American universities have become one of the most visible political arenas of the Gaza war. During the past two weeks, pro-Palestinian encampments have spread across campuses demanding ceasefires, divestment from companies linked to Israel and changes in university policy. This week, the confrontation escalated sharply. At Columbia University, protesters occupied Hamilton Hall before police entered the building and arrested demonstrators. At UCLA, a pro-Palestinian encampment was attacked by counter-protesters in violent clashes before police later moved in to dismantle the camp. Across the country, more than a thousand people had been arrested in campus demonstrations by the middle of this week. The immediate political argument is predictable: one side says universities are suppressing legitimate protest; another says administrators allowed intimidation, disruption and antisemitism to expand until police intervention became unavoidable. The more useful question is institutional rather than ideological: what does a university owe simultaneously to protest, safety, academic continuity and political neutrality when those objectives begin to conflict?
Universities occupy an unusual political space because they are neither ordinary public squares nor ordinary private organisations. Their purpose depends on unusually broad freedom to argue, criticise and organise, including around positions that administrators, governments or donors may find offensive. Student protest therefore should not be treated as an accidental disruption to university life; historically, universities have been important sites of political mobilisation precisely because students possess dense social networks, access to information and enough institutional autonomy to organise. Columbia’s Hamilton Hall, occupied this week, has itself been occupied repeatedly in earlier protests, including during the anti-Vietnam War demonstrations of 1968 and the anti-apartheid movement in the 1980s. The tactic therefore belongs to a long history of American campus politics rather than emerging uniquely from the current Gaza conflict.
Historical precedent, however, does not automatically determine legitimacy. The fact that earlier student movements occupied buildings for causes now widely regarded as just does not mean every future occupation becomes justified. Political movements have a tendency to borrow legitimacy retrospectively from successful predecessors. Protesters compare themselves with civil-rights activists or anti-apartheid movements because history now vindicates many of those causes. Yet participants in the present do not possess the advantage of knowing how history will judge them. Institutions therefore need rules capable of applying before society agrees on who was morally correct. A university cannot reasonably decide whether students are permitted to occupy buildings according to whether administrators believe the political cause will eventually be remembered favourably.
This is why content neutrality is such an important principle in free-expression governance. Rules should generally distinguish conduct rather than political viewpoint. Students advocating Palestinian rights should be permitted to protest under the same conditions as students advocating Israeli security, fossil-fuel divestment, abortion restrictions or any other political objective. Restrictions become more legitimate when they concern behaviour: blocking access to classrooms, physically threatening others, occupying buildings, destroying property or preventing other students from participating in university life. The moment institutions begin deciding which political cause deserves more disruption, administrators become ideological regulators rather than neutral governors of a shared space.
Neutrality does not mean every form of expression must be permitted. Free speech is frequently discussed as though the choice is between complete freedom and censorship. Institutions actually govern speech through time, place and manner rules constantly. A student can criticise the university publicly without having an unrestricted right to shout through a lecture continuously. A demonstration can occur on a quad without possessing an automatic right to prevent other students entering a library. The political opinion remains protected while particular methods can be regulated. The legitimacy of intervention therefore depends heavily on whether rules are applied consistently and whether the university begins with the least coercive means capable of restoring ordinary institutional functions.
This is where the contrast among American campuses becomes useful. Columbia eventually called police after negotiations failed and protesters occupied Hamilton Hall. UCLA moved against its encampment after severe violence and growing concerns over safety. Other institutions, including Northwestern and Brown, pursued negotiated agreements that persuaded protesters to dismantle or modify encampments without equivalent police confrontations. The variation suggests that escalation is not mechanically determined by the existence of protest itself. Administrative choices, communication, protest tactics and institutional history influence whether conflict remains negotiable or turns coercive.
The comparison matters because police intervention changes the character of a dispute. Before police arrive, the conflict is principally between students, administrators and other members of the university community. Once officers in riot equipment enter campus, the institution converts an internal governance problem into a coercive one. Arrest may sometimes be necessary, particularly when violence or serious unlawful conduct occurs. Yet policing carries a higher threshold because coercion can produce injuries, intensify mobilisation and transform a relatively small group of demonstrators into a nationally symbolic confrontation over state power. Good institutional governance therefore requires recognising escalation costs rather than treating police merely as another administrative tool.
UCLA demonstrates the danger of inconsistent coercion especially clearly. Protesters and journalists criticised authorities for responding slowly when masked counter-protesters attacked the encampment, after which a much larger police presence arrived the following night to dismantle the pro-Palestinian camp. California officials themselves criticised the delayed response to the original violence. Whatever one thinks about the legality of the encampment, a state that appears slow to protect protesters from physical assault and quick to remove those protesters afterwards creates a legitimacy problem because enforcement begins to look selective.
This is why state neutrality must be visible, not merely asserted. Police can genuinely believe both decisions were based on operational considerations. Protesters observe outcomes. One group attacks for hours before order is restored; the next night the encampment itself is cleared. Institutional trust depends on whether authorities can explain that discrepancy convincingly. Once citizens believe law enforcement protects political opponents differently, every subsequent intervention becomes interpreted through that prior.
The same problem appears around accusations of antisemitism. Some Jewish students have reported feeling threatened or excluded during protests, and some rhetoric associated with demonstrations has crossed from criticism of Israel into language targeting Jews collectively. Those incidents should be taken seriously. Antisemitism does not become acceptable because it appears inside a protest against Israeli policy. The opposite error is equally dangerous: treating criticism of Zionism, the Israeli government or the Gaza war automatically as antisemitism can collapse political disagreement into ethnic hatred and make legitimate criticism harder to express. Institutions therefore need specificity. What exactly was said? Was a Jewish student targeted because they were Jewish? Was support for Israel assumed merely from religious identity? Was a protest criticising the actions of a government or assigning collective responsibility to a people?
The distinction mirrors the principle from the India-Canada dispute last year: identity is not conduct. Jewish students are not representatives of the Israeli government. Palestinian, Arab or Muslim students are not representatives of Hamas. Political conflict becomes socially dangerous when governments or movements begin assigning individuals responsibility according to group membership rather than behaviour. Universities have a particular obligation to resist that logic because campuses place people from conflicting political communities in unusually close proximity.
The same principle should govern language around terrorism. Supporting Palestinian rights or a ceasefire does not imply support for Hamas. Conversely, invoking Palestinian suffering does not justify glorifying the deliberate killing or kidnapping of civilians on 7 October. A university capable of sustaining serious political argument needs to preserve enough intellectual space that these distinctions remain possible. Once every participant is forced into binary camps, disagreement becomes an identity test rather than analysis.
This is why institutional pluralism matters more than administrators choosing the morally correct position on Gaza. Universities cannot resolve the Israel-Palestine conflict, and pretending they can produces symbolic politics where every statement becomes another battlefield. Their more realistic obligation is maintaining conditions in which students and academics can argue about the conflict without needing institutional authorities to determine the winner. That includes protecting demonstrations, protecting counter-demonstrations, preventing violence and preserving enough access that teaching and research continue.
The divestment demands complicate the neutrality argument because universities are also economic institutions. Endowments invest billions of dollars, universities procure services and institutions may already apply ethical screens to certain investments. Students therefore reasonably argue that investment decisions are not politically neutral simply because administrators describe them as financial management. If a university chooses not to invest in particular sectors for environmental or ethical reasons, protesters can ask why Israel-related investments should be treated differently.
Yet divestment itself requires precision. What does it mean for an investment to be “linked to Israel”? A company headquartered in Israel? A defence contractor selling weapons to the Israeli military? A multinational firm operating commercially in the country? A fund containing shares in hundreds of companies, some of which have Israeli operations? The broader the category becomes, the more difficult it is to distinguish targeted ethical investment from general economic boycott.
This is another classification problem. Political slogans compress complex portfolios into simple demands. Administrators then need operational definitions capable of being implemented consistently. Universities that agree merely to disclose investments may actually improve democratic accountability because protesters can debate real holdings rather than assumed ones. Transparency can therefore function as a compromise between immediate divestment and institutional refusal to engage.
The history of South African apartheid divestment is inevitably being invoked because student movements successfully pressured universities to withdraw investments decades ago. The analogy has political power but should not substitute for analysis. Apartheid South Africa, contemporary Israel and the Israeli occupation are not institutionally identical cases merely because activists use the same tactic. The legitimacy of divestment needs to be evaluated through the present facts and objectives rather than inherited automatically from the moral success of a previous campaign.
This is a broader problem in political protest. Historical analogy is a mobilisation technology. Comparing a present cause with the civil-rights movement, anti-apartheid activism or opposition to the Vietnam War gives participants an established moral vocabulary and strategic playbook. But analogies select similarities while suppressing differences. Their usefulness depends on which features actually transfer.
Students occupying Hamilton Hall can reasonably study how earlier movements used institutional disruption. Administrators can equally study how earlier police responses escalated conflict. History provides information. It does not provide a ready-made verdict.
This week’s police actions also raise the question of civil disobedience. Protest movements often deliberately violate rules because ordinary compliance produces insufficient political pressure. The classic theory of civil disobedience does not necessarily require activists to claim the violated rule never existed. Participants may instead accept legal consequences because willingness to bear punishment demonstrates the strength of their convictions. An activist occupying a prohibited building can therefore simultaneously argue that the action was morally justified and acknowledge that arrest was legally foreseeable.
Political discourse becomes confused when moral legitimacy is treated as legal immunity. A protest can be morally defensible while involving unlawful conduct. Lawful police action can be legally permissible while strategically unwise. Different standards answer different questions.
This distinction is particularly useful at Columbia. Protesters occupying Hamilton Hall crossed a threshold from maintaining an outdoor encampment towards seizing a university building, barricading entrances and preventing normal use. That materially strengthened the university’s argument for intervention. Whether calling the NYPD was the optimal institutional response remains debatable, but the protesters’ escalation changed the decision environment. Political actors cannot increase disruption indefinitely while assuming institutional responses should remain constant.
Administrators face their own escalation incentives. Once protesters violate one rule without consequence, officials fear other rules will become harder to enforce. If building occupation produces concessions, future movements may rationally adopt building occupation more quickly. Institutions therefore worry about precedent. Again, today’s response changes tomorrow’s expected payoff.
This is the same repeated-game logic operating throughout political institutions. Permissiveness can create precedent; coercion can create precedent too. A university that immediately calls police on peaceful protest teaches future administrations that policing is ordinary. A university that never enforces access rules teaches future activists that escalation carries little cost. Institutional design therefore requires selecting a threshold that preserves meaningful protest while making clear that disruption has limits.
The correct threshold should probably be based principally on physical safety, coercion of others and sustained interference with core university functions rather than political embarrassment. Universities are supposed to tolerate embarrassment. Administrators should not use public-relations discomfort as a security category. A protest becoming internationally visible is not itself evidence that police are necessary.
This distinction becomes especially important because university presidents face pressure from several directions simultaneously. Donors may threaten funding. Politicians may summon presidents to congressional hearings. Faculty may defend protest rights. Parents worry about safety. Students demand action in contradictory directions. Trustees worry about reputation and finances. Administrators therefore operate inside a multi-principal governance problem where no single constituency determines institutional objectives.
That can produce overreaction because the most politically powerful constituencies impose more immediate costs than diffuse academic principles. One donor withdrawing a large gift is more visible than hundreds of students gradually losing confidence in institutional neutrality. One congressional hearing creates national headlines. Administrators therefore have incentives to respond to concentrated pressure similarly to governments responding to concentrated economic lobbying.
The defence against that distortion is strong internal governance. Faculty senates, student procedures, boards and published protest rules can make decisions less dependent on whichever external constituency is shouting most loudly this week. Institutional rules are pre-commitments against panic. They tell administrators what threshold applies before they know which political movement will cross it.
This is why universities should write protest rules under a veil of ignorance. Would administrators accept the same rule if the protesters supported a cause they personally hated? Would student activists accept it if the encampment advocated a policy they considered immoral? If not, the rule probably contains more political content than institutional principle.
The campus protests also expose an American constitutional confusion because the First Amendment restricts government, not every private university. Columbia is private and therefore does not have exactly the same constitutional obligations as a public university such as UCLA. Yet private universities often make contractual and cultural commitments to academic freedom and expression that can create standards broader than the constitutional minimum. Legal permission to restrict speech does not necessarily make restriction consistent with the institution’s educational mission.
This is another example of the difference between law and legitimacy. An institution can possess the legal right to do something while exercising that right in a way that damages its credibility. The relevant question becomes what norms the institution has promised its community rather than the minimum conduct courts permit.
Public universities face another constraint because constitutional free-speech protections apply more directly. They still retain authority to regulate conduct, safety and access, but viewpoint discrimination becomes particularly problematic. That makes consistent rules even more important. Police cannot remove one encampment because state officials dislike its politics while tolerating equivalent disruption from another movement they prefer.
The political reaction outside campuses demonstrates why neutrality is difficult. Republicans increasingly portray the protests as evidence that elite universities have become centres of radicalism and antisemitism. Democrats are divided between politicians emphasising student protest rights and those emphasising safety or support for Israel. President Biden said this week that dissent is essential to democracy but disorder is not, while rejecting both antisemitism and Islamophobia. The formulation captures the institutional balance rhetorically; operationalising the boundary between dissent and disorder is where the real difficulty lies.
The presidential election makes every campus decision more politically valuable because parties can use images from protests to support broader narratives about American society. Conservatives can present occupied buildings as evidence of liberal institutional collapse. Progressives can present riot police arresting students as evidence of authoritarian suppression. Each narrative selects the image most useful to its coalition.
This creates a dangerous incentive for performative escalation. Protesters gain visibility when confrontation becomes dramatic. Politicians gain content when confrontation becomes dramatic. News organisations gain attention when confrontation becomes dramatic. Administrators are almost the only actors whose institutional objective should be making the situation less dramatic.
That means restraint can be an active governance strategy rather than passive weakness. Berkeley’s comparatively hands-off approach this week demonstrates the possibility. Administrators there allowed a peaceful encampment while maintaining communication and reserving intervention for clearer safety threats. Reuters reported that the University of California’s own policies, influenced partly by experience with earlier police violence, emphasise avoiding law enforcement where possible.
The lesson is not that every campus should copy Berkeley mechanically. Different protests involve different levels of disruption and risk. It is that non-intervention can be conditional rather than absolute. Administrators can tolerate tents while prohibiting building occupation, maintain dialogue while enforcing access routes and intervene quickly against violence regardless of which side commits it.
This layered approach resembles escalation control in international relations. Begin with communication. Establish red lines. Use limited administrative sanctions when necessary. Preserve coercive force for situations where lower-cost mechanisms fail. Every level remains available, but moving upwards requires evidence that the previous level cannot protect institutional objectives.
The benefit is reducing the probability that policing itself becomes the principal political event.
This matters because arrests can radicalise previously peripheral participants. A student moderately sympathetic to a cause may become much more politically engaged after witnessing a friend arrested. Collective identity strengthens through shared confrontation. Police therefore can suppress physical protest while increasing the movement’s political salience.
This is the repression-mobilisation paradox. Coercion raises participation costs, which can weaken movements. It can simultaneously increase grievances and solidarity, which strengthens motivation. The net effect depends on context.
Authoritarian governments confront the same problem at much higher stakes. Democratic institutions should possess an advantage because they can rely more heavily on legitimacy and negotiated compliance rather than force. Universities, which exist specifically around deliberation, should possess an even stronger comparative advantage in negotiation.
That does not mean endless negotiation. An institution unable ever to enforce a boundary eventually loses credibility too. The objective is graduated enforcement: coercion proportional to the conduct being addressed and preceded where feasible by clear warning, negotiation and opportunities for voluntary compliance.
The protection of counter-protesters needs the same standard. Pro-Israel students have the right to demonstrate against encampments. They do not have the right to physically assault participants. Pro-Palestinian students have the right to criticise Israel. They do not have the right to block Jewish students from spaces because of assumed political identity. Rules need to follow behaviour across ideological lines.
This is harder than issuing statements because behaviour requires investigation. Who threw the first object? Who blocked whom? Was a slogan political advocacy or a direct threat? Universities need evidence before imposing serious sanctions. That takes time, while social media demands immediate moral judgement.
The speed mismatch between institutions and digital politics therefore becomes another source of conflict. A twenty-second video can circulate globally before administrators determine what occurred before or after it. Political communities form conclusions immediately. Any later investigation appears biased if it contradicts the initial narrative.
Institutions need procedures capable of resisting that acceleration. Reliability is more valuable than instantaneous response when facts remain contested. This is the same principle that should govern wartime casualty claims, election fraud allegations and corruption investigations.
Evidence takes longer than outrage.
That delay is not necessarily institutional weakness.
Sometimes it is due process.
Student journalism at Columbia demonstrated the informational value of having independent observers physically inside contested institutions. With much external press access restricted, student reporters documented negotiations, the Hamilton Hall occupation and the police operation from within campus. Their work became an important source of information about events outsiders otherwise saw mainly through official statements and short social-media clips.
This is another reminder that free media functions as institutional infrastructure during conflict. Journalists do not solve disagreements. They expand the evidence available for evaluating them. Administrators and protesters should therefore be particularly cautious about restricting legitimate reporting during confrontations. Institutions possessing coercive authority need scrutiny most when they are using it.
The Gaza protests themselves will eventually decline, whether through semester endings, negotiated agreements, policing or changes in the war. The more enduring question is what universities learn from them. Institutions that respond only by writing stricter rules may reduce future encampments while failing to understand why students mobilised. Institutions that romanticise activism may fail to protect people whose education was disrupted or who experienced intimidation. Learning requires preserving both signals.
Students are saying that university investments and institutional relationships have political consequences they want greater influence over. Jewish students are signalling that some campus environments have become socially threatening. Palestinian, Arab and Muslim students are signalling that they believe their communities’ suffering is minimised institutionally. Faculty are raising concerns about police on campuses. Administrators are signalling that indefinite occupations can become impossible to govern.
All of those are data.
The job of institutional leadership is converting conflicting information into rules people can live under without pretending everyone will agree on the underlying war.
That may be the most important distinction. Universities do not need consensus on Israel and Palestine to maintain institutional order. They need consensus on enough procedural rules that people who profoundly disagree can continue sharing the institution.
Liberal democracy operates through the same principle at national scale. Citizens do not need one interpretation of history, religion or justice. They need procedures strong enough that disagreement does not require coercive domination.
Campuses are therefore miniature constitutional systems. They allocate rights, property, authority and sanctions among people with incompatible preferences. Protest tests those systems because activists deliberately create enough disruption that authorities must respond.
The quality of the institution becomes visible in how it responds.
Too little constraint and other members lose access to the university.
Too much constraint and protest rights become symbolic only.
Selective constraint destroys neutrality.
Inconsistent policing destroys trust.
The solution is not finding a perfect midpoint. It is creating rules based on conduct, enforcing them consistently and escalating coercion only when the protection of safety and core institutional functions genuinely requires it.
The American campus crisis is therefore not primarily a referendum on whether the students are correct about Gaza.
History will continue arguing about that.
The immediate institutional question is whether universities can remain places where people are allowed to be profoundly, angrily and publicly wrong about politics without becoming either ungovernable or coercively silent.
That requires more discipline from protesters than the slogan “free speech” sometimes implies and more restraint from administrators than the phrase “law and order” sometimes implies.
A functioning institution needs both.
The protester needs enough freedom to challenge the university.
The university needs enough authority to keep functioning while being challenged.
The police should enter only when the gap between those two can no longer be managed by less coercive means.
The hardest part is recognising that threshold before politics decides it for you.
