Winner of the New Statesman SPERI Prize in Political Economy 2016

Tuesday 11 June 2024

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





















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
































































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. 

Tuesday 4 June 2024

Should voting in this election be about punishing the Conservatives, signalling to a future Labour government or something else?


I didn’t write about the election last week because I didn’t think there was anything of interest to say that I haven’t already said. At the moment at least the Conservatives are saying anything that might shore up its elderly core vote, however silly, unfunded or poorly thought through those proposals may be. But two recent articles in the Guardian about voting strategy are sufficiently interesting to write about.

I have always advocated tactical voting under the UK’s FPTP system, because I view voting in an instrumental way (how can I achieve some end) rather than an expressive way (voting as a statement about oneself). Actually I would put it more strongly: the right way to vote in a UK General Election is to vote to achieve a better social (or social group) outcome, and if you can do that but you instead vote for the party whose policies are closest to yours you are being a little selfish, anti-social and irresponsible. If you disagree, please read this and tell me why my logic is incorrect.

An article by Jonathan Freedland suggests a very different, backward looking voting strategy, which is to punish the government that has done so much harm over the last fourteen years. The only punishment the voter can enact is to not vote Conservative under any circumstances, but more specifically to vote tactically to ensure the Conservatives have the worst possible result in terms of seats won.

I agree with Freedland about how bad the government has been over the last 14 years. We all know how poor economic performance has been over this period. Hopefully in a later post I will try and evaluate (in money terms for the typical household) how much of that is down to government policy. Add to that the collapse in public services, the increase in child poverty, the corruption, the endless and blatant lying, vote rigging, cruelty towards minorities and incompetence, and there hasn’t been a government as bad as this in my lifetime.

I also have to admit that the idea of using my vote to punish the government for all this is emotionally attractive. With a largely supportive print media, and the BBC under its influence, this government has avoided accountability for its mistakes for so long it deserves to do disastrously in this election. Although the 14 years have seen five Prime Ministers and countless different ministers, most of those changes have been the result of internal rivalries or the desire of Conservative MPs for self-preservation rather than accountability for mistakes. Indeed one of the many things that has made this government uniquely bad is how it has ignored offences that would have in past led to resignations.

However, the idea of voting based on how much you dislike the last government has its obvious flaws. To take a recent example, many may have voted against the last Labour government in 2010 because of their failure to regulate the financial sector sufficiently to reduce the impact of the Global Financial Crisis on the UK economy. However this ignores the fact that the Conservative opposition were constantly criticising Labour for too much regulation. Inevitably the backward looking performance strategy for voting choice tends to focus on what the government did rather than what the opposition might have done if it had been the government. [1] It is much better to use the past to inform a judgement about how well political parties will behave in government.

So although the punishment strategy is emotionally appealing, and may well govern how many will actually vote, I don’t think it is persuasive as a strategy for how people should vote. Luckily in this election my own preferred strategy, and Freedland’s punishment strategy, amount to doing the same thing, which is where possible to vote tactically against the Conservatives.

The second article, by Sonia Sodha, discusses a voting strategy that departs from tactical voting against the Conservatives. The idea proposed by some on the left like Owen Jones is to vote for other, more progressive parties than Labour, even where this might lead to the Conservatives winning the seat. She mentions three reasons often given for doing this. The first is that Labour are no better than the Conservatives, but this is obviously false. The second is that Starmer cannot be trusted because he has dropped most of the policies he won the leadership promoting. As Helen Lewis describes here, Starmer is ruthless about winning and therefore prepared to adapt his positions to that end. But that makes him much more like the average Conservative politician, rather than worse than Sunak and his ministers.

The third, more interesting, argument goes as follows. The outcome of the election is bound to be a Labour majority, so not voting for them in the (now many) Lab/Tory marginals will do no harm. [2] On the other hand voting for a more progressive party would send the next Labour government a message that it cannot take the more progressive vote for granted. Sodha attacks this argument by questioning the politics of the more progressive alternatives to Labour, but as I frame the argument this is beside the point, which is to influence what the next Labour government does.

This is an example of a class of departures from tactical voting against the greater evil that I discussed in my post mentioned at the beginning. But as I also said in that post, most such arguments don’t stand up to scrutiny. Does this one? For the moment let’s assume that a Labour win with a comfortable majority is 100% certain.

Labour in opposition have pursued a clear strategy to win, which is to place Labour in a policy space just to the left of the government, while avoiding any policy differences that might deter Conservative voters from 2019 switching to Labour. An obvious example of the latter is to avoid the subject of Brexit. As government policy has moved significantly to the right while the electorate has not, together with the government’s terrible performance, Labour’s strategy for winning is perfectly sensible, even though it places Labour to the right of most voters on some issues.

The downside to that Labour strategy is that you don’t give your natural core supporters very much positive to vote for, so those voters may not bother to vote or vote for more progressive parties. In other words Labour’s strategy in opposition already assumes that some potential Labour voters will vote for more progressive parties. In that case it makes sense, if Labour are bound to win, for those on the left to discourage a Labour government continuing this strategy when in power by maximising the vote of other progressive parties.

In practice I think that signal is pretty weak. As I argue here, if a Labour government acts in anything like the cautious manner its election campaign suggests a large percentage of those who voted for it will become impatient and disillusioned and this will show itself in large increases in support for the Greens and LibDems a year or two after the election. That, rather than any voting patterns in this election, is what will influence a Labour government.

Still, under the assumption that Labour will win this election comfortably, for progressives to vote for a party whose policy platform is closer to theirs will almost by definition do no harm. However I think there is a better strategy that would do more social good than sending a weak message to the Labour leadership. It involves thinking not about a Labour government, but the opposition to it. 


What happens to the Conservative party after its defeat? The most likely scenario, as I’ve argued elsewhere, is that it continues in much the same policy space as it currently is. If Labour is successful this will tend to keep the Conservatives out of power, but if Labour makes any big mistakes or if accidents happen then the Conservatives will return to power, and we are likely to see another long period like the last fourteen years. In other words, the pattern that began in 1997 will be repeated (where the ‘accident’ was the subprime crisis in the US). In a country where the right wing press has such an influential role, it is foolish indeed to assume a very socially conservative, economically very right wing party can never win an election.

How can this depressing long term future be avoided? The radical way to avoid it would be to take political power away from wealthy media barons and money more generally, but I doubt that Labour governments will have the courage to do that. In addition reform from within the Conservative family (press, MPs, members, donors) is unlikely even if the Conservatives end up with around 100 seats after 4th July.

What could be enough is if the Conservatives lose so badly that they are no longer the main opposition party in the eyes of the media or voters. Being the official opposition gives you much more visibility and influence than being a third party, as any Green or LibDem member will tell you. (The exception is of course Reform and Farage, but again that reflects the power of the right wing press.) The party that can challenge the Conservatives for this official opposition role are not the Greens but the LibDems. If the Conservatives were no longer the official opposition, or had to share that role with another political party, that just might be enough to make Conservative members and newspapers ask whether the party has become too right wing and too socially conservative.

I don’t think this outcome is likely [3], but current polling suggests it is possible with strong tactical voting. Furthermore it has a greater impact than voting Green (say) would have on a future Labour government, so it seems to me to be a more effective and progressive strategy for those on the left to follow. It involves doing similar things to Freedland’s punishment strategy, because both involve voting in marginals against the greater evil, but it comes from a forward looking perspective. But there is an interesting little twist that this strategy adds in some seats where both Labour and the LibDems could plausibly unseat a Conservative MP.

For example, my own constituency is traditionally Conservative, with the LibDems as the main challenger and with Labour in clear third. However, according to the FT model although all three parties today have that same ranking they are very close in terms of number of projected votes. There are likely to be a number of seats like this, where the LibDems have traditionally been second to the Conservatives but where Labour’s popularity has moved the projected Labour vote closer to the LibDems. Without tactical voting, or voting for a more progressive party, the Conservatives could well retain the seat.

In these circumstances should I vote LibDem or Labour, assuming no clear guidance from constituency polls and where projected vote shares based on modelling and national polls is very imprecise. Suppose also that I have very little information about or preferences between the Labour and LibDem candidates. If Labour wins it would add one to an expected large Labour majority, but if the LibDems win it would strengthen their role as an alternative opposition to the Conservative party. If a Labour government has to worry as much about a LibDem opposition as a Conservative one, this would push the UK political discourse in a more socially liberal direction. The best social outcome, as well as the best way to punish the Conservatives, would be to vote LibDem rather than Labour in these particular circumstances.

I therefore remain convinced that tactical voting against the Conservatives where relevant remains the best option for this election. However where it is not clear whether Labour or the Liberal Democrats have the best chance of defeating the Conservatives, it also makes sense in this particular election and if the polls remain as they currently are to vote for the LibDems. This is not because the LibDems have better or worse policies than Labour, but because the real prize in this election would be to deprive the Conservatives of clearly being the main opposition party after 4th July. 

[1] An exception may be among voters whose views are well away from the political centre, who may be tempted to use their General Election vote to punish moves towards the centre, or other failings, by the main party who might otherwise get their vote. For example, preventing Diane Abbott or Faiza Shaheen from being Labour candidates during the election campaign cannot be justified in electoral terms (it will almost certainly lose Labour votes and possibly seats), and looks much more like a factional witch hunt. However unless that voting strategy achieves some change in the party they want to punish, it just represents another example of expressive voting.

[2] There are a small number of seats where the LibDems or Greens have a good chance of winning against Labour, and where there is no chance of splitting the progressive vote and letting the Conservative candidate win. In these seats tactical voting does not apply. In addition tactical voting is irrelevant in safe Labour seats, so voters are free to vote in a more expressive way.

[3] 1997 suggests that the LibDems do well in terms of seats when the Conservatives do badly. Tactical voting will mean they will do better in terms of seats than their national vote share would suggest.