The Audit: 2023

How have my predictions fared this year?

Every December I write a post looking at how my analysis fared over the course of the year and what lessons I’ve learnt. It has two benefits. First, it reminds me that I need to stay focused on trying to be right rather than pushing pet causes. Secondly, it forces me to consider what I got wrong and why. Dad will be doing one too.

Elections

I only made one set of election predictions this year – on the UK locals in the Spring. Next year they’ll be a lot more opportunity with locals and a general election here, plus the US race for the Presidency. (There are also elections in Russia but predicting the winner there is not much of a challenge.)

My predictions for vote share in the locals (Lab 38%; Con 26%; Libs 18%; Others 18%) were pretty close to the actual result (Lab 35%; Con 26%; Libs 20%; Others 20%). And I also correctly anticipated a strong “anti-Tory” tactical vote. This meant I predicted more Tory seat losses 850-900 than other pundits. But I was still too cautious on how bad it would be for them (they lost 1061), as I slightly underestimated both the scale of support for smaller parties and the level of tactical voting.

I’m fairly happy overall as I picked out the key trends in advance. I also tried to make predictions for all 230 councils, which netted out reasonably close to the result (albeit again too cautious on how bad it would be for the Tories) but with plenty of individual errors. This was a valuable learning experience given I want to do seat level predictions for the general election next year, as it showed the importance not just of local issues – like congestion charging in Cambridge which stopped the Liberal Democrats making progress in nearby councils as they did elsewhere in the South-East – but also local demographic shifts which can be quite pronounced between elections.  

Wider political trends

Retrospective assessment of trend analysis is inevitably more subjective than for quantitative predictions, but I think I’ve done reasonably well this year. On the question of whether the Tories would recover under Sunak I’ve been much more bearish than the median commentator view from the start and that has, so far, been born out. This is what I wrote a few days after he took over in October 2022:

“The initial polls since he took over show a small shift back towards the Conservatives but with Labour leads still in the mid to high 20 point range. Sunak himself is far more popular than his party – and is more or less even with Starmer on the “best Prime Minister” question. The reason for this divergence is that he does much better than Truss or Johnson with liberal/remain voters but they are not currently saying they’ll vote Conservative. My guess is that the numbers will converge. Sunak will become less popular with liberal voters as he takes decisions designed to appease the right of his party but the Labour lead will continue to fall as things calm down from the craziness of the past few months. But even if that Labour lead drops to 15pts or so it would still be a dire electoral position. He needs at least a 2 point lead in an election to salvage a small majority.”

And that is more or less the story of what’s happened this year. In January I set out why the Tory’s political strategy wasn’t going to work and an alternative suggestion:

“It would require disassociating himself from the party far more than he currently willing to do, and presenting himself as the mechanism for restoring competence and ethics to government.”

Which I still think would have been worth trying but in reality I’m not sure there’s anything Sunak could have done to substantially improve his party’s position. Voters don’t tend to pay all that much attention and when they notice something – like they did with the Johnson/Truss farrago in 2022 – it tends to stick for a long time. Plus incumbents all over the world, regardless of party ideology, are struggling given the post-Ukraine inflationary surge and ongoing pandemic recovery. When you spend most of your time writing about one country there is always a danger of being too parochial and missing global trends, something I want to spend more time on in 2024.

Nevertheless, Sunak certainly hasn’t made life easier for himself. He spent much of 2023 setting traps that he then walked into. The one on Rwanda and “illegal” migration was so obvious that I still can’t quite understand why they didn’t see it. Back in May I wrote a post about the Illegal Migration Act called “Collision Course” which concluded:

“If either the Rwanda plan and/or the Bill are blocked by the courts, Braverman and friends will push for ECHR withdrawal to be in the manifesto. And at the point Sunak will have to choose between upsetting them plus friendly media and at least some of his core vote, or doing something he knows to be a very bad idea and that would be condemned by other parts of his party.”

The collision actually happened earlier than I expected because the Supreme Court ruling was much more critical of the government position than most saw coming. That has led to new “emergency” legislation which, if passed, would effectively overturn the ruling by disapplying international law. Something that is likely to dominate the first few months of political coverage next year. I suspect Sunak will survive it, as rebel Tories have no viable alternative leadership candidate or any desire to force an election, but in doing so he will have wasted more of his small amount of time in office, and given Reform more of a platform.

In July I set out – in a post called “Wipeout” why I thought the Tories were at risk of a much bigger election loss than most people seem to realise. (Professor Will Jennings is one of a handful of analysts who has been saying similar things all year).

“My position since the Johnson/Truss calamities of last summer has been that the Tories are in very deep trouble. We could yet see some swing back, but there is no reason to assume we will, and quite a lot of reasons to think things will yet get worse.”

And I gave five reasons why:

• Ongoing failure – both economically and in public services.

• Sunak’s clear weaknesses as a campaigner and particularly his inability to hide his peevishness and sense of entitlement.

• Vote efficiency and proportional swing – i.e. Labour’s vote going up in the places which will maximise their seat total under first past the post.

• Tactical voting – as seen in the local elections.

• The rise of Reform (“It’s not hard to think of things that could push Reform towards 10% of the vote: a couple of big defections; further prominent failures on immigration; the return of the more charismatic Nigel Farage to lead the party; and so on.”)

Since then we’ve seen more examples of this ongoing failure – though also one important countervailing trend: wages are currently rising more than inflation albeit after 15 years of stagnation. We’ve also seen plenty of Sunak’s weaknesses as a political performer. And Reform rising towards 10% in the polls as Farage builds up his profile again (and I suspect they will do well in the Wellingborough and Blackpool South by-elections coming early in 2024). Moreover, the Tory vote share in polls has fallen since July and the word “wipeout” is now starting to make more appearances in mainstream coverage.

Of course this may all change next year, and we may still see swing back, but on the basis of current trends and data there’s still no reason to assume it.

I’m less happy with my more immediate response pieces to breaking news, where I have fallen into the trap, a few times, of making incorrect assumptions about Sunak’s reasons for doing things. For instance in a piece I wrote last month I said that, in binning Suella Braverman and bringing back David Cameron, Sunak “had finally made a choice” between his dual political strategies of wanting to look competent and sensible to traditional Tory voters versus crudely aggressive immigration and culture war announcements to appeal to their caricature of new voters acquired in 2017 and 2019.

In fact he had done no such thing. I overinterpreted the Cameron announcement which wasn’t really about political strategy at all but more that Sunak wanted a foreign secretary to deal with more high profile meetings and summits to let him concentrate on domestic policy. Likewise Braverman was finally dismissed because she wasn’t collegiate rather than because of her views.

One of the difficulties in writing about Sunak is that he doesn’t seem to have any political strategy so there’s no good framework through which to explain his decisions. What’s really annoying is that I noted this lack of strategy around the random set of announcements and briefings we got in September:

“There’s always a temptation to try and identify the clever strategy behind the apparent incoherence. There isn’t one. This is what happens when a Prime Minister tries to take every decision themselves without proper advice, or time to do any of them justice, while also going to G20 meetings and organising an AI summit. You can get away with centralising decision-making as Chancellor because you have more space and less to focus on. But not as Prime Minister. This isn’t three dimensional chess, it’s a bad hand of solitaire.”

But I then proceeded to assume he was being strategic a month and a half later. Perhaps the answer here is not to try and do immediate analysis. My strength is in using data to look longer term rather than the kind of post-match commentary that plenty of other people with much better contact books than me engage in. I don’t do much of it but when really big stories break I feel compelled to respond. Occasionally I might have something useful to add, but one lesson from this year it’s usually a better idea to wait a bit. Of course most things that seem important day-to-day turn out to be entirely irrelevant and forgotten within a few weeks.

Policy trends

My posts on policy issues tend to look at current challenges rather than make predictions for the future. But I have made some. In my post on the NHS in January I asked if it was in a death spiral and wrote:

“Nevertheless, there are real limits to short term solutions for this winter. Equally we know that things will calm down once we get into spring and summer purely due to the drop off in seasonal illnesses. The problem is that “calm down” in this context will still mean running over maximum capacity, as the system was last summer, rather than dangerously over capacity. Things might get marginally less hellish for NHS staff but no one will be getting a breather. This makes it hard to rebuild and strengthen the system in time for next year.

The nightmare scenario, and one I think is sadly very real, is that the system, already deteriorating pre-covid, was knocked so off balance that it will not be able to recover. Moreover the worse things get, and the longer the delays to people getting treated, the worse peoples’ health will become which creates additional pressure. At the moment waiting lists are still growing and pressure on social care is still getting worse. If I’m right – and it seems likely – that staff morale is cratering, leading to falling productivity and worsening retention, then that will put ever increasing pressure on remaining staff. There are vicious cycles all over the place.”

It's hard to properly assess this until we see how early January, typically the worst period of the year for the NHS, goes. We can, though, say that elective waiting lists have gone up and A&E waits are more or less at the same level as this time last year. As are delayed discharges. And as per my post at the weekend there is growing evidence of suppressed demand, leading to worse population health that is not picked in waiting list figures.

There are a few positives, ambulance response times have improved, largely as a result of new protocols around dropping off patients, and staff retention looks a little better than earlier in the year. But really the only thing giving me hope is that the government haven’t prioritised the crisis, which means one that did might be able to make some progress.

If anything my August post on the criminal justice system was even bleaker:

“Add all this together and we can see that the criminal justice system is stuck in a precarious balance of incompetence. If the police increase the number of people charged, which is necessary to restore public confidence and deter repeat re-offenders, then that will overwhelm a court system that already has a massive backlog, and a prison system that already doesn’t have enough places…. Which means the next year of debate about crime in the run up to the election is going to bear no relationship at all to reality.”

In it I noted that: “even with this permitted level of crowding the [prison] system is essentially out of space. And it would already be over the limit had it not been for the courts running slow because of the pandemic and barrister strikes.”

Two months later it did run out of space leading to a spate of news stories and new legislation to reduce the number of shorter sentences, and to allow people to be jailed in other countries. Not that this was a hard prediction to make given it was clear it was about to happen from the government’s own data. Similarly, following my analysis of the perilous state of university funding in July, we are seeing more universities having to make big cuts.

I do worry that I focus too much on failure and that my tendency is to notice risks rather than opportunities, though I try to often solutions to problems as well as diagnosis. It’s hard at the moment to see a way out of our doom loop, unless politicians embrace more radical proposals than they currently seem willing to do. Yet Britain has been in a bad place before, with contemporary commentary forecasting perpetual misery, and things have improved. It’s precisely when things are at their gloomiest that the potential routes out are hardest to spot. In particular my knowledge of technology is limited. My general instinct is not to get drawn in by tech utopia hype but I worry that means I am missing genuinely important shifts, and not just regarding AI.

Lessons

Generally I think this audit is a positive one. Regular readers shouldn’t have been surprised by anything that’s happened in politics this year, or by big policy failures when long-running problems eventually exploded into news stories. If you think I’m being too soft on myself please do say in the comments or by email – it’s entirely possible I’ve erased some big error from my mind out of embarrassment.

But as ever there’s plenty of room for improvement. So this year’s lessons:

• Predictions at seat/council/state level require a close reading of not just local issues but also changes in local demographics.

• Focussing all the time on the UK means my writing can get overly parochial and miss European/Global trends (i.e. incumbents are doing badly almost everywhere, something that should feature more in analysis of UK politics). In 2024 I’ll do more on broader trends.

• Immediate “post-game” commentary on political events is not my strength as it tends to involve second guessing the motivations of people I don’t have contact with. I’m going to avoid doing this – there’s plenty of it elsewhere after all – unless I have a particular insight because of people I know or experiences I’ve had. Post-election analysis is an exception because that just involves reading data.

• I will keep highlighting failings in policy and governance, but I do want to try and write some more positive pieces next year! Less around politics, which is not filling me with optimism, and more about broader trends elsewhere.

As ever thanks for subscribing and reading and, in return, I will keep trying to make my writing as useful as possible in 2024!