Winner of the New Statesman SPERI Prize in Political Economy 2016

Thursday 17 July 2014

Public Investment and Borrowing Targets

Often fiscal rules, designed to keep a lid on public deficits or debt, exclude borrowing for public investment from any deficit target. This is true of the UK government’s fiscal mandate, which seeks to achieve a cyclically-adjusted current budget balance within five years. The idea, in simple language, is to only borrow to invest. What could be wrong with that?

Most of the time public investment is not like private investment. A successful private investment will generate future income which can pay back any borrowing. A successful public investment project may raise future output, and this may increase future taxes, but there is no sense in which we would only undertake the project if we could be sure of paying off the borrowing with these extra taxes. A public investment project should be undertaken if discounted future social benefits exceed its costs. This cost has to be paid for by higher taxes at some point, so the question is simply when taxes will increase to do so.

In thinking about when to raise taxes, the obvious principle is tax smoothing. If taxes are distortionary, it is better to spread the pain. So if we need some additional public spending for just this year, one way to pay for it is to borrow, and use higher taxes just to pay the interest on that borrowing. That smooths the distortion over time. This is true whether the public spending involves consumption or investment. In contrast, if we are planning to raise public spending permanently, taxes should be raised by the amount of the increase in spending, and no borrowing should take place. Again this is true whatever the form of the additional expenditure. Now it is true that public investment projects tend to be temporary, while additional public consumption can be permanent, but the principle here is how taxes are distributed, rather than the nature of the spending.

This simple application of tax smoothing takes no account of distributional issues. If we believe that government consumption only benefits those paying taxes at that time, we might want taxes to rise with a temporary increase in government consumption rather than being smoothed. Why should future generations pay for the consumption enjoyed by the current generation? Here public investment would be different if it benefits both current and future generations. So from a distributional point of view, it might make sense to treat government consumption and investment separately. There are two problems here though. The first is that the distinction between public investment and consumption in the statistics does not necessarily follow this distributional logic. Education is classed as consumption. Second, how in practical terms do you allocate taxes paid to benefits received from public investment? (I touch on this here.)

One of the key points that Jonathan Portes and I stress in our discussion of fiscal rules is that rules have to balance optimality when governments are benevolent against effectiveness when they are not. One feature of periods of austerity is that public investment often gets hit hard. The reason this happens may also reflect intergenerational issues. To the extent that public investment benefits future generations, they are unable to complain when it is cut.

This can be one reason why rules sometimes use current balance targets rather than targets for the overall deficit. If public investment does not influence the target, it need not be cut. (This does not seem to have worked with George Osborne, as the victims of flooding found out!) However such rules are inevitably incomplete, because they say nothing about the overall level of public debt. In the case of the last Labour government, there were two rules: one involving the current balance over the cycle (only borrow to invest), and one specifying a total debt ceiling. There was an implicit target for public investment implied by the conjunction of the two rules, but it is unclear how sensible that implicit target was.

Jonathan and I suggest that the simpler and perhaps most effective way of preventing public investment being squeezed in times of austerity is to have a specific target for the share of public investment in GDP. Of course this target should also influence any overall deficit target, but if you want to protect public investment, it seems best to do so explicitly. If you do that, then it makes more sense to have just one target for the overall deficit (primary or total) that includes borrowing to invest, rather than a target for just the current balance.


  1. Very well explained. But I wonder: if we want to move from the model of borrowing then paying the interest costs to the model of paying up-front (ultimately very similar if capital spending is steady), how do we avoid one cohort of taxpayers paying twice over in the transition?

    i.e. in previous years we may have spent £30bn a year in interest on borrowing-financed capital, and further down the line - in your model - we would have moved to instead paying £30bn more each year in non-borrowed capital spending. But if we closed the overall deficit immediately, wouldn't we (a particular cohort of taxpayers) be paying £60bn a year - an intergenerational unfairness?

    1. If I understand correctly I agree. In fact any programme that involves reducing long run debt is intergenerationally unfair, which is partly why I do not think this kind of fairness should over influence what we do. As I've written in the past, if you insist on Pareto optimality in debt policy you have a recipe for deficit bias.

  2. “So if we need some additional public spending for just this year, one way to pay for it is to borrow, and use higher taxes just to pay the interest on that borrowing.” But the reality is that TOTAL government investment is pretty constant from year to year, thus the latter reason for borrowing scarcely arises.

    “If taxes are distortionary…” Where does that idea come from? Taxes CAN BE distortionary (e.g. a tax on red cars but not other cars). Conversely taxes can be very distortion-free: e.g. a flat percentage tax on everyone’s income. But even where taxes ARE DISTORTIONARY, that’s normally for what we believe to be good reasons (e.g. much higher tax on whiskey than soft drinks).

    “Why should future generations pay for the consumption enjoyed by the current generation?” It’s not PHYSICALLY possible for people in 2020 to produce steel and concrete used for constructing a bridge in 2014. That involves time travel. Building a bridge in 2014 is inevitably paid for by people in 2014, or earlier.

    “One feature of periods of austerity is that public investment often gets hit hard.” True, unfortunately. Personally my definition of lunacy is “cutting public investment in a recession”, but we all have our favorite definitions of the word.

    And my final word on public borrowing is that it is all pointless or nonsense (to use technically correct economics jargon). Milton Friedman advocated a “zero borrowing” regime, as does Warren Mosler.

  3. "Most of the time public investment is not like private investment."

    Why not focus on the exceptions, for a change, sustainability investments?
    Investments in
    - solar energy (either direct or indirect via wind, tides etc.),
    - geothermy (tapping earth's subsurface heat),
    - circular economy,
    - strengthening rather than exploiting the waste processing capacity of ecosystems,
    - etc..
    That requires long-time perspective and rethinking the discounting calculus itself, hurdles only a government can take.

  4. Interesting approach. However the suggestion of having a target for the share of public investment in GDP seems to me as easily contestable. Public investment is a long term strategic variable that in a market economy should be well founded on "externalities", "market failures", ..."social benefits" as stated in 2nd paragraph

    1. I absolutely agree, which is how you arrive at a value for the target. What the target does is to protect these microeconomic evaluations against short term political pressure.

  5. A couple of remarks, touched on above, where I think more clarity of discussion is needed.

    1) Public investment may, but may not, result in additional income. This is true obviously for direct income - hospitals in the UK are sources of future costs rather than revenue - but can also be true even when national income as a whole is taken into account. Thus hospitals may increase national income by reducing losses due to ill health - but there's no guarantee. More clearly, a new History faculty building at a University will almost certainly not increase income in any sense at all, but that emphatically does not mean it is not worth doing. This means that even in a recession, public investment must be carefully tested against the future claims it makes on both tax revenue and national income. The argument, which I generally like, that the right time to invest is when the private sector is not using the available resources, does not protect us from costly white elephants.

    2) I would like to underline the temporality issue raised by Ralph Musgrave. The extent to which current spending impoverishes future generations (and current saving enriches them) is far from clear in my mind and I have yet to see a good treatment of the question (if you can recommend one I would be grateful). So the amount of labour, concrete and electronics mixed together now does not necessarily change the quantities available in ten years time. (This doesn't apply to the consumption of non-renewable resources but the day I see proper economic treatment of that issue I will feel the economists' work is largely done!)

    So in sum, it's complicated. I do think public investment is good, especially when private activity is not 'pulling its weight' but the considerations are far from simple.

  6. This cost has to be paid for by higher taxes at some point, so the question is simply when taxes will increase to do do do investment planning


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.