2020 was an odd year, for obvious reasons, and as a result of those obvious reasons I conducted and published an audit in April about two big mistakes I had made that had already borne fruit: the first was my long-term failure to spot that Keir Starmer was the prohibitive favourite in the Labour leadership race, and the second, and more important, was my failure to predict the coronavirus and to adequately describe the risks with the government.
As a result, this year’s self-audit is going to involve rather more self-congratulation than usual, because I have already outlined a number of my mistakes (though I saved a big one right for the end of the year).
One of the consequences of a December election is that I didn’t have as much time to think about some of the things I got wrong both in the election itself, but more importantly in the run-up to the election, and I quite deliberately and explicitly planned to revisit them after the local elections in May 2020.. While there weren’t any elections in May 2020, and I will therefore also revisit these conclusions in December 2021, I want to touch upon them here.
I thought Labour’s position among Remainers would hold up fine
The biggest single miscalculation I made in 2019 happened right at the start of that year. I was of the opinion that Labour could continue to indefinitely hold onto Remain voters while having a pro-Brexit position.
My analysis was that Jeremy Corbyn’s general social liberal bona fides explained why he was able to hold onto so many Remain voters despite holding a pro-Brexit position, and crucially, explicitly and repeatedly saying that he held a pro-Brexit position, and that I saw no reason why this dynamic would change. That various Remainers kept describing Labour’s position as “unclear”, which simply wasn’t true but allowed them to keep supporting a pro-Brexit Labour party was, I thought, a stable equilibrium, however unstable it appeared.
A month later, seven largely obscure backbenchers split from the Labour party and formed the Independent Group. A significant chunk of Remain voters followed them away from the Labour party, though in the end the better-organised Liberal Democrats proved the main beneficiaries in the local and European elections. That forced Labour to move to a full-fat Remain position.
It’s still to be frank not wholly clear to me why the defection of seven MPs – who, and I mean no disrespect to the politicians in question by saying this, were not household names and who still could not be identified by the average voter – did what Corbyn himself repeatedly and openly stating that Labour’s preference was for Brexit to happen did not do.
However I think it does tell us something useful, which is that if something looks unstable, it probably is unstable. Throughout 2018, growing numbers of Remainers were getting restive about going ahead with Brexit, but weren’t willing to back an anti-Brexit party. I saw the various excuses about this – that Labour’s policy was “ambiguous” or “confusing” or “unclear” – as a sign that Remainers would continue to prioritise their commitment to beating the Conservatives over their views on Brexit. But I should have instead seen it as an unstable equilibrium that could not hold.
Just as I didn’t think that the polls made sense in 2015, when voters consistently said they did not want Ed Miliband in Downing Street or Ed Balls at the Treasury, but then claimed they would vote Labour, I should have concluded that Labour’s continued hold over Remainers had to give at some point – either through Remainers deciding actually they were fine with Brexit or with them abandoning the party.
The problem with this is that it doesn’t allow me to make more precise predictions: we can say, for instance, that there is a similar instability to the polls at the moment, which show Keir Starmer with a decisive lead over Boris Johnson in job approval but not a Labour lead over the Conservatives. But we can’t really say definitively which way that ambiguity will resolve itself: other than to say it seems likely to me that it will, one way or the other.
I overestimated the Liberal Democrats in December 2019
I found this particularly alarming because this was the exact same mistake I made in 2015, when I had the right idea about the Conservative-Labour battleground but wildly overestimated the performance of the Liberal Democrats in their constituencies. I really shouldn’t be making the same mistakes (not least because there are so many new ones left to make).
In 2017, I was right about the Liberal Democrats, but only by coincidence – a modest revival in their fortunes, at least as far as winning seats was concerned, did endanger the Conservative majority, but it visibly had very little to do with their own performance.
To make matters worse, my long-term predictions about the Liberal Democrats and their midterm performances were a lot better than my general election ones throughout the last five years. In those elections, and indeed when the Liberal Democrats have been grumbling about their leader – a constant of Liberal Democrat politics in my time covering them – I have consistently said that the party’s biggest challenges are the perception that voting for them risks “the other lot” getting in, making it much easier for them to gain seats in elections voters care less about, like local elections and by-elections, than it is in general elections. In general elections, they are essentially dependent primarily on the performance of the Labour leader.
Yet come general elections, I’ve tended to abandon this analysis, not explicitly, but implicitly, in favour of predictions that gave the Liberal Democrats a lot more agency over their own fate: and the results have been disastrous.
I think 2017 – the only one of the last three general elections in which my predictions about the Liberal Democrats have been right – offers the clue into what I’ve been getting wrong, which is to treat the Liberal Democrats’ fortunes as a statistically independent variable from the big two political parties.
I’m not saying that the Liberal Democrats have no role to play in their own electoral fortunes. If in 2019, every activist in Kensington had been in Wimbledon they’d probably have 12 MPs, for instance. But the Liberal Democrat contribution to their own success probably made the difference between 12 MPs and 13 in 2017 (a better national campaign might have saved Nick Clegg) and 11 MPs and 16 in 2019 (not being the Revoke party would surely have at the margin helped the likes of Stephen Lloyd hold on in Eastbourne, Tom Brake hold on in Sutton, and meant that Jo Swinson and Laura Gordon got enough tactical Tory votes to win in East Dunbartonshire and Sheffield Hallam). These might all have changed the party’s mood music but not, in any appreciable way, changed the general election result. Even in 2015, the coalition was a major problem for the Liberal Democrats, but what made the difference between 20 MPs and just eight was fear of a Labour government propped up by the SNP.
I reached these conclusions for different reasons – in 2015, I visited Liberal Democrat constituencies before the SNP attack line happened, and in 2019, I overcorrected having underestimated the willingness of Remain voters to actually switch from Labour at the start of the year. But the shared problem was in ripping up my pre-existing analysis of how the Liberal Democrats’ performances are shaped.
In future, I intend to make much greater efforts to stick to my midterm analysis of the Liberal Democrats: and to assume that, absent a compelling reason as to why, the party’s performances in general elections have much more to do with how Labour is performing than how they are performing.
I intend to revisit this again next year, in addition to some of my other mistakes from late this year. Speaking of this year’s mistakes, what did I get right and wrong in 2020?
I was right on the plans to loosen restrictions for Christmas
As I wrote in my April post-mortem, I didn’t spend enough time explaining that the government’s strategy was risky and spent too much time opining on whether the thinking behind that risk was correct or not.
On the whole, I think I did a better job in the latter half of 2020 in explaining that, yes, there were huge holes in the government’s lockdown regulations, and that these existed because of the large gap within the government about how to tackle the pandemic, between Rishi Sunak’s “learn to live with it” approach and Matt Hancock’s “lockdown until medical treatments advance” strategy. I’m also glad that, while I wrote a bit about the government’s thinking as far as the Christmas unlocking was concerned, I wrote a lot more explaining the logistical and health challenges of the unlocking.
I owe one of the more alarming parts of this year’s self-audit to Ailbhe, who reminded me on our podcast that in February, I had told her that I expected we would all spend some time being asked to work from home. I find this alarming because it shows a failure to align my private predictions with my written analyses.
I’m pleased that I was one of the earliest and most vocal critics of the United Kingdom’s original plan to ease restrictions over the Christmas period, not just because that analysis was validated by events – the risk of spreading coronavirus cases from cities to towns and rural areas proved too great to carry on with the unlocking – but also because it addressed both of the mistakes of the first half of 2020: I had opted not to see family at Christmas, and I focussed on explaining the risks rather than opining on the benefits of the calculation. (Though it’s important to do both!)
I was wrong about how long the coronavirus crisis would last
I thought that the era of lockdown would be one that took much of the decade, and would reshape life in ways we couldn’t accurately predict.
And from March to June, this underpinned much of my writing about the crisis – my initial criticism of the ideas that eventually became Eat Out To Help Out wasn’t just that they were a subsidy for the coronavirus itself, but that it was a mistake for governments to seek to artificially stimulate a return to pre-crisis ways of living and consuming which would not sustain after the crisis ended.
That was pretty much all wrong, though I’m not too bothered by this. I’m pleased that from July onwards I started to write regularly that the advances in palliative treatment and vaccination research meant that the lockdown era was likely to come to a comparatively swift end. In both cases I listened to and reflected the consensus among scientists and researchers and I would do so again.
I was wrong about Eat Out To Help Out
One reason why I didn’t take much time to consider the individual health risks of government stimulus measures in the summer is that I didn’t think they would work very well and they were happening too early in the pandemic. Had I thought they were going to succeed, I would also have been critical of them on health grounds.
However, Eat Out To Help Out was a success, both in terms of increasing footfall to restaurants and almost certainly in terms of increasing the number of coronavirus cases in the general population. Not only was I wrong about the effectiveness of the scheme but my wrongness meant I didn’t write enough about the health risks of the scheme, because, having concluded that it would not achieve the desired economic effect, I did not bother to also criticise it on health grounds.
I’m not sure why I didn’t bother to follow this up: perhaps I thought that the audience for a blog going “this won’t work, and you shouldn’t do it anyway” would be too small. But I think in general it’s good intellectual hygiene to carry the thought all the way to the end even if you do so over multiple articles and I will aim to do better at this in 2021.
I predicted Dominic Cummings’ revolution wouldn’t come to much
If you want, as I and Cummings both do, for the United Kingdom to focus on its strengths of science and education, at some point you do need to accept the electoral trade-off between bashing elites and ever-more repressive policies on immigration and helping our universities, large and small, the renowned and the new, to develop and grow. I have never believed that, when forced to choose between the electoral sugar hit of campaigning against “Mickey Mouse degrees”, restrictive spousal visas and the like, Cummings would choose to prioritise his stated mission of making the United Kingdom “the school of the world” over his electoral objectives.
I also thought that his combative nature would mean that his time in Downing Street would end similarly to his time at the Department for Education, where he helped make Michael Gove highly unpopular with the public at large, despite overseeing the continuation of a schools policy that was popular under New Labour and that continued to produce good results as far as grades and Ofsted are concerned.
A year later, Cummings left Downing Street at a time not genuinely of his choosing, most of his closest allies in Downing Street have followed him or are set to do so, and his changes failed to deliver.
What I think I did get right here was to take Cummings seriously: if you read his stuff and equally importantly the stuff he eludes to, it was pretty clear that there was a gap between his expressed aims and his real-world political approach. While I didn’t expect or predict the precise cause of his troubles, I did correctly identify the outline, that his combative behaviour and in the end the insufficient commitment to policy over politics would limit his effectiveness.
I thought we would leave with no deal
I wrote more times than I care to think about that we would leave with no deal. My reasoning was that Boris Johnson had consistently prioritised sovereignty for Great Britain, including at times when the risk to his own political advancement was large.
I use the term “Great Britain” quite deliberately, because he had no history of any serious engagement with or concern about Northern Ireland’s constitutional status and I expected that he would, as a result, be happy to sign up to a withdrawal agreement that deepened the sea border between Northern Ireland and Great Britain.
The sovereignty-maximising option was a no-deal Brexit, and, unlike Theresa May, I thought he was committed enough to it for it to actually happen. In the end, there was a EU-UK trade deal, with the British government making a series of concessions to get one.
I’m relieved that I noticed that I had got it wrong when Angela Merkel made an overcooked maximal demand on regulatory alignment, allowing the United Kingdom to accept the European Union’s original position and declare victory, and that the United Kingdom’s increasingly shrill rhetoric was a sign not of deadlocked talks but of progress, but it’s a bit like saying “well, once I hit the wall and was forcibly ejected from the car through the window, I realised I had been driving carelessly”: I think I get some points for observation but not enough to pass the test.
So what did I get wrong?
Something I neglected is that the main reason why I never believed Theresa May’s no-deal rhetoric – the visible and tangible absence of necessary infrastructure to actually manage a no deal Brexit – remained true throughout 2020.
Something I think has been validated in spades in 2020 is my belief that eventually, the politics does catch up with the policy: Rishi Sunak, the author of much about the coronavirus response that the public dislikes, is still very popular but his ratings have declined (while Matt Hancock, the author of much of what it likes, has not experienced a similar decline). At every stage of the pandemic, Johnson has opted to go with Sunak’s “ease restrictions, let the economy do its thing approach” up until the point the hard policy cost of that – rising numbers of coronavirus cases and overloaded hospitals – has begun to bite, and he has then U-Turned towards lockdown. That has been true throughout the pandemic and I think has wider implications for analysing Johnson in particular (in the end, a government led by him will always swerve away from difficult conversations with the electorate) but also governments in general (if the policy reality and the political positioning do not match, then eventually the political position will change to meet the policy reality).