The Machine That Eats Prime Ministers

We need reform (no, not THAT reform)

Sir Keir Starmer has resigned amidst a leadership crisis.

“The question my party is asking now is whether I am best placed to lead us into the next general election. I have heard the answer of my parliamentary party to that question. And I accept that answer with good grace.” _{SIR KIER STARMER}

His successor will be the seventh prime minister in ten years, and the conventional account of Starmer’s failure is already solidifying. He was wooden. There were U-turns. He broke promises. He was cold. He was procedural. He was ideologically lacking.

But is this the story of a man, or a system?

Starmer’s crisis management failures

From a crisis communications perspective, one common thread has run through the UK government over the past 23 months: loss of narrative control.

Starmer was elected on a promise of change and competency after years of political chaos, but came to be perceived as directionless and lacking in ideology. In a Reuters article, “Unloved Starmer quits as UK PM after just two years”, his downfall is described as rooted in having “no big idea,” characterised by policy U-turns and an inability to explain what his government is fundamentally for. (Calling him “unloved” is a bit much, Reuters… “unpopular” is right there). As reported by AP, political science professor Rob Ford said Starmer’s government was “the antithesis of what he said he was going to be about, and it’s very hard to survive that.”

Starmer’s team failed to control the narrative across multiple issues. They let their opponents define the crisis first on issues such as immigration, energy, public services, welfare, and controversial appointments. So these issues became multiple independent subsets of outrage rather than being integrated into a coherent ‘national repair’ approach. Not getting ahead of critics and clearly defining your narrative both gives the narrative to the opponent and makes you the story in terms of your part in it.

Starmer’s team also over-relied on credibility and ‘being competent’, which only works when the promises you make translate into real results. Any campaign is somewhat based on promised competence, but if you’re not delivering on that after the election, restating competence is then perceived as evasiveness. Especially when you’re visibly mishandling mistakes. Starmer’s “I wasn’t told” defense in response to Lord Peter Mandelson failing security vetting during his appointment as US ambassador, for example. A defence that you’ve not been fully briefed about the extent of someone’s ties to Epstein comes across as out of touch at best; at worst, not in control of your own administration.

Furthermore, Starmer’s team failed to build emotional permission for hard choices. Voters can tolerate economic pain and difficult policy decisions when leaders transparently explain the trade-offs early and repeatedly. When you struggle to define your policy agenda, that makes the implementation even more challenging, and the focus on economic growth that never really materialised substantially damaged trust. The administration pursued policies that appeared designed to placate a demographic that would never vote Labour, which was a strategic miscalculation that alienated the base without winning the centre. For example, the decision to means-test the winter fuel payment, resulting in its removal from millions of pensioners who had previously received it, was a communications and political disaster that damaged the government’s relationship with older voters in a way that is extremely difficult to recover from.

The Southport Incident

To really understand what ended Starmer’s government and the broader issues that fed into it, we have to go back to 29 July 2024, when three children were stabbed to death in Southport. Before their names were even released, a lie circulated on X (formerly Twitter) stating that the attacker was a Muslim asylum seeker who had arrived on a boat. The real suspect, Axel Rudakubana, was a Cardiff-born British citizen, but he was also a minor, so the police couldn’t immediately correct the narrative. By the time the truth was shared, the lie had been shared millions of times, and riots had broken out across over 25 towns, with hotels housing asylum seekers set on fire and mosques attacked. Elon Musk declared civil war ‘inevitable’ and said Starmer was responsible.

The government’s response was robust, with over 1,800 people arrested… But Starmer never spoke to the fears beneath it all.

Starmer addressed the riots as a law-and-order problem, which misaligned with the perception of a significant portion of the public, who were feeling that the country was out of control and that the government didn’t understand them, creating a gap between the official reassurance and the lived reality.

This is a classic crisis management issue (perception vs. reality). Managing a crisis is not the same thing as showing people that you understand why they’re afraid or angry (or whatever else they might be showing you that they’re feeling). Procedural competence, when used in isolation, comes across as indifference.

But blaming Starmer misses the issue

Although all of the narrative around why Starmer failed is partly true, none of it is sufficient, because it doesn’t explain why the job of Prime Minister has seemingly become impossible for everyone who holds it.

Promises instead of policy

One of the broader problems the UK is facing is that the role of Prime Minister has itself become a crisis amplifier since the Brexit vote. Each Prime Minister’s central promise has been less of a governing philosophy than a crisis response to the failure of the government before them.

The conditions that ended Starmer’s role as Prime Minister were built long before he made any mistakes. The ‘machine’ that consumed him is the same one that got the five Prime Ministers who came before him. The UK has moved from David Cameron to Theresa May to Boris Johnson to Liz Truss to Rishi Sunak to Keir Starmer in a decade. Six prime ministers, each of whom inherited a set of unresolved structural crises, reduced those crises to a slogan, and collapsed when the slogan met reality. Any system that produces predictable, undesirable outputs is one that is failing.

David Cameron ran on “Stability,” and we got the Brexit chaos and a resignation that, in itself, was a broken promise. Theresa May’s promise was “Brexit Means Brexit,” and her stint as Prime Minister ended with votes of no confidence after Parliament rejected her EU withdrawal agreement three times. Boris Johnson said he’d “Get Brexit Done”, didn’t (at least, not in the way the slogan suggested), and resigned amidst scandal and accusations. Liz Truss promised “Growth” and released a mini-budget that is still affecting perceptions of the UK’s fiscal credibility (she was in office for 44 days). Rishi Sunak ran on “Stability and Unity,” but alienated right-leaning voters and catastrophically lost. Starmer ran on “Change and Competence” and delivered neither.

The issue with these promises as campaign slogans is that they’re crisis communications in themselves, rather than the less emotional and adversarial communications you’d expect from campaign communications. Each one addressed a largely single-issue sentiment created by the government that came before it, and each was designed to be believed in the short term rather than delivered over the long term, creating the conditions for its own failure. Messaging to the electorate, rather than your party base, requires that the underlying message and data hold up under scrutiny. They didn’t.

The contribution of online outrage

Combine the above with the current situation in which online outrage often shapes public opinion faster than institutions can respond…

Political crises are being processed through short clips, memes, influencer commentary, and partisan outrage loops long before official explanations land. Legacy media still commands higher trust, but 70% get news online, and 51% from social media. Online distribution models heavily reward conflict, novelty, scandal, and identity confirmation, and misinformation blurs the line between lies, mistakes, and contested judgments, collapsing them into a single entity. In an Electoral Commission report, 80% of people encountered misinformation on social media in 2025.

Coming back to the Southport incident, Rudakubana had nothing to do with the grooming gangs scandal, asylum seekers arriving by small boat, or the immigration debate in general, but in public perception, these narratives, combined with the false information spread on X, rapidly fused into a single story. The accurate information came too slowly, and with social media, the information that comes first often becomes ‘real’, even when it’s inaccurate.

On that note, looking at Starmer’s failure itself, there’s somewhat of a perception gap between how effective he was and how effective he’s perceived to be. He did sign major trade deals with the United States, India, and the European Union; he kept the UK out of a war, pushed back Trump... But voters want faster results. The bar has moved.

That 80% of people encounter misinformation online matters for how governments are judged. Misinformation degrades the conditions under which any information can be trusted. A government mismanaging a genuine error creates oxygen for false claims; when politicians lie or appear evasive, they make their later truths significantly less believable. A post blaming the Prime Minister for a murder gets millions of views; a careful assessment of his trade policy gets none. Platforms have no incentive to distinguish between them, and neither does a press that is chasing the same fragmented, attention-depleted audience.

The distribution models of evidence-based media now also reward the same dynamics as social platforms, so we have a media ecosystem that structurally disadvantages nuance and rewards the kind of emotionally charged narrative that governments are least equipped to compete with. The Reuters Institute stated in its 2025 Digital News Report that although evidence-based journalism should be thriving amidst the uncertainty we are currently facing, it is, in reality, weaker because news consumption is fragmented and trust is strained.

The Trust Reservoir

It’s that trust depletion that is affecting the UK government the most. Political trust functions as a reservoir. That reservoir matters a lot during crises. When it is full, governments can absorb mistakes and explain difficult decisions and ask voters to accept short-term pain for long-term gain. When it is depleted, any mistake is perceived as evidence of bad faith. And low or no trust in MPs is prevalent; according to NatCen, 45% of people surveyed in 2024 said they “almost never” trust governments of any party to put the nation above the party’s interests, and 58% said they “almost never” trust politicians to tell the truth under pressure.

These are not numbers that any single government created, nor are they ones that any single government can quickly reverse. In the 2024 NatCen report, NatCen Senior Research Fellow Sir John Curtice shared the following:

“The next government will not simply face the challenge of reviving Britain’s stuttering economy and its struggling public services. It will also need to address the concerns of a public that is as doubtful as it has ever been about the trustworthiness and efficacy of the country’s system of government. Addressing some of the policy challenges will help in that endeavour. However, it is likely to require much more than that – in particular, a style and manner of governing that persuades people that the government has their interests at heart after all.”

Starmer’s government did not do this.

But it’s not just Starmer’s problem because Starmer’s team inherited a massive trust and legitimacy deficit, and those things are needed to persuade people to accept difficult decisions (even when they’re necessary). The combination of weak delivery, overpromising, relying on single-issue slogans, scandal-driven media logic, algorithmic outrage, and misinformation can cause legitimacy to collapse even further, making any policy harder to implement because people stop believing that the government is acting competently or in good faith.

Starmer ran on competence. When competence was not perceived as being delivered, restating it was experienced as evasiveness. When evasiveness was perceived, the press highlighted U-turns, and the story became about the government’s lack of control rather than its policy intentions. The government then spent enormous political capital protecting a fragile image, which permanently damaged its credibility with the public. Each step in this chain made the next step worse.

When institutions panic, they make the wrong decisions. Starmer’s administration appeared to prioritise its relationships with internal party factions and external political optics over its relationship with the electorate.

This type of mistake underlies many comms failures, where institutions stop at things like, “If we use a slogan like ‘competence,’ people will trust us.” There might then be an assumption that they will be effective without sufficient strategy to be effective, or risk management to predict areas where efficacy will be blocked, so they then don’t think, “But when we fail to deliver, the electorate will see us as evasive.” So they miss the knock-on effects of, “And then the press will highlight our U-turns, and the story will be about our lack of control rather than our policy intentions,” and “then the government will have spent enormous political capital protecting a fragile image, while permanently damaging its credibility with the public.”

So, what’s next?

Andy Burnham is likely next. He’s spoken about the ‘poisonous’ nature of modern politics, but the machine that consumed Starmer won’t stand down for him. There will be new outrages, and we might be having this conversation again in 2–3 years.

Better communication won’t necessarily help Burnham, just as better messaging wouldn’t have been enough for Starmer because the issue was that Labour failed to build the conditions in which its story could be believed. Trust is rebuilt not through good communication alone but through the accumulation of kept promises and transparent explanations of trade-offs.

And a willingness and ability to correct errors before they become scandals.

So the question is whether Burnham’s policy platform will be more coherent and, more broadly, whether any individual leader can survive in a system that has been structurally set up to destroy them by the past decade of predecessors.

Surviving this system requires recognizing that the system itself IS THE crisis. To avoid being eaten by the machine, the next leader needs to remove its fuel. Part of the solution is to treat communication as part of the democratic process rather than as a damage-control tool. I.e., being transparent, getting ahead of the narrative, and correcting falsehoods before they metastasise, rather than waiting until the lie has been shared millions of times. Separating facts from judgments and making that separation visible to the public. Showing delivery in people’s actual lives, not in statistics that feel abstract and distant.

And making promises that can actually be delivered.

Because it’s action, not words, that matters the most for the next Prime Minister. The electorate is not waiting to be persuaded by better rhetoric.

The system eats ambiguity and nuance. When trust is depleted, it eats Prime Ministers, too. Six of them, in 10 years. For the next one, the question is whether they will be sufficiently different in their relationship to the truth and the public to survive.