In May, the Labour government confirmed that HS2 will run at 320 km/h rather than the 360 km/h originally specified. When the cut was first reported in March, the Daily Mail’s verdict was that Britain’s flagship railway might now be “potentially... slower than in developing countries like India, Morocco and Indonesia.” Underneath the negative reactions was an assumption that is wrong twice over: that HS2 has been turned into an embarrassing, budget version of Britain’s high-speed ambitions, and that the same is true of the proposed HS2 Lite further north.

It helps to start with the original 360 km/h specification. The 2013 business case for HS2 promised the “best and fastest railway in the world”.

Not the most useful. Not the best value. The fastest.

And that is before considering that HS2’s main purpose was capacity relief in the first place. Speed mattered. But it was supposed to serve the railway’s wider purpose — capacity, connectivity and economic growth — rather than become an end in itself.

James Stewart’s 2025 review, commissioned by the government to investigate HS2’s cost overruns, traced part of the problem to “gold-plating”: building to unnecessarily demanding standards rather than asking what the railway actually needed to do. Lovegrove’s subsequent review of the Civil Service’s role in HS2 lists gold-plating among the programme’s “original sins”.

And gold-plating was hardly HS2’s only problem. The programme suffered from changing objectives, inefficient contracting, poor cost estimates, weak risk management and a failure to understand how much work was being handed to contractors before designs were sufficiently developed. Costs were repeatedly underestimated, while the programme became increasingly complex and difficult to manage.

The obsession with speed was not cost-free. HS2 was designed with exceptionally generous curve radii for unrestricted high-speed running, alongside demanding specifications for track, power, junctions and other infrastructure. Oakervee found that the trains themselves were not the main problem: manufacturers told the review that speeds would need to fall to around 250 km/h before rolling-stock costs became meaningfully cheaper. The bigger costs came from the infrastructure associated with a speed that was exceptionally ambitious, but offered limited practical value.

And yet Oakervee found little evidence that the programme had regularly stopped to ask whether these standards were actually worth what they cost. Britain fixed the ambition first and spent the next decade trying to justify it.

So what happens when you remove the obsession with speed?

HS2 runs at 320 km/h.

That is not slow.

320 km/h is the top operating speed of France’s TGV and Japan’s fastest Shinkansen services. HS1, which already carries Eurostar trains out of London on HS1, tops out at 300 km/h. China and Indonesia have services capable of 350 km/h, but 320 km/h is hardly some embarrassing second tier of high-speed rail. Nobody calls the TGV “TGV Lite”.

More importantly, the government has not suddenly rebuilt HS2 as a cheaper railway. The alignment remains the same, and the infrastructure has largely been designed around the costly higher speed anyway. The government's initial assessment says the principal savings come from avoiding the risks and costs of certifying the railway for a speed not currently operated anywhere in the world, alongside reduced testing and commissioning requirements.

The government estimates the change could save between £1 billion and £2.5 billion, while adding about three minutes to the London–Birmingham journey.

Three minutes.

That is not Britain abandoning high-speed rail. It is Britain finally asking whether 360 km/h was worth paying for.

Now apply the same question before the concrete is poured rather than after, and you get the proposal commonly labelled “HS2 Lite”.

After Rishi Sunak cancelled HS2 north of Birmingham in 2023, Andy Burnham, then Mayor of Greater Manchester, joined the West Midlands in commissioning a study, chaired by former Network Rail chief Sir David Higgins and led by Arup into a cheaper Birmingham–Manchester connection. Burnham is now Prime Minister, and his government has set a long-term objective to build a "lite" line between the two cities.

Except: Lite compared with what?

The Birmingham–Manchester leg of HS2 does not exist. There is no railway being downgraded, no 360 km/h service being replaced and no half-built line being stripped back. The Higgins–Arup proposal is therefore not a scaled-back version of a railway Britain had already committed to build. It is a proposal for a railway Britain has yet to decide how to build.

And it is deliberately less extravagant. The proposal envisages trains running at around 300 km/h and looks for savings through simpler infrastructure and greater use of land and assets already acquired for HS2. Its proponents estimate that it could cost up to 40 per cent less than the original Birmingham–Manchester leg.

The obvious response is that this must mean the north is getting less, but a cancelled railway is a strange benchmark for judging what Britain should build next. HS2 Lite is better understood not as a downgrade, but as a proposal that tries to learn from HS2’s mistakes.

That is why the language matters. Calling it “Lite” makes the original scheme the default and everything cheaper a compromise. But the original scheme is not the thing being delivered.

Iain Docherty and Jon Shaw have made a similar point about British infrastructure more broadly: things treated as fairly ordinary infrastructure elsewhere are often presented in Britain as “world class”, with the label becoming an excuse for greater complexity and cost. Look at high-speed rail elsewhere in Europe. Italy built the Bologna–Florence high-speed line through the Apennines, with around 94 per cent of the route in tunnels. They built difficult railways because they wanted railways.

Britain somehow ended up wanting the fastest railway in the world as well.

There is an obvious political incentive here. “We are building the fastest railway in the world” is a good headline. “We have specified this railway to the level actually required” is not.

One sounds ambitious, the other sounds competent.

That helps explain how 360 km/h survived for so long, and why 320 km/h is now being presented by some critics as a national embarrassment. It also explains why “HS2 Lite” sounds like a consolation prize when there is no original northern railway to be consoled for losing.

The problem with HS2 was never that Britain tried to build high-speed rail. It was that it repeatedly confused the most extravagant possible specification with the best possible railway.

320 km/h is not a failed 360 km/h railway.

A £1–2.5 billion saving is not evidence that Britain has given up.

And a Birmingham–Manchester railway designed at 300 km/h for substantially less than the cancelled scheme is not “HS2 Lite” simply because the number on its specification sheet is smaller.

The only embarrassment of the whole ordeal is that it took Britain thirteen years, multiple reviews and one cancelled phase to discover that “fastest” and “best” are not the same thing.

This is the second heatwave of 2026. The first, between May 23rd and 30th, was the earliest ever observed in France, a spring 2026 that is now officially the hottest on record, with a mean temperature 1.69°C above normal. And that’s before we even reach July and August.

The entire timeline is so familiar it is almost redundant: the weather maps turn ruby red, authorities issue orange and red vigilance alerts, ministers issue statements urging people to close their blinds and drink water, experts appear on television, civilians suffocate. Then it ends, the news cycle moves on, until another blistering heatwave emerges, a little hotter, a little longer, a little deadlier.

This heatwave should not be interpreted as an anomaly. It is a trajectory. The heatwave of August 2003 killed 15,000 people in 24 days, a death rate of 638 per day, peaking at 3,540 excess deaths on August 12th alone, a figure that exceeded the daily mortality rate of Covid-19. A Senate report called it “a veritable sanitary earthquake” for which the country was “manifestly unprepared.” Nearly 23 years have elapsed. That diagnosis is still applicable.

Take hospitals. At the Robert-Ballanger hospital in Seine-Saint-Denis, temperatures in south-facing rooms have been touching 45°C, with patients requesting early discharge. The improvised response, including survival blankets taped over windows, damp sheets in front of desk fans, blocks of ice standing in for cooling systems, is the standard. Only 7% of French schools are equipped with functional air conditioning, with an estimated 1,300 temporary closures already attributed to extreme heat. There is no legal requirement to close a school when interior temperatures become dangerous, despite the WHO considering temperatures above 30°C a genuine health hazard. France’s newest hospital, built at a cost of €1.3 billion, will have cooling only in intensive care and operating theaters — in other words, is obsolete before it even opens.

The political response has been predictably inadequate. Rival parties race to propose air conditioning, with the righ tabling mandatory cooling units in schools and care homes and the far-right promising €20 billion in zero-rate loans for refrigeration equipment while the far-left pretends to be Earth’s guardian angel while promising to end nuclear energy on their first day in office, which would, whether they admit or not, precipitate the return of fossil fuels, as we saw in Germany in the 2010s.

But mass mechanical cooling in hospitals already under severe budget pressure would worsen their finances and, by expelling heat externally, worsen the very urban heat islands that make heatwaves deadly.

One may wonder, then, if not air conditioning, what is the solution? One country provides part of the answer: Singapore.

The small Asian country is not an obvious reference point for a French summer: it sits one degree north of the equator, in permanent tropical heat, with humidity that makes Paris in June feel mild. But it is precisely because of these very conditions that it is the right comparison: Singapore has spent fifty years engineering urban livability in conditions far more extreme than anything France currently faces and has done so not with air conditioning as a first resort.

Its Green Plan 2030 commits to an additional 1,000 hectares of green space, with every resident within a ten-minute walk of a park. Its Skyrise Greenery program, enacted in 2009, has introduced over 300 hectares of vegetation into new urban developments through rooftop gardens and vertical green facades. While many find this architecture visually pleasing, it is, most crucially, effective: greening building facades reduces energy cooling loads by 10 to 31%. Cool paint coatings, tested in public housing estates, reduce ambient temperatures around buildings by up to 2°C. The city has also launched Cooling Singapore 2.0, a research programme specifically designed to map urban heat sources at the neighborhood level and feed the findings directly into planning decisions.

The contrast with France and, indeed, other European countries, is due to political apathy. France’s cities are built of stone and concrete that absorbs heat during the day and releases it at night, which is why Parisian nights in June no longer cool and Parisians — like myself — are suffocating. Its streets are overwhelmingly impermeable, preventing the evaporation that makes a forest floor ten degrees cooler. Most lowest-income private renters live in properties rated F or G on the energy performance scale, describing buildings that trap cold in winter and trap heat in summer with equally devastating efficiency. Countless people, especially students, live in “chambres de bonne,” small rooms under the zinc-rich roof that are rented out and particularly exposed to the heat.

Lastly, its urban planning framework contains no binding requirement for green coverage, cool surface materials, shading standards or the kind of systematic heat mapping that Singapore has made fundamental.

Whether it is France or any democracy in the West, there is an impossibility of preventing the inevitable. It is thus tragically unsurprising that, despite knowing that heatwave episodes alone are projected to cost a staggering €206 billion cumulatively between 2026 and 2030, politicians are deferring.

What Singapore demonstrates is that dense, hot cities can be made liveable through planning discipline applied consistently over time: green infrastructure, urban cooling treated as a public utility, heat resilience designed into the core of new developments. None of this requires a tropical climate or an authoritarian government, it merely requires decision-making, which European democracies seem increasingly incapable of doing.

While this heatwave is unprecedented, it is clear that its disorienting records will be broken, once more, in the near future, and again incessantly over the next century. While there are still ways to limit the the total temperature increase above pre-industrial levels, there is no longer any serious scientific debate about whether the damage already done can be undone because it cannot. The heat baked into the atmosphere by a century of emissions is not reversible. We will face more summers like this one. The question is whether, by the next heatwave, we will have decided to act to adapt. Every year of delay means more dead, more billions spent patching a crisis that planning could have absorbed and more records broken that will feel increasingly inevitable.

In April 2024, Malawi’s government, led by President Lazarus Chakwera, made a telling decision that revealed its true economic priorities. With elections approaching, the Reserve Bank of Malawi (RBM) reversed an IMF-mandated devaluation and fixed the official exchange rate. The logic was straightforward: devaluation was pushing up import prices, import prices were fuelling inflation, and inflation was losing votes.

Fix the rate, shield consumers, win the election.

Chakwera lost anyway. In September 2025, voters chose his predecessor Peter Mutharika by a landslide — 56.76% vs. 33.01% of the vote — with the economy as the dominant issue. The exchange rate fix had failed even on its own political terms, and left behind a structural wreckage that Mutharika now inherits.

This is not primarily a story about one government’s cynical monetary policy, but rather a story about what happens when a state tries to use an exchange rate anchor it cannot actually sustain.

The mechanics are worth understanding. When Malawi re-fixed its rate, it created an immediate gap between the official rate and the parallel market, which was running roughly 100 per cent above official for essential goods at its peak. This is not an abstract distortion. It meant that ordinary import-dependent households — the vast majority of the population — were being pushed onto a black market to access basic goods, while the formal rate offered them nothing.

Meanwhile the tobacco sector, which benefits from periodic step-devaluations and conducts dollar-denominated transactions, remained largely insulated from the kwacha (Malawi’s currency) depreciation eroding household purchasing power. Yet even here, the desperation of the FX position is visible: the RBM imposed a 30 per cent export surrender requirement to claw back foreign exchange from its primary beneficiaries, imposing distortions even on the sector the arrangement was designed to protect. The fix looked like consumer protection. It functioned as a subsidy to export capital at the expense of import-dependent labour.

Subscribe now

Credible exchange rate commitments require transparency, enforceability, and sufficient political cost of abandonment. Malawi’s arrangement failed all three. The Reserve Bank published no intervention data, giving private actors no basis for confidence in the rate. The government's willingness to re-fix the exchange rate after already agreeing to float under IMF Extended Credit Facility conditions demonstrated that the commitment could be abandoned on demand. And the ECF terminated without completing a single review, removing the one institutional constraint that gave the arrangement any external credibility. With gross reserves at just 0.7 months of imports — far below any recognised adequacy threshold — the Reserve Bank was not defending a rate so much as declaring one.

The consequences were predictable. Inflation peaked at 35 per cent in early 2024 and has since fallen to around 24 per cent as of early 2026 — a sign of some stabilisation — but still deeply damaging for a country where nearly three-quarters of the population lives below $2.15 a day as of 2025. The parallel market spread deterred foreign direct investment not only because of Malawi’s underlying economic weaknesses, but because investors could not reliably assess effective returns when profit repatriation might occur at either the official or the parallel rate. The currency distortion was generating uncertainty on top of genuine economic weakness, compounding both.

The IMF’s prescription — a unified market-clearing rate — is technically correct. The problem is that it is insufficient on its own. Malawi arrived at this crisis through a combination of fiscal dominance, reserve depletion, and absence of genuine central bank independence. A rate unification that is not preceded by fiscal consolidation, reserve rebuilding, and meaningful RBM autonomy will simply face the same political pressures that have hampered every previous stabilisation attempt. The structure that produced the April 2024 decision has not changed. Only the government has.

Mutharika inherited an economy where public debt has returned roughly to the levels that preceded Malawi’s major debt relief in 2006, in which the ECF remains terminated, and in which the distributional coalitions that benefit from exchange rate distortion are still politically powerful. The tobacco sector’s interests have not changed. The political incentives to prioritise short-term price stability over long-term monetary credibility have not changed. The question is whether a new government, without an electoral horizon for another four years, has the window to do what its predecessor could not.

There is a narrow case for cautious optimism. Chakwera’s loss demonstrated that economic mismanagement carries electoral consequences; the electorate delivered the punishment the IMF’s terminated programme could not. Mutharika has both the political capital of a landslide and the institutional memory of having governed before, providing a rare political window for change. The interest rate cut from 26 to 24 per cent in March 2026 is modest but suggests a willingness to begin normalisation. Yet these advantages are opportunities rather than achievements. The underlying constraints — weak reserves, high debt, and powerful interests that benefit from distortion — remain largely unchanged.

Optimism has to be conditional on the sequencing being right. Rate unification without fiscal consolidation is not reform; it is exposure. Malawi’s monetary problem is downstream of its fiscal and institutional problems, and treating the exchange rate in isolation will not resolve it.

The IMF knows this.

The question is whether the new government does too, and whether it has the political appetite to act on that knowledge before the next electoral cycle begins to foreclose options once again.

When a bank is fined for misconduct, the press release follows a reliable script. The institution expresses regret. It announces new internal controls. It commits to a culture of compliance. Politicians cite the penalty as proof that the system works. And then, usually within a few years, something very similar happens again.

This is not a coincidence. It is a pattern. And it points to something more troubling than individual bad actors or under-resourced regulators. Across sectors — from finance to employment law — organisations routinely satisfy the letter of the law while resisting its spirit. The gap between formal compliance and substantive change is not a failure of enforcement design. It reflects deeper structural pressures that shape how regulation operates in practice.

Large organisations hire compliance officers, establish internal review processes, and document procedures meticulously, not primarily to change behaviour, but to demonstrate, if challenged, that they took the rules seriously. Some become adept at compliance theatre: immaculate paperwork, impressive policy manuals, and reform that rarely extends beyond the filing cabinet.

Shauhin Talesh’s research on U.S. insurance intermediaries shows how this works in practice: firms reframe anti-discrimination law around litigation risk, bulletproofing workplaces against legal challenge while leaving deeper inequalities untouched. The letter of the law is observed. Its spirit is quietly neutralised.

The problem is not confined to firms. Research on regulatory prosecutors in Brazil found that overstretched officials routinised standard workloads just to create space for more meaningful engagement. Organisational pressure had impacts beyond the regulated because it reshaped enforcement itself. The people responsible for holding firms accountable were operating under conditions that made accountability harder to deliver.

Regulators face their own structural constraints. They are often technically outgunned, dependent on data supplied by the very firms they oversee, or on industry experts whose careers are intertwined with the sector they evaluate. In complex industries like finance, oversight frequently relies on proxies — the status of an expert, the reputation of an institution — rather than independent verification of what is actually happening. Overlapping jurisdictions blur responsibility further, making it genuinely difficult to determine whether substantive objectives are being met, or just formally gestured at.

Deterrence models assume firms rationally calculate the costs and benefits of non-compliance. In reality, those calculations are shaped by uncertainty, limited knowledge, and resource constraints, on both sides. Even carefully designed enforcement strategies falter when regulators lack the expertise or access to detect violations comprehensively. And when violations are detected, the penalty is often calibrated to make headlines rather than change behaviour.

There is a subtler problem too, one that rarely surfaces in policy debates. James Kwak, writing about the 2008 financial crisis, identified what he called cultural capture: the process by which regulators who work closely with an industry over years — attending the same conferences, drawing on the same expert pools — gradually absorb its worldview. They begin to see industry objectives as aligned with the public interest rather than in tension with it. Oversight becomes negotiation. When regulatory expectations then clash with organisational culture, firms may deliver what scholars call “unnatural” compliance: outward conformity without meaningful change. By the time a serious violation occurs, the regulator may genuinely not have seen it coming, not because they were corrupt, but because proximity had quietly narrowed their field of vision.

· · ·

Some theorists have argued for more collaborative, iterative approaches — responsive regulation, meta-regulation — built on trust and dialogue rather than fixed penalties. The instinct is right. But it carries its own risk: that collaboration slides into capture, that engagement becomes co-option, that goodwill and information transparency, both of which can be strategically managed, end up serving the industry’s interests in the name of working with it.

Substantive compliance is difficult not because regulators fail to design effective enforcement mechanisms, but because organisations adapt to them. Information gaps, institutional incentives, and cultural pressures systematically constrain regulatory authority. The obstacle is not the rulebook. It is the relationship between the regulator and the regulated, and the organisational cultures on both sides that determine whether rules produce change or just paperwork.

Governments announce crackdowns. Regulators publish strategies. Companies issue commitments. The audience applauds.

The question worth asking is not whether the rules are strong enough. It is whether anyone involved — regulator, firm, or politician — has a genuine interest in finding out.