Why following the path of the last Labour government will not work this time

 

The Blair/Brown government waited until they had won a second general election before raising taxes ‘on working people’ to accompany a significant increase in the proportion of GDP spent on public services. In the initial years after becoming Chancellor Brown stuck to Tory spending plans, which together with unexpectedly buoyant tax receipts meant debt to GDP fell substantially. This was despite eighteen years of Conservative government leaving public services in a very poor state. Many argue the reason why the last Labour did this was because they felt they needed to demonstrate prudence to the media and markets.

Are Starmer and Reeves following a similar path? In one sense not: Reeves increased tax significantly in her first budget and was therefore able to substantially increase public spending relative to the previous government’s plans. In terms of aggregate numbers rather than previous plans total current public expenditure fell from 33% of GDP to 32% in the first two years under Brown, whereas under Reeves it is projected to be slightly better than roughly flat.

Unfortunately keeping public spending roughly flat as a share of GDP when starting from a position which is awful (and arguably considerably worse than the one Brown inherited) is not going to please anyone. Does Reeves, like Brown, feel constrained because she worries that more substantial increases in public spending and taxes will upset the media and markets?

The Rock and Roll politics podcasts of Steve Richards are always worth listening to, and recently he suggested an alternative Labour government that we could compare today’s to. Not Harold Wilson’s, but that of MacDonald (PM) and Snowden (Chancellor) of 1929-31. He is not predicting the Labour party will split in the same way it did back then. Instead he points out that the 1929-31 Labour government, in the face of an economic crisis, chose to stick to economic orthodoxy because it wanted to appear ‘economically responsible’, even though it had promised voters change. This was despite a clear but radical (at the time) alternative path available, and in retrospect one far superior, which was the Keynesian path that the New Deal in the US eventually took.

He suggests that Starmer and Reeves also hunger to appear economically responsible by following orthodoxy, and that this is preventing them fulfilling their promise to voters of change, just as it did for MacDonald and Snowden. Once again economic ‘respectability’ defeats real ambitions for change.

It is a very provocative comparison, but one that loses some of its appeal once it is examined in more detail. Reeves is indeed insistent that her fiscal rules are not negotiable, just as Snowden insisted on a balanced budget, but today we do not have a global recession caused by inadequate aggregate demand. A balanced budget made no sense in the 1930s depression, just as Reeves’ golden fiscal rule wouldn’t, but we are not in or close to that kind of recession today. The golden fiscal rule does make sense in the situation the UK is in today where demand is strong and inflation is above target. In other words it was George Osborne, not Rachel Reeves, that was making the mistake that MacDonald and Snowden made.

Perhaps this is taking the comparison too literally. The current UK economic crisis is a dual one of crumbling public services and stagnant growth, and a key mistake the Labour government is making is keeping taxes too low. But is their reluctance to raise taxes primarily due to a hunger for economic respectability? It seems to me it is more about a hunger for electability, and thinking (incorrectly in my view [1]) that higher taxes will lose them votes because it will add to the current cost of living crisis.

Perhaps a better way to make the comparison work is not to think about particular types of economic issues, but think instead about radical change versus incremental change. Here orthodoxy means incremental, slow and hardly noticeable change, of the kind that the MacDonald government did do. Change that will not upset the horses of the establishment media and the markets. That might include one-off but not so great tax increases, of the kind that Reeves made in her October budget. It is also consistent with modifying the dreadful falling debt to GDP supplementary fiscal rule rather than abolishing it completely, as she should have done.

Where I do agree with Steve Richards is that incremental change is not what most people want. Voters recognise the huge hole the UK is in, both in terms of living standards (past and prospective growth) and public services, and they are impatient to see a government leading the way out of that hole, rather than just ensuring the hole doesn’t get any bigger. This is why the Labour government has become so unpopular so quickly, because it has shown no sign of the radical actions that voters were looking for.

But I still think the more relevant comparison is with the last Labour government, because the question those like myself, who want something more radical in terms of tax increases and additional public spending, have to answer is why incremental change worked (in electoral terms) for Blair and Brown but cannot work for Starmer and Reeves. Here Steve Richards makes the key point. In 2001, when Labour won a second term, there were effectively only two alternative parties: the Conservatives and the Liberal Democrats. The continuing problems with public services did cost Labour in the run up to the general election, where they lost many votes to the Liberal Democrats. But they lost far fewer votes to the Conservatives, because the opposition was the same party offering similar policies to the one Labour had replaced, and it made little sense to replace Labour with the party that had run public services down in the first place.

In addition, two other factors meant that in 1997 Labour could get away with keeping to Tory spending plans in the first years of government. The first is that obviously Tony Blair had more charisma and better presentational skills than Starmer, and arguably a better communications team behind him. The second is that Labour’s inheritance on growth and living standards was far better.

Economic growth, and therefore growth in living standards, was steadily rising during Labour’s first term, after stalling in the mid-1990s following the ill-fated ERM entry and subsequent recession. [2] In that context. Blair and Brown’s argument that they were ‘fixing the foundations’ in their first term to prepare for better public services later on made some sense to voters. In contrast Labour’s growth inheritance from Sunak and Hunt has been dire, so they will be very unlikely to replicate what happened in the first term of the last Labour government. Forecasts show very modest growth in living standards over the next four years. [3]

So the key differences between the 1997-2001 Labour government and now is that today Labour’s inheritance on growth is very different, and voters today can vote Reform. [4] This helps explain why, while Labour held its 1997 vote share for the first two or three years in government, Labour today have very quickly lost theirs. While xenophobia may represent the core of Reforms popularity, right wing populism generally only wins when that is combined with economic disenchantment. Incremental change will still leave Labour at the next election defending weak growth in living standards and seriously underfunded public services, and those are the conditions in which right wing populism thrives.

[1] The ‘cannot raise taxes because of the cost of living crisis’ idea is fundamentally flawed because it misunderstands that crisis. The UK savings ratio, the average proportion of income UK consumers are saving, has been steadily rising since mid-2022. That indicates that plenty of consumers are able to buy what they need and still increase their savings. The rising cost of food and energy has created a crisis for those poorer voters who are not able to save. There are plenty of consumers who can easily bear higher taxes, while still avoiding increasing taxes on those who cannot.

[2] GDP per capita growth exceeded 2% in every year from 1993 to 1996, following the large devaluation and lower interest rates as we left the ERM. Initially the devaluation hit living standards because of the higher cost of imported goods, but this effect was largely over by 1997.

[3] The first term of the last Labour government also saw a concerted effort to reduce poverty, but that didn’t really start until 1999.

[4] The threat on the left is probably not very different. It is true that the Greens are now a more feasible choice for those on the economic left, while social liberals can still vote Liberal Democrat. However the LibDems in 2001 were more left wing in economic terms than they are now.

The IMF on the UK’s fiscal rules

 

Press coverage of the IMF’s annual report on the UK economy in both the FT and Guardian chose to headline their comments about ‘refining’ fiscal rules. Those headlines might have been interpreted as saying the IMF thought Reeves could raise spending without raising taxes. If you thought that then you would have been disappointed. Nothing the Fund said does that. The Fund effectively endorsed the current fiscal rules, which for reasons I will discuss below is a little disappointing coming from one of the few institutions that has expertise on these matters. But for those wanting higher public spending (which is most voters), their message is one I would agree with: its tax, not the fiscal rules, that has to change.

What they do comment on is how what happened in March might be avoided in the future. Reeves has sensibly committed to just having a tax changing budget in the Autumn, but the OBR is legally obliged to do two forecasts a year. Unfortunately, Reeves and the Treasury believe that each forecast has to show the government meeting its fiscal rules. After the Autumn budget those rules were only just met, so that meant that the OBR’s March forecast had a good chance of showing the government breaking its rules. Reeves decided, wrongly in my view, to make disability payments the residual item of spending that would adjust to make sure the rules held.

Singling out any one item of public spending to be the residual to be adjusted to ensure your rules hold in the OBR’s March forecast is a daft way to make fiscal decisions. Furthermore, adjusting spending or tax on a six monthly basis to get a fiscal forecast to hit a target is unnecessary. I noted at the time Charlie Bean (ex BoE, MPC and OBR) saying that a grown up approach would just promise to adjust policy in the Autumn budget to meet the fiscal rules, but following the Truss debacle politicians and civil servants were so scared of the bond market this didn’t happen.

The IMF mentions some ways of avoiding what happened in March happening again. I can interpret them as involving three possibilities. The first is to formalise this grown up policy. The OBR could still produce two forecasts each year, but in March they would omit their final chapter on assessing whether the fiscal rules were met. Instead the Chancellor would simply commit to making the necessary policy adjustments in the Autumn. The second option is to get the OBR to forecast just once a year, to coincide with the Autumn budget. A third possibility mentioned by the IMF involves “de-emphasizing point estimates of headroom in OBR assessments of rule compliance”.

The second option, the OBR producing just one forecast a year, is the one favoured by The Institute for Government (see Gemma Tetlow here). I have no problem with this, because issues involving fiscal sustainability are about the long term, not short term. I find the media’s discussion of monthly deficit numbers alternatively funny and exasperating. However the Treasury might want a six month OBR forecast to at least frame the pre-budget discussions. While plenty of other macro forecasts exist, the OBR’s detailed analysis of the public finance data is unique. It would be politically impossible for the OBR to do a forecast for the Treasury in secret.

I’m also not sure how de-emphasizing point estimates would work, because they could always be derived even if the OBR hid them. As Gemma Tetlow points out, the Bank has done this for some time and everyone still focuses on the point forecast. Saying that the fiscal rules would be formally met if there was a 40%, rather than 50%, chance of hitting the targets would rightly be seen as a major change in the ruIes.

I still think a viable option is to formalise the grown-up policy, because at some stage politicians and the Treasury need to start treating the bond market as grown-up too. (The media, I fear, never will.) By March next year Reeves should feel she has established enough credibility to be able to say that she would make the adjustments needed to meet the fiscal rules in the Autumn, without fear of spooking the bond market. To be honest I don’t think the bond market would have been spooked even if she had done this last March.

The two issues that I was disappointed the IMF didn’t comment on are the supplementary fiscal rules, and the change to a three year horizon.

The main fiscal rule, and the one driving policy at the moment, is the ‘golden rule’, which says current spending must be matched by taxes at some date in the future. The IMF rightly says that this rule “helps preserve space for investment”, which is one reason why it is a good fiscal rule. But there are two supplementary rules: a rule for welfare spending and a rule saying the debt to GDP ratio must be falling at some date in the future. In my view both are a waste of time.

Yet the IMF says that the falling debt to GDP rule “safeguards fiscal sustainability”, implying that the golden rule alone fails to do that. By fiscal sustainability they mean that debt to GDP remains stable or falling in the longer term. This is just incorrect, as the IMF must know. Getting debt to GDP to be falling in a few years time does not ensure sustainability. It is quite possible for the rule to be continuously met and yet for debt to GDP to rise steadily, because current increases more than offset any expected future falls.

As I have argued consistently (and to my knowledge no one has established otherwise) the golden rule is the best we can do to ensure sustainability, as long as taxes matching spending is defined sufficiently to allow space for an average level of investment spending. Periods of unusually high public investment never threaten fiscal sustainability. The falling debt to GDP rule therefore adds nothing useful. Yet what it does do, if it happens to bind when the golden rule doesn’t, is encourage governments to cut back on useful investment. It therefore partially negates a key advantage of the golden rule.

Putting the same point another way, any rule that looks at government debt is just looking at one side of the balance sheet by ignoring government assets. It is like saying that when you take out a mortgage on a house your financial position has reached crisis point because your debt level has exploded. If instead you required the GDP share of public sector net wealth to fall over time, you can formally show that this is simply double counting the golden rule. Again, you don’t need a supplementary rule.

The IMF does note that the fiscal rules are changing from looking five years ahead to instead look just three years ahead. All the IMF say about this is that this change is “expected to make the rules more credible, while allowing time to adjust gradually to shocks.” I agree that the Treasury expects this three year horizon to be more credible, but the IMF ignores the basic problem in moving from five to three years ahead, and that is the business cycle.

The current deficit target in the golden rule is not cyclically adjusted. This is potentially a problem because in economic downturns the deficit rises and vice versa. However as long as the fiscal rules looked five years ahead, this cyclicality was never going to be a problem, because the OBR would always be forecasting a return to trend output after five years. But that is no longer true if the target is just three years ahead.

You don’t need a full blown recession for this to become a problem. For example, at the moment the OBR is forecasting steady growth from 2026 onwards. But supposing Trump’s tariffs or other factors depress the global economy in 2026, and this is expected to continue into 2027. This impacts the UK, such that the OBR forecasts made in the Autumn of 2026 show the UK economy still 1% below trend in 2029. In 2026 the fiscal rules will apply three years ahead, which is 2029.

That may well mean that the current balance will not be expected to be at target in 2029, but will be by 2031. To cut spending or raise taxes in those circumstances, just before an election, would be crazy from both an economic and political point of view.

Did the Treasury not think about this possibility? If they did, why did Reeves still take this risk? One of the features of fiscal rules under the Conservatives was endless tinkering for short term political gain at the expense of economic logic, and unfortunately the change from a 5 year to a 3 year horizon is I believe another example. To be quite frank, I just wished they had asked me about this change before they did it. [1]

The scenario above might not happen of course, or it may happen at some other time in the future. But that it might well happen means this rule change is another example of Reeves erecting hurdles that she might fall over in the future. Labour’s tax pledges are the biggest of course, but at least there was a political justification for those. Changing the target horizon from five to three years has no such excuse. The bond markets were never going to be spooked by her October budget with its small adjustment to the falling debt to GDP rule, so she didn’t need to throw them a bone, particularly one they have no interest in. [2]

[1] If that sounds pompous, I am one of the very few in the UK who has written an academic paper about both the theory and practical issues behind fiscal rules.The paper recommends using the golden rule, with a target five years ahead. I advised the previous Labour Chancellor but one to adopt this rule, advice he accepted. Even if the current Chancellor was determined to make this change to ‘enhance credibility’, I would have pointed out that it should have involved moving to a cyclically adjusted target for the current deficit.

[2] Bond markets don’t really care about the details of fiscal rules anyway. Just look at the number of times the last government changed those rules, none of which produced any reaction from the markets. As I set out here, markets are mainly interested in where the central bank will set interest rates in the future, and how uncertain those expectations are.