A year and a half without moving
Support has been remarkably stable. Fifty-nine percent of citizens support the student protests and 34% oppose them - and that balance has barely moved for a year and a half, from 61% in December 2024, through a 64% peak in February 2025, to 58% in September 2025 and 59% now, even as the protests are held far less often than at their height.
In a polarised country where most questions swing with events, this one has not. The near-total absence of "don't know" answers says why: on the protests, almost everyone has taken a side, and the larger side has proved remarkably fixed.
In a polarised country where most questions swing with events, this one has not. The near-total absence of "don't know" answers says why: on the protests, almost everyone has taken a side, and the larger side has proved remarkably fixed.
Bigger than any single party
The movement is also bigger than its electoral vehicle. Support runs to 99% among student-list voters, but also to 70% of undecided citizens and 56% of those who say they will not vote at all - and above 80% among people who inform themselves through social networks and portals. It is majority in every age group except the over-65s, and at every education level except those with only primary schooling.
Support collapses only among ruling-party voters (5%) and audiences of the pro-government commercial channels (9%). In other words, the protests are not a partisan project that happens to be popular; they are a broad civic current that a party list has since given electoral form.
That distinction matters: far more people support the protests than would currently vote for the student list, so its support extends well beyond its electoral base, leaving room to convert more of that support into votes.
Support collapses only among ruling-party voters (5%) and audiences of the pro-government commercial channels (9%). In other words, the protests are not a partisan project that happens to be popular; they are a broad civic current that a party list has since given electoral form.
That distinction matters: far more people support the protests than would currently vote for the student list, so its support extends well beyond its electoral base, leaving room to convert more of that support into votes.
Doubting the official account
The survey also measured how citizens view the state's response to the protests, and here their assessments differ sharply from the official account. Sixty-two percent of citizens are worried about the use of police force against protesters (42% very worried), against 34% who are not; concern peaks among the young, around four in five of whom are worried, and ruling-party supporters are the only group in which a majority are unconcerned.
On the most contested episode - the disturbance in the crowd at the 15 March 2025 Belgrade rally - a majority (52%) read it as a deliberate tactic to intimidate demonstrators, against just 7% who accept it as a justified security measure. On both counts, the public's instinct is to distrust the state's version of events.
On the most contested episode - the disturbance in the crowd at the 15 March 2025 Belgrade rally - a majority (52%) read it as a deliberate tactic to intimidate demonstrators, against just 7% who accept it as a justified security measure. On both counts, the public's instinct is to distrust the state's version of events.
Responsibility runs upward
Citizens also hold a settled verdict on the tragedy that started the movement, and it has scarcely moved since the day it happened. Asked what caused the Novi Sad collapse, 43% of citizens place themselves at the far end of the scale marked "lack of government oversight and corruption", against just 14% who attribute it to accident or technical failure; only ruling-party supporters lean toward the latter.
That reading is not new. Back in December 2024, citizens already assigned political responsibility to government ministers (48%) and the President (39%) ahead of city (35%) and national (33%) government, and criminal responsibility to the contractors (67%) and the supervisory bodies (58%).
Eighteen months on, the judgement is unchanged: responsibility is seen to run upward, to those who were meant to oversee the works. That conviction - that sixteen people died because the state did not do its job - is what holds the movement together, more than any single demand.
That reading is not new. Back in December 2024, citizens already assigned political responsibility to government ministers (48%) and the President (39%) ahead of city (35%) and national (33%) government, and criminal responsibility to the contractors (67%) and the supervisory bodies (58%).
Eighteen months on, the judgement is unchanged: responsibility is seen to run upward, to those who were meant to oversee the works. That conviction - that sixteen people died because the state did not do its job - is what holds the movement together, more than any single demand.
Aligned with the movement
Put together, these findings describe a public that is not merely sympathetic to the protests but aligned with them on the substance - on the demand for accountability, on the fear of force, and on the reading of the Novi Sad tragedy - and a state whose account of events a majority does not believe.
Its durability and broad support are what let it move from the street into electoral politics, where the student list now leads the field. The gap between what citizens believe happened and what the authorities maintain has only widened.
Its durability and broad support are what let it move from the street into electoral politics, where the student list now leads the field. The gap between what citizens believe happened and what the authorities maintain has only widened.
Source: CRTA public opinion survey, conducted with the Democracy Action Lab (Stanford CDDRL); fieldwork by Pulse Research. Face-to-face (CAPI), n=2324, representative of the adult population of Serbia (excl. Kosovo), 10-24 June 2026, margin of error ±2.08%. The support-over-time comparison draws on earlier CRTA surveys; some waves used telephone (CATI) rather than face-to-face interviewing. The responsibility figures are from CRTA's December 2024 survey (face-to-face, n=1030).
