather than infinite life type models, and in those models REq does not hold, but the connection seems indirect.

  16. I find this in New York Magazine interesting regarding all the ridiculous nonsense Niall Ferguson has been spreading so widely -- including but not limited to the FT. It's had to understand why anyone would take him seriously about anything.

    It might be worth a look at the FT but in the interim, here's a good paragraph from NY Mag's Jonathan Chait.

    "The most recent example of “correct politicalness” is the humiliation Ferguson suffered when the Financial Times was forced to correct his recent column there. Ferguson, the FT concedes, “incorrectly stated that at no point after May 2010 did business confidence sink back to where it had been throughout the past two years of Gordon Brown’s premiership.” Ferguson likewise committed an act of extreme deception. He wrote, “Weekly earnings are up by more than 8 per cent; in the private sector, the figure is above 10 per cent. Inflation is below 2 per cent and falling.’” That passage would seem to convey that wages had grown sharply and stayed well above the rate of inflation. In fact, the opposite is true. As the FT concedes, “Real wage growth was negative from 2010 until September 2014.” Ferguson defended fiscal austerity on the basis that it produced higher wages, when in reality it produced lower wages."

    There are links in that piece to the FT.

  17. "To put the point in distributional terms, if we paid back most government debt within a generation, we would be giving that capital to later generations without them making any contribution towards it."

    And we are giving them pollution, nuclear weapons, an ecosystem in disarray, and much, much more.

    Since future generations are not the ones to decide about which capital is provided, why should they pay?

  18. It does not make any sense to discuss the optimal difference revenues and expenditures without discussing their content. This is as true in the long run as in the short run.

    If government typically has few investment opportunities with NPV greater than the borrowing rate and taxes fall mainly on consumption and one thinks that taxpayers are irrationally under-saving, then it would be good for the government to run surpluses so the private sector could invest more than it saves. Is this what Mr. Osborne has in mind?

    Now if Mr. Osborne simply means that as a rule of thumb (because of the way government borrowing costs and opportunity costs of inputs into government expenditures move when NGDP is below trend) deficits should be higher during recessions than during booms, then he's just a conventional Keynesian and welcome to the club.

  19. "That policy would imply running deficits of around 3% of GDP, which is a long way from a surplus. "
    The UK's current account deficit is bigger than this. That net flow out of the country has to be matched somehow, via either public or private debt.
    This strategy is unsustainable, the deficit is too small leading to high private sector debt, with the associated high interest payments.

    Look at the household dissaving!

  21. "That policy would imply running deficits of around 3% of GDP, which is a long way from a surplus."
    Hmm, it just so happens 3% is the arbitrary SGP values? Just a coincidence. Ho hum.


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.