Winner of the New Statesman SPERI Prize in Political Economy 2016


Tuesday 26 September 2023

Challenges to the strong golden rule: MMT and bond market paranoia

 

The idea that responsible fiscal policy involves matching plans for future taxes to projected day to day government spending, often called the golden rule, has a long history. It is often expressed, as Shadow Chancellor Rachel Reeves did recently, as saying governments shouldn’t plan to pay for current spending byborrow. There is a lot to discuss on how you should implement this golden rule (like what matching taxes to spending actually means), but here I want to focus on the principle behind the idea.


Furthermore, I want to look at a strong form of golden rule, which says that following the golden rule is all responsible fiscal policy should do. If you have a fiscal rule that implements the golden rule, you not only don’t need any additional rules, but you shouldn’t have any additional fiscal rules. In particular, as I have persistently argued, you should not have any aggregate rule that restricts public investment. [1] From memory, no UK government has ever implemented this strong form of the golden rule, because even when they followed that rule they also had additional fiscal rules.


In this post I’m going to look at two very different objections to this strong golden rule, coming from MMT on the one hand and on the other coming from what often seems like bond market paranoia. But first we need to add a big caveat, which is that the golden rule should not apply in recessionary periods [2] In recessionary periods short term nominal interest rates are either at or could soon be at their lower bound. At that lower bound monetary policy can no longer provide a predictable stimulus to the economy, and so fiscal policy should take its place. [3]


Outside of recessionary periods short term interest rates can rise and can therefore do the job of stabilising the economy. Fiscal policy therefore no longer needs to step into that role. Here I would argue that something like the strong golden rule should apply. [4] Both MMT and those worried about the bond market would disagree.


Reasonable and unreasonable MMT


MMT has two arguments against the golden rule, which I will call reasonable and unreasonable. The unreasonable argument is that interest rate increases do not reduce aggregate demand and inflation, and therefore fiscal policy has to play the macro stabilisation role at all times. It is an unreasonable claim because it contradicts the large amount of evidence that higher interest rates do reduce aggregate demand and inflation, evidence that you will find in the academic economic literature.


Estimating the impact of higher interest rates on aggregate demand and inflation is quite hard, in part because central banks put interest rates up when inflation rises. This simultaneity problem means eyeballing data just doesn’t count as evidence. For me, some of the most convincing evidence comes from studies that look at unexpected monetary policy changes, such as here for example. Until MMT can provide studies that are of a similar quality that come to a very different conclusion, then arguments that higher rates don’t reduce demand and inflation are just baseless assertions. [5]


The reasonable MMT objection is that even though interest rates can stabilise the economy, it would be better if fiscal policy did so. You could argue, for example, that fiscal policy had a more predictable impact on aggregate demand than interest rate changes. I have discussed the relative merits of fiscal or monetary stabilisation elsewhere, but the key point is that today, in almost all advanced economies, central banks use interest rates to target inflation. Given this, and the reasonable view that this works, the idea that fiscal policy should take this role instead is not a description of how economies actually work, but instead a suggestion of an alternative way they might work..


A standard half-truth that MMTers often state is that the only constraint on fiscal decisions is the demand for real resources and therefore inflation. This would be the case if interest rates were held constant when inflation rises, but they are not. So the constraint on higher government borrowing is not real resources and inflation, because the central bank will raise interest rates which will divert resources from the private to the public sector, and keep inflation under control.


Perhaps the MMT argument should be, for the economy as it actually operates, that the real constraint on fiscal policy is the level of interest rates. So why isn’t this a disincentive enough to prevent governments using borrowing to match higher spending or lower taxes where inflation is either stable or rising? Why do we need a self-imposed financial constraint of the golden rule as well?


Why higher short term interest rates are not a sufficient constraint on fiscal policy


The problem is that the implications for interest rates of breaking the golden rule may not be clear or felt for some time. Imagine, for example, that a year before an election a government permanently announces increased spending in lots of popular ways, and pays for that increase by borrowing, thereby breaking the golden rule. Interest rates may not immediately rise, because demand fluctuates from year to year for countless reasons other than fiscal decisions. Interest rates will be higher than they would otherwise have been, but most people do not follow central bank decisions that closely. The temptation for governments to use borrowing to finance popular spending increases ot tax cuts to gain votes therefore remains.


The golden rule provides a more direct and immediate check on government fiscal plans and decisions. If the government finances higher current spending or less taxes by borrowing, that will tend to increase inflation and therefore lead to higher interest rates. Rather than wait for interest rates to rise, breaking the golden rule gives us advance warning that this could happen.


So by sticking to the golden rule outside of recessionary periods, the government is in effect committing to take decisions on day to day spending and taxes which will not put upward pressure on interest rates. Nice to have, but hardly the most important commitment the government makes. The golden rule certainly does not represent an externally imposed financial constraint on what the government can do. Instead it is just good practice. The interesting question is whether there are exceptional situations other than recessionary periods where that good practice should not be followed. [6]


What about the bond markets


Following the golden rule alone does nothing to ensure that the government debt to GDP is stable or falling, or that the total (current+investment) government deficit hits some target. Any period of high public investment could mean higher debt to GDP, although public sector net worth would not change. Any fiscal rules that targets the level or change in government debt, or the total (rather than current) deficit, are not following the strong golden rule.


I suspect governments have departed from the golden rule mainly because of a feeling that they have to target government debt in some way. In my view this is a consequence of mistakenly looking at just one side of the balance sheet. Governments need to start thinking about their net worth, not their debt. (Arguably a growing failure to invest in public services across Europe at least is a consequence of this mistake.) But those who worry about the bond market (or say they do) would say we need additional targets to placate that market.


Does that market need placating? If government borrowing is rising because it is investing, why would any rational trader worry about this additional borrowing? The situation is just like a firm borrowing to invest, except that unlike a firm the government will never be forced into bankruptcy because it can create its own currency. As a result, bond holders will always get their money back. The chance that a country like the UK that is following the golden rule will choose to default on their debt is zero.


What can happen in rational bond markets is that interest rates on government debt (long rates) can increase, making it more costly to borrow to invest. Whether this happens depends on what bond traders think will happen to future short rates. If they think that higher public investment means that the Bank will have to raise rates in the future, current long rates may rise even if current short rates do not.


Whether long rates do rise following high public investment financed by borrowing depends a lot on the nature of that investment. Investment that enables more green energy will tend to lower future energy prices, putting downward pressure on future inflation [7]. But even if high public investment puts up long term interest rates, this increase is likely to be modest.


What about irrational bond market panics? The first point is that I cannot think of any bond market panic that occurred when a country was following the strong golden rule and borrowing to productively invest. The second point is that in the case of a disorderly bond market the Bank is likely to step in, as they did after the Truss fiscal event.  


This is why the Truss fiscal debacle provides no justification for going beyond the strong golden rule. That fiscal event broke the golden rule, and the predictable result was higher interest rates. More importantly, it greatly increased uncertainty in markets for sterling assets, because it was unclear whether tax cuts would be accompanied by spending cuts, and even if they were, whether such cuts were credible. The Truss event provides a good example of why governments should follow the golden rule, but does not provide any evidence that we need additional targets like falling debt to GDP. [8]


As Toby Nangle writes, "fear of bond market monsters is a bad reason to delay investment in vital infrastructure or the green transition." Imposing fiscal rules, like falling debt to GDP, that go beyond the golden rule because of fears about the bond market are both pointless and risk delaying that vital investment. 


[1] Paying for investment through borrowing not only makes intuitive sense, because it’s what consumers and firms do, but it also makes sense on fairness grounds. The benefits of investment are spread over time, so the taxes to pay for it should be too. If taxes only have to rise to pay the interest on borrowing this means that future as well as current generations pay for the investment. Of course this definition of public investment may apply rather more widely than the national accounts definition.


[2] I define ‘recessionary period’ as either a recession is likely to happen in the near future, a recession is happening or the economy is recovering from a recession. This terminology is deliberately broader than just when output is falling.


[3] In a recessionary period should we worry that the bond market will react negatively to all the extra deficit spending that fiscal stimulus is likely to create. (IMF evidence suggests fiscal stimulus could well reduce the debt to GDP ratio in a recession, but we don’t need to make that argument here.) This bond market paranoia was the public basis for UK austerity that began in 2010. Bond market paranoia in a recession is completely misguided because of Quantitative Easing: the central bank buying government debt to keep long term interest rates low. In the unlikely event that traders stopped buying UK government debt, the central bank would step in to buy government debt. We saw that happen during the pandemic as well as after the Global Financial Crisis.


Indeed we could go further. In a recession any central bank that allowed long term rates to rise because of worries in the bond market would be failing in its duty to stabilise the economy. In addition, a responsible central bank is likely to step in and buy government debt whenever the bond market acts in a disorderly manner, as it did after the Truss fiscal event. Having a central bank in charge of its own currency is our safeguard against bond market panics, whether we are in a recessionary period or not.


[4] One caveat would be that the institutional framework exists to check that public investment projects generate an adequate social return. If that institutional framework didn’t exist, then you might want to set some limits on public investment as a kind of second best. But that does not apply in most advanced economies. It is silly to use aggregate fiscal rules to kill off dozens of good public investment projects just because you are suspicious that the odd lemon might get through.


[5] There is an element of conspiracy thinking involved in this claim. It requires all the central banks that are currently changing interest rates to control inflation to be participating in some kind of mass delusion, where none have noticed that what they are doing has had no impact or is making things worse, or those that have noticed are silenced for some reason. Equally it requires that mainstream academic economists are so indoctrinated by the idea that higher rates reduce inflation that they will not allow contrary evidence to be published.


[6] A war is an obvious example, but some have plausibly argued that climate change (and specifically a green new deal) could be another


[7] To the extent it reduces national vulnerability to volatile oil and gas prices, this reduces uncertainty which will also put downward pressure on long term bond rates. However any investment in domestically produced goods is a claim on resources, so for this reason short term interest rates might be slightly higher if the economy is otherwise strong.


[8] Indeed, Kwateng did say that he would stick to the falling debt to GDP target, and that did nothing to prevent rates rising.

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.