Joe Biden withdrew from the 2024 presidential election today and endorsed Vice-President Kamala Harris, ending an extraordinary three-week crisis inside the Democratic Party and pushing the United States into territory no major modern presidential campaign has experienced this late in an election year. Biden had already won almost all of the roughly 4,000 pledged Democratic delegates through the primary process and was expected to be formally nominated at the Democratic National Convention in August. Instead, after a disastrous debate performance against Donald Trump on 27 June intensified longstanding concerns about his age and capacity to campaign effectively, Democratic lawmakers, donors and party figures increasingly concluded that the risk of continuing with him had become greater than the risk of replacing him. His departure therefore raises a question much larger than which Democrat now runs against Trump: how should political institutions remove or replace leaders when the very advantages of incumbency make acknowledging decline extraordinarily difficult?
Incumbency usually represents one of the strongest assets in democratic politics. Sitting presidents possess universal name recognition, enormous media attention, established donor networks, a record they can campaign on and the symbolic authority of the office itself. Parties therefore have strong incentives to protect incumbents from internal challengers because competitive primaries can expose divisions and weaken the eventual nominee. Biden benefited from precisely this logic. Serious Democratic challengers largely stayed out of the race, he accumulated delegates overwhelmingly and the party moved towards November assuming that the election would be another Biden-Trump contest. The system behaved rationally under normal assumptions: replacing a sitting president who had already defeated Trump once appeared riskier than allowing that president to seek re-election. The problem is that institutions optimised for stability can become slow to respond when the leader they protect deteriorates faster than expected.
This is an organisational problem far beyond American politics. Companies, governments and universities frequently build structures around successful leaders because success creates trust, loyalty and institutional dependence. The longer someone occupies the centre of an organisation, the more information reaching decision-makers becomes filtered through people whose careers, relationships or identities are tied to that leader. Bad news becomes harder to communicate. Subordinates fear appearing disloyal. Allies interpret criticism as helping competitors. The leader themselves possesses the strongest incentives to believe temporary problems can be overcome. Incumbency therefore produces not only political advantage but informational insulation. The very network that helps an incumbent govern and campaign can make the organisation slower to recognise when continuation has become the greater risk.
Biden’s age was not newly discovered during the June debate. He is 81 years old, already the oldest president in American history, and public polling had shown persistent concern about his age long before the campaign crisis. Democratic officials nevertheless largely treated the issue as something voters would overlook once faced with the alternative of Donald Trump. The June debate changed the political equilibrium because it transformed an abstract concern into a common public observation. Millions of voters, donors, journalists and politicians saw the same performance simultaneously. The information was no longer privately distributed across individuals wondering whether others shared their concern. It became common knowledge.
That distinction is crucial. Suppose many Democratic politicians individually believed Biden had become a weak candidate, but each also believed most colleagues remained confident in him. Publicly challenging the president would then be politically expensive because the dissenter could appear disloyal while accomplishing little. Once a nationally watched event convinces everyone that everyone else has seen the same weakness, the coordination problem changes. Lawmakers begin speaking publicly, donors threaten to withhold money and media coverage shifts from whether concern exists to how many officials will act on it. The debate therefore mattered not simply because Biden performed badly but because it synchronised information across the party.
This is similar to how financial crises can develop. Depositors may individually worry about a bank while continuing to leave their money inside because they believe others remain confident. A sufficiently visible shock changes expectations collectively. Withdrawals begin, and every withdrawal becomes further evidence that confidence has changed. Political support can experience the same cascade. One Democratic member of Congress calls for Biden to withdraw. Another now pays a lower reputational cost for saying the same thing. Senior party figures begin questioning the campaign. Donors become more willing to pause contributions. The process becomes self-reinforcing because the probability that Biden leaves rises with every person publicly behaving as though he might.
This does not mean the party suddenly discovered an objective medical fact unavailable previously. The decision was fundamentally electoral and institutional, not a diagnosis. Biden remains president and has not resigned the office. The argument made by many Democrats was that he no longer appeared capable of conducting the campaign required to defeat Trump, not necessarily that he was constitutionally incapable of performing presidential duties. Those are separate standards. A president can be capable of making decisions with advisers, reading briefings and exercising executive authority while being less capable of travelling constantly, giving unscripted speeches, debating effectively and persuading undecided voters through an exhausting national campaign. Electoral fitness and governing fitness overlap without being identical.
This distinction also explains why demands that Biden resign the presidency automatically because he abandoned the campaign are too simplistic. A political party choosing someone else for the next four-year term is making a forward-looking judgement about electoral viability and future capacity. The Twenty-Fifth Amendment and presidential resignation involve the president’s current ability to discharge the powers of office. Treating the two as identical would turn every decision not to seek re-election into evidence that a president must leave immediately. The correct question is whether specific evidence demonstrates inability to govern now, not whether the party believes another candidate should govern from January 2025.
The more difficult democratic issue concerns the primary process. Millions of Democratic voters participated in primaries whose overwhelming outcome was Biden’s nomination. He is now releasing delegates less than a month before the convention, and those delegates will select a replacement voters did not evaluate through a normal competitive primary. That creates an obvious tension between procedural democracy and emergency adaptation. Biden’s critics could reasonably argue that voters were never given a meaningful choice because major alternatives were discouraged from running. His supporters could reasonably respond that primary voters knowingly chose him and that replacing him after elites became nervous risks allowing donors and politicians to override ordinary party members.
Both concerns are legitimate because American political parties are unusual institutions. They are not government agencies, but their nominations determine who can realistically become president. They therefore possess private organisational rights while performing a public democratic function. Party rules give delegates mechanisms for responding when a presumptive nominee withdraws, and Reuters reported that Biden’s delegates are now free to support another candidate. Harris begins as the obvious frontrunner because she is vice-president, has Biden’s endorsement and is already part of the Biden-Harris campaign organisation, but the nomination is not automatically inherited constitutionally. The party must still decide according to its rules.
This raises a broader question about when elite judgement should override a completed mass-selection process. Democracies normally protect voter decisions from retrospective revision by party officials. If leaders could simply replace nominees whenever polling deteriorated, primaries would become advisory rather than meaningful. Yet a rigid rule making replacement impossible after a candidate becomes incapable, dies or withdraws would also be absurd. Political systems therefore require escape mechanisms while making those mechanisms difficult enough that they remain exceptional.
This is exactly the function of institutional flexibility under emergency. Constitutions contain succession procedures. Companies have boards capable of replacing chief executives. Parties have delegate rules. The existence of an escape mechanism does not imply ordinary selection procedures are meaningless. It means institutions recognise that uncertainty continues after selections are made. The difficult part is preventing exceptional flexibility from becoming routine elite discretion.
The best safeguard is transparency. Democratic leaders should explain why circumstances changed, what rules now apply, who possesses decision-making authority and how alternative candidates can compete. The more compressed the timeline becomes, the more important procedural clarity becomes because citizens have fewer opportunities to participate directly. Harris may become the nominee rapidly because the political incentives favour consolidation. That outcome would be more credible if it emerges through publicly understood delegate commitments and visible support rather than through the impression that a small group privately determined succession.
The financial structure makes Harris uniquely advantaged. Reuters reported that the Biden-Harris campaign possessed approximately $91 million in campaign funds, and because Harris was already named on campaign filings, legal experts indicated that much of that money could probably remain available to a Harris campaign, though disputes were possible. Other potential candidates would face more complicated funding and organisational challenges. The party therefore is not choosing among candidates on a blank institutional field. Existing campaign infrastructure heavily favours continuity through the vice-president.
This demonstrates how path dependence shapes political succession. Biden selected Harris as running mate in 2020. That decision placed her inside the national executive, campaign structure and constitutional succession hierarchy. Four years later, those institutional advantages make her the easiest candidate around whom the party can coordinate under extreme time pressure. Decisions made for one political purpose create options years later that leaders could not forecast completely when making them.
Vice-presidential selection therefore matters more than campaigns often admit. Running mates are frequently discussed through electoral geography, demographic appeal or ideological balance. Their constitutional function is succession. A vice-president should be someone the president and electorate can plausibly imagine assuming national leadership unexpectedly. Biden’s withdrawal is not constitutional succession, but the same organisational logic applies: Harris already possesses national recognition and executive experience because she has spent four years one heartbeat from the presidency.
Her path is not politically effortless. Harris struggled during her own 2020 presidential campaign, has experienced periods of weak approval and now must inherit both the achievements and liabilities of the administration in which she serves. A successor drawn from within government cannot credibly campaign entirely as a break from that government. This is the continuity-change dilemma. Harris benefits from Biden’s organisation and record while needing to persuade voters that replacing the candidate changes enough about the election to address concerns that forced Biden out.
This challenge becomes sharper because Donald Trump spent the week consolidating Republican unity at the national convention in Milwaukee. Republicans formally nominated him for a third consecutive presidential campaign and selected JD Vance, the 39-year-old senator from Ohio, as his running mate. Vance used his convention speech to emphasise industrial decline, working-class communities and economic nationalism, presenting a Republican Party increasingly organised around Trump’s populist rather than traditional business-conservative politics. The contrast is striking: Republicans spent the week demonstrating unusually high internal consolidation around one leader while Democrats ended it replacing theirs.
Trump’s position also illustrates the opposite side of incumbency-like personalisation. He is not the sitting president, but he has effectively dominated the Republican Party since 2016. Former rivals including Nikki Haley and Ron DeSantis appeared at the convention supporting him, while Vance himself had once been a severe Trump critic before becoming one of his strongest allies. Political organisations can therefore become so centred around one figure that previously independent factions either adapt or leave. This creates enormous coordination strength while increasing long-term succession risk.
The assassination attempt against Trump on 13 July, one day before this week’s period began, intensified that personalisation further. Trump entered the convention after surviving a shooting that wounded his ear, killed a rally attendee and injured others. The image of him raising his fist after the attack became immediately central to Republican mobilisation. American politics therefore entered this week under extraordinary emotional conditions before Biden’s withdrawal transformed the race again. The combination demonstrates how rapidly events can overwhelm campaign strategies constructed over years.
This is why political forecasting is particularly fragile. Campaigns invest hundreds of millions in models, polling, advertising and turnout operations while outcomes can be transformed by a debate, an assassination attempt or a candidate withdrawing. Politics contains fat-tail events: low-probability shocks with consequences much larger than ordinary daily variation. Organisations need enough resilience that one unexpected event does not make their entire strategy obsolete.
The Democratic Party’s response can therefore be evaluated partly as a test of institutional adaptability. Continuing with Biden simply because the primary process had already selected him would preserve procedural continuity while potentially ignoring new information about electoral risk. Replacing him recognises changing information but creates procedural legitimacy costs. Neither option is costless. Good institutions often choose between competing failures rather than between one obviously correct and one obviously incorrect path.
The strongest argument for Biden’s withdrawal is ultimately one of comparative expected risk. Democratic leaders increasingly believed the probability of losing to Trump under Biden had become unacceptably high, particularly given their view that another Trump presidency poses serious risks to abortion rights, executive constraints and American democratic institutions. Once party elites interpret an election as unusually consequential, their tolerance for candidate weakness declines. A candidate who might be acceptable in an ordinary election becomes replaceable when expected costs of defeat are considered extreme.
That logic itself deserves scrutiny. Parties can justify almost any internal intervention by describing the opponent as uniquely dangerous. If every election becomes existential, leaders acquire broader permission to circumvent ordinary processes in the name of protecting democracy. Emergency politics can therefore erode the procedures it claims to defend. The appropriate safeguard is proportionality: how extraordinary is the intervention, how strong is the evidence of necessity and how quickly does the system return to ordinary democratic procedure afterwards?
Biden voluntarily withdrawing reduces part of this concern because he was not formally removed by party officials. Pressure was intense, but the legal and party decision remained his. This matters institutionally. A leader voluntarily surrendering expected power after concluding that continuation harms their coalition demonstrates something different from an organisation forcibly reversing a voter decision. The distinction does not erase elite pressure. It preserves the principle that the candidate ultimately controlled his own nomination until relinquishing it.
There is also a personal dimension that political analysis should not sentimentalise but should recognise. Powerful leaders almost never possess incentives to conclude that their own departure is organisationally optimal. Biden had spent more than half a century in national politics, sought the presidency repeatedly, defeated an incumbent Trump in 2020 and believed his administration produced major legislative achievements. Asking someone with that history to accept that the coalition now considers them its principal electoral liability requires a profound reversal of identity. Succession problems are difficult partly because institutions are rational abstractions operated by human beings whose status and self-conception depend on remaining inside them.
This is why strong institutions need norms making departure honourable rather than humiliating. If stepping aside is always interpreted as weakness or personal defeat, leaders fight harder to remain. Systems become safer when relinquishing power can itself produce prestige. George Washington’s decision not to seek indefinite rule became foundational partly because retirement was transformed into evidence of republican virtue. Modern democratic politics rarely reaches that symbolic level, but the mechanism remains relevant. Leaders need political narratives allowing them to leave without believing every achievement has been invalidated.
Biden’s statement framed the decision around what he considered the interests of his party and country. Whether one regards that framing as sincere, strategic or both, it provides the required face-saving institution of exit. Democrats can honour his presidency while selecting someone else. Biden can endorse Harris rather than appearing overthrown. The party can present succession as continuity with renewal. Political narratives are not superficial when they reduce the personal cost of necessary institutional transitions.
The United States also demonstrates why term limits solve only one type of succession problem. The Twenty-Second Amendment prevents presidents from being elected more than twice, ensuring no individual can legally accumulate indefinite presidential tenure. Biden was still eligible for a second term. The constitutional limit therefore did not answer whether an eligible incumbent remained the party’s best candidate. Formal term limits protect against permanent rule; they cannot substitute for organisational judgement about capacity.
Ageing democracies may confront this problem more frequently because political careers are becoming extraordinarily long. Biden and Trump are both substantially older than the median American voter, while congressional leadership historically has also included many politicians serving into their seventies and eighties. Longer life expectancy and stronger incumbent advantages allow successful politicians to remain powerful for decades. This can preserve expertise while slowing generational turnover.
The trade-off resembles institutional memory elsewhere. Experienced leaders understand bureaucracies, foreign relationships and legislative bargaining. Younger leaders may bring different information, energy and generational perspectives but lack comparable networks. Political systems need renewal without destroying accumulated expertise. Succession should transfer knowledge rather than requiring every generation to rebuild institutions from zero.
Vance’s selection illustrates the Republican approach to generational transition differently. At 39, he is the first millennial placed on a major-party presidential ticket and could potentially carry Trumpism into a post-Trump Republican Party. His ideological trajectory suggests that succession is being managed through doctrinal inheritance rather than moderation back towards the pre-2016 Republican establishment. Trump is therefore not simply selecting a deputy; he may be helping define what survives politically after him.
Democrats now face the same question under much greater time pressure. Is Harris simply the person best positioned to finish one electoral cycle, or does her candidacy represent the beginning of a generational transition in the Democratic Party? Those questions cannot be answered this week, and treating her nomination as inevitable before delegates decide would move beyond the chronology. The immediate point is that Biden’s withdrawal has reopened a succession process the primary calendar appeared to have closed months ago.
This should also encourage reconsideration of the American primary system itself. Presidential nominations once depended much more heavily on conventions and party elites. Reforms after the chaotic 1968 Democratic convention shifted power towards state primaries and caucuses. The modern system increases mass participation but makes replacing a presumptive nominee late extraordinarily difficult because there is no standing institution with obvious democratic legitimacy to reconsider the choice. Party leaders possess information and organisational responsibility but weak formal authority to override voters.
A more elite-controlled system would adapt faster.
A more voter-controlled system possesses stronger participatory legitimacy.
Again, adaptability and participation trade against one another.
There may be no institutional design that maximises both completely. The current crisis exposes which failure mode the United States chose decades ago: voters dominate nomination under normal conditions, leaving parties with awkward emergency procedures when circumstances change after primaries conclude.
The lesson should not necessarily be returning to smoke-filled rooms. It should be designing clearer contingency rules before the next crisis. What happens if a presumptive nominee becomes incapacitated after winning primaries? Should delegates automatically become free? Should a replacement election occur digitally or through state parties? Who sets deadlines? How should campaign funds transfer? Institutions perform better when emergency procedures are written before anyone knows which faction benefits from them.
This is the veil-of-ignorance principle applied to party rules. Democrats redesigning succession procedures after Biden’s withdrawal would know exactly who benefits. Rules drafted years earlier would possess greater neutrality.
The larger democratic lesson is that leadership selection cannot eliminate uncertainty. Voters make decisions using information available at one moment. Candidates age, become ill, encounter scandals or reveal weaknesses later. Institutions therefore need enough stability that choices matter and enough flexibility that obviously changed circumstances can still be processed.
Too much stability creates rigidity.
Too much flexibility creates elite manipulation.
Biden’s exit sits directly on that boundary.
The next several weeks will determine whether Democrats manage the transition in a way that voters perceive as legitimate. Harris begins with significant institutional advantages and Biden’s endorsement, but her authority ultimately needs to derive from the party’s formal nomination process and then from voters in November. No endorsement transfers presidential legitimacy automatically.
That is the distinction between succession and inheritance.
Political office in democracies cannot be inherited merely because one leader chooses a preferred successor. Endorsements coordinate support. Delegates decide nominations. Citizens decide elections.
Biden can open the door for Harris.
He cannot make her president.
That separation is important because his departure has already demonstrated the limits of individual authority. One month ago, the Democratic nomination appeared almost entirely settled around an incumbent president. Today it is open because political institutions updated around new information and the incumbent ultimately accepted that change.
This is a sign of institutional disorder.
It may also be a sign of institutional resilience.
Resilience does not mean preventing crises. It means absorbing them without breaking the constitutional system.
A presidential candidate has withdrawn.
The president remains in office.
Delegates will select another nominee.
The general election remains scheduled for 5 November.
Power will still be transferred in January according to constitutional rules.
The surrounding politics are chaotic.
The underlying state is not.
That distinction deserves attention in a week when American political rhetoric increasingly describes every unexpected event as evidence of systemic collapse.
Democratic institutions are supposed to allow political organisations to admit mistakes.
The Democratic Party has just made an extraordinary admission: the candidate it spent more than a year preparing to nominate is no longer the candidate it believes should contest the election.
That should invite criticism about why the problem was not confronted earlier.
It should also invite recognition that institutions capable of changing course before failure occurs are stronger than institutions that preserve leadership simply because reversing course would be embarrassing.
The key question is what Democrats learn from the delay.
If the lesson is only that one debate went badly, the analysis is too shallow.
The deeper failure was organisational: an incumbent’s advantages, party loyalty and fear of internal conflict made honest reassessment unusually expensive until public evidence became overwhelming.
That mechanism exists in governments, corporations and political movements everywhere.
Leaders accumulate power partly because institutions coordinate around them.
Eventually the institution must decide whether coordination is serving the organisation or merely preserving the leader.
Joe Biden answered that question today by stepping aside.
The Democratic Party now has to prove it can replace personal continuity with institutional continuity quickly enough to compete in November.
That is not merely an electoral test.
It is what succession looks like when a political system discovers, very late, that stability and inertia were no longer the same thing.
