Liz Truss became prime minister on 6 September. On Thursday, 20 October, she stood outside Downing Street and announced that she would resign.
Her premiership lasted 44 days, making it the shortest in British history. She entered office promising a fundamental change in economic policy based around lower taxation and faster growth. Her government’s mini-budget then triggered severe financial-market turmoil, the pound fell sharply, government borrowing costs rose, mortgage markets were disrupted, Chancellor Kwasi Kwarteng was dismissed, and his successor Jeremy Hunt reversed most of the economic programme that had defined Truss’s leadership campaign. By this week, Conservative parliamentary discipline itself was visibly deteriorating.
The political verdict seems obvious. Truss’s premiership failed with extraordinary speed.
The institutional verdict is more complicated.
Britain is now preparing for its third prime minister this year. Boris Johnson won an enormous general-election majority in December 2019, yet resigned this summer after a succession of scandals and ministerial resignations destroyed support within his own party. Conservative members selected Truss to replace him. Six weeks later, Conservative MPs effectively concluded that her premiership had become unsustainable too.
From one perspective, this looks like astonishing instability. Britain once associated frequent changes of prime minister with countries whose political systems it considered dysfunctional. Now the governing party is preparing to select another leader while continuing to govern through the parliamentary majority obtained almost three years ago.
Yet there is an important distinction between political instability and institutional instability.
Political instability means the people controlling government change frequently, governing coalitions fracture or policies reverse rapidly. Institutional instability means nobody knows which rules determine who governs, whether constitutional procedures will be respected or whether political disputes will be resolved through mechanisms outside the established system.
Britain currently has a great deal of the first.
It has relatively little of the second.
Nobody expects Liz Truss to remain inside Downing Street because she disputes the legitimacy of her removal. Nobody expects the army to determine her successor. Nobody expects rival Conservative candidates to organise supporters to occupy parliament. There is no serious dispute over whether the next prime minister will possess legal authority once the Conservative leadership process concludes.
The political actors are unstable.
The rules governing them remain comparatively stable.
That distinction matters because democracies are often judged visually. A country changing leaders repeatedly looks chaotic. A country retaining the same leader for twenty years looks stable.
But longevity can conceal fragility.
Vladimir Putin has dominated Russian politics for more than two decades. Xi Jinping has consolidated extraordinary authority in China. Numerous authoritarian rulers have remained in office for decades.
Their personal political stability tells us relatively little about how smoothly their systems would manage an unexpected succession or serious internal conflict.
Britain is currently demonstrating the reverse phenomenon. The individual at the top can become almost disposable while the broader constitutional system continues operating.
That may look humiliating for British politics.
Institutionally, there is something healthy inside it.
A political system should ideally be able to recognise failure and change leadership without requiring the entire constitutional order to enter crisis.
This is essentially a political error-correction mechanism.
Governments make mistakes because politicians are human and policy operates under uncertainty. The relevant question is therefore not whether a system can somehow produce leaders who never fail. No system can.
The more useful question is how expensive correcting failure becomes.
In an authoritarian system, removing a failing leader can be extraordinarily difficult because political authority has become personalised. Challenging the leader may threaten the careers, wealth or physical security of elites surrounding them. Institutions therefore acquire incentives to conceal mistakes rather than correct them.
A parliamentary democracy can operate differently.
The governing party may decide that keeping an unpopular prime minister threatens its own electoral survival. MPs therefore acquire incentives to remove the leader before voters remove all of them.
That incentive is not noble.
It is still useful.
Self-interest can produce accountability.
Conservative MPs did not suddenly overthrow Truss because they developed a philosophical commitment to fiscal credibility. Many became increasingly alarmed because opinion polls suggested the party was heading towards electoral catastrophe.
Their motivation may be survival.
The institutional result is that a prime minister who lost political authority could not simply continue governing for another two years because the formal general-election deadline had not arrived.
This is one of parliamentary government’s advantages.
Presidential systems provide executives with fixed terms. That creates stability because a president cannot usually be removed merely for becoming unpopular or politically ineffective.
But fixed terms also make replacing a failing executive much harder. Impeachment generally requires serious misconduct rather than simple incompetence.
Parliamentary systems trade some executive stability for greater flexibility.
A prime minister remains in office only while enough legislators are willing to sustain them.
That can produce revolving-door politics.
It can also prevent prolonged executive paralysis.
Again, institutional design involves trade-offs rather than perfect solutions.
Britain is currently experiencing the cost side of parliamentary flexibility. Conservative leadership contests have become substitutes for national elections in determining who leads the government. Theresa May, Boris Johnson, Liz Truss and whoever follows her have all occupied the premiership during the Conservative period of government, while only some initially entered Downing Street directly following a general election victory.
This creates a democratic legitimacy problem even if the constitutional mechanism is clear.
The Conservative Party won the 2019 general election under Boris Johnson, with a manifesto and political coalition closely associated with his leadership. Truss subsequently won an internal Conservative contest on a substantially different economic programme. Now another prime minister will take office with yet another political programme.
At what point does the parliamentary mandate become too detached from the electoral choice that created it?
There is no precise constitutional answer.
Britain elects MPs rather than prime ministers directly. The Conservative majority therefore remains legally entitled to govern while it retains the confidence of the House of Commons.
That is parliamentary democracy.
But constitutional authority and democratic mandate are not identical.
The further government policy moves from what voters reasonably believed they were choosing in 2019, the stronger the political case for asking voters again.
This is particularly important because Truss’s premiership demonstrates how much policy can change without a general election.
Conservative members selected her partly because she promised substantial tax reductions and a more aggressively supply-side economic programme than Rishi Sunak had advocated during the leadership campaign.
Fewer than 200,000 party members participated in that final leadership vote.
Those members therefore exercised extraordinary influence over the economic direction of a country of roughly 67 million people.
The arrangement is constitutionally permissible because they selected the leader of the parliamentary majority.
It is still democratically unusual.
This is where political parties occupy an ambiguous position. They are private membership organisations, but when one party controls parliament, its internal procedures can determine who exercises public executive power.
Party constitutions therefore have constitutional consequences.
The British constitution is partly being operated through the Conservative Party’s rulebook.
That should make parties themselves objects of democratic scrutiny.
Who gets to choose leaders?
How quickly?
What happens when members and MPs prefer different candidates?
How much policy change can occur without returning to voters?
These are no longer merely internal organisational questions when the party controls government.
The current leadership contest demonstrates the problem particularly clearly because Boris Johnson himself has considered returning, only weeks after his resignation. Rishi Sunak is also attempting to succeed Truss, while Penny Mordaunt has entered the contest.
As of tonight, Johnson has withdrawn, leaving Sunak in a very strong position, but the episode itself is remarkable.
A prime minister can be forced from office by his own party, replaced by another prime minister who collapses within six weeks, and then immediately become a plausible candidate to return.
This is not necessarily unconstitutional.
It does make political accountability look strangely reversible.
If removal from office produces no durable political consequence, what exactly did the removal mean?
The answer may be that parliamentary systems separate fitness to lead at a particular moment from permanent political exclusion.
Johnson lost enough support to continue governing in July.
That does not legally prevent Conservative MPs from deciding three months later that circumstances have changed.
Democratic politics allows rehabilitation.
Voters can re-elect defeated politicians.
Parties can restore former leaders.
There is nothing inherently wrong with that.
But repeated leadership recycling can also indicate that a party’s underlying problem is institutional rather than personal.
Perhaps the Conservatives are searching for a leader capable of solving contradictions that no leader can easily solve.
Their 2019 electoral coalition combined traditional affluent Conservative voters, Brexit-supporting working-class constituencies, economic liberals, social conservatives and voters primarily attracted by Johnson’s promise to “Get Brexit Done”.
Those groups do not possess identical economic interests.
Truss’s programme appealed strongly to one Conservative tradition: lower taxes, smaller government and supply-side reform.
The energy crisis and cost-of-living pressures simultaneously created demands for enormous state intervention.
The government therefore attempted something internally difficult: reducing taxation while spending vast sums protecting households from energy prices.
The fiscal contradiction became visible almost immediately.
Changing prime ministers can change how the contradiction is managed.
It cannot make the contradiction disappear.
This is why leadership crises often distract from coalition crises.
A party blames the leader because replacing one person is easier than renegotiating what the party itself stands for.
But if the electoral coalition contains incompatible expectations, every successor eventually confronts the same trade-offs.
Who pays for public services?
How large should the state be?
How quickly should debt fall?
How much redistribution is acceptable?
How should Brexit’s economic consequences be managed?
What level of immigration does the economy require?
These are substantive conflicts.
Leadership contests cannot resolve them through personality alone.
This is also why Truss’s failure should not be reduced entirely to bad communication. Governments frequently describe unpopular policies as communication failures because that preserves the assumption that the policy itself was sound.
Sometimes communication genuinely matters.
Markets and voters need to understand what governments are trying to accomplish.
But communication cannot rescue arithmetic.
If investors believe borrowing will rise unsustainably, another press conference does not necessarily solve the problem.
Credibility requires either different numbers or a convincing mechanism explaining why the existing numbers will work.
The government’s eventual reversal demonstrates this.
Jeremy Hunt did not simply explain the mini-budget more clearly.
He dismantled most of it.
Policy changed because the government’s assessment of what was politically and economically sustainable changed.
That is error correction again.
The humiliating part is that correction came almost immediately after the government had insisted the original programme represented necessary economic transformation.
Yet embarrassment should not become an argument against reversing mistakes.
One of the worst incentives in politics is the belief that changing course demonstrates weakness.
If evidence changes, policy should change.
The alternative is allowing reputational sunk costs to determine national policy.
Truss’s government initially appeared trapped by exactly that problem. Reversing the mini-budget meant admitting that the programme defining her leadership had failed.
Eventually the economic and political cost of maintaining it exceeded the reputational cost of abandoning it.
By then, however, the reversal created another problem: if the economic programme on which Truss won the leadership contest had been removed, what was the political purpose of her premiership?
She answered that question on Thursday by resigning.
There is something almost mechanically rational about the sequence.
A leader receives authority to implement Programme A.
Programme A becomes impossible to implement.
Programme A is abandoned.
The leader concludes the basis of their authority has disappeared.
The speed is extraordinary.
The logic is not.
This is where Britain’s unwritten constitution again demonstrates both strength and weakness.
There was no formal legal mechanism saying, “A prime minister whose fiscal programme collapses must resign after 44 days.”
Political pressure performed the function.
Cabinet resignations, parliamentary rebellion, market reaction, polling and party rules combined until continued leadership became impossible.
Conventions and incentives substituted for codified removal procedures.
That flexibility allows rapid adaptation.
It also makes the system opaque.
Citizens cannot always identify exactly who removed their prime minister.
Was it Conservative MPs?
Financial markets?
The cabinet?
Opinion polls?
The media?
All contributed.
Diffuse accountability is the cost of flexible political correction.
This brings us back to the argument I made three weeks ago about markets. Liz Truss was not formally removed by bond traders. Markets changed the economic price of implementing her policies. Conservative politicians then changed their political assessment of whether supporting her was sustainable.
Economic information passed through political institutions.
That distinction remains important.
Markets did not become sovereign.
Parliament remained sovereign.
But sovereign institutions cannot operate as though economic feedback does not exist.
This may be one of the most useful ways to think about modern democratic government. Political authority exists inside networks of constraints that are not themselves democratic: interest rates, exchange rates, energy prices, demographics, geography and technological capacity.
Democracy decides how governments respond to those constraints.
It cannot vote them away.
Truss attempted a policy based on one interpretation of Britain’s economic constraints.
Markets and institutions supplied evidence that the interpretation was not credible enough.
The government reversed.
The prime minister resigned.
Now the Conservative Party will choose again.
That sequence can be described as chaos.
It is chaos.
But there is a difference between a system experiencing chaotic politics and a political system losing the capacity to resolve chaos.
Britain still knows how the next prime minister will be chosen.
The opposition can demand a general election.
Conservative MPs can decide whether their government survives.
Voters will eventually judge the entire period.
Nobody needs to invent a new constitutional mechanism this weekend.
That institutional continuity should not excuse the extraordinary political instability. Britain cannot change prime ministers every six weeks without damaging policy continuity, international credibility and public trust.
Stability has value.
Long-term reforms require governments capable of sustaining decisions.
Businesses need predictable policy.
Civil servants need political direction.
Allies need reliable counterparts.
The ability to remove leaders quickly is useful only if it eventually produces leaders capable of governing.
Error correction without convergence becomes oscillation.
That may be the next danger.
A thermostat that constantly switches between extremes is technically responding to information.
It is not regulating temperature well.
Britain needs more than another correction.
It needs a political equilibrium capable of lasting long enough to address inflation, energy costs, public services and weak economic growth.
The next prime minister therefore faces a very different challenge from Truss.
Truss promised transformation.
Her successor will probably promise stability.
That itself demonstrates how rapidly political demand can change.
Six weeks ago, Conservative members chose disruption because they believed Britain’s economic model required radical reform.
Today, after experiencing disruption, the party appears desperate for predictability.
Political preferences respond to outcomes.
That is democracy’s feedback mechanism operating inside a governing party rather than through a national election.
The unanswered question is how long that can remain democratically sufficient.
The Conservatives possess the legal right to choose another prime minister.
Whether they still possess enough political legitimacy to make another major change of government without asking the electorate is more debatable.
Those are different questions.
Constitutions tell us what governments may do.
Democratic legitimacy also asks what they should do.
Britain’s institutions have demonstrated remarkable capacity to remove leaders who lose political authority.
The next test is whether those institutions can produce a government that regains enough authority to stop needing the mechanism.
Liz Truss lasted 44 days.
That is evidence of extraordinary political failure.
The fact that Britain can replace her peacefully, constitutionally and almost immediately is evidence of something else.
A political system should not be judged only by how rarely it makes mistakes.
It should also be judged by how difficult it makes correcting them.
Britain corrected this one remarkably quickly.
The uncomfortable question is why it had to make it in the first place.
