(Writing this post in multiple parts because 1) it ended up being longer than I intended and 2) work has got in the way of my finishing it)
In the financial year of 2018-2019, the UK government borrowed £24.7bn. For the following year it was £103.7bn and forecasts are that it could be as much as £250bn for the current financial year. Some people think that this is a problem, though in my experience they rarely seem to explain why. The folk understanding is that if the UK government borrows one pound today, then at some point it must raise a pound’s worth of tax revenue to settle this debt. Neoclassical economists tend to write equations into their models which look a little bit like this:
If g is government spending in period t, R is government revenue, and r the real effective yield on government debt, then this says that when t is at infinity, government debt must be zero. Most of the time, we need this to make sure that the model actually has a solution, but the danger is that this is taken literally as a serious concern. Infinity is a very long time in the future. I am no astronomer, but my understanding is that our sun will burn away its hydrogen stores and become a red giant in around 5 billion years. At this point, I suspect that ensuring that all those who have lent to governments are made good will not be a priority. Furthermore, it is even possible that the economic system which has endured so far for around 300 years will not last another 5 billion. And again, at the point of transition from capitalism to something else, where perhaps the concepts of government and debt are as alien as the concepts of serfdom and the divine right of kings are to us today, I hardly think the revolutionary vanguard will pause to honour the capitalist state’s financial commitments. It is actually highly suspect to claim that government debt reflects a claim on societies that they will in any way have to honour one day.
Perhaps a more relevant concern about debt levels is that, as Simon Wren-Lewis regularly argues, higher debt means higher interest payments. And this means higher taxes to pay for them. Well, maybe. What if you make the payments by issuing more debt? Then there is exponential growth in the stock of outstanding government debt. The reason we should care about this is because at some point, the government will run out of people willing to buy up its debt, and be forced into a default. Knowing this, investors will be very wary of lending to governments which look like they are on this path (so-called “bond vigilantes”) and start selling of its debt early. One way of answering this is to say that exponential growth in government debt is only a problem if exponential growth in the private wealth needed to hold it if slower. That is why the relevant statistic is the debt-GDP ratio. In the figure below we can see the difference a 1 percentage point change in the interest rate makes when growth is 2%. In fact, so long as your interest rate is lower than your GDP growth, there is a debt-GDP ratio that the economy stabilises at for any government deficit. This should keep the bond vigilantes at bay.
So where does the UK sit right now? Bond yields are in fact negative in nominal terms, and at no point on the yield curve do they even reach 1%, let alone the Bank of England’s 2% long-run target. To be clear, this means that in real terms investors are paying the government to hold their money. It means that if interest rates stay anything like where they are, and real GDP growth is above zero in the long run, then there is no major concern about the UK government being able to find buyers for its debt.
Does this mean that the UK does not have to worry about its debt? Unfortunately not. Although the UK’s real interest rate has been negative for some time, there is no guarantee that it will remain so. In fact, the gold standard estimate for estimating equilibrium real interest rates, the Laubach-Williams measure, has the rate for the UK at about 1.5%, and the real equilibrium growth rate at less than 2%. Given the UK economy’s shocking post-financial crisis record on productivity, this is too close for comfort. If the real interest rate overshoots the growth rate, then a medium-term budget surplus is required to maintain a non-explosive debt path. If this would have been a faff at the 80% debt-GDP at which the UK went into the covid crisis, it is probably mildly more of one at the over 100% ratio at which it looks set to leave it. One should not exaggerate what this means (mildly higher taxes, or mildly lower spending, to keep the UK on a non-explosive debt path) but nevertheless, it seems likely that governments will want to do something to reduce the burden of the covid debt.
Lest I continue sounding like George Osborne on a particularly pessimistic day, none of this justifies immediate government fiscal retrenchment, for (at least) three reasons. Firstly, debt is never a short-term problem. When the debt-GDP ratio is on an explosive path, the government usually has plenty of time to back out of it. We are living in a continuous world and will therefore have plenty of warning before this becomes a problem. When I had a proper job I used to write about economies that genuinely had debt problems. Even they have ample corrective action available. For this reason, countercyclical fiscal policy is more important than debt sustainability when there is an output gap (i.e. spare capacity in the economy). A shift return to potential output would make it easier to reduce the debt burden through higher taxes without jeopardising growth. We should recover from the coronavirus crisis, stabilise the economy, and when full employment is reached make some tweaks to get the debt-GDP ratio on a gently downward-sloping path. The mistake the coalition government made was to start this fiscal consolidation before this stabilisation occurred, resulting in a flatlining of the recovery and an unnecessary prolonging of the 2008 recession.
Secondly, and relatedly, additional government spending may not add to the debt burden at the moment. De Long and Summers show that if monetary policy is at its effective lower bound (i.e. monetary policy has maxed out its capacity to stimulate the economy) and if a faster recovery from recession now means even small amounts of additional long-term growth, then government spending is self-financing. Specifically, this is because when monetary policy is maxed out, government spending does not lead to higher interest rates, lower consumption and lower investment, and so has a larger effect on the economy than usual. Furthermore, with a larger economy, we get bigger tax revenues. This situation is rare in the context of the global history of capitalism, but we are undoubtedly in that situation right now (if not, in my view, since 2007). And this is before we consider the pandemic. Much government spending is aimed at preserving job matches that currently exist, and supressing the virus, a second wave of which would drive the UK deeper into recession. In other words, it is highly likely that additional borrowing now, if spent on the right things, will actually lower future government debt.
Thirdly, imagine the worst case scenario: a world in which interest rates had risen above the growth rate and the UK has not yet got around to reducing its debt burden. This would be a world in which the Bank of England was raising interest rates, selling off debt, or both. It would only do this if the economy was at or near full employment. This would be a much better world in which to raise the taxes required to pay off the covid debt than the one we currently inhabit, and would have smaller effects on the economy. Furthermore, the effects of this fiscal policy itself would reduce the interest rate and lessen the economic burden.
At this point I should probably deal with the objections from Modern Monetary Theory. This is a resurrection of the opinion, by some dizzyingly clever scholars (chief among them Stephanie Kelton) that governments will full control over a fiat currency cannot become bankrupt. The constraint on government deficit spending is not debt, but inflation. The argument goes something like this: money is an IOU. It is a promise by the government that when it comes to paying your taxes, you will be able to write down your liability to the government by an amount up to the amount of it you can get your hands on. Similarly, government debt is an IOU. Because of this, the distinction between the issuing of debt and the issuing of currency is a false one. Governments can never go bankrupt because they always have the option, via the central bank, of swapping out one IOU for another (printing money and buying back government debt). Note that this implies that central banks and sovereign governments would need to cooperate. This is both revolutionary and anodyne: it means, on the one hand, that the government can keep on issuing debt indefinitely. But this also leads down the road of infinitely large growth in the money supply as governments print more and more money to buy up their own debt. And this is where the inflation constraint comes in. Formally, the MMTers say that some taxes are required to “soak up” money and prevent exponential money supply growth in excess of GDP growth. Informally, this means that if government debt is on an explosive path, the government will eventually be forced to choose between hyperinflation and higher taxes. So this leads us, boringly and inexorably, back to the New Keynesian/neoclassical view: government debt and deficits are not inherently scary, but in the long run, you cannot just let them grow faster than the economy forever.
So given that, firstly, the UK’s debt has risen, and secondly, that this will generate some economic burden in the future, it makes sense, once the economic recovery is secured, to think about ways of reducing this debt faster than would happen simply from economic growth and low real interest rates. One way would be to run a budget surplus, or smaller deficit. Politically, this would be toxic. The Conservative government was elected on a promise of ending austerity and “levelling up” the country. I happen to think that they lack the ambition and intellectual apparatus to do this, but even so, I think it will be difficult for them to actually double down on shrinking the state any further. There are also very good humanitarian and welfarist reasons not to advocate spending cuts. There are many reasons I could list but here are two: the BMJ linked austerity to stalling life expectancy growth. The austerity under the coalition involved cuts to sure start centres which we now know did vital work preventing children getting infectious diseases or being injured. Some tax rises might close the deficit, but frankly given the long-run starvation of public services in the UK, see for example my previous post about healthcare spending, there will rightly be pressure to spend any additional revenue. The temptation therefore will be to inflate the debt away. A period of higher inflation will depress the debt relative to GDP (technically, this is a reduction in the real interest rate.) The jargonistic term for this is “financial repression”, and in this paper Reinhart and Sbrnacia show that this played a significant part in the erosion of the real value of the UK’s post-war debt. Many will argue that, having tried a politically and economically painful alternative of austerity with the GFC debt overhang, the best course of action is to revert to this tried-and-tested method.
What I want to argue here is three things: firstly that it’s not obvious that inflating away government debt is an option for developed markets in 2020. Raising inflation has been the explicit economic goal of central banks since the global financial crisis, and they have by-and-large spectacularly failed. Secondly, this reflects a crisis in macroeconomic theory. Thirdly, however, that this weakens, rather than supports, the case for knee-jerk reactions to developed markets’ rising debt levels…