Sometimes the most important political shifts are the ones that didn’t happen. At last year’s Reform conference it seemed set to become the dominant party of the right. It was 14 points ahead of the Tories, while Kemi Badenoch was struggling badly. There was a wave of Conservative defections, and the right-wing press looked poised to back Nigel Farage.

But now that lead has shrunk to four points. Reform’s leadership has been caught up in a series of financial scandals, with extraordinary new revelations published by Channel 4 this week, that appear to show Farage and his policy chief, James Orr, conspiring to break the law on political donations. Orr and another key aide, Dan Jukes, have had to stand down. On top of this their policy offer has become even more extreme; there is widespread infighting; Farage himself looks increasingly fed-up; and his attempt to force the other parties into fighting an attention-grabbing summer by-election failed.

One might have expected the Tories to see a substantial polling bounce as a result. But that hasn’t happened either. They are only two points higher than September last year and have performed dismally in recent elections. Badenoch has significantly increased her popularity, particularly with current and former Conservative voters, but that has not led to any material shift in voting intention. Instead Farage’s lost votes have mostly gone to Rupert Lowe’s Restore, an even more aggressively ethnonationalist party. Lowe is not going anywhere either. Last week he received a £500,000 donation from tech multi-millionaire Simon Galbraith to go alongside the loud and fervent backing of Elon Musk and his X algorithm.

This is a disastrous position for the British right. They are supported by almost 50% of voters but are so deeply split that this could translate into a small minority of seats at the next election. Their situation is even worse than it looks in the polls because the Labour and Lib Dem vote now operates as something close to a single bloc. There are almost no seats where they are in competition with each other and most of their voters are willing to vote tactically.

In the rest of the post I’ll run through the right’s options for resolving this problem. Can there be a merger or pact between the Conservatives and Reform? Can one party gain dominance and become the obvious choice? Would a leadership change make any difference? Could the press force the issue? Or is Andy Burnham going to luck out and go into the next election against a completely split opposition?

Read more

If you’ve been away over August you can look at our recent posts here - including ones on why “suicidal empathy” is such a dangerous idea; the myth of the “Boriswave”; why Iran could be American’s Suez; and much more.

Subscribe now

The problem with writing about Donald Trump is that one can expend an enormous amount of energy trying to explain moves he doesn’t understand himself. He has a set of underlying, and mostly wrong, beliefs about the world – mainly consisting of the idea that the US is being taken advantage of by everyone else. Beyond that, he’s susceptible to influence from almost anyone willing to apply the requisite level of flattery, takes almost no time to understand the historical context required for good decisions, and actively punishes those who tell him truths he doesn’t want to hear.

This is a particular challenge for foreign policy analysts, who like to write about grand strategies and world orders. Saying “this guy is in obvious cognitive decline, ill-informed and doesn’t know what he’s doing” over and over is not very satisfying.

One option is to assume that Trump’s erratic behaviour is part of a cunning plan to discombobulate opponents and get his way: a tactic straight out of The Art of the Deal. This is effectively what James D. Boys argues in US Grand Strategy and the Madman Theory. The book’s opening chapters offer an engaging intellectual history of the theory’s development and use under previous presidents. Most people associate the idea with Richard Nixon, but Boys convincingly argues that Nixon was influenced by President Eisenhower’s veiled threats to use nuclear weapons to end the Korean War.

The use of unpredictability as a way to improve negotiating strategy was developed by Cold War theorists including Thomas Schelling, Daniel Ellsberg and, most important, Henry Kissinger, who ended up working with Nixon. Whereas the eminently reasonable Eisenhower had used his secretary of state, John Foster Dulles, as his “bad cop” to create some uncertainty about how far the US would go, Nixon and Kissinger swapped roles. White House tapes captured them discussing Kissinger going around the world, encouraging co-operation from other states by telling them he couldn’t control his boss.

Boys sees Trump as consciously taking up this tradition in actions that have transformed US foreign policy, from the imposition of tariffs to threatening Greenland and yelling at Volodymyr Zelensky in the Oval Office. There is no doubt that Trump has, at various times, thought this was what he was doing. He clearly believes that being as bombastic as possible is the best way to negotiate, and often says that’s his plan, in the style of a reality TV contestant being interviewed about their tactics by the producers.

There was an early example during his first presidential campaign when he refused to rule out using tactical nuclear weapons against Islamic State: “I’m never going to rule anything out … I wouldn’t want to tell you that because at a minimum, I want them to think that maybe we would use them”. During the 2024 campaign, he told the Wall Street Journal that he wouldn’t need to use force against China because President Xi “respects me and he knows I’m fucking crazy”. (That these things are contradictory seems to have escaped him.) His aides regularly invoke Madman Theory as a justification for his behaviour.

The author expresses irritation with commentators who use discussion of Madman Theory as a hook to call Trump mad, noting that the whole point of the approach is to appear as if you’re a lunatic while actually being strategic. But the problem is not that Trump is mad, but that he’s utterly predictable. His ramblings on his Truth Social platform may be incoherent, but the broad framework in which he operates is consistent: he is a bully. Where he thinks he can take advantage of the US’s immense military and economic strength to undermine, belittle and force fealty, he does. When he faces any kind of serious consequences for his actions, he backs down. Constantly asserting how tough you are, to the extent of namechecking Madman Theory, is not what someone who was actually trying to be strategically unpredictable would do. Nixon never once acknowledged the idea in public (though ultimately his use of it failed to end the Vietnam War, because no one really believed that he was mad).

While other states find Trump hard to deal with, they have figured out that the best way to deal with him is through a combination of flattery, pretend compliance and waiting him out. As he neither cares about outcomes nor can pay attention to anything for more than a few weeks, he can be effectively managed this way. To justify his focus on Madman Theory, Boys goes out of his way to give Trump the benefit of the doubt, at great cost to his broader argument. In the discussion of tariffs, he accepts that the president’s decision to impose them on Canada was “in part, to stem the flow of fentanyl across the northern border” – even though this is an obvious nonsense. Less than 1 per cent of illicit fentanyl arriving in the US comes via Canada.

Boys describes the deal done with Ukraine to secure access to rare earth minerals as payment for military support as if it were an actual achievement rather than an obvious ruse to keep Trump quiet. Likewise, the bombing of Iran’s nuclear sites last year is described as evidence of “Trump’s willingness to follow through on his threats”, despite his failure to actually destroy the nuclear sites, in the face of his claim to the contrary. Indeed, he ultimately forced Israel’s Benjamin Netanyahu to back off from further bombing runs.

On none of these matters have Trump’s tactics worked. The US trade deficit is unchanged. Greenland has not been sold. Ukraine has not been forced to agree a peace deal with Russia. Iran’s nuclear capability remains. The current conflict with Iran, which began after Boys’s book was completed, is the clearest evidence yet that other states have seen through Trump’s bluster and faux belligerence. Despite numerous threats of armageddon, the Iranians have not backed down, and they called Trump’s bluff because he appeared unwilling to accept the economic consequences of escalation. China has taken a similar approach to managing trade wars, forcing concessions on the export of chips while finding ways to circumvent tariffs.

Perhaps the only way one could argue that Trump has achieved a long-term American ambition is in forcing European countries to up their defence spending, albeit nowhere near as far or as fast as they have promised to do so. Boys claims this as an example of Madman Theory’s efficacy, with threats of abandonment encouraging compliance by other NATO members. But there is no unpredictability here. Trump has made it clear that he has no meaningful interest in European security. It’s not a bluff, just hostility. And the cost of such disdain for the interests of former allies is that they are starting to exclude the US.

Ukraine no longer depends on the Americans, which has removed any leverage Trump had to force a peace deal on Russian terms. Increasingly, “middle powers” are profitably co-operating on matters of defence and security. While the US remains the strongest nation in the world due to its military spending and the dollar’s role as the world’s reserve currency, Trump has diminished its global standing for very little benefit. It is tempting to ascribe this to madness, but really it’s just incompetence from a president with limited capabilities.

For some, a reduction in the US’s footprint is a positive, even if it is through unconventional and often embarrassing means. Trump has found tentative backing in the past from “realists” who have spent years complaining about US imperial overreach and idealistic world-building missions. One of the most influential, Elbridge Colby, occupies a senior role in the renamed Department of War. Others, such as John Mearsheimer, have bemoaned Trump’s incoherence and inability to execute while praising his scepticism of liberal interventionism and willingness to criticize Nato’s expansion into eastern Europe.

Unfortunately for the realists, in his second term the president has found that he quite enjoys using the military, causing some dismay to those who thought he was an ally – not least to his vice president, J. D. Vance, and America Firsters in the media, such as Tucker Carlson. Those who weren’t already disillusioned have found the Iran conflict a chastening experience – it is hard to imagine a more obvious example of a pointless intervention, with unclear and unrealistic objectives that changed, and still change, day to day.

In First Among Equals, Emma Ashford does not try to defend Trump’s behaviour in office, instead setting out a realist programme more consistent with the objections to American adventurism that he expressed during the 2016 campaign. Ashford objects to the rhetoric of America First because it doesn’t sit well with potential allies: no one likes a bully. But her proposals, which she calls “realist internationalism”, are in that vein, seeking to scale back US military objectives and costs. As the US’s brief moment as the sole global hegemon comes to an end, Ashford wants “middle powers” to pick up the slack so that US taxpayers see more of their money spent on their domestic needs.

The so-called realists have been bolstered in recent decades by the failure of more ambitious American strategies. Russia, China and other autocracies have not been converted to liberal democracy via capitalism, as many hoped in the years following the Cold War. The post-9/11 invasions of Afghanistan and Iraq are now acknowledged by almost everyone to have been disasters on a historic scale, undermining the US’s moral and military authority at enormous expense in terms of lives and cash. The US’s humanitarian interventions have a chequered record: they stopped Slobodan Milošević from massacring Kosovar Albanians, but failed in Somalia and Libya. The realists also argue that the conflict in Ukraine is a direct result of overreach in the 1990s and 2000s, as the US and western allies unnecessarily provoked Russia by interfering in its “sphere of interest” – though others would note that Vladimir Putin’s own imperial fantasies might have more to do with it.

In practical terms, Ashford calls for a phased withdrawal of the US military from Europe and the Middle East over ten years, arguing that the strategic basis for such a large presence there (the Cold War and oil) no longer holds. She contends that this will give the Europeans plenty of time to get their act together and pay for their own defence, especially as Russian forces have been so heavily depleted by the war in Ukraine. In the Middle East, she wants most bases closed, with just a small naval force left to safeguard shipping lanes. China is now the main strategic risk to the US and so, she says, more focus is needed there, but with the emphasis on containment and building up regional allies to take up more of the slack in the future.

Realism is a useful lens insofar as it encourages policymakers to think about the world as it really is and to set viable objectives. Trump’s Iran fiasco is another reminder, if one were needed, of why it is a bad idea to use military force when there’s a lack of clarity of purpose in assessing likely outcomes. We might think that it would be good to help the Iranian people to remove a parasitical and brutal regime, but there is no practical way to do so via military means.

The problem with realism, though, especially in the rigid formulation set out by Ashford, is that it assumes that a narrow assessment of interests is always the right course, regardless of circumstances. Ukraine is an obvious example. Though the author is reluctant to say so explicitly, the logic of her position is not only that the US shouldn’t be supporting Ukraine now, but also that it shouldn’t have in the first place, given the lack of direct American interest in the outcome.

In a section entitled “No one loves a political realist”, Ashford notes that “realist arguments about the war in Ukraine – and about the limited interests that the United States has in that conflict – have not sat well with a commentariat populated by those who idolize the liberal ideas of the unipolar moment”. But she doesn’t explain why they are wrong. Without the support provided by the Biden administration in the first few years, Ukraine may well have collapsed, leading to an emboldened Russia. Avoiding that was unquestionably a good thing. To complain that it violated a neat approach to grand strategy is itself a failure of realistic thinking.

One doesn’t need to think that the US can or should respond to every humanitarian crisis or autocratic horror to think that it should help out when it reasonably can and when the objectives are clear and achievable. Seeing the world as it is can mean asserting caution in the face of moral outrage; it can also mean acknowledging that the idealistic solution is sometimes the right one too. Bill Clinton’s decision to intervene in Kosovo also violated a narrow understanding of US interests, but it worked and it saved countless lives. Realists would no doubt argue that the short-term success was not worth the effect it had in bolstering the misplaced belief that the US should intervene everywhere. But that’s just an argument for taking every situation on its merits and accepting that logical consistency isn’t always achievable in a messy and unfair world.

The same applies to the realist view of relations with allies, which is where there is the clearest overlap with Trump’s foreign policy. Ashford argues that the liberal interventionist obsession with permanent alliances, particularly Nato, leads to the US being landed with heavy costs and responsibility as others seek to freeload, while making it difficult to withdraw troops from regions that are no longer a strategic priority. The author approvingly quotes Glenn Snyder: “Alliances have no meaning apart from the adversary threat to which they are a response”. But while it is true enough that alliances and partnerships can outlive their purpose, calcify unhelpful beliefs and constitute a false economy – Netanyahu’s hold over multiple presidents is perhaps the most obvious example – Snyder’s line is manifestly untrue.

Alliances that go beyond the transactional, built on history and shared beliefs, have a value beyond the question of immediate threat. They create relationships that allow for a rapid response to new challenges that may not have been predicted. In the aftermath of 9/11, Nato’s alliances led to a genuine outpouring of support for the US (even if that support was misused by George W. Bush). Such alliances also help to build backing for broader non-military goals. One has to hope that, at some point, a president is going to acknowledge the costs of climate change and the level of international co-ordination required to manage them.

The diminution of alliances under Trump has led not to a stronger and more focused US, but to an increasingly isolated one that is spending no less on its military and simultaneously forfeiting any moral authority that remained after Afghanistan and Iraq. This is in the interests of neither the US nor the world. Again, genuine realism requires a bespoke approach to alliances that acknowledges a wider range of factors and is open to taking different approaches in different circumstances.

While Trump is president, there won’t be coherence in US foreign policy, beyond the occasional reassertion of his long-standing bugbears. The world will keep trying to work around him, probably successfully. The bigger question is what happens next. A Vance foreign policy would no doubt be more obnoxious than Ashford would like, though it would be more consistently “realist” in the sense of trying to reduce the US’s global commitments. But what about the other side? One hopes that a prospective Democrat president would try to find a role for the US that acknowledges the limitations of intervention and approaches global challenges with a greater level of humility, while understanding that there is a place for US leadership, and that the country’s extraordinary power can still be used for good in the right circumstances.

Opposing action to tackle climate change is getting trickier. Back in the early 1990s you could just deny it was happening. A group of fossil fuel companies set up a front organisation called “Information Council for the Environment” which put out ads with headlines like: “Some say the earth was warming. Some also said the earth was flat” and “Who told you the earth was warming….Chicken Little?”. Oil firms put out slightly subtler adverts in their own names, while trying to block the 1997 Kyoto Agreement, explaining that the science was terribly uncertain about everything, though their own unpublished research showed otherwise.

During the 2000s and 2010s that became an unviable strategy as a strong scientific consensus developed around the reality of global warming. So companies, commentators and politicians who wanted to argue against moves to combat climate change shifted to a new strategy. Now they emphasised the value of oil and gas, and that there was no reason to rush precipitously given uncertainty about the effects. After Joe Biden announced a major plan to tackle climate change, the number of these ads on social media shot up, falling just as fast once Trump won re-election.

While none of this has stopped countries from taking any action, it has helped to limit it, particularly in the US. But in the last year or so there has been another shift. The impact of climate change is now hitting rich western countries hard. It is no longer possible to discount this as a future problem. This summer we have seen extraordinary wildfires, thousands of heat-related deaths, and floods. It is only going to get worse. Public concern is rising and support for net zero remains solid in the UK, despite most press coverage being opposed.

So opponents of action are having to change course again, with a surge of anti-net zero articles and op-eds over the last year or so. The new approach is to argue the damage is now done so we should accept the inevitable and focus on dealing with the consequences rather than waste money by ameliorating further change. It was well illustrated in a recent piece by Allister Heath, editor of the Sunday Telegraph:

“Climate change is real, but Britain can do almost nothing about it. The war against global warming has been lost, with only grievous collateral damage to show for it….It’s time to accept this sobering reality, cut our losses, cease tilting at (and building) windmills, ditch the despicable con that is net zero, and learn to cope with, and whenever possible, even enjoy, a warmer, drier, more volatile Mediterranean climate, if that is indeed to be our fate.”

Heath is a particularly hyperbolic writer but he is using standard arguments: there is nothing Britain can do by itself; trying will mean irreparably damaging our economy; and will waste time when we should be focusing on adaptation.

This sequence of arguments perfectly matches a theory put forward by the influential economist Albert O. Hirschman in his 1991 book “The Rhetoric of Reaction”. Going back to the expansion of suffrage he shows that those wishing to stop popular progressive policies tend to acknowledge the underlying problem while seeking to prevent action via three forms of argument:

Futility: the reform will make no meaningful difference (Britain can’t change the climate alone)

Jeopardy: the reform may help but will harm some other valuable existing achievement (it will trash our economy)

Perversity: the reform will aggravate the problem it is meant to solve (emissions will go abroad / it will make adaptation harder)

In the rest of the post I’ll look at why these superficially attractive arguments aren’t true when it comes to climate change. But also at how this rhetorical structure keeps reappearing across numerous policy debates. When you have seen in once you see it everywhere. Progressives and centrists use it too. Though these types of criticisms can be true, they are nearly always overstated, yet can be very effective at killing popular policies.

Read more

I remarked to Sam that Donald Trump’s war with Iran seemed to be the stupidest war ever and he challenged me to back it up. I couldn’t because history is full of stupid wars. In any case, how does one measure stupidity to allow comparisons to be made?

Nonetheless, our conversation concluded with the thought that it might still be instructive to explore why this war is so stupid. And at least to compare it with other stupid wars of our time waged by Western countries to help understand the potential consequences for the US and the Middle East. Autocracies have their own ways of waging stupid wars, as Putin demonstrated in Ukraine with, in this case, human and material costs far higher than those of the Iran War. But democracies are supposed to have mechanisms for challenging such stupidity by subjecting the big decisions on war and peace to intense scrutiny. The need for secrecy may be used as an argument against letting too many see the plans, but there ought to be opportunities for the various agencies involved in any war-making to comment on the wisdom and feasibility of any proposed action.

For a war to be stupid it does not have to lead to the worst outcomes or be the most brutal in its conduct. For my purposes (and this is not an academic paper) three conditions have to be met for a war to count as stupid:

1. The war is launched even though there is a high probability that it will fail and even be counterproductive;

2. This probability could and should have been known in advance;

3. Later decisions compound the original mistake, even as it becomes apparent that the war is going badly.

Certainly Iran meets these conditions for a stupid war. It was set in motion on the basis of a flawed strategic logic and without thinking through the likely consequences. The possibility that the decapitation attack would not lead to benign regime change but instead to harmful retaliatory action, including the closure of the Strait of Hormuz, was evident to anyone who had studied this conflict. The Wall Street Journal reports how unequivocally the second condition was met:

‘Then-Director of National Intelligence Tulsi Gabbard outlined assessments from America’s intelligence apparatus. According to people familiar with the meeting, she told Trump that killing Iran’s supreme leader would likely usher in a more hard-line regime open to acquiring nuclear weapons. Tehran would rush to close the Strait of Hormuz, destabilizing the global energy market. And it would strike U.S. forces and partners in the Middle East, prompting questions about American resolve and reliability with allies.’

The administration had no right to be surprised by how things turned out.

But where this war really excelled in its stupidity was in meeting the third condition in its response to the original setbacks. This reflects a feature of the Iran war: it has been waged by stupid people, not only Trump but also his Secretary of Defense/War Pete Hegseth.

Read more

Last week Reform’s treasury spokesman Robert Jenrick set out proposals to cut welfare spending on working age adults in half. There are, to put it politely, some highly questionable assumptions in the analysis and costing (see footnote).1 But what really caught my eye was a piece Jenrick wrote for the Sunday Telegraph to launch his plan in which he criticised the current arrangements as an example of “suicidal empathy”.

I had been wondering when this phrase would escape social media confinement and make it into mainstream political discourse (it also featured on Radio 4’s “Moral Maze” last week). It was coined a few years ago by Gad Saad, a Canadian professor of marketing, who has embarked on the now familiar and lucrative career path of utilising an academic title to provide flimsy intellectual justifications for various prejudices. The basic concept is that empathy has become maladaptive in modern western societies leading people to direct assistance towards inappropriate recipients. This supposedly does such harm to truth and self-preservation that it will lead to societal suicide.

Saad has built up a following on the radical right due to his prophesies that European countries are on the verge of civil war thanks to immigration (and, as per Jenrick’s piece, spending too much on welfare). But his phrase really took off when Elon Musk started using it regularly on X and elsewhere to explain why wokery is going to end western civilisation.

If it was just another unpleasant right-wing meme it could, perhaps, be ignored. But the reason it is catching on is that it is based on an old set of ideas about how empathy, or related concepts like pity, can be harmful. These raise genuinely interesting philosophical and political questions. What Saad and Musk have done is develop a viciously ideological version to attack their enemies. It is not the first time philosophical discussion of empathy has been hijacked by the political right into a crude dismissal of suffering, with dire consequences.

In the rest of the post I’ll look at the antecedents of the phrase, how it has evolved, and why Saad’s version is particularly misleading and dangerous. Then I’ll use a few examples, including Reform’s welfare proposals, to show how it leads to inevitable misdiagnosis of problems and the wrong solutions. But also how a variant of the idea, properly applied, can lead to very different policy approaches to ones Musk or Jenrick would endorse.

Read more

These are my answers. There were a number on China which I have not dealt with here but will try to look at in a separate post.

Topics covered include:

• Lots on Russia and Ukraine: the situation on the front line; Russian casualties and mobilisation; the Wildberries attacks; Ukraine’s postwar situation; the challenge of a ceasefire for Ukraine’s allies; Russia’s threat to Europe and much more.

• UK military spending: how should it be spent? What’s the right amount? How do we get more bang for our buck?

• The Israeli elections and the impact of a Netanyahu loss for peace in the Middle East.

• The ongoing shambles in Iran, and the possible outcomes, as well as challenges for the Iranian economy.

• What I might have done in an alternative career and which historical figures I’d like to challenge at croquet…

Russian and Ukraine

Jonathan Tapp notes my view that setbacks can accumulate and add to the pressure on Putin without there being a cliff-edge that would force him into an immediate change of course, but then asks what would it take to convince him to shift him from his current approach? Ben raises one possibility of Ukraine replicating the success of its recent counteroffensive in the Oleksandrivka area elsewhere along the front. Against this, James McCoy asks whether Ukraine could be approaching a tipping point where Russian military advantage could start to increase exponentially?

Winston Reigh, noting the extent of Russian casualties, and how many families must have been affected and that this must raise suspicions about the extent of the losses. Staying with the topic of Russian public perceptions, Carol Gamm asks about general support for more mobilisation.

Jurai Kriz notes my difference with Mark Galeotti. Mark (who I take very seriously) seems to suggest that Ukraine’s attacks on Russian infrastructure are a bad idea as they embolden Kremlin hawks.

The issue that prevents serious discussion of a ceasefire is Putin’s determination to acquire all of Donetsk, although it would be naive to assume that if we reached the point where Ukraine was obliged to concede on this matter that this would be the end of all of Putin’s territorial demands.

Putin is pressing for the subjugation of Ukraine and he does not need emboldening to take punitive action. He has been doing it for years. It is hard to blame Ukraine for fighting back or assuming that it requires badly hurting Russia to get Putin interested in a ceasefire (which Ukraine continues to propose).

Many among the Russian elite question whether it is worth continued and growing economic pain along with the heavy casualties to get hold of land that matters far more to ordinary Ukrainians than it does to ordinary Russians. But Putin has shown no sign of bending, apparently fortified by the optimistic stories told to him by the Russian general staff about their advances. He may also be confident that for now however much Ukraine can hurt Russia, Russia can hurt Ukraine more.

Elite criticisms are becoming more regular although not enough to be regime threatening, and the banning of Yabloko, the closest Russia has to a ‘peace party’, from next month’s rigged elections means that there is no way for ordinary Russians to make their feelings known. A serious crisis - a ‘cliff edge’ - would force Putin’s hand but we have yet to reach this point. A widespread perception of imminent crisis might be sufficiently persuasive, if Putin accepted that he would be better off taking the initiative before he is seen to have no choice other than waiting until the situation is desperate.

One can read commentary predicting crises in the oil sector, the wider economy, and at the front. I’m always nervous about doubting Russia’s ability to adapt to challenges, but claims from earlier in the month that the Kremlin was on top of the refined oil problem have been undermined by recent Ukrainian strikes. In a recent post I discussed the regular attacks on Wildberries warehouses, which continue (see picture above), and their implications for the VTB bank, which is already carrying a lot of debt. This is a problem that can be handled by the Central Bank but only by adding to inflation. These are the ways that a cliff edge might be avoided but at the cost of making the longer term problem worse.

The biggest issue is the front line.

Read more

Topics covered include:

• Will/should Burnham call an early election?

• Have we passed peak Reform?

• Can Restore become an electorally viable party?

• Was Rishi Sunak the last Tory prime minister?

• Why isn’t Burnham talking about the climate crisis?

• If Remain had won how different would British politics be?

• What would happen if we introduced compulsory voting?

• What’s going on with the civil service?

• When will we get a coherent policy on universities?

• What fiscal options are open to the new Chancellor for the budget?

• Is the NEET problem really about the labour market or education?

• Is reindustrialisation the wrong objective for Britain?

• What’s going to happen in the US midterms?

• Could Project 2025 happen here?

• Is socialism finally breaking through in America?

• Can Jean-Luc Mélenchon win the French presidential election?

Andrew: Do you think Burnham will go for an early election given the poor performance of Reform and the Tories currently?

I don’t think there will be an election this year, nor do I think Burnham’s team have a definite plan for one in 2027. I suspect there is a concerted effort not to talk about it internally. The moment discussions go beyond a few people something inevitably leaks and there would be a press frenzy.

But the prospect will certainly be on peoples’ minds given the weakness of the opposition parties – Reform is struggling with multiple scandals and the Tories are clearly nowhere near ready for an election.

An even bigger reason to go early is the fiscal challenge facing the government in the next few years. The spending review will focus minds on the latter point, as ministers will have to set out spending cuts for election year, or announce further chunky tax rises that would go beyond hitting the rich. Real wages are also projected to fall over the next few years. So there will be a temptation to go soon in the hope of finishing off Farage and creating the space for better economic conditions by 2032.

Of course the downside is the risk of losing, and the high likelihood of a lower majority even with a win. A loss would have serious consequences for many of the most vulnerable people in society. That’s a high price to pay so if I were Burnham I would want three criteria to be met before agreeing:

Read more

The chapter reproduced here comes from another collection, this one edited by Hal Brands of the John Hopkins School of Advanced International Studies (SAIS). It covers all aspects of the interaction between AI and great power rivalry, military competition, economic statecraft, and competing models of governance. It has been published as Geopolitics of AI: Power, Conflict, and the Future of Global Order by Johns Hopkins University Press and is available on Project MUSE, https://muse.jhu.edu/book/143760.

Subscribe now

All actions take place in something virtually akin to dusk, which in addition, like fog or moonlight, gives objects an exaggerated size and a grotesque view.

— Carl von Clausewitz, On War

The concept of friction is one of Carl von Clausewitz’s enduring contributions to military thought. It was his way of integrating a practitioner’s perspective into his theory of war, showing the importance of understanding all that could go wrong. Friction was “inconceivable unless one has experienced war.” Recognizing its importance was how “real war” could be distinguished from “war on paper.” Friction was evident when apparently simple tasks became unexpectedly hard to achieve because of accumulating difficulties—“countless minor incidents” and “chance events.” Friction was why levels of performance always fell “short of the intended goal.” This was not about occasional disappointments but an essential feature of war.

Of all of Clausewitz’s concepts, this is the easiest to recognize and least challenged. Histories of campaigns are full of examples of armies being caught out by changes in the weather, unexpectedly difficult terrain, delayed supplies, or bouts of illness. Although this was not Clausewitz’s term, the “fog of war” is closely as-sociated with friction. This refers less to the physical constraints that may afflict an army on the move and more to the cognitive dimension. Vital aspects of a situation may be obscured or confused because so much is going on. The wider context in which a war is being conducted may be imperfectly appreciated, so that even what is known with apparent accuracy can still be misinterpreted.

When the benefits of modern information technology, including artificial intelligence, for military operations are discussed, reference is regularly made to their potential for lifting the “fog of war.” Commanders may still have challenges when it comes to designing and executing their strategies, but compared to Clausewitz’s day, their situational awareness should be excellent. Thus, one recent article on how AI is being applied to modern warfare describes the aspiration of connecting “sensors, weapons and other systems across all branches of the US military, delivering the information to commanders on one live interface—a single pane of glass to see right through the fog of war.”¹

In this chapter I will first consider the concepts of friction and the fog of war, then look briefly at Vietnam, and then the “Revolution in Military Affairs” (RMA), as discussed in the 1990s, which demonstrated some of the problems with attempts to penetrate the fog, before turning to more recent experiences with AI. The old forms of fog have not gone away, though their impact may be reduced, and there are standard problems of human cognition that can never be wholly eliminated. But there are now two new forms of fog. The first reflects the extent to which the information environment has become a battleground in itself, which means that belligerents have an incentive to generate forms of digital fog. The second is a function of the demand for speed in moving from detection of targets to their destruction, which AI can facilitate but creates risk of mistakes in target identification. In both cases, the problem is not so much AI, as the issues would exist in its absence, but of flawed strategic concepts.

Friction and the Fog of War

Clausewitz stressed the importance of friction because the inevitable elements of chaos and unpredictability in any military campaign required mental preparation. If the margins of error were slight, then a small amount of friction could thwart success. His remedy for friction was to have a clear plan, stick to it, avoid distractions, and not to try to be too smart.

While his appreciation of the unpredictable and confusing aspects of war still resonates, his doubts that better intelligence would help seem dated. Clausewitz saw intelligence as a source of friction rather than a remedy. In principle one might assume that better information about the enemy and its dispositions might pro-vide clarity rather than confusion. But he thought it inherently unreliable: “Many intelligence reports in war are contradictory; even more are false, and most are uncertain.” The rush to push forward new information encouraged a tendency toward pessimism, full of imagined perils: “War has a way of masking the stage with scenery crudely daubed with fearsome apparitions.” These got in the way of systematic thought and might lead to an unwarranted loss of confidence in the original plan. Better than jumping to hasty conclusions was to rely instead on the “laws of probability” and a “knowledge of men and affairs and . . . common sense.”²

Clausewitz’s concept of friction combines the physical difficulties connected with moving and employing armies with natural uncertainties about the true state of affairs and confusions of the mind. This latter aspect is often conveyed by the vivid metaphor of the “fog of war.” Although this term is often attributed to Clausewitz, it is not one he used. The idea can be found once, when he explains how “three quarters of the factors on which action in war is based are wrapped in a fog of greater or lesser uncertainty.”³ Elsewhere, as Eugenia Kiesling observed in 2001, he uses “fog” in a much more literal sense, as a meteorological event that contributes to friction.⁴

No matter its origins, the idea of the “fog of war” is compelling. While the metaphor may seem straightforward, it can refer to two different types of fog. The first is the fog that impedes perception; the second is the fog in the mind that impedes comprehension. The first form appears in 1896 in a short book with that title by Colonel Lonsdale Hale. The fog of war referred to “the state of ignorance in which commanders frequently find themselves as regards the real strength and position, not only of their foes, but also of their friends.”⁵

The origins of this state of ignorance can be actual fog, or the swirling smoke and dust of battle generated by detonations and explosions, or just too much to follow and too much going on to be able to keep track.

The second form of fog afflicts the brain, with its mental maps and scripts, expecting events to follow a certain course, only to be surprised when they do not. Poor or inadequate information might contribute to this fog, but that will not be the only cause. Flawed analyses of situations will make their own contributions. Understanding what makes organizations work, what makes different, and often hostile, countries tick, and why humans under certain forms of stress act the way they do is always hard.

This second type of fog, as a problem of comprehension more than perception, was reflected in former Secretary of Defense Robert McNamara’s commentary on his role in the Vietnam War. In a documentary made with Errol Morris about what lessons might be learned from Vietnam, entitled The Fog of War, McNamara explained,

“There’s a wonderful phrase: ‘the fog of war.’ What ‘the fog of war’ means is: War is so complex it’s beyond the ability of the human mind to comprehend all the variables. Our judgment, our understanding, are not adequate. And we kill people unnecessarily.”⁶

And then both types of fog may be aggravated by human or organizational frailties. Most important in war, which is why the fog here may be more severe than in other areas of human affairs, is that adversaries will seek to add to the fog—through concealment and deception, by looking for ways to surprise and confuse.

Information Superiority

The best way to deal with the most literal forms of war’s fog is to find ways to improve perception. Generals have progressed from directing a battle from the highest ground, to taking a close look at developments using telescopes and bin-oculars, to receiving signals and images from radars and satellites, and now taking feeds from ubiquitous drones. Even with the fog of the mind, there is always a hope that better information will make a difference. Espionage and code breaking might illuminate enemy intentions and capabilities while encouraging expertise on different societies. Information needs to be sifted, assessed, collated, and interpreted, and then incorporated into operational plans and the development of strategy.

The use of the scientific method to achieve this was developed in earnest by the British during the Second World War, for example in forming tactics for dealing with the German U-boat campaign in the Atlantic or identifying targets for bombers. After the war, it was picked up and taken much further by the United States using new analytical techniques, facilitated by improvements in computing.

Yet still the experience of the Vietnam War demonstrated there were limits to how much could be reliably known about enemy movements and plans during a counterinsurgency campaign, as well as the more fundamental problems of comprehension that later came to haunt McNamara. His demand for metrics—notoriously “body counts”—created a false impression of precision. The methodology meant that what could be counted was given exaggerated importance while what could not be counted, which might be more important in assessing situations, tended to be disregarded.

Kevin Baker describes one example from Vietnam. There was a collection program known as Operation Igloo White that used 20,000 acoustic and seismic sensors to track North Vietnamese supplies moving along the Ho Chi Minh Trail. These fed signals to computers at the Nakhon Phanom air base in Thailand, which analyzed the sensor data and sent aircraft to strike the convoys where the analysis predicted they would be found. Because the system sensed but could not see, the North Vietnamese realized how it could be tricked. They “played recordings of truck engines, herded animals near the sensors to trigger vibration detection, and hung buckets of urine in trees to set off the chemical detectors.” This led to many USAF strikes and claims of thousands of trucks being destroyed, although no wreckage of any was found.⁷

In this way, the Vietnam experience warned of the possibilities of deception but also the dangers of self-deception, caused by overoptimistic reports from the field and then consequentially misleading analyses presented with more confidence than they deserved. Forms of analysis with a scientific veneer could crowd out forms that were more qualitative and subjective.

While the shortcomings of the conduct in Vietnam were later acknowledged, new technologies encouraged the view that they might be overcome in future wars. The military should benefit from the collection of more accurate information, speedily analyzed and employed to great effect. In 1978 Under Secretary of Defense William Perry declared as an objective “to be able to see all high value targets on the battlefield at any time, to be able to make a direct hit on any target we can see, and to be able to destroy any target we can hit.”⁸

The next significant war fought by the United States encouraged a more optimistic view about the ability to penetrate the fog of war. After Iraqi forces had been overwhelmed in Operation Desert Storm in 1991, which served as a revelation about the capabilities of precision technologies, there was enthusiasm about the idea of a Revolution in Military Affairs. This was the result of the interaction between systems that collected, processed, and communicated information with those that applied military force. A so-called system of systems would make this interaction smooth and continuous.⁹ The underlying idea was that warfare could be more contained and discriminate, geared to disabling an enemy’s military establishment with the minimum necessary force.

The 1998 paper Joint Doctrine for Information Operations defined information superiority in largely warfighting terms as “the capability to collect, process, and disseminate an uninterrupted flow of information while exploiting or denying an adversary’s ability to do the same.”¹⁰ This could be achieved by means of “excel-lent sensors, fast and powerful networks, display technology, and sophisticated modeling and simulation capabilities.”¹¹ As Vice Admiral Arthur K. Cebrowski, United States Navy, and John H. Garstka argued in 1998, such a force would have “a dramatically better awareness or understanding of the battlespace rather than simply more raw data.” This could make up for deficiencies in numbers, technology, or position and speed up command processes. Forces could be organized “from the bottom up—or to self-synchronize—to meet the commander’s intent,” leading in turn to “the rapid foreclosure of enemy courses of action and the shock of closely coupled events.” There would be no time for the enemy to follow Boyd’s now-famous OODA Loop.

The most enthusiastic proponent of the potential transformation was Admiral William Owens, who argued that the “system of systems” would mean at last that the “fog of war” could be dispelled. He noted that in Desert Storm the United States had a real information advantage—after all the country had been able to move whole divisions without the Iraqis noticing. Yet even so, “vision was still spotty, usually late, and hardly clear.”¹² The classical fog of bad weather was present, aggravated in this case by smoke clouds from the massive oil field fires deliberately set by Iraq. The huge amount of force used was necessary to compensate for these deficiencies.

Looking forward, Owens wanted a system of “unprecedented fidelity, comprehension, and timeliness.” He “imagined a future conflict in which the commander has instant access to a live three-dimensional image of the entire battle-field displayed on a computer screen, an image generated by a network of sensors including satellites, unmanned aerial vehicles, reconnaissance aircraft, and special operations soldiers on the ground.”

At the same time, the commander would have “instantaneous access to information about the US military force and its movements” and would be able to relay orders to his subordinates, including by video teleconferencing. Their weapons would be precise.

The vision was clear, and in retrospect we can see how it was followed. Owens presented it in a compelling and authoritative way, and one can detect its influence, even to this day.

The Watts Critique

One of the most important early critiques of this line of thinking came from Barry Watts.¹³ He took on the claim that “foreseeable advances in surveillance and information technologies will sufficiently lift the fog of war to enable future Amer-ican commanders to ‘see and understand everything on a battlefield.’”

His analysis was complex and erudite, involving detailed case studies, including that of the Gulf War. It deserves to be recalled because it provides a helpful framework for assessing similar suggestions that AI might also dispel the fog of war.

Watts started with a careful consideration of Clausewitz’s concept of friction, which he then refined to come up with some criteria for identifying what this might mean in practice. The revised edition of his monograph came out after the start of the campaigns in Afghanistan and Iraq. Against this backdrop, he noted that here the challenge had not been so much one of defeating the Taliban, al-Qaeda, and Iraqis in straightforward military engagements. Instead, the “most troubling manifestation of general friction in America’s most recent conflicts has been achieving long-term political ends after the cessation of major combat operations.” This posed a set of challenges quite different from those imagined by Owens.

In looking forward, Watts asked “whether Clausewitzian friction will succumb to the changes in leading-edge warfare that may lie ahead, or whether such im-pediments reflect more enduring aspects of war that technology can but marginally affect?”

In his analysis, he stressed the physical as well as the cognitive sources of friction, resulting from the movement of soldiers and machines across challenging terrain in harsh conditions, while suffering all the stresses and hazards of actual combat.

Whatever the form of combat, the basic reason why friction could not be eliminated was because humans were involved: “Because at least one of the root sources of friction lies, when all is said and done, not in the weapons we wield but in ourselves. The presence of humans in the loop—with all the diverse frailties, physical and cognitive limits, purposes, and decisions which their presence and participation entail—alone seems sufficient to render friction impossible to eliminate entirely and, in all likelihood, extraordinarily difficult to reduce greatly in any permanent sense.”¹⁴

From this, Watts challenged the presumption that “knowing everything that is going on” in the battlespace is a problem that technological advances in sensor technologies and information systems will eventually “solve” once and for all. This was not, he insisted, a technical problem. However rich the information environment, humans could cope with only so much of what was available: “The greater the stress, the more data will be ignored, noise mistaken for information, and information misconstrued, and the greater will be the prospects for confusion, disorientation, and surprise.”¹⁵

At key moments, what was needed would simply not be available. And however much information was gathered, prediction would still be problematic. This reflected not only the limits on our understanding of reality but also the nonlinear dynamics of conflict. This, Watts argued, would not change “no matter how much information and processing power technological advances may one day place in human hands.”

He also saw three risks that might flow from the embrace of “digital networks and synthetic environments.” First came “the potential for subtle manipulation of our situation awareness by a sophisticated adversary, to say nothing of simply confusing ourselves with an overabundance of information. Indeed, it is entirely possible that applying information technologies to future war will expand, perhaps enormously, the potential for deception.”¹⁶

Second, the more “transparent” the battlespace became, there would be more pressure on human participants to “make life-or-death decisions in shorter and shorter spaces of time.”

Lastly, Watts stressed that however much it might be possible to reduce friction through technological advances, the key question was which side could do this most effectively. There was a “relative balance of friction.” War is a competitive exercise. The issue was not the absolute level of fog but which side was affected most, and what each could do to intensify the fog faced by the other.

An Artificial Reality

Watts anticipated that some of the problems faced at the turn of the century with the management of information might be eased because of new technological developments. Artificial intelligence would seem to come into this category. Are the sources of friction he assumed to be permanent susceptible to the extraordinary speed with which new systems can cope with vast quantities of information?

“AI” is an umbrella term for many distinctive types of machine learning with a variety of applications. In principle, systems that can bring together information from a wide range of sources and illuminate complex situations should limit the fog of war. Nonetheless, discussions of the impact of AI on contemporary conflict identify two ways in which AI might add to the fog of war rather than limit it. The first reflects the ability of an adversary to generate an artificial reality, a set of fake propositions that if accepted widely could undermine national security. The second concerns not so much the consequences of enemy deception but the self-deception that can result from overreliance on flawed systems. After briefly considering the first set of issues, I will focus on the second, for this is the area of greatest continuity with past efforts to penetrate the fog.

It is an evident feature of contemporary conflicts that the information space is soon polluted with various forms of fabricated information.¹⁷ An account of the 2025 Indo-Pakistan conflict reports a series of deepfake videos, including Pakistan’s Prime Minister Shehbaz Sharif apparently conceding defeat and footage showing simulations of combat as the real thing. Not all the fakery was AI generated, and it was probably a small proportion of the total. There were old tricks such as claiming that footage demonstrated a devastating blow to the enemy when it was taken from another conflict at another time. Asad Ali Shah notes the issue is not that the “disinformation ecosystem depends on advanced tools” but instead that it “thrives on emotionally charged content, visual ambiguity, and rapid dissemi-nation, often through trusted figures or institutions. AI simply supercharges the reach and realism of these efforts.”¹⁸

He worries that false claims can “provoke public outrage, inflame nationalism, and pressure political and military leaders into escalatory responses.” After suggesting ways to limit the harm done by misinformation, he observes: “The fog of war may be inevitable, but in the digital age, it is no longer created by smoke and chaos alone. It is algorithmically generated, virally shared, and frighteningly persuasive.”¹⁹

How serious is this issue? The problem lies not in identifying the activity, which is ubiquitous and intense, but in assessing its impact.

Those on the side promoting the disinformation might find it comforting, boosting morale through encouraging unwarranted optimism about the actual state of affairs. For real strategic effect, it needs to be taken seriously on the adversary side, causing confusion and dismay. Successful propaganda requires a good understanding of the target society so that the information imparted is sufficient and credible to subvert official narratives. This is only likely to happen if the veracity of these narratives is already in question. In practice, it is hard to shape the perceptions of an adversary society by inserting false or misleading images and texts, though it might be possible to accentuate divisions already present by feeding the concerns of the disaffected and anxious, in effect supporting a domestic opposition. Russia has undertaken many information campaigns, backed by bots and numerous posts on social media, often with little to show for their efforts.²⁰ This is not to be complacent about the risks, but it is to urge caution when assuming that shifting the perceptions of other societies will ever be straightforward. Brett V. Benson and Brett J. Goldstein explore the idea that AI can be used to generate a fog that will affect not only the military sphere but a whole country, undermining its ability to wage a war. They warn that adversaries “can poison the training data of models on which countries rely, thereby altering their output, or launch AI-enabled influence operations to sway the behavior of key officials. In a high-stakes crisis, such manipulation could limit a state’s ability to maintain credible deterrence and distort or even paralyze its leaders’ decision-making.”²¹

But their analysis, using a scenario of Chinese pressure on Taiwan, asserts the possibility of reshaping, and notes the dire consequences of success, without ever quite explaining the mechanisms through which such external interventions would work. They imagine hostile parties simulating “authorities giving false orders to stand down or divert to alternative communication channels, which could open a window for an adversary to gain ground.” This would lead to “a cognitive fog of war.”

Again, while there are risks of enemy manipulation of sensors or information sources—and we have examples from the Middle Eastern wars and Ukraine—or one side hacking into the systems of another in order to mislead and confuse, achieving this on a scale that would be transformational in terms of the conduct of a war would be unlikely. The ways in which opinion is formed, both at the elite and popular levels, are complex. There would need to be some staggering lapses of attention for so much external interference to have so much influence without the state being alerted to what was going on.

Speed Kills

This takes us to the second and more important contribution of AI to the “fog of war,” because it reflects an established stream of strategic thought. Clausewitz’s core theories were focused on warfare between peers, and that has been true of much thinking since, including the RMA. Aspirations to information superiority encouraged strategic concepts based on high-paced, high-intensity precision warfare.

The RMA reflected a wider doctrinal shift, emphasizing maneuver over attrition, placing a premium on speed, on the assumption that early blows would leave the enemy disoriented and confused. A fast operational tempo was essential to this approach, requiring targets that were picked with accuracy and hit with precision.

The underlying belief was that the way to win a war was to identify targets of high military value and then hit them remorselessly and rapidly. To explain their importance, another Clausewitzian concept was employed. This was the idea that an enemy force contained a “center of gravity” that if hit decisively might cause a more general collapse. There were different views about what might constitute a center of gravity—from unusually strong military formations to command structures to institutions of political control to alliance relationships. Behind the concept were often simplistic notions about how the enemy might be conceived as a politico-military system, so that striking the right spot could cause the whole system to unravel. It was also formulaic, as if the same approach could work against opponents with different values and social structures, with possibilities for resilience and retaliation that were not easily accommodated by the underlying American strategic theory.

The greater the confidence in a clear view of the battlespace, the greater the reliance on military strategies that could benefit from that clarity, even though they might not be the best ones to meet political objectives.

The emphasis on speed meant the challenge was to find better ways to maximize the impact of initial strikes by reducing the time taken from detection of targets to their destruction. As the military potential of AI began to be explored in the 2010s, this was the challenge it was expected to grasp. It could bring order and efficiency to a process that involved many different types of sensors feeding in information. It was not only possible to acquire and corroborate information through many means but also to present it in ways that its meaning could be quickly understood, enabling it to be effectively integrated with military operations. It could address at least one of Watts’s concerns—about how so much incoming information might overwhelm mere humans.

The so-called Third Offset Strategy, promoted by Deputy Secretary of Defense Robert Work in 2014, assumed that the best way for the United States to gain advantage over adversaries with larger forces was to develop operational and organizational constructs that would let the United States make decisions faster than peer adversaries, disorienting them as they struggled to match the faster operational tempo of the Americans. Out of this came Project Maven. Set up in 2017, it was intended to prevent operators from drowning in data by locating and collating essential information, picking up on patterns that might otherwise be missed and opportunities that might otherwise be lost. It addressed not only the old fog of war problem, in which vital information could not be seen, but also the new problem, in which the information was present but could not be found or turned into something militarily useful.

After a staff revolt at Google about Maven, the contract for its development went to Palantir. The company went through many iterations of the project, working closely with the military. Maven “fuses satellite imagery, drone video feeds, radar data, and signals intelligence into a single interface, allowing operators to classify targets, recommend weapons, and generate strike packages in near real time.”²²

The achievement was to offer operators a single interface so that it was not necessary to check with many different systems while putting together a case for a particular target. Once the case was made and accepted, then only a few clicks would set in motion the target’s destruction.

As Baker notes,

“The AI underneath the interface is not a language model, or at least the AI that counts is not. The systems that detect targets in satellite imagery, fuse data from radar and drone footage, and track objects across multiple intelligence sources are computer vision and sensor fusion. They predate large language models by years.”²³

The key thing about Maven is that it makes the process more efficient: “It consolidated the systems, compressed the time, and reduced the people.”²⁴ A 2024 study of army exercises using the system showed that, compared with the Iraq War targeting cells, roughly 20 people could do the work of more than 2,000 soldiers.²⁵ An account of a demonstration of Maven “showed the ease with which a user could turn a structure into a ball of flame with a ‘left click, right click, left click.’” This was a process that would previously have taken hours.²⁶

This was the culmination of an aspiration that had long existed, reflecting a strategic theory that emphasized the speed of multiple early blows. Because of the apparent quality of the information and the form of its presentation, it would be hard to challenge it, especially in a headquarters operating at a high tempo.

Yet the system could not provide perfect information. In a 2024 assessment some problems were noted. Because so much of the military real-life data came from the Middle East, algorithms honed on desert conditions could struggle when identifying objects elsewhere. When compared with human analysts identifying targets, who could achieve 84% accuracy, Maven achieved only about 60%.

“Some-times the system confuses a truck with a tree or ravine. Tanks are generally the easiest to spot, but with objects such as anti-aircraft artillery, or when snow or other conditions make images harder to parse, the accuracy rate can fall below 30%.”²⁷

Thus, the literal fog of war—meteorological conditions—remained a factor. Two further problems were noted. One was the risk of an enemy poisoning training data or hacking software updates. And then came the problems resulting from not being able to fully grasp the processes behind the options offered on the interface:

“Algorithms can lose accuracy over time, and because their decision-making is opaque, they’re harder to test than other military technologies. All this means that when it comes to using AI in the field, commanders will have to judge whether military need and the broader context justify the risk of errors. Con-fusing a wave with a warship is one thing; at worst, that might result in a mis-sile crashing harmlessly into the sea. Using AI targeting on a crowded terrestrial battlefield, with civilians in the vicinity, is another matter entirely.”²⁸

Many such problems were evident from the extensive use of AI by the Israel Defense Forces (IDF) to identify human targets inside Gaza in the campaign that followed the Hamas attack on October 7, 2023.29 The Israeli intelligence network is impressive and extensive. It listens in on phone calls—the Lavender profiling system can identify potential Hamas suspects based on age, gender, location, communication behavior, and so on—uses facial-recognition technologies, and employs a large language model of millions of intercepted Arabic conversations. All this enabled the IDF to scale up rapidly the number of human targets identified as potential Hamas agents. While some were discarded as not really being fighters, many were accepted without question. Although Lavender was said to make mistakes some 10% of the time, in practice the checks were cursory. As The Economist noted, the sprawling target banks maintained by the IDF meant the information was out of date. One analyst observed that “when civilians were killed in strikes, we would go back and check our information. Often it was because Hamas had used that building in the past but moved on and families had moved in.”³⁰

According to Mhajne,

“The apparent precision and objectivity of AI imparts unwarranted confidence and legitimacy to its utility as a force multiplier. The appearance of neutrality helps justify strikes that would otherwise provoke greater scrutiny, reinforcing policymakers already predisposed toward loosened safe-guards and expansive definitions of threats.”³¹

But the core problem was not the use of AI per se. It could have served as a starting point for commanders seeking out militants. The core problem was the disinterest in protecting civilian life while the search was on. It lay in the increasingly relaxed rules of engagement, tolerating high levels of civilian casualties, especially when aiming for senior Hamas leaders. The high levels of Palestinian civilian casualties reflected the strategic choices made by the Israeli leadership.

Epic Fury

The potential and problems of AI were demonstrated with Operation Epic Fury, launched on February 28, 2026. Taking advantage of Maven, the US military was able to attack more than 1,000 targets in the first 24 hours. Unfortunately, one of

the sites bombed was an Iranian school, leading to at least 175 deaths, most of them children. This was the result of dated intelligence, which had not been checked. If it had been, the presence of a school would soon have been revealed.

The situation was not helped by a Pentagon culture under Defense Secretary Pete Hegseth, who has emphasized “lethality” in the armed forces over “tepid legality,” so there are now far fewer people available to assess risks of civilian harm. CENTCOM’s civilian-harm team is at one-third of its pre-Hegseth staffing. So while humans stay in the loop, “some insiders acknowledge that the increas-ing scale and tempo of strikes has created incentives to give computers greater lati-tude in actually firing on the targets that they have generated.”32

The problem again was not reliance upon AI but that it was supporting a model of warfare with which the United States had been working for some decades, one that prioritized the rapid elimination of enemy capabilities. Because of the reluc-tance to use ground forces, everything depended on the impact of these early strikes, as there was no other way of putting pressure on the regime and getting it to negotiate on American terms.

The Pentagon boasted about how many targets had been hit and how much had been destroyed, but they could not hide the lack of evident strategic gains. The Iranian regime had lost layers of leaders but was still in power. Its military capa-bilities had been severely degraded, but it was still able to launch missiles at Israel and neighboring countries. Its control over the Strait of Hormuz was creating a major energy crisis and threatening to throw the international economy into recession.

Explanations for the failure pointed to the expectation that the first strikes would be so decisive that the Iranians would soon capitulate. The options avail-able to Iran should it survive the initial strikes were no mystery. There was just unwarranted optimism that the regime would not get a chance to implement them. When it did, the US options for mitigating their effects turned out to be inadequate.

With a sound strategic concept, AI can be a valuable and effective tool. Ukraine has shown how it can help to integrate many sources of intelligence to identify the positions of Russian troops, their equipment, and their supply routes. At a tactical level, Ukraine has made an effort to reduce the time from detection to destruction. But this was in a familiar context of a sparsely populated front line that operators knew well. Over more than four years of war, there have been regular technical and tactical innovations. The system has matured. This does not mean the fog had been eliminated. In its original, basic sense, it was still present. The front line was transparent, constantly observed by drones, creating a kill zone in which anything spotted was likely to be instantly hit. But low clouds and fog could still provide opportunities for small groups to move forward. There were efforts to confuse sensors using decoys and electronic warfare. And when Russian forces suddenly lost their connectivity in early 2026, when their Starlink no longer worked, they were, at least for a while, effectively blinded.

Clausewitz saw friction as an inevitable part of war, a feature to be accepted and considered in all plans. Since the early decades of the nineteenth century, the practice of war has changed considerably. Ways have been found to ease many, although not all, of the forms of friction with which Clausewitz was familiar. But with new systems and modes of warfare come new sources of friction. This includes the imperfect perception and cognition known as the “fog of war.” This has been presented as a challenge to military technologists. They have sought ways to blaze lights through the fog so that all becomes clear. The problem is that this has become an end in itself, encouraging strategies that can exploit whatever clarity can be achieved. In Clausewitz’s time, the problem was a scarcity of reliable information; now the information is plentiful, but it can be contradictory, dated, manipulated, and excessive. This creates the potential for different forms of fog. This should be expected and managed appropriately. It is not just a question of having humans “in the loop” as they try to make sense of AI-generated presentations of options, but to be aware that not all that appears on their screens can be trusted, that vital information may be missing, and that the course of a war will be shaped by factors that are intangible. In some instances, speed really matters, but at other times a rush into action can lead to costly mistakes.

Military strategy should be geared to political objectives, and that will require dealing with features of the operational environment that are difficult to know, and learning an appreciation of the priorities and options of adversaries that can be hard to grasp. As Clausewitz understood, the best strategies do not pretend that friction and fog can easily be overcome but accept their likely presence and plan accordingly.

Notes

Epigraph. Carl von Clausewitz, On War (Princeton University Press, 1976), 119–20.

1. Katrina Manson, “AI Warfare Is Already Here,” Bloomberg, February 28, 2024, https://www.bloomberg.com/features/2024-ai-warfare-project-maven/.

2. Carl von Clausewitz, On War, 117–18.

3. Clausewitz, On War, 101.

4. Eugenia C. Kiesling, “On War Without the Fog,” Military Review, September–October 2001, https://www.clausewitz.com/bibl/Kiesling-OnFog.pdf.

5. Lonsdale Hale, The Fog of War (Aldershot Military Academy, 1896).

6. The Fog of War: Eleven Lessons from the Life of Robert S. McNamara, directed by Errol Morris (2003), film. James G. Blight and Janet M. Lang published a book drawn from the film with a similar title: James G. Blight and Janet M. Lang, eds., The Fog of War: Lessons from the Life of Robert S. McNamara (Rowman & Littlefield, 2005).

7. Kevin T. Baker, “AI Got the Blame for the Iran School Bombing: The Truth Is Far More Worrying,” Guardian, March 26, 2026, https://www.theguardian.com/news/2026/mar/26/ai-got-the-blame-for-the-iran-school-bombing-the-truth-is-far-more-worrying.

8. Antoine Bousquet, The Eye of War: Military Perception from the Telescope to the Drone (University of Minnesota Press, 2018), 2.

9. William A. Owens, “The Emerging System of Systems,” Proceedings 121, no. 5 (May 1995), 35–39, https://www.usni.org/magazines/proceedings/1995/may/emerging-system-systems.

10. US Joint Chiefs of Staff, Joint Publication 3-13, Joint Doctrine for Information Operations (GPO, October 9, 1998), GL-7.

11. Arthur K. Cebrowski and John H. Garstka, “Network-Centric Warfare—Its Origin and Future,” Proceedings 124/1/1,139 (January 1998).

12. Bill Owens and Ed Offley, Lifting the Fog of War (Farrar, Strauss & Giroux, 2000).

13. Barry D. Watts, Clausewitzian Friction and Future War, rev. ed., McNair Paper 68 (Institute for National Strategic Studies, National Defense University, 2004).

14. Watts, Clausewitzian Friction and Future War.

15. Watts, Clausewitzian Friction and Future War.

16. Watts, Clausewitzian Friction and Future War.

17. Yara Elbehairy, “Fog of Truth: How AI Is Rewriting the Rules of War and Propaganda,” Washington Eye, June 23, 2025, https://thewashingtoneye.com/fog-of-truth-how-ai-is-rewriting-the-rules-of-war-and-propaganda/.

18. Asad Ali Shah, “AI, Deepfakes, and the Fog of War—Disinformation in the 2025 India-Pakistan Conflict,” IFJ Blog, June 25, 2025, https://www.ifj.org/media-centre/blog/detail/article/ifjblog-ai-deepfakes-and-the-fog-of-war-disinformation-in-the-2025-india-pakistan-conflict.

19. Ali Shah, “AI, Deepfakes, and the Fog of War.”

20. Lawrence Freedman and Heather Williams, Changing the Narrative: Information Campaigns, Strategy and Crisis Escalation in the Digital Age (Routledge for International Institute for Strategic Studies, 2023).

21. Brett V. Benson and Brett J. Goldstein, “The Fog of AI: What the Technology Means for Deterrence and War,” Foreign Affairs, January 6, 2026, https://www.foreignaffairs.com/china/fog-ai.

22. Michael Brown, “The First AI War: How the Iran Conflict Is Reshaping Warfare,” Forbes, March 21, 2026, https://www.forbes.com/sites/mikebrown/2026/03/30/the-first-ai-war-how-the-iran-conflict-is-reshaping-warfare/.

23. Baker, “AI Got the Blame.”

24. Baker, “AI Got the Blame.”

25. Emelia S. Probasco, Building the Tech Coalition: How Project Maven and the U.S. 18th Airborne Corps Operationalized Software and Artificial Intelligence for the Department of Defense (Center for Security and Emerging Technology, August 2024), https://cset.georgetown.edu/wp-content/uploads/CSET-Building-the-Tech-Coalition-1.pdf.

26. Cybele Mayes-Osterman, “The US Is Waging AI-Assisted War on Iran: Here’s How,” USA Today, April 1, 2026, https://www.usatoday.com/story/news/politics/2026/04/01/us-ai-assisted-war-iran/89299007007/.

27. Manson, “AI Warfare Is Already Here.”

28. Manson, “AI Warfare Is Already Here.”

29. Anwar Mhajne, “Regulating Military AI: Gaza and Beyond,” Survival 68, no. 2 (April–May 2026): 25–40.

30. “How America and Israel Built Vast Military Targeting Machines,” Economist, March 11, 2026, https://www.economist.com/international/2026/03/11/how-america-and-israel-built-vast-military-targeting-machines.

31. Mhajne, “Regulating Military AI.”

32. “How America and Israel Built Vast Military Targeting Machines,” Economist.

In the standard telling of postwar British politics 1979 is the pivot. Before then, so the story goes, there is a broad consensus between Labour and Conservative governments around Keynesian economic policy, active industrial strategy and keeping unemployment as low as possible. This approach falters in the 1970s as inflation becomes unmanageable and relations between trade unions and government deteriorate. Thatcher arrives, bolstered by an ideological cabal, determined to drive inflation down and smash the unions, regardless of the consequences. For the right this was necessary to save the country, for the left it was a disaster, responsible for many of today’s woes. When Andy Burnham speaks of “40 years of neoliberalism” he means 47 years.

Those with a more detailed knowledge of political history know that it is not quite that simple. The Labour government preceding Thatcher were forced into cutting spending, after borrowing money from the IMF, while trying to manage inflation. James Callaghan’s famous “in all candour” speech to the 1976 Labour party conference warned delegates the country could no longer spend its way out of a recession.

During the 1980s European countries with left-wing governments were also forced into similar policies, albeit without the level of privatisation or industrial strife that Thatcher was willing to support. In 1983 François Mitterrand had to abandon his pseudo-Keynesian economic policy to keep France in the European Monetary System. Some form of shift away from the postwar approach happened everywhere - leading to the rise of “third way” social democratic parties in the 1990s.

But the true story in the UK goes much further back than this, to the unplanned postwar revival of the City and its global consequences. It shows that much of what is termed “Thatcherism” came about as the Conservative government adapted to circumstances outside its control, which were often poorly understood by the Prime Minister herself. The end product was not at all what she had hoped to create. New Labour’s attempts to modify the Thatcherite approach were also distorted by structural economic changes that were not fully appreciated at the time.

This is not just a matter of historical interest. A simplistic understanding of Thatcherism on both right and left means today’s economic debates often miss the point, and risk repeating the same mistakes.

Read more

If you want to ask anonymously you can email it to samf@substack.com (but please use the comments otherwise as it’s much easier for us).

We always look to answer the most popular questions, so do “like” the ones others have asked that you’re most interested in. We’ll post our responses next week.

Subscribe now

If you’re a new subscriber here some of the things we’ve covered recently:

Sam has written on subjects including:

• Burnham’s policy agenda - with posts on the risks around devolution and the problems with introducing vocational education pathways at 14.

• Why Farage and Reform are in serious trouble (despite his win in Clacton).

• The myth of the “Boriswave” and why the right is being dragged towards ethnonationalism.

• Why Kemi Badenoch’s improved ratings don’t seem to be helping the Tories.

• The education attainment gap that never closes and what to do about it.

Lawrence has covered:

• Russia’s growing difficulties as the war in Ukraine continues - including Putin’s energy crisis; his troubles in Crimea; and whether he’ll need to launch another mobilisation.

• Ukraine’s issues with a reshuffle in their top jobs and the challenges of Russia’s war on civilians.

• The ongoing fiasco in Iran and tensions between the US and Israel.

• The problem with sensational fictional scenarios about war from the Thucydides trap to Annie Jacobsen’s nuclear disaster.

As always we’ve had some great guest contributions - including barrister Emma Chamberlain on the best way to do wealth taxes and Jack Watling talking about his time on the frontline in Ukraine.

If you’re not signed up as a paid subscriber yet we tend to paywall most of our posts so if you want access to all this sign up below. It’s the price of a coffee a month, or less if you go for an annual subscription. We’re now recommended by 630 other substack writers and have over 92,000 subscribers (free and paid), so we must be doing something useful.

If you’re already a paid subscriber, thanks so much for your support, which allows us to do all this. We’re entirely reliant on word of mouth for marketing so please do share posts and spread the word if you like what we do.

Subscribe now

Share

The longer the Russo-Ukraine war goes on the more it turns on questions of sustainability. The fight for territory continues as both sides probe for weak points and try new tactics and technologies. But increasingly substantial resources are directed at strikes against economic and social targets, seeking to weaken resolve and capacity. Each side is causing the other pain, and the pain will only get worse. The question for both governments is whether the pain is sufficient to cause a reappraisal of strategy or a readiness to change their conditions for accepting a ceasefire.

The Russians have air superiority. Their assaults on Ukraine well exceed the number of attacks in the other direction. In July Moscow launched a daily average of 172 strikes on Ukraine, compared to 28 on average launched by Ukraine. The recent Russian attacks have been particularly intense (one on Kyiv on 5 August is pictured above).

In a statement on 9 August President Volodymyr Zelenskyy said Russia had launched 61 missiles – including 56 with ballistic trajectories – more than 1,560 attack drones and 1,540 guided aerial bombs against Ukrainian cities just over the past week. July was the deadliest month for civilian casualties since 2022. At least 1,396 civilians were killed and 7,978 wounded in the first half of this year, up by well over a third on the first half of 2025.

Ukraine will have to absorb more intense and deadly Russian missile attacks over the coming months. At the same time, however, it has found new ways to cause pain for Russians, targeting not just energy infrastructure but other economic bottlenecks.

In the rest of this post I’ll discuss the similar but still distinctive strategies being followed by both sides as they try to make life unbearable for their opponents.

Read more

Over summer 2024 anonymous right-wing accounts on twitter, who support Reform or more extreme ethnonationalist parties, started using the term “Boriswave” to describe the brief but sharp rise in net migration from 2021-2024. A popular account called “Drukpa Kunley” pushed it particularly hard using racist imagery: “groups of Africans standing or walking along country lanes in the middle of nowhere – Boriswave”.

It was an inspired bit of propaganda, simplifying a complex set of different phenomena (humanitarian visas to Ukrainians and those leaving Hong-Kong, a post-pandemic surge in student arrivals, and a large increase in health and care workers), into a single word that neatly assigned blame to a political opponent.

The mainstream press started using it fairly quickly and by summer 2025 it had become familiar enough to appear without any qualifiers or explanation. Reform’s leadership began using it early on, as it helpfully emphasised the distinction between them and the Conservatives. Despite the term’s origins on the ethnonationalist right, even some Labour MPs couldn’t resist a phrase that blamed a political enemy. By September “Drukpa Kunley” was celebrating:

“Widespread adoption of the term ‘Boriswave’ in mainstream politics is a big victory for the British Online Right so-called - where the term originated last year. Internet influence is easy to overstate but it shows how low energy mainstream politics is [that] it can have this influence.”

As net migration is now plummeting towards zero, and small boat numbers have fallen almost 50% in the past year, the mainstream right have become increasingly obsessed with the “Boriswave”, as a way of keeping immigration at the top of the news agenda. Because the vast majority of those that arrived during that period did so legally the focus is not on deporting them immediately but ensuring they cannot get indefinite leave to remain (ILR), and ultimately citizenship.

For a time it looked like Shabana Mahmood’s reforms to ILR would stop many from staying, but pushback from Labour MPs is likely to lead to major concessions (though it’s not yet clear how far these will go). This has led to another bout of wailing and gnashing of teeth on the right. It’s got to the point where supposedly serious commentators are proposing retrospectively removing citizenship from people who’ve done nothing wrong and committed no crimes, simply because they arrived during 2021-24. See this, for instance, from David Shipley in the Spectator:

“Whatever Labour do, the Boriswave will have to be reversed. The next government will have no other choice. And if that means cancellations of settlement, and even of acquired citizenships, then so be it. We should be honest about this – it’s time for Reform and the Tories to announce that they will reverse the Boriswave, whatever it takes.” (My italics)

This fearmongering is usually justified with reference to the supposedly huge cost of accommodating all these people on a permanent basis. Shipley says it will “bankrupt” the country. Reform has talked of a £700 billion bill, £20k per household.

So in the rest of the post I’ll look at who the arrivals actually were; what we really know about the costs and benefits; and the dangerous policy implications of overreacting to a largely invented problem.

Read more

In early 1947 President Harry Truman was faced with a conundrum. The British government had told him that they could no longer afford to support the Greek government in its civil war with the communists and Turkey as it sought to resist Soviet pressure over access to the Dardanelles. Could the US please take over the funding. Truman accepted the challenge. He sought the support of the Republican leader in the Senate, Arthur Vandenberg, so that this could be a bipartisan effort. Vandenberg explained that to get support for such a bold measure, which would mean yet more entanglements in the affairs of war-prone Europe, he had to ‘scare the hell out of the American people.’

This led to one of the most important speeches of Truman’s presidency. Instead of making a prudent case to support a couple of European countries in trouble he offered something much more far-reaching. ‘It must be the policy of the United States,’ he declared, ‘to support free peoples who are resisting attempted subjugation by armed minorities or by outside pressures.’ Thereafter US leaders spoke regularly about the dangers of global communism with rhetoric that was, as Dean Acheson put it, ‘clearer than truth.’ Selling the threat worked. The commitment was accepted and the funding authorised. It was followed later in the year by the Marshall Plan to support European reconstruction. The Cold War was now underway.

Since then it has become accepted practice to ‘sell the threat’ when it comes to persuading the public of a potentially controversial move in foreign policy. We saw this when the American and British governments argued for military action against Iraq in 2002-3. Reasonable concerns about what Saddam Hussein’s regime might be up to with chemical weapons eventually led to talk of nuclear weapons - ‘we don’t want the smoking gun to be a mushroom cloud’ as Condoleezza Rice put it. This episode revealed one obvious risk with overselling threats - the huge loss of credibility when it turned out there was no smoking gun. When the same intelligence agencies correctly warned about the Russians being about to invade Ukraine in 2022 they were doubted because they had been so wrong twenty years before.

In a rare discussion of the pros as well as the cons of ‘threat politics,’ Brett Rosenberg observes that stressing threats can unify the country against external dangers and help rally allies as well as mobilise support for challenging measures to deal with the threat. But she also notes how the unifying effect can be short-lived and how policy-makers can get trapped. Having ‘emphasized or inflated a threat, it is a lot harder for them to dial it back down or to keep it confined to the areas they apply it to.’ Overstating the danger may lead to performative responses, thereby aggravating conflicts and precluding policy options that presume a degree of rationality from the adversary. The public might take the threat too seriously, as with the ‘Red Scare’ and McCarthyism that began as the threat of global communism was stressed in the late 1940s.

It also makes it very difficult to get measured appreciations of the real challenges posed by potentially hostile actors. Unfortunately careful analyses which place threats in context doesn’t excite the media or bring in the votes. Debates about defence and security are rarely framed in terms of sensible preparations for vague and uncertain contingencies. They are about whether the country faces a dire emergency.

I have discussed this in connection with the current UK government’s attempts to build support for higher defence spending, at a time when the security situation has deteriorated, in terms of finding the right balance between panic and complacency. Oversell the threat and then the natural question is shouldn’t we be doing much more? Undersell the threat and then shouldn’t resources be diverted to more important priorities?

This is also an issue for those who write and comment on these issues. Gaining attention requires dramatising issues and talking up dangers. An author is more likely to attract notice and readers with a ‘wake up call’ than with bland assurances that things are not too bad. In my book The Future of War: A History I looked back at many examples of such wake up calls, noting how predictions about future war are usually forms of advocacy. Many of these warnings sink without trace but some authors do break through and manage to frame public debates. In this post I’ll look at three individuals who have pushed issues onto the agenda by seeking to ‘scare the hell’ out of their audiences - William Le Queux on German spies in Britain in 1909-10; Graham Allison on nuclear terrorists and then a coming war with China; and more recently Anne Jacobson on nuclear war and now biological war.

Read more

Taxes on wealth seem the only place left to find the money, and there have been plenty of hints from his advisers that this is their preferred solution. There are lots of ideas floating around about how this could be done - and the pitfalls to avoid. So we asked one of the country’s leading experts on wealth taxes for a plan on how to do it right.

Emma Chamberlain is a barrister specialising in tax and a fellow of the Chartered Institute of Taxation, as well as a visiting Professor of Law at the University of Oxford and LSE. She is the co-author of the Wealth Tax Report produced by the Wealth Tax Commission in 2020. Emma is writing for us in a personal capacity and these views are her’s alone.

The case for a roadmap on taxing wealth

The debate on how to extract more money from the wealthy has already started in advance of Budget 2026: wealth taxes, land value tax, exit taxes, restricting lifetime gifts, capital gains tax (CGT) on death, higher rates of inheritance tax (IHT) on death. They are all being discussed in newspapers, think tanks and no doubt the Treasury.

This post argues that rather than redesigning the whole system or introducing a new wealth tax – expensive, slow and uncertain in yield – we should improve the existing taxes on wealth (CGT, IHT, stamp taxes, council tax) within a published ten year roadmap. This is also the best way to secure more revenue.

No successful company changes its financial objectives every year, yet Britain does precisely that with its taxes. We have had corporate tax roadmaps in 2010 and 2024; individuals and families deserve the same predictability. Pre-Budget speculation about taxes on wealth deters wealthy people from moving here or investing in the UK at all, reducing the available income for public services.

Major reform of capital taxes cannot succeed without some level of political consensus: if Labour taxed all lifetime gifts and the Conservatives promised to reverse it, people would simply wait for a change of government – exactly what happened the last time a Labour government tried to do this in 1974. The search for solutions on funding social care, and the suggestion of a 10% flat rate charge on death, may be the opportunity to build a better system more generally.

A roadmap requires a hard look at the capital taxes and their reliefs to see whether they achieve their purpose. Why must you survive seven years before a lifetime gift is exempt? Should inherited wealth be taxed differently from income? Above all we need coherence. CGT, IHT, the residence regime and property taxes interact, and reforming each in isolation – an exit tax here, a relief cap there – produces double and triple charges, and a sense of being “picked on” that drives mobile taxpayers abroad, and leads to protests outside Parliament.

Change can come in steps, but people need to know where they are going. I summarise at the end a possible roadmap for change.

Property taxes: fix what exists

Britain already taxes property more heavily than any other OECD country. The problem is not the quantum but the design: council tax is regressive and based on 1991 values; stamp duty (SDLT) is an economically damaging transaction tax with arbitrary cliff-edges; business rates deter improvement for fear of higher taxes.

The temptation is to sweep all three away and replace them with a land value tax (see this by Dan Neidle) under which only the value of land and not the buildings on it are taxed. Economists often like the idea because it encourages development and efficient use of land. But it should be resisted. LVT is conceptually clean but administratively hopeless.

Read more

I was on a family holiday when Andy Burnham set out his plans for vocational education for 14-16 year olds. It led to a lively discussion (we know how to have fun).

My twins, who are 16 and have just finished their GCSEs, were keen on the idea, having spent two years sitting with classmates who were disengaged by academic subjects. Why force them to take history or French when they could be learning something that will help them get a job?

It was a good reminder of why politicians keep coming back to the idea, and why it polls so well. It seems obvious that some people are just better at practical pursuits than reading textbooks. And employers want it too. They’re constantly complaining about an overly academic curriculum. We have worker shortages in construction and trades. It all seems to make sense.

But there are good reasons why vocational education at 14 has never taken off, despite being a popular idea that reappears every decade. Some of these are practical: secondary schools have limited space and are designed for classroom teaching not learning how to fix a boiler; good vocational training is expensive because it requires kit and more one-to-one support; and the teaching workforce is heavily oriented towards academic subjects. Most schools already offer some vocational options alongside GCSEs, but they tend to be in things like music and sports science which can be run by staff also teaching the related academic course.

If these were the only problems, though, they could be fixed by a government willing to find the money to build new facilities and train more teachers. The bigger question is whether it’s the right thing to do. Does encouraging young people down a vocational route that early in their education artificially restrict their life chances? Is it useful, or even possible, to teach employment-related skills to fourteen year olds? Is this what the education system should be for?

In the rest of the post I’ll look at what’s being proposed, reading between the lines of a somewhat vague announcement, and then at why I disagreed with my twins. I’ll finish with what really needs to change if we’re serious about improving technical education. Along the way I’ll tell you about the most awkward meeting I had while an adviser at the Department for Education.

Read more

Like many people trying to follow the Russo-Ukraine War I spend a lot of time staring at maps of the front line. These are often extremely complex, with red and blue areas shaded according to the degree of Russian or Ukrainian control of particular areas, red and blue arrows indicating lines of advance or retreat, and a lot of cross hatching to show where both sides have presence but neither are truly in control.

Because they are often zooming in on a small section of the front, sometimes no more than a town centre, it can at times be difficult to get a true sense of scale. Minor movements can appear more important than they actually are.

But then Zoom out and it all looks very different. Put the contested areas of Donetsk, for example, on a full map of Ukraine and it does not look to be such a big deal.

This perhaps helps explain why Vladimir Putin is prepared to squander so many lives to reach an administrative boundary that would allow him to say he has achieved a vital objective. Imagine him looking at the map from Moscow, presiding over his own vast country, and you can understand why he expects his forces to traverse the relatively tiny distances yet to be covered to the Donetsk boundary. As he continues to be fed encouraging news from his general staff about the steady progress of Russian forces, no wonder he keeps on claiming that victory is round the corner.

The Ukrainians are also asked why they bother trying to hold on to this marginal piece of land when the fighting could end if they just handed it over to the aggressor. They answer that it is not marginal but sovereign territory where people still live. Russia will struggle to take the two remaining big cities of Kramatorsk and Sloviansk, but if they were handed over then the land beyond Donetsk would be much harder for Ukraine to defend. Future Russian offensives would be comparatively straightforward. And another offensive would be likely even if there is a ceasefire. One big concession will only whet Putin’s appetite. He is still talking in expansive terms about how much of Ukraine should properly belong to Russia which is why he still demands in negotiations what he has failed to take in battle.

Yet Russia’s offensive is stalling. Despite all the effort, and huge losses, progress this year has been minimal. If he can’t get Kyiv to capitulate then battle is the only way forward. But if there are not sufficient forces to sustain the current offensive, then more will be needed. And this is why the issue of mobilisation has been raised, including by President Zelenskyy, who regularly warns that Moscow is preparing to call up hundreds of thousands of extra troops.

In the rest of this post I’ll discuss the reasons why Russia has a manpower problem and how it is struggling to overcome it without mobilisation. I will then consider whether Russia is better prepared now for a partial mobilisation than the last time one was ordered, in September 2022, before explaining why Putin might still hesitate before he gives the next order, and how the prospect of mobilisation is contributing to a broader sense of crisis.

Read more

Number 10 North is already up and running opposite Manchester town hall (affectionately christened “Manc-a-largo” by some civil servants). As expected it has been given responsibility for devolution, with teams moving across from the Treasury and local government department. Louise Haigh and Angela Rayner, who are the two ministers most responsible, will spend a portion of their working week there. Several senior staff from the Greater Manchester team that supported Burnham when he was mayor, and know the benefits and pitfalls of devolution, have moved across. This is going to be the defining project of his premiership.

As regular readers will know this is an agenda I support. Much of my book was about the problems caused by an overcentralised state, which has gummed up the centre of government while leaving us without the capacity to deal with problems at local or regional level. A couple of years ago I wrote a report about how to decentralise public services, for which I interviewed several of the Greater Manchester officials now working in Number 10 North.

So I’m excited by the prospect of a prime minister who genuinely believes in pushing power out of Westminster. But I’m also worried about how it happens. Doing it badly would discredit the idea for a generation or more, as well as harming the economy and an already fragile public sector.

One obvious risk is that it is oversold as the answer to everything. Devolution, like any substantive policy, involves trade-offs. There are big potential benefits in freeing up central government to be strategic, joining up services at local level, and giving people a greater sense of control over their area. But that inevitably means creating disparities, as some places will be better governed than others. Some mayors have done a bad job; so have some local authorities. We will be able to learn from what works and what doesn’t - but there will be failures.

Devolution also means additional complexity, as different areas adopt different approaches to things like public health policy or apprenticeships. In some areas life will get harder for businesses and other institutions that need to interact with the state and have, to date, been able to focus their attention on one Whitehall department. In selling devolution, Burnham and his ministers need to be honest about the costs as well as the gains.

But as well as these inevitable costs there are choices about how to devolve that will affect how well it works, and could lead to the whole project collapsing. In the rest of the post I’m going to look at five of the most important choices. How they are made will tell us whether a decentralised England is being set up to succeed.

Read more

On 15 July President Zelenskyy sacked the Minister of Defence Mykhailo Fedorov. Then, after a tumultuous few days, on 21 July he dismissed Commander-in-Chief Oleksandr Syrskyi.

Fedorov was sacked after proposing that Syrskyi should go because he was blocking vital reforms. Zelenskyy’s initial view was that, if he had to choose, then Fedorov had to go because he didn’t want to change the commander-in-chief. Syrskyi had the twin advantages of having no political ambitions and being ready to follow the president’s wishes. But then popular protests over Fedorov’s dismissal (pictured above) made Syrskyi’s departure inevitable. So in a matter of days Zelenskyy has lost both the top civilian and the top general in charge of the war effort.

Fedorov wants his old job back but for now he has been replaced by Yevhenii Khmara, a major general who had been acting head of the Security Service of Ukraine (SBU). The new commander-in-chief is General Mykhailo Drapatyi, previously commander of joint forces. He has had valuable experience for one so young (he is 43), is well respected and is widely seen as representing a fresh start.

For casual observers of the war this turmoil may come as a surprise because this year has seen Ukrainian successes in stalling Russian advances at the front, denying supplies to Crimea and causing fuel shortages around the country. Both Fedorov and Syrskyi can claim to have contributed to these successes. So why the tension between the two and how did it end up with both losing their positions? And what does it mean for the wider war?

Read more

There are many considerations in appointing a cabinet: rewarding loyal supporters, balancing factions, keeping rivals away from the spotlight, and, ideally, competence. They tell you a lot about what new prime ministers are worried about and how they intend to govern.

One thing we learnt from yesterday’s appointments, and some of Burnham’s other early moves, is that he has a much stronger theory of power – a view about how he wants to use the office and what he wants to prioritise – than his recent predecessors.

Throughout his time in office Keir Starmer was strangely passive, waiting for decisions to be presented to him and refusing to settle disputes between ministers. His government looked buffeted by events and u-turns. It is telling that in valedictory interviews he has talked of scrapping the two-child limit as a signature achievement, given it took eighteen months of lobbying by MPs and a mass revolt on welfare to make it happen.

Burnham, though, has indicated not just that he has priorities but that he is willing to make the institutional changes required to make them happen. John Healey’s appointment as Chancellor was part of a wider move to shift the locus of economic policy towards No. 10. I said back in May that I didn’t think Ed Miliband would be Chancellor, not just because of criticism from the right but also because he could create an alternative power base, given support within the party.

Healey isn’t a pushover, and certainly isn’t going to sanction a big increase in borrowing, but he is not going to threaten Burnham. And his views on economic policy are broadly in-line with the prime minister’s. He was responsible for regional economic policy under Gordon Brown, and has since argued that fiscal devolution hasn’t gone far enough. And he pushed for more social housing when he was the minister responsible back in 2009. He is, of course, keen to find more money for defence.

Notably, Healey was appointed after Burnham had given an interview on economic policy, saying that he thought there was flexibility in the fiscal rules and hinting at changes to income tax thresholds. Not the behaviour of someone who plans to sit back and let the Treasury manage the economy. This morning VAT on electricity bills was scrapped. This is the first in what will be a series of “cost of living” announcements that had been lined up by Burnham’s team well before Healey was offered the job (as per my assessment of his policy agenda from a few weeks ago).

In the rest of this post I’ll look at why changes to the No. 10 machine point in the same direction and then at what the rest of the cabinet appointments tell us about Burnham’s likely priorities and concerns, as well as the challenges to come.

Read more

After a brief respite the Iran war is back on. The question at the heart of the latest exchanges of fire is who is to control the Strait of Hormuz, to which the current answer is ‘nobody’ because both sides can deny control to the other and it is not obvious how either can change the situation solely through military action. Iran is betting that Trump’s concerns about the midterms and inflation will lead him to agree to almost anything to make the war go away, as he appeared ready to do when the memorandum of understanding (MOU) between the two sides was signed on 17 June. The US is betting that by denying Iran the economic benefits that came with the MOU, and by battering it with waves of air raids, it will be coerced into a more accommodating position.

Trying to work out how this will unfold is challenging because on the one hand there is a hubristic regime in Iran, pushing its luck, and on the other an erratic one in the US that sends a different message every day.

Some of what Donald Trump says can be dismissed easily. When he suddenly had the bright idea on 12 July that the US would not only control the Strait of Hormuz but charge a 20 percent fee on all cargo shipped, it was obvious that this was unlikely to last for more than a few hours. So it proved. It is only because he is president that the media felt obliged to treat this as if it was a serious proposition. It did not require the ‘shipping operators and logistics experts’ the New York Times consulted to know that the fee ‘would significantly increase the cost of transporting oil and other products through the crucial waterway.’

This article noted that ‘Mr. Trump did not explain exactly how the 20 percent fee would be calculated’ as if he had made a calculation rather than pulled a number out of the air because it sounded like an appropriate rake off for almost any deal in which he might have been engaged.

Read more