Winner of the New Statesman SPERI Prize in Political Economy 2016

Monday 14 March 2016

Does public investment have to pay for itself?

A key distinction between Labour’s new fiscal rule and Osborne’s fiscal charter is that the former allows borrowing for investment. When supporters of the fiscal charter treat borrowing as if it was inherently sinful (‘Labour will borrow forever’) it is natural to remind them that firms often borrow to invest and grow, and consumers invariably borrow if they buy a house. We also note that public investment can enhance economic growth. But this can lead to a confusion about whether such investment has to ‘pay for itself’.

When a firm borrows to invest, it hopes to make enough profits to pay back the borrowing. There may be forms of public investment which could raise future GDP (and therefore income and spending) such that eventually taxes rise by enough to pay at least the interest on the borrowing that made the investment possible, or even start paying back the borrowing itself. But there are two other important reasons why it makes sense to borrow to invest.

The first involves intergenerational equity. Suppose we have a public investment project which significantly enhances the quality of life, but there is no pecuniary benefit: GDP does not rise. So taxes will have to rise at some point to pay for that borrowing. But who should pay those taxes? When we are talking about investments that are long-lived, the obvious answer is those that benefit from the investment, which means future generations as well as the current generation. That can happen if investment is paid for by borrowing rather than raising current taxes.

This helps answer a point that is often raised, which is what should count as public investment and what should not. With this reasoning it makes sense to borrow whenever the social benefits of public spending are long lasting. When the benefits are short lived, spending should be paid for by higher taxes. So the relevant metric for what should count as investment in this context is who benefits most. While paying doctors or teachers more may have some knock on benefits for the future, the main beneficiary will be today’s doctors or teachers. The benefits of new schools and hospitals are longer lasting. 

The second reason for using borrowing to pay for investment is if the increase in investment is a one-off. As taxes are distortionary at the margin, it makes sense to smooth those taxes over time. Once again, that can be achieved using borrowing.  

If a lot of public investment does not pay for itself, wouldn’t borrowing only to invest mean that debt just went on increasing and increasing? What matters here is the debt to GDP ratio. If you want to keep that ratio constant, and you always run a zero current balance, then that tells you how much investment you can do. The numbers are fairly simple to work out. If the economy grows in nominal terms by 4.5% on average, and debt is 80% of GDP, net investment could be around 3.5% of GDP to keep the debt to GDP ratio constant. Osborne plans net investment over this parliament averaging 1.6% of GDP.

This leads to one final, important point. You cannot have separate goals for all three of debt to GDP, the current balance, and public investment. In Labour’s new rule, the commitment to reduce borrowing as a share of trend GDP over the lifetime of a parliament, coupled with the zero current balance target, puts an upper limit on the amount of investment the government could do. Whether that is a sensible upper limit in economic or political terms, and what you do if it is not, I will leave as an exercise for the reader.


  1. You say "While paying doctors or teachers more may have some knock on benefits for the future, the main beneficiary will be today’s doctors or teachers."

    Surely not. The MAIN beneficiaries will be today's patients or pupils. Doctors and teachers will benefit, certainly, but that is not why they are, or should be, paid more.

    But more generally, what is wrong with working with IRR including social costs and benefits in the calculation? In general, I would argue there should be two 'drivers' to encourage investment: would the social rate of return exceed other uses of the resources; and is there a risk of crowding out private uses of those resources? These are not independent, of course, as private rates of return should rise (or have risen) as the 'crowding-out' limit approaches.

    This is, of course, diametrically opposed to the 'Swabian' model of household accounting: does the government have 'spare' income it can use to invest? But I hope that is not a drawback from the economic perspective. If it is, as it seems to be, a drawback at the political level then economists need to work harder to explain, honestly, what may not be obvious to others. To any who shout 'technocrats' we could point out that many changes, now viewed as positive, have emerged because experts advocated them, politicians listened and the public was eventually convinced.

  2. Investment is not the same thing as expenditure, and I think it is unfortunate that politicians muddy the waters on the two.
    By definition, an investment is intended to make a return. If it does not, then it is a bad investment.
    Expenditure is intended to benefit the one who pays out (and, of course, it will also benefit the one who receives) which is a good thing, but it does not make it an investment.

  3. What matters here is the debt to GDP ratio. If you want to keep that ratio constant, and you always run a zero current balance, then that tells you how much investment you can do. The numbers are fairly simple to work out. If the economy grows in nominal terms by 4.5% on average, and debt is 80% of GDP, net investment could be around 3.5% of GDP to keep the debt to GDP ratio constant. Osborne plans net investment over this parliament averaging 1.6% of GDP.

    Sorry Simon the debt to GDP ratio is irrelevant.

    Instead of linking it to some debt to GDP ratio you should be linking it to the unemployment rate.

    What if the fiscal deficit was 3 per cent of GDP and rose to 5 per cent of GDP as new large-scale public works programs stimulated employment and unemployment fell from 6 per cent to 5 per cent?

    Would we consider that a deterioration that made any sense?

    What if the fiscal deficit fell from 5 per cent of GDP to 2 per cent of GDP as government austerity impacted poorly on economic growth and unemployment rose from 5 per cent to 6 per cent?

    Would we consider that an improvement that made any sense?

    You have to realise that the fiscal balance at any point in time is of no particular interest in itself, and must be interpreted in the context of what is happening in the real economy.

    What actually tells you how much investment you can do is the unemployment and underemployment rates.The size of the deficit should be linked to the umemployment rate and not the interest rate.

    1. What you say may be roughly true at the moment, but it is not generally true. Between 1970 and 2000 the OECD debt to GDP ratio almost doubled, and I do not think that was due to unemployment.

  4. Why is it that taxes have to rise to pay for borrowing?
    The UK is a sovereign currency issuer, and issues money before borrowing it, and taxing it.
    According to the sectorial balances chart produced by the OBR, the UK has a massive trade deficit, and growing private debt. Trying to balance the books in a neoliberal manner will keep the economy in stagnation.
    Surely we need a whole new paradigm - Keynesian reconstruction. Deficit spending for every day, and PQE for investment in infrastructure, until we are productive enough to get the trade deficit down.
    Taxes could be aimed only at foreign bought assets and monopolies, and those receiving corporate welfare - they do not need to rise generally, especially not for those on middle class and low incomes. What about getting rid of VAT?
    In the Abba Lerner and Keynes tradition, the aim should be well being and high levels of employment, then "the deficit can take care of itself."
    Having a goal of fiscal rules is monetarist and neoliberal, a cloak for the goals of the rich.

    1. Neoliberalism, at least as I understand it, is about a smaller state not balancing the books. Fiscal rules are a response to clear evidence that some governments suffer from deficit bias.

    2. I received my impression of neoliberalism partly as Margaret Thatchers assertion "there is no such thing as public money, only tax payers money, the State has no source of money other than money which people earn themselves. If the State wishes to spend more it can do so only by borrowing your savings or by taxing you more. It is no good thinking that someone else will pay—that "someone else" is you.

      Thatcher assumed this wrongly from Hayek, who wrote the paper "the denationalisation of money." His aim was to increase private lending and reduce government spending. No wonder Keynes thought he was "a logician, by several steps, was on his way to bedlam."

      This is why I think that the idea of balancing the books is misleading the public. It is giving people the idea that governments need the money of their citizens to spend, which is just plain wrong.

      I noticed that you have written a piece on MMT, I am glad that you express here that governments uniqueness is already established, and we do not need MMT economists to tell us.

      Keynes himself stated in his Treatise on Money that there is "Money Proper, and bank Money." One is an asset to the economy and one a liability. It th UK at the moment, private household and business have about 1.5 trillion in debt, which must be causing stagnation as Thomas Palley writes in his book about the GFC. The growth is debt driven, another housing bubble primed by the help to buy scheme.

      Surely governments do need to spend through PQE, and deficit spending (as well as controlling bank lending for housing), to reduce the ratio of incomes to debt? Preferably by creating jobs?

      Without any polemic, I would say that balancing the books is a neoliberal idea if it is based on Thatchers beliefs. But even with Keynes insight, balancing the books is not always good aim.
      If the economy has a large trade deficit, and bank money is being created for mortgages and not for industry, there is a potential for another financial crisis, so people need to earn government money to pay their debts as well as thrive. Lets face it - the banks are not lending to industry.
      Predicting how much or how long government needs to do this to reach a healthy economic state can be a little difficult.
      Reading an article by Francis Coppola, - she states that governments cannot control deficits because the sectorial balances dictate automatic stabilisers. Any attempt by government to cut deficits while a trade deficit exists would increase private bank lending, credit cards, wonga loans, etc. Either that or increases in poverty and food banks.
      Surely we are in to much of a depressed state to balance the books?

  5. I am often left uneasy by arguments based on the idea that public investment financed by taxation is paid for by those currently alive (or at least by the adults), while if financed by borrowing it is paid for by future generations. I know that in a simple OLG (over-lapping generations) model, there is a "burden" on some future young generation, when the government bonds which paid for the investment are finally liquidated. What bothers me is how far the results of that model apply in a less stylised world.

    The basic resource point is that if there is reasonably full-employment then, ignoring the foreign sector, the real resources for the investment have to come from either private consumption of those currently alive (old and/or young) or cuts in government consumption expenditure which benefit those currently alive, or from cuts in private investment. Similarly, the future GDP at any moment in time is not diminished by whether or not there are bonds to be paid off. There is only a redistribution between those alive at the time, some of whom may be elderly, some young, some wealthy, some poor.

    Among more realistic examples than the simple OLG are: some currently elderly people pay taxes; some bonds owned by retired people are bequeathed to their heirs, rather than being sold to their younger contemporaries; taxes to pay interest on previous borrowing are not a net cost to future society, since the interest-earners are members of the same society. If we do not simplify by splitting current and future societies into 2 homogeneous generations, can we still meaningfully talk about future societies paying for the current investement if it is financed by borrowing but not by taxation? I don't know, but am prepared to learn if others do know.

    1. "The basic resource point is that if there is reasonably full-employment then, ignoring the foreign sector, the real resources for the investment have to come from either private consumption of those currently alive (old and/or young) or cuts in government consumption expenditure which benefit those currently alive, or from cuts in private investment. "

      Well said. This has been one of my hobby-horses for years. Of course, costs and benefits can be transferred between different groups of people (or economic actors if that sounds better) at any given time, and sometimes these different groups may be identified by their ages, or generations. But transfers across time are investments not consumption and the income stream (or non-income benefits) cannot be brought forward or deferred.

      This is perhaps simplest to see when considering pensions. There is a strong current of opinion that argues that unfunded pension rights are imposing an intolerable burden on the future. But in fact the burden on the future, for a given level of pension payments, is identical whether funded or unfunded. We are not squirrels piling up nuts that we can eat in our old age. We are piling up pieces of paper, or electrons, that give us the 'right' to demand part of what future generations will produce WHEN they produce it. But the amount that is produced in the future, in total, does not depend on those pieces of paper or electrons. And our 'right' to have some of it (I speak as a pensioner) will only be recognised by the future pensioners if they see it as part of an equitable settlement that will bring them benefits in their turn (or if they think it's the right thing to do, of course, but hey, we're economists here).

    2. Perhaps this will help:

      But Anonymous14 March 2016 at 08:08 you are absolutely right that just because these transfers happen in a stylised OLG model does not mean they are important in reality. Ideally we would look at these issues with detailed intergenerational accounting.

  6. Funding investment via borrowing certainly SEEMS to impose some of the cost on future generations because they have to pay back the debt. Trouble is that when an investment has been funded via borrowing, the next generation inherits THREE items: 1, the physical investment (where for example its infrastructure), 2, an asset in the form of government bonds, and 3, the obligation to pay back owners of those bonds. So on balance, the next generation inherits an asset equal in value to the investment.

    That reflects a brute physical reality: if the present generation of working people expend person hours, steel, concrete etc constructing infrastructure, then it’s the present generation that bears the full cost. You can’t build a bridge in 2016 with steel produced in 2026.

    The exception to that rule occurs where a country borrows from ABROAD to fund an investment, as my namesake Richard Musgrave pointed out in the American Economic Review in 1939.

    Re the idea that funding via borrowing enables taxes to be smoothed, a weakness in that argument is that the amount that countries now spend on infrastructure and other investments (e.g. education) is so large, that specific items, like the new Forth road bridge are peanuts compared to total amounts invested. Thus large infrastructure items can easily be funded out of tax.

    Re public investments that DON’T pay for themselves (to some extent true of education and NHS hospitals), I suggest the best policy there is for us to decide at election time what % of GDP goes to health, education etc. Having done that, the amount of investment should be whatever maximises output or minimises costs in health and education. I.e. I’m not convinced that any sort of pre-conceived ideas as to how much should be invested in new schools and hospitals makes sense.


  7. Is it fair to say that the cost of any investment is always going to be borne by someone in the world at the time the investment is actually made? As in foregoing current consumption by diverting the use of real materials and labor to that investment? If that is the case then future beneficiaries of that investment are really not "paying for the investment". They might be willing to fork over a chunk of their current production to those who actually bore the costs of the investment in the past, maybe to honor commitments made by their predecessors, but I think there is a distinction there. So I think I agree with Ralph Musgrave in his comment on a previous post- that when lack of cash is the only impediment to investment then it doesn't make sense for a government to borrow to "fund" it.


    2. The real burden on future generations from consumption today is the opportunity cost of lost higher production, foregone by not instead using those resources for investment. Taxation and interest distribute that burden across individuals.

  8. "When a firm borrows to invest..."

    This, unfortunately, is a most fundamental error in analysing the fiscal power of a country like the UK.

    Unlike a firm or individual, the government of the UK is the monopoly issuer of the pound. As the monopoly issuer of a currency which floats in value relative to other currencies and is not backed by any metal or other commodity, the UK can issue as much of its own currency as is needed to purchase the goods and services the government needs.

    So, the UK government has the unrestricted ability to fully fund the National Health Services, provide tuition free education, etc.

    The national debt of the UK is nothing more than the total amount of pounds previously issued but not yet taxed out of existence. The UK will always be able to pay any debt denominated in pounds.

    What is important is whether the real resources are available for sale in the pounds which the UK government can produce without restriction at the time those real resources are needed.

    Underfunding education, health services and other important government programs is guaranteed to deprive future citizens of the goods and services they are going to need.

    It is foolish and short-sighted, to say the least.

    1. The Government creates currency, but not value.

    2. I wrote the subsequent point for you. The idea that just because the government can create money it does not have to have fiscal rules is nonsense. Just think about what would happen if the government stopped issuing debt and created money instead.

    3. It's right to say that what is important is the availability of real resources but wrong to argue that government can puchase as much as it needs just by issuing currency.

      Both taxation (permanently) and borrowing (temporarily) remove purchasing power from the private sector, so allowing government to buy goods and services without increasing competition for the resources needed to produce these. Just issuing money adds to that competition, so will tend to increase prices.

      When there is unused capacity, this need not cause concern as high elasticity of supply constrains price rises, which may even be desirable when inflation is very low. But this will change as the economy expands and government ability to puchase will be limited by private demand for tightening resources.

      At present, there is a strong case for fiscal expansion supported by money creation but that case is weakened if we do not explicitly acknowledge that issuing currency has its limits.

    4. The government Treasury creates, as the currency "issuer" the money first by spending its "unit of account", into existence, out of thin air. Only then is there "money" available in the non-government sector, for the currency "users", to buy the gilts.

      Those Gilts can also be used in a monetary policy control function if required, to add or drain "reserves" (government spending) from central bank (CB) accounts of commercial banks, to control a base interest rate, via tools like LIBOR. You have to do an add reserves before you can do a drain reserves. But there are simpler, less costly ways of controlling a base interest rate.

      So your sentence "Just think about what would happen if the government stopped issuing debt and created money instead." Is not applicable to the UK or the USA; or any other sovereign floating fiat currency economy. The government does not have to issue Gilts, it does it voluntarily.

      BUT BUT BUT, your sentence is applicable to the Eurozone, where member states issue debt instead of money.

  9. It's a good and useful discussion but I think you've missed an intermediate step. Government investments that only increase GDP but do not pay for themselves can still be very beneficial. If the government constructs a road that does not quite pay for itself in terms of taxes, it would surely pay for itself more than twice over in terms of net benefits for the country as a whole. My point being that the government, unlike private companies, cares about the benefits that accrue to the population, even the part that is not taxed.

    So in addition to supporting investments that improve well-being but not GDP, the gov. should make investments that provide a net benefit for the country. With a downward adjustment for the cost of having to impose distortionary taxes, of course. A benefit to this argument is that it can even be used to convince companies of the gains from higher public investment.

    1. Absolutely, and apologies if I appeared to say otherwise. That was the whole point of considering who should pay for such investments.

  10. I'm worried about the idea that we should ask future generations to pay for our initiatives. Haven't we already, through our degradation of the natural environment, imposed enormous liabilities on future generations?

    1. What pays for it is real resources used now. Please please for the love of god stops this nonsense I am one of the 'future generations' and we are suffering from austerity.

      "enormous liabilities on future generations?"

      Issued in the state's own currency. My point is the government does not have control over that. All government spending will come back as tax if there is no saving in the spending chain. It is a private sector asset - savings.

      MMT talks about the build up of excess savings in the non-government sector - these 'savings net of investment' - and it allows them to build up first in the causal chain. That means that the non-government sector can be seen as 'pushing' money onto the government sector. This is the view that MMT takes and bases its economic model upon - the non-government sector is the one doing the 'net saving' which the government sector then has to react to.

      The question is: does anybody get to say NO and make it stick.

      Always remember that the only reason you can't spend as much money as you want is because the bank will bounce your cheques (or the central bank/regulator will declare the bank insolvent) *and* the courts and enforcement officials will back that decision up - with force if necessary.

      Nobody gets to say 'no' to a sovereign government spend request, because if you do then you are simply replaced with somebody who will say 'yes' - due to the ownership/power over the currency issuer.

      There is *nobody* in this system that can bounce a government cheque and there is nowhere else for anything HM Treasury spend to go other than to other accounts at the Bank of England. So it all just bounces back and forward intra-day and settles up nicely at the end of the day (with DMO borrowing back from the banks like any other bank does if they are short of reserves to hit the arbitrary end of day 'clearing' figure).

      Everything is always fully funded because the money can't go anywhere else. It's like sitting on a water bed. And that's the key point.

      When you look at France, the money there can leak out to Germany - because they are in a fixed exchange system called the Euro.

    2. Patrick: you make an important point (and random is just MMT obsessed so he fails to see it). The problem is that if you say all public investment has to be paid for out of current taxes to compensate future generations for the losses we are imposing on them, the result will be that investment will not happen, which adds to these losses.

    3. Thanks. I'm trying to understand. I can see that there may be a ceiling to the level of tax that the present generation is prepared to pay, such that if some of the burden is not passed onto future generations then important public investment will not happen. Presumably, though, there is also a ceiling to the level of tax that future generations will be prepared to pay. By requiring that some of the tax that future generations pay go towards repaying our debts we are surely reducing the amount that remains for them to invest in projects that seem to them to be important (investments which, thanks to improved technology, are likely to be more efficient than the investments that we may make.) I can quite see that resources are being wasted if we fail to implement fiscal stimulus to escape economic stagnation, and that the plight of future generations may be worsened by our failure to do this, and also that future technological improvement depends on present investment. This, though, it seems to me, is not an argument from 'intergenerational equity.'

  11. On the question “what should count as public investment and what should not” you answer “it makes sense to borrow whenever the social benefits of public spending are long lasting”. I agree with the principle here but it raises the question of how this aligns with the public accounts. The ONS now treats as investment R&D but not skills formation, although it is clear that the benefits from that are also long lasting.

    So should McDonnell’s ‘borrow to invest’ promise be interpreted as applying just to Gross Fixed Capital Formation or should it be broader than that? I would favour the broader view for the reason you give but it risks leaving Labour open to the charge of ‘fiddling the books’ by treating as investment some of what the ONS treats as current spending. So if that wider view is to be taken, then it will need to be explicit, transparent and bounded, with any current spending treated as investment (e.g. for skills development) handled through a distinct funded programme for visibility.

    1. place time limits,like banks do 1,2,3,4,5 years terms a government could make such borrowing over any period of time,although i think a robust criteria would be wise.

    2. In a way my post can be seen as a starting point for trying to do just that.

  12. Whether the old CBA meant much, it was a useful challenge mechanism for investments especially on transport schemes. Does Osborne's pay now approach not mean investment programmeslead to higher taxation now and in the future?

  13. You are of course correct to state that the commitment to reduce the debt to trend GDP ratio places an upper limit on the amount of public investment that could be undertaken. With the current numbers, that gives enough headroom to double investment so I don’t expect that to constrain priority spending on this in the next Parliament. As and when the debt ratio has been substantially reduced then it might do so but a future government could then adjust the rule if required.

  14. A very good piece only one other distinction to make that between assets that can be paid off for the next generation and those that can't. take landlines and drainage the cost to both incurs a standing charge that pays off the debt the actual cost and then some more,this is a tax and if given a life span per unit,then in time it will free up money to create new investment because it will not drag on economic activity but boost it

  15. "But this can lead to a confusion about whether such investment has to ‘pay for itself’."

    I'm quite careful there in not even trying to make the claim that the increased tax revenues will mean that investment pays for itself. That's actually a view I ascribe to Murphy, not me. On the reasonable grounds that he does actually say that happens.

    The point I make is that the investment itself must be an economically sensible one to make in and of itself, before we start to think about how we're going to finance it. Which is really pretty conventional of me, isn't it?

    1. I agree, and if I have misinterpreted what you wrote my apologies.

  16. Thanks Simon, there is being developed on this blog a series of cogent discussions that could greatly improve the understanding of macroeconomic policies for the governance of Britain today and in the future. For me an important part of this is the comment section as it questions and expands upon proposed policies.
    Thanks to all of you.
    I must now get these sorted into a set distilled arguments with which I can better challenge the 'Swabian economists' that make up the majority of the general populous (perhaps most of those who have not studied economics) and journalists. These need to be good and easily communicated as New Labour allowed the 'Swabian economic' fallacy to flourish for so long without challenge it's become a near universally accepted fact amongst the general public.
    I still can not comprehend why Labour let that happen, but that's another question and of a different discipline.


  17. Simon I hope you do teach a few of your MIT friends what actually goes on over here, even if you believe in sticky price optimisation models.

    Krugman says:

    "t’s surely worth noting that other advanced countries, with much more generous welfare states, aren’t showing anything like the kind of social collapse we’re seeing in the U.S. heartland. ..."

    Krugman clearly has not ventured north of Oxford to our former industrial heartland, but I would imagine you have. You would also agree that key to the reason why lnorth western European continental countries have not had such effects is very related to things like their centralised wage bargaining systems- (which MIT economists oppose on the basis of theory!) rather than social security provision.

  18. Have you heard the latest public 'investment' proposal?
    A tunnel linking Manchester to Sheffield.
    At a cost of 6 billion pounds.
    6 billion pounds to enable you to go into a hole in the ground in one Northern city and come out the other end in basically the same thing.
    How on earth is that worthwhile?
    You might as well dig a pit at each end and let people go in and climb out again.

  19. Simon, you reference your 2012 post several times in responses to comments. This is certainly a dense post that repays rereading - several times - but I am not convinced that you have to get into the details of r>g to determine the time-effects of debt.

    Surely the issue is simpler: failure to invest will reduce economic resouces available to future generations. So the obligation is to ensure that sufficient output is put into investment (irrespective of who owns or finances the investment) to maintain or increase future output. This investment may very well be devoted, in part, to reducing the physical resources consumed in output, of course.

    Who pays for the investment, and who benefits from the future output it generates, is (like most important economic questions today) a distribution issue that macroeconomics tends to assume away. But that distribution is among actors today (and for future outputs, actors alive when that output is produced) and not between generations and time periods.

  20. If investment payed by deficits:
    we are left with :
    i) road
    ii) increased money supply
    iii) debts

    iv) if in times of underemployment; additional employment with satisfied citizens that improved their lives and their children.

    If payed by increased taxes:
    i) road

    What a difference the type of financing investment makes:
    3 to 1 benefit.
    And debt bonds is usually left unpaid by future generations which is why governement debt keeps growing indefenetly. Even the interest on debt is payed by more debt.

    Does anyone know a generations that payed off previous generation's debts? Since ever? I do not know. Even Germany is piling on debts of previous generations and does not pay it off even tough there is large trade surplus that keeps accumulating.

    Government debts are never payed off by any of the future generations.
    Future generations are left with roads to use, increased money supply and more employment to enjoy while debts are forever postponed. Free lunch.


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.