Winner of the New Statesman SPERI Prize in Political Economy 2016

Wednesday 8 May 2024

When are large and persistent increases in debt to GDP justified?


In this post I showed a chart of UK government debt to GDP since 1900. It starts off at below 50%, then rises sharply to 200% during WWI. It rose to 250% during WWII, but then fell steadily during the post war golden age, going below 50% by the mid-1970s. The next sharp rise is after the Global Financial Crisis, followed by a smaller rise during the pandemic, getting to current levels of around 100% of GDP.

This brief chronology suggests that debt to GDP (and more recently also central bank issued reserves to GDP) rise sharply in extreme crises. But climate change is an extreme crisis facing every country, so why are most governments (maybe US excluded) resisting the idea that they should be running large deficits to pay for measures to reduce the extent of climate change? Is there a clear way of deciding when it is OK to allow large and persistent government deficits and when it is not?

In this post I set out why there is. To make things simple, I will assume the case for higher spending is overwhelming, so the issue is only how it should be paid for. The obvious alternative to allowing this additional spending to generate persistently large deficits is to substantially increase taxes. There is a standard economic proposition called ‘tax smoothing’, which states that it is better to smooth taxes over time than have erratic increases or decreases. In economic terms most taxes have distortionary costs, where the costs associated with a unit increase in taxes are likely to increase with the level of taxes, so it is less costly to smooth taxes over time. It is also intuitive: if a government offered people the chance of paying no tax for 5 years followed by the certainty of paying double normal tax for the next five, then most people would reject this offer.

A key point is that while there are clear benefits from keeping taxes smooth rather than allowing their path to be dictated by erratic movements in government spending, and occasionally periods of very high spending, there are no costs to erratic movements in government deficits, government debt or bank reserves. So it makes perfect sense to try and smooth taxes, letting government deficits take the strain of both erratic and sometimes high levels of government spending.

It is tax smoothing that justifies allowing debt or reserves to rise substantially during periods of unusually high spending. However note that the higher government spending that justifies a high deficit is temporary. If it was permanent, tax smoothing would not apply, and taxes would need to rise in line with government spending, with no increase in the deficit. Even if the expenditure has long lasting benefits, such as better education for example, that doesn’t imply it should be funded by borrowing or money creation if that higher spending is permanent. (This abstracts from Keynesian arguments, which we consider below.)

The centrality of tax smoothing to thinking about deficits and debt is why many papers written on fiscal rules, including my own (see here for example) often start by setting out that idea. Tax smoothing also tells us why balanced budget rules are bad economics. As government spending is inevitably lumpy and/or erratic, balancing the budget would match that with a lumpy, erratic path for taxes. More generally it shows why fiscal rules should, if possible, [1] always take the long view, to avoid ‘shocks’ to government spending being translated into a bumpy path for taxes.

Tax smoothing also implies that in a downturn or recession the ‘automatic stabilisers’ (e.g. lower taxes, more unemployment benefits) should be allowed to operate, but obviously tax smoothing can say nothing about countercyclical fiscal policy, because the argument for such policy comes from Keynesian macroeconomics. For moderate downturns or recessions any deficits that come from either the automatic stabilisers or fiscal stimulus will be matched in later years by boom periods where we get automatic surpluses because taxes are unusually high, and also perhaps because of fiscal contractions to moderate demand. In other words the normal business cycle should not lead to persistently higher debt or reserves.

Unfortunately history tells us that occasionally we get very deep recessions, and these are not matched by very large booms. Not only will the deficits caused by the automatic stabilisers be particularly large in these recessions, but fiscal stimulus is likely to be essential because interest rates will hit their lower bound. This is of course what happened during the Global Financial Crisis.

Tax smoothing implies that large deficits during the GFC, to allow for both the automatic stabilisers and fiscal stimulus, were sensible. Once the recovery from a deep recession is complete, it makes sense to - if required - return any deficits to normal levels using spending or taxes. The great sin of 2010 austerity was starting fiscal consolidation in the middle of the recession rather than when the recovery was complete.

Investment is also a lumpy activity: sometimes you need a lot and sometimes not so much. Tax smoothing implies that it would be a mistake to match annual public investment with taxes, which is why I argue that public investment should be paid for by borrowing or issuing reserves, with taxes only matching some long term average of investment expenditure. That is why it makes sense to exclude public investment from any medium term deficit target, and also why the falling debt to GDP rule is a bad target.

Tax smoothing accounts for and justifies large increases in debt during wars and, together with Keynesian policy, implies persistently higher debt after a very deep recession. What about a global pandemic? The justification at the time for large deficits was that this too created a deep recession. In a recent post I questioned whether the extent of deficits run up in the UK and other European countries was justified, but this is really an argument about the extent of the additional spending on the furlough scheme, not tax smoothing.

How does tax smoothing apply to spending to reduce the extent of climate change? It has become increasingly clear that such spending is temporary, in the sense that green energy is likely to be cheaper once the capacity to produce, store and distribute it is in place. So we are talking about a green transition requiring temporarily higher public and private spending to change how we produce energy. Even if that spending lasts for a decade or so the tax smoothing argument once again applies. Spending to green the economy should not be ‘paid for’ by higher taxes, but instead should see a rise in the ratio of debt or reserves to GDP. [2]

How quickly should debt be brought down after one of these crises? In the stylised smoothing model where people live forever (or care about future generations as if they live forever, by leaving appropriate bequests to children) the answer is debt should never be brought back down. Even in more realistic models the answer is that debt should be brought back down very slowly indeed, as I helped show here. If you think that the current generation is more likely to be selfish than caring about future generations, then how quickly debt is brought down will also depend on how much the original spending helps future generations.

The idea that it is politically responsible to keep government debt to GDP constant or bring it down is deeply ingrained in parts of the population and much of the media. This idea is not entirely baseless, as some politicians have in the past increased deficits for their own political advantages, rather than because it made sense in economic terms. But it is equally important to recognise when it is economically irresponsible not to allow large deficits to fund temporary but essential expenditure, particularly if that expenditure might not happen if it had to be paid for by higher taxes. Spending to green the economy and reduce the extent of climate change is a clear example of that.

Note that the tax smoothing argument applies to public spending to green the economy whether that spending is classed as public investment, day to day public spending or involves tax breaks to incentivise firms or consumers to reduce their carbon footprint. The key reason why this spending should lead to higher deficits rather than higher taxes is that it is temporary as part of a green transition.

How can we be sure that a government that says debt is increasing because of measures to green the economy is not in reality abusing this idea to just avoid having to put up taxes to pay for other areas of permanent spending? Checking this in an economy like the UK is perfectly feasible, but it needs to be done by an independent body that is very familiar with the details of government spending and accounting. This sounds like a worthy task for a fiscal council, like the Office of Budget Responsibility in the UK.

[1] Portes & Wren-Lewis discusses what ‘if possible’ means in this case. The more a government (with the help of a fiscal council) can be trusted to abide by the spirit of the rules, the more long term these rules can be.

[2] A carbon tax alongside green investment and subsidies still makes economic sense, which if other taxes remain unchanged would reduce the size of additional deficits. However the political problems with implementation may mean such taxes remain well below effective levels.

No comments:

Post a Comment

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.