Peru had a president on Wednesday morning. By Wednesday evening, that president had attempted to dissolve Congress, been impeached, arrested and replaced by his vice-president.
The speed of the collapse makes the sequence almost difficult to process. President Pedro Castillo appeared on television on 7 December and announced that he was temporarily dissolving Congress, establishing an emergency government and preparing to rule by decree while a new Congress was elected and the country’s institutions were reorganised. Congress had been preparing to debate another attempt to impeach him later that day. Rather than face the vote, Castillo attempted to remove the institution conducting it.
The plan failed almost immediately. Ministers resigned from his government. Peru’s Constitutional Court rejected the move. Crucially, the armed forces and national police announced that they would continue respecting the constitutional order rather than support Castillo’s dissolution of Congress. Lawmakers proceeded with the impeachment vote and removed him from office. Castillo was subsequently arrested, and Vice-President Dina Boluarte was sworn in as president.
It is tempting to make the political judgement extremely simple: Castillo attempted a coup, the institutions stopped him, democracy won.
The first two propositions are easier to defend than the third.
Castillo’s attempt to dissolve Congress outside the existing constitutional framework was an attack on Peru’s democratic institutions. But the fact that the attempt failed does not mean Peru’s democracy is healthy. If anything, Wednesday’s extraordinary events are the latest manifestation of a political system that has spent years approaching constitutional breakdown from several directions simultaneously.
Peru has experienced repeated clashes between presidents and Congress, impeachments, resignations, corruption scandals and rapidly changing governments. Castillo himself survived two previous attempts to remove him before Wednesday’s vote. Congress is deeply unpopular. Political parties are fragmented and weakly institutionalised. Presidents can win national elections without constructing reliable legislative coalitions capable of governing afterwards.
The political system therefore creates incentives for the executive and legislature to treat one another less as competing democratic institutions and more as existential threats.
That context helps explain Castillo’s actions.
It does not justify them.
This distinction between explanation and justification has become increasingly important to me because political analysis often loses it precisely when institutions enter crisis. Castillo’s supporters can correctly identify serious problems with Congress. His opponents can correctly identify serious problems with Castillo. Neither side acquires unlimited constitutional authority simply because the other behaves badly.
Congress being unpopular does not give the president the right to abolish it.
The president being incompetent or allegedly corrupt does not give Congress unlimited licence to remove him whenever it dislikes his politics.
Democratic institutions need to constrain one another according to rules even when everyone involved has legitimate grievances against everyone else.
Otherwise institutional conflict becomes reciprocal escalation.
Congress tries to remove the president.
The president tries to remove Congress.
Each side cites the other’s previous behaviour as justification.
Eventually the constitution becomes less important than which institution can act first.
Peru came extraordinarily close to that point on Wednesday.
The reason Castillo failed is also instructive. He possessed the presidency but discovered that holding an office and controlling the state are different things.
A president can announce emergency government on television. The announcement becomes politically meaningful only if other institutions obey.
The military has to follow orders.
Police have to enforce them.
Civil servants have to continue administering the state.
Courts have to accept the new arrangement or become incapable of resisting it.
Political allies have to remain loyal.
Castillo lost most of those institutional battles almost immediately.
This is a powerful reminder that political authority is relational rather than personal. A president does not physically possess the powers written beside the title. Those powers exist because networks of institutions recognise the office and follow lawful instructions issued through it.
Once Castillo attempted to operate outside the constitutional framework, many of those institutions stopped treating his instructions as legitimate.
His formal position became surprisingly empty.
This is the difference between personal power and institutional power.
Personalised leaders often attempt to make the state dependent on their authority. Institutionalised states make leaders dependent on the state.
Peru demonstrated the second principle on Wednesday.
The armed forces’ refusal to support Castillo may have been the decisive moment. Latin American political history contains too many examples of militaries deciding political succession directly. Peru itself experienced military government during the twentieth century and Alberto Fujimori’s 1992 autogolpe, when the president dissolved Congress with military backing and concentrated power in the executive.
Castillo’s move inevitably invited comparison with Fujimori.
The crucial difference was that Castillo did not obtain the coercive support required to make his announcement real.
That is institutionally encouraging.
The military possessed enough power to determine whether Castillo’s attempted dissolution could be enforced and chose not to use that power on his behalf.
But we should describe this carefully. A democracy is not healthy because generals happened to choose the constitutionally correct side on one afternoon.
The long-term objective should be a system in which military political preferences are irrelevant.
The armed forces should defend constitutional order without becoming arbiters of which politicians deserve power.
There is a fine distinction between refusing an illegal order and deciding political outcomes independently.
Wednesday appears much closer to the first.
That restraint matters.
It also demonstrates how democratic stability depends partly on professional institutions whose members obey offices and rules rather than individuals.
This principle extends beyond the military. Castillo’s own ministers began resigning after his announcement. The attorney-general rejected the move. Constitutional institutions refused to recognise it. Congress continued meeting.
The state effectively communicated that the president could not unilaterally redefine the constitution simply because the constitution had become inconvenient.
That is exactly what institutional constraint is supposed to accomplish.
The harder question is why Peru keeps producing situations in which constitutional restraint is tested so dramatically.
Castillo’s presidency was troubled almost from the beginning. A former rural schoolteacher and trade-union activist, he entered national politics as an outsider and narrowly defeated Keiko Fujimori in the 2021 presidential election. His support was particularly strong in poorer rural and Andean regions, where many voters felt excluded from Peru’s economic and political establishment.
His election therefore represented something important.
A politician from outside Lima’s traditional elite reached the presidency by mobilising citizens who believed the existing political system did not represent them adequately.
That democratic significance should not disappear simply because Castillo later behaved disastrously.
This is where political analysis needs to separate the legitimacy of the constituency from the quality of the representative.
Castillo can fail without proving that the voters who supported him were wrong to feel excluded.
If anything, his failure may make the problem worse.
People who voted for an outsider because they distrusted the political establishment may now see their representative arrested while the deeply unpopular Congress remains.
They may interpret the sequence not as institutions successfully defending democracy but as the establishment successfully removing the politician they elected.
That perception is already contributing to demonstrations in parts of the country. Castillo’s removal has triggered protests, particularly in regions where he maintained stronger support, with demonstrators demanding variously his release, new elections or the dissolution of Congress.
This does not make Castillo’s constitutional breach legitimate.
It means constitutional legality alone may not be enough to restore political legitimacy.
Again, the two concepts diverge.
Dina Boluarte is constitutionally president because she was Castillo’s vice-president and succeeded him after Congress removed him.
That provides a legal answer to who governs.
It does not automatically provide a political answer to why citizens should trust the system that produced her.
Boluarte therefore inherits a problem much larger than Castillo’s failed self-coup.
She inherits a state in which enormous numbers of citizens distrust both the executive and legislature.
Replacing the person at the top cannot repair that by itself.
Peru’s institutional crisis is partly produced by extreme political fragmentation. Presidents frequently lack stable congressional majorities, while parties themselves can be weak, personalistic and short-lived. This makes executive-legislative bargaining difficult because there are fewer durable organisations capable of enforcing agreements over time.
Strong political parties are frequently unpopular because they appear bureaucratic and self-serving.
Weak political parties create their own problems.
If legislators possess little long-term attachment to coherent organisations, coalition agreements become less reliable. Politicians switch alliances more easily. Presidents struggle to construct legislative programmes. Elections become heavily personalised around individual candidates.
Political outsiders can rise quickly.
They can also arrive in office without the networks or parliamentary organisations required to govern.
Castillo embodied that problem.
Winning an election requires enough voters.
Governing requires enough institutions.
Those are different resources.
An outsider can be excellent at the first and terrible at the second.
This is one reason anti-establishment politics so often disappoints. Voters correctly identify that existing institutions perform badly and choose somebody who has spent little time inside those institutions.
The outsider then discovers that lack of institutional experience can itself become a governing weakness.
Appointments become chaotic.
Coalition management fails.
Technical capacity is limited.
Scandals multiply.
The outsider blames the establishment for obstruction.
The establishment points towards the outsider’s incompetence.
Both may contain truth.
The resulting failure can reinforce the original distrust that produced the outsider.
That is a vicious cycle.
Weak institutions produce outsider politics.
Outsiders govern badly because institutions are weak and they lack governing capacity.
Their failure further weakens trust in institutions.
Another outsider becomes attractive.
Breaking the cycle requires more than finding a better president.
It requires stronger parties, clearer constitutional rules, more professional administration and political incentives that reward cooperation rather than permanent confrontation.
Peru’s impeachment mechanism deserves particular scrutiny in that context. The constitution allows Congress to remove a president for “permanent moral incapacity”, a phrase sufficiently broad that it can become a political weapon during conflicts between legislature and executive.
Accountability mechanisms need flexibility because constitutions cannot predict every form of presidential misconduct.
Too much ambiguity creates instability.
If presidents believe Congress can remove them whenever enough legislators become politically hostile, they acquire incentives to protect themselves aggressively. If Congress believes presidents may attempt to dissolve it when threatened, legislators acquire equivalent incentives.
Mutual distrust becomes self-reinforcing.
This is a constitutional version of the security dilemma.
Each institution takes defensive action.
The other interprets it as offensive.
Both become less secure.
A better institutional settlement would make removal possible for genuinely serious misconduct while raising enough procedural barriers that ordinary political disagreement cannot repeatedly become an existential struggle over presidential survival.
That is difficult to design.
Presidential systems inherently contain two institutions claiming democratic legitimacy: the president and legislature are both elected.
When they conflict, neither can simply claim to represent “the people” exclusively.
Peru demonstrates the danger of pretending otherwise.
Castillo was elected nationally.
Congress was elected too.
Both possessed democratic legitimacy.
Neither owned it.
This is precisely why Castillo’s attempt to dissolve Congress was so serious. His electoral mandate did not give him superior ownership over popular sovereignty.
A president saying “the people elected me” is true.
Legislators can say exactly the same thing.
Democracy is plural.
Authority is distributed intentionally.
The same principle should constrain Congress. Legislators cannot treat presidential elections as temporary suggestions that can be overturned whenever political conflict becomes inconvenient.
Checks and balances become illegitimate if every check is used maximally all the time.
Institutions need forbearance: the willingness not to exercise every legal power to its fullest possible extent simply because the text technically permits it.
This is one of the most difficult democratic norms because politicians receive immediate advantages from using powers available to them while the long-term cost of institutional escalation is distributed across everyone.
The United States has experienced versions of this problem around executive orders, judicial appointments and congressional procedure.
Peru’s version is much more destabilising because presidential removal and congressional dissolution sit close to the centre of political conflict.
Wednesday was what happens when forbearance disappears completely.
Castillo faced an impeachment vote.
He attempted to eliminate the institution voting.
That is not democratic self-defence.
It is an attempt to make one’s own political survival superior to constitutional procedure.
The fact that Castillo originally entered office through a democratic election makes the lesson more important rather than less.
Democracy can produce leaders who later attack democratic constraints.
Election is not democratic certification for everything that follows.
This is one of the weaknesses in treating democracy purely as majority rule. A leader can win 51% and then use that victory to weaken courts, legislatures, media or electoral institutions.
The original mandate was democratic.
The subsequent concentration of power may not be.
Liberal democracy therefore requires both popular authorisation and institutional limitation.
Elections answer who governs.
Constitutions answer what governing does not entitle them to do.
Castillo crossed that boundary.
But if Peru wants Wednesday to become more than a successfully defeated self-coup, its political class needs to confront why citizens were willing to elect someone like Castillo in the first place and why many remain willing to defend him afterwards.
It is not enough to say those voters were deceived.
Many had genuine reasons to feel excluded.
Peru has achieved significant economic growth over previous decades and is a major producer of minerals, yet economic opportunity and state capacity remain distributed extremely unevenly. Rural and indigenous communities often experience the state differently from wealthier citizens in Lima.
A macroeconomic success story can therefore coexist with deep political alienation.
This is the same distinction that appeared in Chile.
Aggregate performance does not guarantee perceived fairness or representation.
If political institutions repeatedly appear incapable of incorporating marginalised regions except through outsider candidates, the system will keep generating anti-establishment politics.
The solution cannot be suspending constitutional rules whenever the outsider misbehaves.
It has to be making constitutional politics itself more representative and effective.
Boluarte now faces that task under terrible conditions. She must establish authority without appearing simply to be Congress’s preferred replacement for Castillo. She must respond to protests without converting legitimate dissent into another security problem. She must determine whether early elections are necessary to restore legitimacy.
As of tonight, none of those questions is resolved.
That uncertainty should remain in the post.
What is already clear is that Peru’s institutions survived one extremely serious attack this week.
Survival is not the same thing as health.
A patient surviving a heart attack has demonstrated resilience.
Nobody concludes the cardiovascular system is therefore functioning perfectly.
Peru’s democracy resisted Castillo because Congress, courts, ministers, police and the military refused to accept his attempt to dissolve constitutional government.
That deserves recognition.
The same political system has nevertheless cycled through repeated presidential crises and produced profound public distrust.
That requires reform.
Both conclusions can be true simultaneously.
Perhaps that is the lesson of Wednesday.
Democratic institutions do not prove their value by producing perfect leaders.
They prove it partly by remaining capable of stopping leaders when those leaders decide their own electoral legitimacy is more important than the rules that created it.
Pedro Castillo was elected president.
The constitution made him powerful.
When he attempted to place himself above that constitution, the same institutions that had made him president stopped obeying him.
That is not evidence Peru’s democracy is finished.
It may be evidence that there is still enough democracy left to repair.
