The impact of austerity in the UK

When they do their forecast evaluation report, the OBR also look at the impact of fiscal policy on GDP. Here is the relevant chart from their report published yesterday.

There is a useful innovation compared to previous years, which comes close to an something I suggested a few months ago. The orange bar shows the impact of fiscal measures implemented in that year. The total effect of fiscal policy is this plus the impact of previous fiscal policy actions unwinding.

Suppose for example that fiscal policy reduced GDP by 1% in year 1, but its impact on the level of GDP was expected to decay by half in year 2. If no fiscal policy was enacted in year 2, then fiscal policy in year 1 would increase growth in year 2 by 0.5%. Why might the impact of fiscal policy on the level of GDP decay over time? The obvious explanation is that monetary policy ensures that it does by stabilising the level of GDP. This assumption is problematic when interest rates are stuck at their lower bound, which is why it is useful to publish the within year estimates as well as the total estimates,

You can see when this matters from 2012 onwards. We have a run of years where the total impact of austerity on growth is zero or positive, but only because of the unwinding of previous austerity. If monetary policy, or anything else, had not been able to offset earlier fiscal tightening, then instead the impact of austerity would be to reduce growth in all those years. In that (extreme) case the level of GDP in 2016/17 would be over 4% below what it would otherwise have been without any fiscal tightening from 2010/11.

As the OBR’s assessment of fiscal impact is in their publication on forecast errors, they naturally talk about whether there is any relationship between the two. This year they included this chart.

It is important to understand what we are looking at here. It is not whether there is a correlation between fiscal consolidation and GDP. As we have seen the OBR assumes there is, and indeed their calculations were the source of my estimate that the average household had by 2013 lost a total of £4,000 worth of resources as a result of austerity. The foolishness of austerity in 2010 was not that the OBR underestimated its impact, but that it left us vulnerable to negative shocks because interest rates were at their lower bound. The shock that hit in 2012 was the Euro crisis and the impact of austerity there.

What the chart above might tell us is whether the OBR have in fact underestimated the impact of austerity i.e the numbers in Chart E are too small. Each year there are hundreds of potential reasons for forecast errors, of which underestimating the impact of austerity is just one. So the best we can expect, if the OBR are underestimating the impact of fiscal policy, is a negative relationship going through zero between the two variables in Chart D but with lots of random variation on top. That is what we see in Chart D.

The economic and political cost of UK austerity

The UK now has a surplus on the government’s current budget. George Osborne tweeted “We got there in the end — a remarkable national effort. Thank you.” This has been a remarkable period in UK macroeconomic history, but not in the way Osborne thinks. A majority of economists have always been against trying to reduce the deficit when interest rates are stuck at their lower bound, a majority Osborne chose to ignore. So what has been the cost of this “remarkable national effort”?

The first time I looked at this I did a very simple calculation. The OBR estimate (here, chart E) that fiscal consolidation took just under 1% out of the economy in 2010/11 and over 1% in 2011/12. I wanted to get a simple estimate that no one could suggest was too high. As actual output was pretty flat until 2013, I assumed that output was 2% lower in 2011/12 (1% from the previous year plus the additional 1%) as a result of fiscal consolidation, remained 2% lower in 2012/13, but then fully recovered by 2013/14. That gave a total output loss of 5%, which is almost £4,000 per household.

I think we can now do things a little more scientifically. (If you are not into these sorts of calculations, you can skip to the paragraph starting £10,000.) I originally took the OBR estimates which had embedded in them a declining influence on GDP over time, based on historical experience. I think it is wrong to use these, because the reason that the impact of fiscal consolidation normally declines is that monetary policy counteracts it. This didn’t happen after the Great Recession because interest rates were stuck at their lower bound and QE was pretty ineffective. The OBR have now provided estimates of the ‘direct’ effect of fiscal consolidation, that take out the impact of the decay from past consolidation. See here for a detailed discussion.

The Impact of UK fiscal consolidation on GDP

Fiscal impacts	10/11	11/12	12/13	13/14	14/15	15/16	16/17	17/18
Direct impact on growth	-0.8	-1.4	-0.6	-0.7	-0.3	-0.5	-0.2	0.0
Impact on level of GDP, no decay	-0.8	-2.2	-2.8	-3.5	-3.8	-4.3	-4.5	-4.5
80% decay	-0.8	-2.0	-2.2	-2.5	-2.3	-2.3	-2.1	-1.7
Cumulated loss		-2.8	-5.0	-7.5	-9.8	-12.1	-14.2	-15.9

The table above starts in row 2 with the direct impact of fiscal consolidation (the orange bars in Chart E). There are reasons for thinking these numbers are too low, because they still embody some within year offset from monetary policy, but lets go with them. Suppose there was no tendency for GDP to rebound from these impacts (like kicking a ball each time it stops). The third row computes the total impact on the level of GDP in each year.

Assuming zero decay from fiscal consolidation is too strong, even when interest rates are at their lower bound. For example the impact of tax or transfer cuts are likely to be greater in the short term than the longer term. QE had some impact. So row 4 assumes a decay of 0.8 i.e. only 80% of the fiscal consolidation remains in aggregate demand the following year. This is very crude and no substitute for a proper model based estimate, but I do not know of any recent model based estimates so it is the best we can do. The final row shows the accumulated loss of output: the total cost of fiscal consolidation over the whole period. The final figure suggests the national effort to reduce the deficit cost over 15% of GDP, which when GDP is around 2 trillion and there are 27 million households, gives over £10,000 per household.

There is a big objection that has been made to this calculation. If GDP had been 2% higher in 2016/17, say, the Bank of England would have raised interest rates because that level of GDP would have been inflationary. In other words I should be using a much higher decay factor as we come closer to 2017. However there is an even stronger counterargument to that. I argued here that austerity was a cause of the productivity slowdown that began in 2012. By delaying the recovery for three years, austerity made firms put productivity enhancing projects on hold, and we have seen no sign as yet of any catch-up. I think it is reasonable to assume that the productivity slowdown caused by austerity led to a reduction of at least 2% of GDP from the supply side by 2015. That nullifies the argument that the bank of England would have had to raise rates if austerity had not happened. [1]

£10,000 for each household is an average figure, but we know that austerity did not fall evenly, but was concentrated on those at the bottom end of the income distribution. It is certain that cuts to social care and the NHS cost lives: it is just a question of how many thousands of lives we are talking about.

And then there is the political cost of austerity. The Coalition government, and particularly our current Prime Minister, has used immigration as a scapegoat for the impact of austerity. With the help of the right wing press that scapegoating has worked. In particular, as I show here, many people believe that immigration has been bad for public services like the NHS. In reality the opposite is true, but the government and press have succeeded in creating what I call a politicised truth: something that is believed to be true just because politicians and the media keep saying it is.

The government may well have pursued this line even if austerity had not happened, but it gained some of its potency because austerity did lead to pressure on public services like the NHS. That in turn helped create the atmosphere required to gain a majority for leaving the EU. Austerity, for this and other reasons, created the conditions that allowed Brexit to happen. Those who think the UK descended into political madness with Brexit are wrong: the madness started with austerity in 2010.

The final point is that austerity was completely unnecessary. By austerity I mean cutting the deficit when interest rates could not be cut to offset the impact of fiscal consolidation. There is zero evidence that the markets demanded austerity in 2010, and plenty of evidence they did not. Even if the markets had panicked at the size of the deficit, the Bank of England would have bought government debt as part of its QE programme.

The unusual feature of the Great Recession was not just its size, but that for the first time since the 1930s governments started reducing spending in what should have been the recovery period. They have never done that since the 1930s because economic textbooks and state of the art models say it is a stupid and costly thing to do.

Of course the deficit needed to be reduced, but the government could easily have waited for a few years until the recovery was well underway and interest rates were well above their lower bound. The £10,000 per household is not the cost of deficit reduction. If the government had been patient it could have reduced the deficit with no cost at all. Whatever the motive for George Osborne disregarding the lessons of history, his actions have lost the average household £10,000 worth of resources and caused additional ongoing economic and political damage to the economy. Not so much a “remarkable national effort” as a predictable man made disaster.

[1] There is an argument that without austerity interest rate would have increased in 2011, because they nearly were anyway. But that would have been a huge mistake by the Bank, who were panicked by higher inflation. One of the reasons inflation was high was austerity: the increase in VAT. So I think letting austerity off the hook and passing the hook to the Bank of England because of something they might have done is not a very convincing argument.

The macroeconomic cost of Conservative government

During this election period there has been plenty of analysis that looks at how the economy has performed since 2010 (the IFS here for example). All show the UK performing very badly indeed. But how much is that down to macroeconomic policy mistakes, and how much is due to factors outside the government’s control? I will attempt to answer that question in this post, and try to be as conservative as possible.

I will begin with austerity, because it’s a calculation I have already done. The table below is taken from this post.

The first row comes from an analysis done by the OBR (Chart E on page 27). The main negative impact on growth came in the first two years as public investment was cut back sharply, but continuing fiscal consolidation in later years reduced aggregate demand by significant amounts. The key issue is how persistent these impacts are. To see what persistence means in this context, consider a hypothetical example.

Suppose cuts in public investment in 2010 reduce GDP in that year by 1%. Public investment stays at this lower level in 2011. Other things being equal, does GDP stay 1% lower in 2011, or do other components of demand rise to take the place of some of that lower public investment? In normal circumstances the answer to that question would be the latter, because central banks would react to lower GDP by cutting interest rates which would stimulate private spending. However throughout the period examined above interest rates were at their lower bound, so this couldn’t happen. But other factors (e.g. Quantitative Easing) may have crowded in private demand to some extent.

In this calculation I assumed that the impact of fiscal consolidation decayed by a factor of 0.8 each year. The third row therefore gives the impact of austerity on the level of GDP in each year over this period. For example, the OBR estimate there was no fiscal consolidation in 2017/18, so the impact of past austerity on the level of GDP in that year is to lower GDP by 2.1% x 0.8=1.7%. In theory austerity would have had some impact after 2017/18, but interest rates started rising at the end of 2017, suggesting that the Bank thought there was no longer much deficient demand.

However it is also likely that the earlier prolonged period of deficient demand had an impact on how much the UK economy can supply. I examined this here. The argument is that productivity improving investment was lost during the austerity period, and that had a longer lasting impact on UK productivity and the stock of capital. The problem here is attaching numbers to this idea. Empirical estimates can sometimes be very large (for example here), and the IMF study I looked at here is also consistent with austerity (fiscal consolidation in a recession) having significant long term impacts on GDP. But I want these estimates to be conservative, so I will assume that austerity during the 2010-17 period reduced GDP permanently by 1.5%.

The OBR estimate that Brexit will end up reducing UK GDP by 4%. However I need more than just a long run impact. The following is based on a NIESR study by Kaya et al, and in particular their Table TF4. (I’ve done some extrapolation for the initial years.)

GDP impact of Brexit

GDP	2016	2017	2018	2019	2020	2021	2022	2023	2024
%	-0.4	-0.6	-0.7	-0.8	-0.9	-1.2	-1.8	-2.5	-3.0

Again I suspect this is quite a conservative estimate for the immediate impact of Brexit, even though their long run impact (at -5.7% for 2035) is greater than the OBR’s number.

We also need to add something for the pandemic. The UK was hit in 2020 comparatively hard, both in terms of deaths and lost GDP, even though other countries like Italy were hit earlier. Not only did Johnson’s government waste the early months of 2020 with the idea of ‘herd immunity’, but it also waited far too long in introducing lockdowns, which meant when those lockdowns inevitably came they were more severe and prolonged, giving a more sustained hit to GDP. UK GDP fell by over 10% in 2020, compared to just over 6% in the Euro area. I think it is fair to class this as an economic mistake, because the reason the government gave for delaying lockdowns was to protect the economy, whereas in reality they were doing the opposite.

The third and last lockdown extended into 2021. In addition, the failure of the government to give the NHS the resources to bring waiting lists down after the pandemic, coupled with the steady squeeze in health funding that preceded it, began to have a clear macroeconomic impact during the 2020s. While labour force participation returned to its pre-pandemic trend in most other countries, it did not in the UK, and a significant part of that was due to poor health.

The table below collects these three elements together.

A conservative estimate of the economic cost of Conservative government, % GDP

	10	11	12	13	14	15	16	17	18	19	20	21	22	23	24
Austerity	0.8	2.0	2.2	2.5	2.3	2.3	2.1	1.7	1.5	1.5	1.5	1.5	1.5	1.5	1.5
Brexit							0.4	0.6	0.7	0.8	0.9	1.2	1.8	2.5	3.0
Covid											5.0	1.0	0.5	0.5	0.5
Total	0.8	2.0	2.2	2.5	2.3	2.3	2.5	2.3	2.2	2.3	7.4	3.7	3.8	4.5	5.0

From 2011 until 2019 households were over 2% poorer mainly as a result of austerity, but with additions from Brexit after the referendum. By 2024 that had increased to being 5% poorer, mainly because of Brexit. That means that the average household was losing over £4,000 worth of resources (public and private consumption plus investment [1]) in 2024 as a direct result of government decisions. The Conservatives like to accumulate these things, so adding up the losses over all fifteen years comes to (in today’s prices) a massive £35,000 loss of resources for the average household.

Is there any way of comparing these numbers with the UK's actual performance, either compared to history or other countries. Comparing GDP per capita growth to a trend growth line based on post-war data would give a much bigger gap, but that comparison is misleading because there were signs UK growth was slowing down before the financial crisis, and this fits with a gradual reduction in underlying growth in other countries. Unfortunately all the major economies beside China undertook austerity from 2010, so international comparison are little help here.

However, John Springford has compared growth in the UK since 2016 with a doppelgänger based on other countries, and he estimates the UK has grown by 5% less than these other countries suggest it should. If we combine my estimate for 2024 for Brexit and post-pandemic health we get 3.5%, which given the uncertainties involved is consistent with Springford's analysis. 

 

A UK government that enacts policies that reduce GDP by around 2% during its time in office is pretty unusual. To reduce it by 5% is extraordinary, but then since WWII we haven’t had a government that has cut public spending in a recession when interest rates were stuck near zero, or one that deliberately raised trade barriers with our largest market.

The way these numbers are constructed it looks like the consequences of three bad mistakes, but I think it goes deeper than that. What connects them all is crass economic incompetence. In each case expertise was ignored because it didn’t fit in with ideological or political objectives. As I have sometimes said, mistakes made by politicians because they have followed the expert consensus are understandable and to some extent forgivable, but mistakes made because politicians ignore the expert consensus have to be owned by those politicians.

This propensity of Conservative governments to ignore the economic consensus and as a result make very costly mistakes is not unique to this period, as my recent discussion of monetarism showed. What is really alarming is the failure to learn from these mistakes, or even recognise them as mistakes. This isn’t just the natural reluctance of politicians to admit error, but goes far deeper. The Conservatives have created through the right wing press, pressure on the BBC, think tanks and rich donors an alternative reality for themselves, where disasters are seen as triumphs never to be questioned. Which is why in this election they are plugging tax cuts despite crippled public services, refusing to recognise the costs of Brexit and where even the delayed pandemic lockdowns are seen as a mistake.

As a result, as things stand any future Conservative government will be likely to continue to make serious economic policy errors that cost most UK households a substantial amount in lost income and resources.

[1] The idea of household resources (GDP divided by the number of households) is less familiar than, say, household income, but in my view it is a better measure of underlying welfare. It includes, for example, public services like the NHS, which household income does not. It is of course just the household equivalent of GDP per capita. 

How to muddy the waters on impact of austerity

Jamie Murray of Bloomberg writes “U.K. Austerity Critics Must Explain Productivity Link”. To which my immediate reaction is ??!? When have those who criticise UK austerity ever claimed it is responsible for the UK productivity puzzle? Certainly not me, and I’m the only critic quoted in the piece.

The problem is in the framing, set out in the first paragraph. There is apparently a debate between those who blame poor UK growth since 2010 on austerity, and those who blame it on other factors like the Eurozone crisis and a risk averse banking sector. No, there is no such debate. The reason there is no such debate is given in the first line of the next paragraph: “Both are likely to be true to some degree ….” Exactly right! But then the sentence goes on “and the extreme polarization of views reflects badly on economists and commentators, some of whom may be motivated by political ideology rather than reason.” You talkin’ to me?

This is classic “plague on both houses” stuff - both sides are too extreme (probably because they are ideologically motivated), so let’s take a more reasonable middle view. The only problem in this case is that to do this you have to misrepresent one side. There is very little disagreement among serious economists that UK austerity in 2010 and 2011 reduced UK GDP. The OBR, using rather conservative multipliers, thinks that GDP in 2010 was about 1% lower, and in 2011 about 1.5% lower as a result of fiscal policy, with significant knock on effects into later years. The estimates by Jorda and Taylor involve similar magnitudes, but perhaps if you added in large expectations effects you could double the OBR’s numbers, although whether you could then also assume no offset from monetary policy becomes an important issue. Whatever you do, there is no way you are going to get close to the fact that GDP is currently around 15% below its pre-recession trend. Put simply, the productivity puzzle matters more. But that does not mean austerity is unimportant: just 1% of GDP completely wasted is a big deal.

Now it could be that the productivity puzzle is not completely independent of austerity, and Murray’s article actually makes some suggestions about possible links. But the bottom line is that, among serious economists, austerity is highly unlikely to explain all or indeed most of poor UK growth. So this side of the 'debate' does not really exist. But the other side does exist, if only in the shape of the UK government. They want to avoid as far as possible any discussion of exactly how much austerity reduced GDP, and instead they want to claim that austerity produced the recovery! It is that side of the debate that has some explaining to do. Given the links between austerity and productivity the article discusses, a much fairer headline would have been 'impact of UK austerity may be even worse than critics suggest'.

While I’m in complaining mode, two other things that I get annoyed at. Murray says a good place to start is the OBR’s analysis of how its forecast changed between 2010 and 2014. He is not the only one to focus on the OBR’s forecast errors when addressing this ‘debate’. But this makes no sense, because the 2010 OBR forecast already had their estimates of austerity (see above) embodied within it! The government loves to play this trick, but good journalists should not follow their lead.

The second point is that critics of UK austerity would never argue that external factors were not important, if only because a key external factor was the impact of austerity in the Eurozone. As this Commission analysis shows, backed up by similar analysis at NIESR, austerity was a key factor behind the second Eurozone recession, and most of that austerity came from core rather than periphery countries. The argument that this was all a necessary response to the debt crisis just does not wash after the obvious impact of OMT. 

What is really going on here is this. The intellectual debate over austerity is largely over, and the right side won. Good commentators who do not have an axe to grind know this. But they also know that the political debate went the other way. Writing about this paradox would appear politically one-sided. Much better to portray critics of this political outcome as much more extreme than they actually are, and get to something like the truth by being more ‘balanced’. It is not a game I like, and I like even less being misrepresented.

The impact of austerity in the UK and Eurozone

When I wrote this recent post on the impact of austerity in the UK, I was slightly nervous about the numbers. I suggested that the 2011 and 2012 cuts in spending on goods and services alone would have reduced GDP in 2012 by between 1.25% and 2.5%, but this was a back of the envelope calculation using simple aggregates and static multipliers, and it left out the impact of important tax increases and transfer cuts. However luckily we today have a rather more systematic analysis from Dawn Holland and Jonathan Portes at NIESR. This adds in the impact of tax increases and transfer cuts, and comes to a figure of over 4% for the impact of 2011 and 2012 austerity on UK GDP in 2012. 

Now at this point I should declare an interest of sorts. Holland and Portes use the Institute’s global econometric model NIGEM. I built the first version of this model. However that was so long ago, and I’m sure that the improvements subsequently made by Ray Barrell, Dawn Holland and others mean that nothing is left of my original construct.

What is especially interesting about the NIESR study is that they also do the analysis for the Eurozone economies. If we extend the calculations to 2013, Greek GDP in 2013 is over 13% lower as a result of 2011-13 austerity, Portugal’s GDP nearly 10% lower and Spain’s 6.7% lower: UK GDP is ‘only’ 5% lower. Now a cut in average incomes of 5% or more is a big deal, and it is why I keep saying that the welfare costs of these measures dwarf other considerations. But we know that this pain is not evenly spread: many people are not much affected by austerity, while others are receiving much larger cuts in their incomes.

The key point, which the NIESR study also puts numbers to, is that this pain is not in any way inevitable. It comes because austerity is being undertaken when monetary policy can do very little to counteract these effects. In more normal times, which means once the recovery is sufficiently underway such that interest rates begin to rise again, this scale of austerity will have a much smaller impact on GDP. In principle, its impact could be completely offset by monetary policy. So the argument that these large income cuts are inevitable because debt has to be brought down at some point is simply wrong.

However the NIESR study does miss one thing out of its calculations. Nowhere is there a confidence fairy that will magically persuade the private sector to spend because government debt is coming down. (As the NIESR study shows, the debt to GDP ratio actually rises because GDP falls by more than debt, but hey that’s a detail – GDP will recover one day, probably.)  Now to believe in fairies you need pretty good evidence, and that is just what we do not have. A few economists think they saw some in the data, but that is not enough - nothing like enough - to justify inflicting this scale of pain on so many. (Others, of course, saw nothing (e.g. pdf).) Unfortunately too many people just wanted to believe in the confidence fairy. Normally we find those who really do believe in 'real' fairies either rather amusing or rather strange. Unfortunately some of those who believed in the confidence fairy were put in charge of running our economies.  

Mediamacro myth makers fight back

This may also be the first in a series!

In a recent blog, David Smith of the Times writes

“One of the most enduring claims about the British economy in recent years is that the then coalition government abandoned austerity in 2012. It is a claim that gives comfort to those who see everything that has happened to the economy through the lens of fiscal policy. Only when austerity was abandoned in 2012, some argue, did the economy begin to recover. Unfortunately it does not fit the facts. It is a myth.”

Chris Giles of the FT tweeted: “The shocking thing about this excellent post is the misinformation that forced @dsmitheconomics to write it”.

Now the reference to myths might make you think David Smith is having a go at yours truly, but I would never be so narcissistic. I know this cannot be the case because I have never said that austerity was abandoned in 2012. In fact I cannot think of anyone who did, but clearly I’m not reading the right people. Of course this could be another example of the straw man trick: to defend position X (plan A continued) against position Y (the pace of austerity slowed), create a third position Z (austerity abandoned) which is a silly exaggeration of Y, and show that Z is false. Ergo X must be true. Remember how critics of austerity had to be wrong because they claimed a recovery would never happen.

What David concludes, of course, is that the pace of austerity slowed from 2012 onwards, which is obvious if you just look at the data. So why does he think this is such a problem for critics of austerity? Again we need a straw man: someone who “see[s] everything that has happened to the economy through the lens of fiscal policy.” Now I’m sure I have never met anyone like that, but if such a person existed then the 2013 recovery would be inexplicable, because austerity was continuing (albeit more slowly).

This is terrible stuff. Every macroeconomist besides those of David Smith’s imagination knows that the economy is influenced by all kinds of factors, or which fiscal policy is but one. So a recovery is perfectly compatible with austerity being a drag on growth, particularly if monetary policy is highly expansionary.

One way of thinking about the impact of a fiscal contraction is that it has its maximum impact on the level of GDP when it happens, but this impact dies away as other forces, like monetary policy, bring GDP back to its ‘natural’ level. Whether that is the right way to model the impact of fiscal policy in a liquidity trap is debatable, but that is how the OBR treats the impact of fiscal policy, and from his post I’m glad to see that David thinks the OBR is an authority on these matters.  

Here is a chart from this OBR document.

The orange bars show the impact the original 2010 plan would have had, and the blue bars what actually happened (and what will happen) as seen in March 2014. The blue bars are the basis for my conservative estimate that austerity cost every UK household on average £4000 worth of resources. Even though in both cases austerity continues through 2012 and 2013, the impact on growth dies away (or even becomes positive), because the negative effect of any new austerity is offset by the impact of earlier austerity dying away.

Harmful austerity does not need to be abandoned before a recovery can happen. Slowing down austerity clearly makes a recovery easier, but that is not the main reason why the mediamacro myth that ‘Plan A’ continued is important. As I wrote here: “Not making it clear that the plan had changed was a serious failure. If that call had been made, the Chancellor would have had to account for why he had allowed deficit reduction to stall, and that in turn would have established quite clearly that previous austerity had delayed the recovery.”

It really is very simple. George Osborne campaigned in 2010 that Labour’s plan to cut the deficit by half in five years was much too slow, and so began a much tougher austerity programme. More rapid deficit reduction was at the centre of that plan. But deficit reduction was allowed to slow from 2012. Why did the media not challenge Osborne on why this was happening? Why did it go along with the fiction that the plan was unchanged? The media has no problem asking Labour politicians to account for why they borrowed too much (allegedly), but when George Osborne borrows much more than he planned, having previously stressed the importance of cutting the deficit quickly, this suddenly becomes unimportant. Strange that.