Winner of the New Statesman SPERI Prize in Political Economy 2016


Monday, 9 March 2020

Fiscal rules: a primer for the budget


Do you want to know why fiscal rules should never involve targets for the debt/GDP ratio, or debt interest, or any stock measure, and why public investment should not be part of a fiscal rule? Read on, but first we have to establish two key principles.

Principle 1: what are fiscal rules for?

The way some economists are tempted to think about fiscal rules is in terms of economic optimality. What is the optimum debt to GDP ratio and what is the best way to get there? This misunderstands what fiscal rules are for. If every government was benevolent (always did the right thing for society) they would follow the optimal policy without the need for fiscal rules. Governments would just consult experts and do the right thing. Why try and formulate the optimal policy into a rule?

Fiscal rules are there to restrain governments from not being benevolent. For many governments the attraction of fiscal giveaways just before an election (and letting monetary policy deal with the consequences for inflation and future generations the responsibility of servicing the extra debt) is too great. Before the GFC the OECD’s debt to GDP ratio was almost double what it was 30 years earlier. In addition right wing governments have discovered that austerity during a recession is a good way of shrinking the state while pretending to be responsible at the same time. When I talk about an irresponsible government below, austerity is the clearest example of a government behaving irresponsibly.

So fiscal rules are part of the apparatus to ensure the government does not give in to either temptation. Good fiscal rules are not overriding democracy but dealing with a negative externality of democracy that occurs because electorates can be fooled. The first best option would be to educate electorates so they could not be fooled, but in a society where the media makes that impossible fiscal rules and fiscal councils are a good second best. Fiscal rules do not override democracy because governments choose to follow them and have the option of ignoring them. No fiscal rule should ever be enshrined in a constitution.

Principle 2: rules should not ignore optimal policy

As my paper with Jonathan Portes explains, fiscal rules are a delicate balance between stopping a government being irresponsible and not stopping them behaving optimally. Too often rules seem designed mainly to constrain governments with little regard to what an optimal fiscal policy would be. Rules that force governments to enact a fiscal policy that harms the economy are bad rules, and are unlikely to be met for very long.

The clearest example of that is when fiscal rules stop a government stimulating economy when interest rates are stuck at their lower bound. Every fiscal rule before the one proposed in our paper had this fault. Every fiscal rule enacted in the UK has had this fault.

Five requirements for a good fiscal policy rule

  1. The rule must mandate fiscal stimulus in a recession where interest rates hit (or might hit) their lower bound.

  2. Debt and deficits are shock absorbers, and fiscal rules should interfere with that as little as possible while still discouraging irresponsible government behaviour. We discuss how to achieve that in detail below, but one clear implication is that fiscal rules should involve deficit targets and not debt targets [1].

  3. Public investment should not be part of any deficit target. To abandon good investment projects to reduce the deficit is a cure worse than the disease. The best way to stop white elephant investment projects is not some arbitrary limits on the share of investment to GDP, but an infrastructure commission with some power.

  4. A fiscal council with teeth is an essential complement to a fiscal rule. Such a fiscal council (which has more power than the OBR) can allow a more optimal fiscal rule.

  5. A good fiscal rule should be designed to last. It should have a structure that can deal with all eventualities. But there is every reason to revisit the deficit target every five or ten years in conjunction with the fiscal council and an open debate.

Requirements (2) and (4) interact. A fiscal rule that allows the maximum discretion for governments to allow deficits to be a shock absorber is a rolling five year target. Future targets with a fixed date become problematic as that date approaches, because any shocks just before the date require damaging short term fiscal adjustments, violating (2). But a rolling target allows the government to cheat, because the target never arrives. A fiscal council that can call out such cheating (and which carries political clout) allows a more optimal policy.

How well have the Conservatives done since 2010?

The obvious point is that in 2010 the Conservatives became the most irresponsible government since the General Theory was written in 1936 by violating requirement (1) and giving us austerity. Ironically by adopting a five year rolling target for the current balance (the deficit less investment) the rule met requirements (2) and (3). They created the UK’s fiscal council, the OBR, but failed to give it a strong enough mandate to effectively call out cheating. Rupert Harrison brought with him the expertise of the IFS in formulating the rule, but perhaps because the IFS did not do macro or perhaps for other reasons requirement (1) was ignored, at the very point interest rates hit their lower bound.

That was the high point for fiscal rules under a Conservative Chancellor! Despite the fact that their main rule excluded public investment, the government cut it sharply anyway, which is why the recovery was delayed from 2010 to 2013. Later on the Chancellor decided to play with fiscal rules as if they were his own political plaything to embarrass the opposition with, and I have lost count of how many there have been. A lot of the blame here lies with the media. Gordon Brown was strongly attacked when he allegedly fiddled with the definition of the cycle, but he kept to his fiscal rule for 10 years and it was important at influencing policy. In contrast Osborne was allowed to change fiscal rules as he liked, with very little comeback.

Where do we now stand? We will see shortly, but I would be very surprised to see any rule that meets the first requirement. To do so would come too close to admitting austerity was a mistake. (Something in a rule that allows the automatic stabilisers to work does not meet requirement (1), because in a recession where interest rates are stuck we need a substantial fiscal stimulus.)

The fiscal rule at the time of the 2019 Tory manifesto was “to have the current budget in balance no later than the third year of the forecast period; to limit public sector net investment to 3% of GDP; and to reassess plans in the event of a pronounced rise in interest rates taking interest costs above 6% of government revenue.” This rule illustrates how far backwards we have come since Osborne’s first fiscal rule. The deficit target involves a fixed date rather than being rolling. The current balance is still the target, but there is an arbitrary limit on public investment. (Osborne went even further backwards for a time by targeting the whole deficit.)

What number should the deficit target (excluding investment) be?

Good fiscal rules should have a structure that is designed to last, and a good way of testing them is to see how they would have done over the past. Including a debt target in a fiscal rule is a sure way of making it not last: this was why Gordon Brown’s rules failed. However, as requirement (5) states, there is no reason why you shouldn’t change the deficit target within the rule every decade or so, to reflect changes to your views about long term interest rates and other factors. One possibility is to give the OBR the responsibility of reviewing the deficit target every five years, taking into account changing views about long term interest rates, trend growth rates and what the ultimate debt to GDP target should be.

This is where debt interest should come in. This issue is discussed in detail in a recent paper by the Tony Blair’s Institute, and here I will only make the key point. If you target, say, is to keep the debt to GDP ratio constant, what the primary deficit (the deficit excluding interest payments) should be depends on what interest rates on debt currently are expected to be and what the growth rate of nominal GDP is expected to be. Interest rates matter because they raise the stock of nominal debt for a given primary deficit, and the growth of GDP matters because it reduces the debt to GDP ratio.

Of course the difference between the actual deficit and the primary deficit are debt interest payments. So far governments have tended not to target the primary deficit. In that case the issue is whether current debt interest rates are different from what they are expected to be in the future. This is possible if interest rates on debt are falling and a lot of debt is long term. It is complicated, which is why a body like the OBR needs to be involved.

Why is a maxima (and perhaps minima) for debt interest payments, as proposed here for example, not a good idea for a fiscal rule? Debt interest is just the debt stock times the average interest cost of debt. As a result, it suffers from the same problems as a target for the debt stock alone. Although some positive shocks to debt are correlated with lower interest rates, not all are. And even if they are, why should the correlation be just what is required to smooth fiscal policy?

So far I have talked about long run targets for government debt to GDP rather than government net wealth. This is simply because it is a more familiar concept. In advising on what the deficit target should be, a fiscal council should look at government net worth as well as levels of debt. A government that sells off a profitable asset in order to reduce its debt is just pulling the wool over people's eyes, and schemes like PFI are in danger of falling into a similar trap. 

How should we define public investment?

Requirement (3) excludes public investment from the deficit target, because that investment benefits future generations. But some current spending has some benefits for future generations too, like education. So is the national accounts definition of pubic investment too limited for a fiscal rule? There are rumours the Chancellor is thinking along these lines.

First, it is important to clear up a common confusion. A bridge built with public money will hopefully benefit many generations to come. But a bridge requires constant servicing, and that cost is part of current spending. Does this make sense? Yes it does. You can think of that servicing cost as the price that every generation has to pay to keep the asset in place. There is no reason the current generation should be exempt from paying that. Equally a new hospital should last for many generations, but every generation needs to pay for the nurses to staff the hospital.

What about education? Education isn’t like a bridge. It is not like a one off expenditure that ends when the bridge is complete. If a government decides to reduce class sizes (say), that requires more sending now but more spending in the future as well. Like servicing a bridge, it is a constant commitment that requires constant resourcing to match it.

Public investment should be excluded from the deficit target not just because it helps future generations, but also because it is a one-off investment that will benefit future generations.

Financing a Green New Deal

This discussion is important in thinking about financing spending to combat climate change. Making energy production green is a clear example of a one-off investment to benefit future generations. For this reason some economists have understood why government debt should rise to pay for greening the economy. As I explain in the link, there is a limit to how much this should be done, because the polluter (the current generation) should pay. But as I said here, no fiscal rule should be used as a pretext to stop vital work to green the economy.

Why not run fiscal policy like monetary policy

Interest rates are decided by central banks, or committees including some outside experts as in the UK. No one is suggesting a similar committee of experts can enact budgets, but could they - as Jagjit Chadha suggests here - set the overall deficit any Chancellor has to achieve in their budget. This committee, perhaps part of an enhanced OBR, could work out what the optimal deficit path is to achieve some long term objective for the debt to GDP ratio. No need for fiscal rules.

Inflation targeting works because everyone wants low inflation, and crucially everyone should realise that how you get it is an issue that involves a great deal of expertise. You need to have a model of how the economy works. Aggregate fiscal policy is not like that. There is no analogy to low inflation. Stable debt to GDP might come close, but many would argue that after the GFC debt is too high, and others would disagree. These views are worlds apart.

The same is true about how you get to any long run target. As I suggested above, macroeconomic theory suggests you should do it slowly but does that mean 50 years or 100 years? We just do not know. As a result, I suspect that at the end of the day all the experts would be able to agree on is general virtues like fiscal smoothing. This virtue can be embodied in a good fiscal rule.

Let me put the same point in a slightly different way. You can have rules for good monetary policy, but I suspect the majority would agree that a committee of experts could do better than this rule. I’m not sure the same applies to fiscal policy if the rule is good.

There is also a political economy issue. I doubt that a committee of experts setting the deficit Chancellors had to achieve would last very long. For that reason I don’t think it would ever be established. Maybe if academics devoted as much time and energy to fiscal policy as they do to monetary policy then in twenty years maybe, but it is not a viable proposition right now.


[1] To see why, think about the impact of coronavirus on deficits and debt. Hopefully the virus will lead to a sharp but short-lived recession, which will at least reduce taxes and raise the deficit and debt. The impact on the deficit in future years will be minimal, so requiring little or no adjustment to fiscal policy in future years. This is the optimal policy response. However the debt/DGP ratio will be permanently higher because of the short term deficit, so a debt target would require a (suboptimally) rapid fiscal response. Deficit targets are better than debt targets at preserving shock absorbers.

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.