Democratic Resistance in the Serbian Elections

On 25 October, the Student List won an absolute majority in Serbia’s parliamentary election.  Aleksandar Vučić conceded within minutes of the polls closing. The significance of this election goes beyond the composition of parliament and lies in what its result will allow Serbia to leave behind, since, like Poland and Hungary, Serbia is one of the states whose courts and other key institutions have been captured by the ruling party. The new majority must now make constitutional guarantees and legal duties effective in institutions the Serbian Progressive Party (SNS) still largely staffs.

What makes the Serbian case distinctive is that, two years after the canopy of the Novi Sad railway station collapsed, a leaderless movement born in occupied faculties now holds a parliamentary majority (OSW). With 8,155 polling stations processed, the Student List had 52.73% of the vote (2,257,863 votes) and Vučić’s United Serbia 39.15% (1,676,138 votes) (Vreme), and according to the final projection by Ipsos and CeSID the Student List will have 141 of the 250 seats, with 103 for United Serbia and six for four minority lists. According to CeSID’s projections, it will have 140 of the 250 seats, with 103 for United Serbia and seven for minority lists. Now that the question of whether SNS could be voted out is settled, the harder problem, and the subject of this piece, is what a new majority may lawfully do with institutional arrangements and informal structures built to outlast those who built them.

Serbia’s history shows why this question is of crucial importance, since the country found itself in a similar situation in 2000: on 5 October of that year, after Slobodan Milošević refused to concede the September election, Ljubisav Đokić drove a bulldozer into the state television building in Belgrade, and Milošević conceded the following day (JusticeInfo/AFP). However, the state he had built largely survived him, beginning with its security apparatus, whose “Red Berets” unit supplied the men who assassinated Prime Minister Zoran Đinđić in March 2003 (NBC News). Many of those who went on to build today’s SNS state had served the old one, including Vučić, who served as Milošević’s information minister from 1998 (Global Voices). This means that winning the election did not restore the Rule of Law, and the students draw on that experience.

What Was on the Ballot

The protests leading to the election were the students’ response to institutional inaction, driven largely by a decade of political pressure and the dominant ruling party’s erosion of judicial independence. The founding moment of the movement came on 1 November 2024, when the canopy of the railway station in Novi Sad collapsed, killing sixteen people. A consortium of two Chinese state companies, China Railway International and China Communications Construction Company, had reconstructed the station building as part of the Belgrade–Budapest railway, a Belt and Road project financed largely by Chinese loans (Vreme), and it reopened only four months before the collapse. The companies and the government insisted that the canopy had not been part of the reconstruction; the authorities had refused to disclose the reconstruction contracts, and “Corruption kills” became the slogan of the protests. In the weeks after the collapse, students blockaded more than 60 faculties behind a Rule of Law demand: publish the documents, prosecute those responsible, and let institutions do their job (BTA). Violeta Beširević called it a demand for government under the Rule of Law rather than rule by one man (Beširević).

What happened over the following months explains why the students came to see an election as the way to get those institutions working again. In May 2025, they called for early elections, and in August, TOK, the Public Prosecutor’s Office for Organised Crime, detained a former construction minister and senior railway officials over financial flows in the station’s reconstruction (N1), after which the police director pulled 17 officers out of the task force tracing that money (N1). In January 2026, as parliament cut short prosecutors’ secondments to the office, TOK stated that the withdrawal of investigators and repeated failures to hand over documents had made obtaining the evidence practically impossible (TOK statement; Danas), a pattern of political pressure that the Commission’s July report confirms.

To contest the election they had demanded, the students built a list of their own: Student List – Students Win, a citizens’ list whose candidate criteria, set by 105 student plenums and open assemblies in every higher-education unit, excluded students, officials of opposition parties and former members of the ruling coalition (Jacobin). Several opposition parties withdrew in its favour (BTA). It asked voters for a mandate to make institutions work, and that is the standard by which it will be judged.

The election result will most certainly change Serbia’s future, but it will also matter in retrospect. Namely, the vote decided who controls accountability for the past. In his last days as president, Vučić announced pardons for 49 people whose names have not been published (N1). To this day, only a handful of names are public, and speculation, along with public hints from Vučić (Radio Free Europe), suggests the pardon included officials investigated and charged with corruption related to the fall of the canopy and other high-profile affairs (Guardian). Two days later, Radar published Sky communications in which Vučić reportedly appears as “Oskar” and a pardon is mentioned as a way of correcting an unfavourable court outcome for the drug trafficker Darko Šarić (N1). These accounts need independent investigation, and nothing so far shows that Šarić is among the 49, although the arrangement they describe, in which legal consequences can be negotiated through closeness to power, is the one that the students’ slogan, “Serbia or Mafia”, put before voters. Vučić also chose the timing and form of the vote, dissolving parliament on 9 September (Balkan Insight) and resigning on 27 September to run for prime minister.

How the vote was held is itself a Rule of Law question, as a BIRN investigation published on 20 October shows: it documented public employees pressed to deliver votes, attend SNS rallies and do party work under threat of transfer or dismissal (BIRN). While these testimonies cannot establish how widespread the practice was, they show the mechanism by which dependence on a public job becomes political obedience, as ODIHR’s interim report of 9 October had already noted (ODIHR). Captured public authority thus helps produce the electoral support later invoked to justify its partisan use, making the thirteen-point margin a measure of how strong the demand for change was, achieved under unacceptable conditions.

Vučić’s election-night concession removed the risk of a contested result (Euronews), and ODIHR’s preliminary findings describe high turnout and generally well-run voting in a contest seriously skewed by ruling-party advantages, misuse of state resources, intimidation and media bias (ODIHR). However, none of this changes the diagnosis of a Rule of Law crisis that reaches the foundations of institutional life. In politically sensitive areas, law has become a thin layer over informal autocracy, legitimising the government abroad while failing to organise public action or constrain those who hold power (Vuković, Spaić; BiEPAG). We therefore use the word “restoration” with some care, since in many areas no functioning legal order is held in reserve, and the students are striving to make law effective in government and institutions, in some domains for the first time.

Why it Matters

With the antecedent of 2000 in mind, the student and later popular movement set out to turn protest into electoral authority (Bloomberg), and because it won power under the existing constitutional rules, it will be expected to exercise that power within them as well. Comparisons with Poland and Hungary illuminate inherited appointments and entrenched power, and also highlight what is specific to the students’ project: practical force for existing guarantees and duties, a goal for which a constitution-amending majority is the wrong yardstick.

What the students face goes beyond laws designed to entrench a ruling party, since political instructions, selective enforcement, appointments, and dependence on public employment can displace the rules while leaving the text untouched. As early as 2022, Beširević observed that Vučić sought to control how every institution functioned (Beširević), resulting in institutions doing routine work yet still failing wherever an investigation, appointment, or regulatory decision threatens a protected interest, making governance through general, publicly knowable rules unreliable. Calling every institution an empty shell would overlook the individual judges, prosecutors, and many universities whose sustained resistance has supported the movement.

We believe three questions must be kept apart: whether a norm is legally valid, whether it is effective, and whether the exercise of power meets Rule of Law standards. A valid norm does not, by itself, guide official conduct, and effective coercion does not make conduct lawful, since arbitrary commands can be enforced very efficiently. In Serbia’s case, where informal authority displaces legal duties, and no effective remedies exist against that displacement, restoration requires officials who treat the law as a reason for action even after an unwelcome phone call, and citizens who can obtain compliance when officials fail them.

European actors share responsibility, as the European Parliament’s July assessment finds a persistent gap between legislative alignment and implementation, which Beširević traces to the EU’s long preference for regional stability over Serbia’s democratisation (Beširević). The gap has also been politically useful: legal reform has earned the government credentials abroad as informal control continued at home, so European scrutiny should examine what institutions do when the law requires action those in power do not want.

The Day After October 25

The transfer of power still runs through constitutional procedure, under which the acting president proposes a candidate for prime minister who must obtain the votes of a majority of all MPs. If SNS wins the presidential election in December, its president could return legislation to parliament and delay it, with the obligation to promulgate any law parliament passes again by a majority of all MPs (Constitution, Articles 109, 113, 120 and 127). Within this constitutional framework, what the new majority can do will depend on the programme on which the Student List was elected.

The Student List’s eleven-point programme puts institutional independence and respect for the Constitution and laws at its centre and proposes no constitutional amendment. It translates the claim behind “Serbia or Mafia”, that public offices should perform their legal functions, into measures that work within the existing constitutional order: a special prosecutorial police unit, action against unlawful wealth, review of public procurement contracts, a functioning media regulator and appointments on professional criteria (Student List programme). Many of the obstacles these measures target are practical, and removing them will require legislation, money and enforceable procedures, guided in each case by the question of what stops a lawful investigation, audit or appointment from taking effect.

The hardest problem is that a single institution can contain compromised leadership, unlawful practices and professionals who have kept their independence. Treating the judiciary or the prosecution service as captured wholesale would destroy the capacity the reform depends on. Yet personal independence is of little use when a prosecutor cannot get the police to comply. Restoration must therefore protect those trying to do their jobs, establish responsibility for obstruction, and give lawful decisions operational effect, with allegations of misconduct assessed individually and the validity of appointments treated as a separate question.

This calls for a distinction between restoring the conditions of legality and retaliating by legal means, what Mark Tushnet calls “constitutional hardball” (Tushnet). Restoration changes the conditions of public decision-making, so officials must give legal reasons and answer for departing from them. Retaliation uses legal instruments to decide beforehand who will lose office, be prosecuted, or be favoured, and is, in that sense, a form of abuse: a legal form put to ends the rule does not authorise. Whether a given measure amounts to one or the other depends on its design and operation, whatever its authors say they intend, and the distinction matters most for personnel, where two separate questions arise. The first concerns removal: a judge or prosecutor appointed under SNS cannot be removed for having been appointed during that period, still less for presumed loyalties, and removal requires grounds established in law, evidence of individual conduct, a hearing and review by an independent body. The second concerns invalidity, which arises where the appointing body was itself constituted in breach of the rules conferring the power to appoint and raises the question of whether the office was ever validly conferred (Kristan). This second question is categorical and should be settled through a competent procedure, with proportionality applying to the effects of the decisions taken in that office and the appointee’s status following from the finding on validity. Where someone holds office lawfully, reform should reduce the room for abuse within that office without inventing grounds for dismissal, a standard that also protects officials whose lawful decisions displease the new government.

The practical agenda should begin where legal duties have been defeated, as when TOK’s lawful instructions simply went unexecuted. We regard this as a violation because those under a duty to provide the means of investigation induced the incapacity, which distinguishes it from a deficiency in institutional capacity (Kristan). The proposed prosecutorial police unit is meant to ensure investigators follow a prosecutor’s instruction even when the executive prefers inaction, but whether it achieves this, or becomes a new partisan force, will depend on who commands it, how its members are appointed, what resources it controls, and who oversees it. Its success should be measured by whether a lawful instruction reliably produces investigative action, including against the powerful.

The pardons and the Sky revelations show another way to hollow out the capacity to investigate power, and accountability therefore requires disclosure and lawful scrutiny of the September clemency decisions, along with the conditions for independent investigation. The constitutional grant of the pardon power leaves open whether a given pardon served a public purpose, and a pardon that complies with the clemency power while protecting a network of private loyalties would be a misuse of power, an act within the letter of a competence that defeats the principle justifying it (Kristan). An electoral mandate neither establishes anyone’s criminal responsibility nor authorises the new government to disregard existing legal acts, so the new majority must rebuild the procedures through which responsibility is established and consequences imposed, procedures that will also bind it.

For the same reason, we do not adopt a stance of equidistance: documented evidence of coercion and an unsupported denial cannot carry equal weight, and legal analysis owes its readers explicit evidential standards, a clear line between institutional diagnosis and individual criminal guilt, and consistent application of both. These standards guard against two equally damaging errors, since requiring a captured institution to certify its own capture would make criticism impossible, and treating every allegation as a proven offence would turn accountability into punishment by political designation.

What Defeat Does Not Undo

Losing the government has still left SNS with a strong presence in the state, since it keeps 103 seats, the interim presidency until the new Assembly convenes and a chance in the presidential election, a victory in which would allow its president to appoint five of the fifteen judges of the Constitutional Court. Its appointees also remain throughout the judiciary, the prosecution service, the regulators and the public broadcaster, and pro-government outlets still dominate national media (OSW).

Vučić has promised a combative opposition, and if he wages it from these positions outside parliament, the danger is a government that wins elections and still cannot get institutions to act. The test for the students’ project is therefore whether judges, prosecutors, and nongovernmental actors can now act without political approval, regardless of which government is in office.

The students’ original demand – that institutions be allowed to do their jobs – remains the right measure of what follows. It is a demand to live and work without political permission: for a public employee, to refuse party work without losing their job; for a prosecutor, to investigate protected suspects; and for a university, to teach and criticise without retaliation. A change in government will matter only if it makes norms effective in these places, under duties that would also bind a Student List government. The SNS’s defeat has made that possibility real. Its realisation will show what institutions can do when the politically powerful want them to act otherwise.