On 20 May 2020, Britain was living under some of the most extraordinary restrictions on ordinary life imposed in peacetime. People could not freely meet friends and family. Weddings and funerals were heavily restricted. Millions were isolated from people they loved because the government said limiting social contact was necessary to control COVID-19.
On that same evening, around forty people were reportedly invited to the garden at 10 Downing Street for drinks.
This Wednesday, Prime Minister Boris Johnson admitted in Parliament that he attended the gathering for approximately 25 minutes. He apologised and said he had believed it was a work event. He also said he understood the anger of people who believed that those making the rules had not followed them.
The legal details matter, and an investigation is already examining gatherings across government. We should not decide the exact legal status of every event before that investigation is completed. But I increasingly think the political importance of Partygate extends well beyond whether Boris Johnson personally broke a particular COVID regulation.
The deeper issue is why governments are able to ask citizens to obey costly rules in the first place.
Most laws are not enforced every second by police officers. Governments depend heavily on voluntary compliance. People pay taxes without an auditor watching every transaction. Drivers stop at red lights when no police officer is present. During lockdown, millions of people remained inside homes even though the government obviously could not place an officer outside every front door.
Political systems therefore run partly on coercion, but also on legitimacy. Citizens follow rules because they believe the authority making those rules possesses some right to make them and because they expect other people to be bound by broadly the same system.
Partygate threatens the second part of that bargain.
This is why the defence that Johnson believed the gathering was a work event may ultimately matter less politically than it does legally. Even if that explanation places the event within some technical exception, the public is not evaluating Downing Street as though it were a courtroom examining only statutory wording. People are comparing their own behaviour in May 2020 with the behaviour of the people who designed the restrictions.
A person who cancelled a birthday celebration because gatherings were prohibited does not necessarily become less angry after being told that drinks in the prime minister’s garden may have qualified as work.
A family that could not visit a dying relative is unlikely to evaluate the distinction only through legal definitions.
This is where procedural fairness becomes politically important. People care not only about whether a rule produces a useful outcome but whether burdens are distributed fairly and decision-makers appear subject to the same standards.
During an emergency, this matters even more because governments ask citizens to accept unusually large sacrifices. The state restricts movement, closes businesses and limits social interaction on the argument that individual sacrifice is necessary for collective welfare. That argument depends partly on reciprocity: everyone is supposed to be sacrificing.
If citizens believe the people imposing the sacrifice privately exempt themselves from it, compliance begins to feel less like collective action and more like obedience.
That is politically corrosive even if the behaviour of a few officials had little measurable effect on national infection rates.
The epidemiological impact of one Downing Street gathering is probably negligible compared with the pandemic as a whole. The institutional impact can be much larger because political behaviour communicates information about the rules themselves.
Leaders do not merely issue instructions. Their behaviour signals how seriously those instructions should be taken.
This connects to something I wrote during India’s second wave last April. Political leaders have a different responsibility from ordinary individuals because their behaviour influences millions of other decisions. A politician standing before a large rally does not simply accept personal infection risk; they communicate what behaviour appears socially acceptable.
The same principle applies here.
If officials who understand the rules better than almost anyone else behave as though informal gatherings are acceptable, citizens can reasonably update their beliefs about how necessary or coherent those rules actually are.
That creates a compliance externality. One official’s hypocrisy can reduce other people’s willingness to obey.
This is why legitimacy has practical value rather than merely moral value. A government that is trusted can implement difficult policies at lower enforcement cost. A government that is distrusted needs more monitoring, more penalties and more coercion to produce the same behaviour.
Trust is therefore a form of state capacity.
Economists might describe something similar as social capital. Political institutions work more efficiently when people expect others to cooperate and believe rules are applied fairly. Once those expectations deteriorate, transactions become more expensive. More verification is required. More enforcement is necessary. Every interaction begins from suspicion.
The same principle operates in taxation. If citizens believe everyone contributes fairly, voluntary compliance is easier. If they believe wealthy or politically connected people can avoid obligations unavailable to ordinary taxpayers, evasion becomes easier to rationalise.
The Pandora Papers illustrated this in finance.
Partygate illustrates it in public health.
Different policy areas, same institutional problem: rules become harder to sustain when citizens believe the people with the greatest power over those rules operate under different practical standards.
There is another reason this scandal is particularly damaging for Johnson: his government made the rules.
This distinguishes Partygate from an ordinary politician being caught violating some minor regulation. Downing Street was at the centre of designing and communicating Britain’s pandemic response. Government press conferences repeatedly explained what citizens could and could not do. Ministers defended restrictions as necessary to save lives and protect the National Health Service.
That creates an informational asymmetry in the opposite direction from the one politicians might prefer. Government officials cannot easily claim they were unaware of the spirit of regulations that their own government was asking the entire country to follow.
Johnson’s statement that he believed the garden gathering was a work event therefore raises another question: if the prime minister himself could not clearly determine whether an event involving drinks in the Downing Street garden complied with his government’s rules, how reasonable was it to expect ordinary citizens to interpret those rules correctly?
This does not necessarily prove the regulations were badly designed. Pandemic rules had to distinguish between necessary workplace interaction and prohibited social gathering, and unusual workplaces inevitably created ambiguous cases.
But legal complexity creates unequal burdens. Government officials have advisers and lawyers. Ordinary citizens often have a website and a press conference.
Rules that become so complicated that compliance depends on technical interpretation can undermine legitimacy even when each distinction has a rational explanation.
This is a problem beyond COVID. Complexity gives discretion to people enforcing rules and advantages to people capable of purchasing expertise to navigate them. The more complicated regulation becomes, the greater the gap between formal equality and practical equality.
Partygate therefore combines two politically dangerous perceptions: the rules were confusing, and the people making them may have enjoyed greater flexibility in interpreting them than everyone else.
Johnson’s political survival will not be determined by Labour alone. Britain has a parliamentary system. The more important question is whether enough Conservative MPs conclude that keeping him as leader threatens their own political interests.
This creates an interesting distinction between public accountability and party accountability.
Voters cannot remove Johnson as prime minister tomorrow. The next general election is not happening this week. But Conservative MPs can potentially change their leader.
That means Johnson’s political position depends partly on how MPs expect voters to behave in the future.
If Conservative MPs believe Partygate will fade, loyalty remains rational. If they believe Johnson has become electorally toxic, their incentives change quickly.
Political parties therefore operate partly as mechanisms of anticipatory accountability. MPs do not need to wait until voters actually remove them. The possibility of future electoral punishment can change their behaviour now.
This is one reason opinion polls and by-elections matter between general elections. They provide information about the future cost of continuing with the current leadership.
Again, expectations become political power.
A prime minister can possess the same formal constitutional authority on Monday and Friday while becoming dramatically weaker if enough MPs change their expectations about whether he can win the next election.
That is why political authority cannot be measured entirely through legal powers.
Johnson still occupies Downing Street.
He still commands a parliamentary majority.
But if Conservative MPs stop believing he can protect their seats, his effective power may deteriorate much faster than the constitution suggests.
The scandal also illustrates why apologies in politics are so difficult. A genuine apology requires accepting enough responsibility to appear sincere. But accepting responsibility can create legal or political consequences. Politicians therefore have incentives to apologise emotionally while limiting the factual admission.
“I am sorry this happened” is safer than “I knowingly broke the rules.”
The result can satisfy nobody. Supporters hear unnecessary weakness. Critics hear an apology designed to avoid responsibility.
Johnson’s statement contained this tension. He apologised for the public perception and acknowledged that he should have handled the event differently, while maintaining that he believed it fell within work-related activity.
Whether that explanation is accurate will matter.
But politically, apologies cannot restore trust simply by existing. They work only if the audience believes the speaker understands what they did wrong.
That is why scandal management is fundamentally an information problem. The public is trying to infer character and future behaviour from incomplete evidence.
Was this one ambiguous event?
Was it part of a broader culture inside Downing Street?
Did Johnson know about other gatherings?
Were rules knowingly ignored?
Did officials believe normal restrictions did not really apply to them?
Each new revelation changes the interpretation of the previous one.
This makes scandals cumulative.
One incident can be dismissed as an error.
A second changes the probability that the first was accidental.
A third begins looking like a pattern.
The political damage therefore does not necessarily rise linearly with the number of revelations. At some point, new evidence changes the story from “someone made a mistake” to “this is how the institution behaved.”
That threshold is extremely dangerous for governments because institutional culture is harder to apologise for than individual error.
There is also a broader argument about whether any of this should matter when Britain faces much larger problems. Inflation is rising. The NHS remains under pressure. Russia is threatening Ukraine. The economy is recovering unevenly from the pandemic. Arguably, the prime minister’s attendance at a garden gathering eighteen months ago is trivial compared with those issues.
In material terms, it probably is.
But democratic politics cannot be evaluated only through material consequences.
Rules themselves are public goods.
A government that weakens confidence in equal application of law damages something difficult to measure but essential to every other policy it wants citizens to follow.
This is why relatively small scandals can destroy political careers while much larger policy failures sometimes do not.
Policy failures can be complicated. Voters understand that governments operate under uncertainty and that outcomes depend on many actors.
Hypocrisy is simpler.
You told me to do something.
You did something different.
That asymmetry produces anger because it violates reciprocity rather than merely competence.
A government can sometimes survive being wrong more easily than being perceived as contemptuous.
There is also an important distinction between hypocrisy and corruption. Corruption usually involves using public power for private gain. Partygate is not principally about financial enrichment. Its political damage comes from status.
The allegation is that proximity to power created access to freedoms unavailable to ordinary citizens.
That is a different form of inequality.
We usually discuss inequality economically: income, wealth, housing, education.
Political systems also create inequalities of permission.
Who can obtain an exception?
Who receives discretion?
Who gets the benefit of ambiguity?
Who is punished when rules are unclear?
A liberal state can tolerate enormous differences in wealth more easily than it can tolerate the perception that legal obligations themselves vary according to status.
The principle of equality before the law therefore matters partly because it prevents political hierarchy from becoming legal hierarchy.
Britain’s institutions are strong enough that Partygate itself is being investigated, journalists are reporting aggressively, opposition MPs can question the prime minister publicly and Conservative MPs can potentially remove their own leader.
That should not be overlooked.
A scandal becoming visible can itself be evidence that accountability mechanisms function.
Authoritarian systems have parties too.
They may simply never become public scandals.
This connects to Kazakhstan last week. The absence of visible political controversy does not necessarily mean better government. Sometimes democratic systems look worse precisely because their failures are observable.
A British newspaper publishing allegations about Downing Street and the prime minister being forced to answer questions in Parliament looks chaotic.
The alternative—nobody being permitted to ask—is considerably quieter.
That is not necessarily better governance.
Democracy makes political failure visible.
Its strength depends on what happens after visibility.
Does investigation establish facts independently?
Do ministers correct misleading statements?
Do parties impose consequences when appropriate?
Do voters eventually receive enough information to judge?
If those mechanisms work, scandal can actually strengthen norms by demonstrating that violating them has a cost.
If nothing happens regardless of what evidence emerges, the opposite lesson is learned.
That is why I don’t think we can yet answer whether Johnson should resign solely from the information available this week. There is an ongoing investigation, and factual questions remain unresolved. If evidence eventually shows he knowingly and repeatedly misled Parliament or deliberately violated rules he expected others to follow, the case for resignation would become much stronger.
But even before those facts are established, the political problem is already real.
Johnson’s government asked millions of people to make extraordinary sacrifices during 2020. Many did so willingly because they believed those sacrifices were collectively necessary.
Some people missed funerals.
Some grandparents did not see grandchildren.
Some families remained separated while relatives were ill.
Businesses closed.
People spent birthdays alone.
Those choices cannot be returned to them.
This is why telling the public that a gathering may technically have been work will never completely answer the accusation.
The issue is not only whether Downing Street complied with the narrowest possible interpretation of a regulation.
It is whether the people imposing extraordinary restrictions behaved as though they understood the sacrifice they were demanding from everyone else.
Political authority is not sustained simply because Parliament passed a law.
It depends on citizens believing that obedience is reciprocal: I accept restrictions because everyone else is expected to accept them too, including the people with the power to write them.
Once that belief disappears, enforcement can still produce compliance.
But something more valuable has been lost.
Consent.
And democratic governments should be much more frightened of losing that than of losing an argument over whether a garden party technically counted as work.
