The Approval Pendulum: What Four Years of Wartime Polling Reveal About Ukraine
From 91 percent to 53 and back. The numbers tell a story about Ukrainian society that most commentators miss.
The Kyiv International Institute of Sociology published a number on March 18, 2026: 62 percent of Ukrainians approve of President Volodymyr Zelensky's performance. Fieldwork ran from March 1 through March 8. In February, that figure had been 53 percent. In January, 61 percent. These three data points alone contain a nine-point rebound, an eight-point dip, and a recovery that slightly exceeds the previous level. The headlines wrote themselves. But the headlines, as usual, missed the real story.
The real story is not one month's bounce. It is four years of wartime polling that track how a society processes an existential conflict in real time, what happens when the initial solidarity erodes, and what those numbers actually measure when millions of citizens are displaced, occupied, or unreachable. This article follows the data from February 2022 to March 2026. Where the data speaks clearly, we report what it says. Where it does not, we say so.
The March Number in Context
Start with what the poll actually found. KIIS recorded 62 percent approval and 32 percent disapproval. That leaves roughly 6 percent in the "hard to say" or "neither" category, a group that rarely appears in coverage but consistently hovers in the single digits across wartime Ukrainian polling. The 32 percent disapproval figure deserves as much attention as the 62 percent approval. It means nearly one in three Ukrainians who could be reached by telephone told a pollster, during an active war, that they disapprove of their wartime president.
In January 2026, approval had stood at 61 percent. In February, it dropped to 53. In March, it returned to 62. The temptation is to read this as a V-shaped recovery and move on. But the trajectory since 2022 does not move in neat shapes. It moves in response to events, to household-level consequences of the war, and to a shifting media landscape where the framing of every diplomatic encounter matters as much as the encounter itself.
How You Poll a Country at War
Before interpreting the numbers, the methodology needs scrutiny. KIIS shifted its primary data collection method after February 24, 2022, moving from face-to-face interviewing to computer-assisted telephone interviewing, known as CATI. The shift was a necessity. You cannot send field interviewers into areas under bombardment or active occupation. But CATI introduces its own distortions.
Telephone polling reaches people who have working phone connections, live in areas with cellular network coverage, and are willing to spend time answering survey questions during wartime. Internally displaced persons, who have changed phone numbers, moved between regions, or lost stable housing, are structurally harder to sample. People in frontline communities answer at lower rates. KIIS acknowledges these coverage gaps in its methodology documentation and applies demographic weighting to compensate, but no statistical adjustment fully eliminates the bias that comes from polling a population in motion.
The occupied territories of Crimea and parts of Donetsk, Luhansk, Zaporizhzhia, and Kherson oblasts are excluded from the sampling frame entirely. This means the poll measures the views of Ukrainians in government-controlled territory, not the full population. The distinction matters less for the approval trend over time, since the exclusion has been consistent, but it matters enormously for anyone who treats the headline number as a national census of opinion.
KIIS is not the only polling operation working in Ukraine. The Rating Sociological Group conducts parallel surveys, sometimes producing results that diverge by several percentage points. The International Republican Institute runs its own polls with larger sample sizes of approximately 2,400 respondents. The Razumkov Centre and the Democratic Initiatives Foundation jointly produce surveys that add further data points. Comparing across these sources reveals that the broad trajectory is consistent even when individual readings differ. The direction of travel is more reliable than any single number.
The Rally That Lasted Eighteen Months
In December 2021, before the full-scale invasion, Zelensky's approval stood at approximately 31 percent. He was an embattled president facing domestic opposition, criticism over unfulfilled reform promises, and sagging public confidence. The invasion changed everything.
By March 2022, multiple polling organizations recorded approval figures around 90 percent. Zelensky's decision to remain in Kyiv, his direct communication via social media, and the Ukrainian military's unexpected success in repelling the initial assault on the capital produced a rally-around-the-flag effect of extraordinary magnitude. A leader polling at 31 percent was, within weeks, polling at 90.
The more remarkable feature was not the peak but the duration. John Mueller's foundational 1973 study on rally effects documented that the typical boost from an international crisis lasts six to twelve weeks before erosion sets in. George W. Bush's post-September 11 approval, which spiked to 90 percent, began declining within months and had fallen to around 50 percent by late 2003. Zelensky's rally held above 80 percent for approximately eighteen months. The liberation of Kherson in November 2022 produced a secondary spike that extended the elevated plateau.
What sustained this unusually long rally was the nature of the threat. Bush faced a discrete attack followed by discretionary military campaigns abroad. Zelensky faced ongoing existential bombardment of his country's civilian infrastructure. Every missile strike on a Ukrainian city renewed the conditions that produced the rally in the first place. The threat was not receding into memory; it was arriving nightly.
There is another factor that political scientists studying rallies often undercount: the physical displacement of the population. By mid-2022, millions of Ukrainians had relocated within the country or fled abroad. Those who stayed, and who remained reachable by pollsters, were by definition the population most committed to remaining. This self-selection effect likely inflated the topline approval during the rally period, though the magnitude is impossible to quantify precisely. KIIS weighting adjustments compensate for some of this, but no weighting scheme can fully correct for the fact that the people who left the country are not the same, on average, as the people who stayed.
The Long Erosion: Mid-2023 to Late 2025
The erosion, when it came, was not sudden. It mapped onto a series of specific developments, each of which converted the abstract war effort into a concrete household-level cost.
The 2023 summer counteroffensive was the first major crack. Months of Western weapon deliveries and training had raised public expectations. When Ukrainian forces failed to breach the Russian Surovikin defensive line in any decisive way, the gap between promise and result became visible. Approval did not collapse, but the downward trend began. The polling showed a shift from the mid-80s toward the low 70s over the second half of 2023.
The dismissal of Commander-in-Chief Valerii Zaluzhny in February 2024 deepened the decline. Zaluzhny had become the second most trusted public figure in Ukraine, and his removal, followed by the appointment of Oleksandr Syrskyi, was read by a significant portion of the public as politically motivated. Polling by Rating Group and the Razumkov Centre showed Zaluzhny maintaining high personal trust ratings even after his removal, creating a visible alternative pole in Ukrainian public life. The appointment as ambassador to the United Kingdom partially defused the tension but did not eliminate it.
Then came the mobilization law amendments. The legislative changes passed through the Verkhovna Rada in 2024 tightened conscription rules, lowered the mobilization age, and expanded the categories of citizens eligible for military service. For families across Ukraine, the war shifted from something they supported in the abstract to something that might take their son, husband, or father. The polling numbers tracked the shift, moving from the 60s into the upper 50s by late 2024 and fluctuating in the 55-to-65 range through 2025.
The erosion was not uniform. Western Ukrainian oblasts, further from the front, maintained higher approval levels throughout this period. Eastern and southern communities closest to the fighting showed steeper declines. Urban-rural splits also emerged, with larger cities retaining higher approval than smaller towns where mobilization took a proportionally heavier toll on the working-age male population. These regional and demographic divergences, visible in the cross-tabulations published by Rating Group and the Razumkov Centre, suggest that the national topline number obscures as much as it reveals about the state of Ukrainian opinion during this period.
The February Dip: 61 to 53 in Four Weeks
The eight-point drop between January and February 2026 was the sharpest single-month decline in the post-rally era. The timing coincided with the most turbulent period in Ukraine's relationship with Washington since the start of the full-scale invasion.
The Trump-Zelensky confrontation at the White House on February 28, 2025, and its long aftermath dominated Ukrainian domestic media throughout early 2026. The minerals deal negotiations, which many Ukrainians perceived as a concession extracted under pressure from an unsympathetic American administration, became the focal point of public frustration. The perception was not that Zelensky had surrendered Ukrainian interests, but that he had been forced into a negotiation from a position of extreme vulnerability, and that the terms reflected that vulnerability.
Yet attributing the entire dip to Washington would be an oversimplification. The February polling captured a cumulative weight: the fourth winter of systematic attacks on energy infrastructure, ongoing mobilization that touched ever more households, and the growing awareness that the war had no visible endpoint. Disentangling these factors from the diplomatic controversy is precisely the kind of analytical work the data does not do on its own. The correlation between the dip and the Washington crisis is visible. The causation is not proven.
The March Rebound: 53 to 62
Nine points in a single month. The March recovery demands explanation, and several competing explanations fit the data.
The most widely cited is a rally-around-the-flag response to perceived external pressure. Washington's approach to the minerals deal and the broader Trump administration posture toward Ukraine may have triggered the same defensive solidarity that the Russian invasion triggered in 2022, albeit at a much lower intensity. When the external pressure is perceived as coming from an ally rather than an enemy, the rally effect tends to be weaker and shorter. But it still registers.
A second factor is Zelensky's own communication pivot. Between February and March, Zelensky shifted from a posture of conciliatory negotiation toward more assertive public messaging. The framing of the minerals deal moved, in Ukrainian media, from "concession under duress" to "tough negotiation from a position of limited options." Whether this shift in narrative drove the poll numbers or merely accompanied them is unclear.
A third possibility is simpler. The February dip may have been an overcorrection that partially self-corrected as the immediate shock of the diplomatic controversy faded. The March number of 62 percent is only one point above January's 61 percent. Read this way, February was the anomaly and March was the return to baseline.
The data permits all three interpretations. It does not select among them.
The 32 Percent: Who Disapproves and Why
Coverage of Ukrainian polling almost always leads with the approval figure. The disapproval number is the buried story, and it reveals more about the fault lines in Ukrainian wartime society than the approval number does.
The 32 percent who disapprove of Zelensky in the March 2026 KIIS poll are not pro-Russian. Survey data consistently shows that opposition to Zelensky and support for Ukrainian sovereignty are entirely compatible positions. These are Ukrainians who support the war effort in principle but oppose specific aspects of how it is being conducted, who have lost family members or neighbors to mobilization and hold the political leadership responsible, or who see Zaluzhny's removal as evidence that Zelensky prioritizes political survival over military effectiveness.
The demographic pattern of disapproval tracks two variables more strongly than any others: proximity to the front and age. Communities in eastern and southern Ukraine that live under regular bombardment or within range of Russian artillery show higher disapproval rates than western Ukrainian cities that are more insulated from daily violence. This runs counter to the casual assumption that western Ukraine, with its stronger orientation toward Europe and deeper skepticism of Russian-era political figures, would be the seat of opposition. The front-line communities are not opposing the war; they are opposing the distance between their lived experience and what they perceive as the priorities of the political leadership in Kyiv.
Age is the second axis. Older Ukrainians, who have experienced more political leaders and have more direct exposure to mobilization through their children and grandchildren, express disapproval at higher rates than younger cohorts. The generation that came of age during or after the 2014 Euromaidan revolution tends to show higher tolerance for Zelensky's wartime governance, partly because their political frame of reference begins with the conflict itself.
Zaluzhny's shadow falls across both variables. His sustained personal trust ratings, even after his removal from command and reassignment to London, represent the closest thing to an organized alternative that Ukrainian politics currently offers. He has made no overt political moves, but his popularity provides a focal point for diffuse dissatisfaction. The 32 percent know his name.
What the Numbers Cannot Show
Every number in this article comes with a caveat that is worth making explicit. An estimated 3.7 million Ukrainians remain internally displaced within the country as of early 2026. Approximately 5.3 million Ukrainian refugees are recorded across Europe. Neither group is adequately represented in telephone polls conducted from Kyiv.
Social desirability bias is a documented phenomenon in conflict polling. When a country is fighting for its survival, the social cost of telling a pollster you disapprove of the wartime leader is not zero. This does not mean the disapproval figure is understated by a large margin, but it means the true level of dissatisfaction is likely somewhat higher than 32 percent. How much higher is anyone's guess.
The absence of elections since 2019 makes the polls carry weight they were never designed to bear. Zelensky's presidential term technically expired in May 2024, but martial law, renewed by the Verkhovna Rada at regular intervals, prevents elections from being held. In this context, polling serves as the only real-time signal of democratic sentiment. It is an imperfect substitute. Polls measure stated preferences in a low-stakes setting. Elections measure revealed preferences with consequences. The gap between the two can be large, and there is no way to calibrate it without an actual election.
KIIS, Rating Group, the Razumkov Centre, and the IRI all know these limitations. Their methodological notes say so. The problem is not that the polling institutes are careless. The problem is that the space between "methodological notes" and "headline numbers" is where most of the public discourse lives, and that space does not tolerate nuance.
The data shows what it shows. Four years into an existential war, a majority of reachable Ukrainians approve of their president, a significant minority does not, and the gap between those two groups fluctuates in response to events that are themselves still unfolding. In the absence of the ballot box, this is the closest thing to a democratic signal that Ukraine has. It deserves to be read with the precision it requires and the humility its limitations demand.
- KIIS (Kyiv International Institute of Sociology), monthly public opinion surveys 2022-2026, kiis.com.ua
- Rating Sociological Group, public opinion polling series, ratinggroup.ua
- Razumkov Centre, socio-political surveys, razumkov.org.ua
- International Republican Institute (IRI), Ukraine Public Opinion Survey series
- Democratic Initiatives Foundation (DIF), joint surveys with Razumkov Centre, dif.org.ua
- John E. Mueller, "War, Presidents and Public Opinion" (1973)
- UNHCR, Ukraine refugee and IDP situation data, 2026
- IOM (International Organization for Migration), Ukraine displacement tracking
- Verkhovna Rada, mobilization law amendments 2024
- Ukrainska Pravda, Kyiv Independent, Babel.ua, reporting on domestic political dynamics 2022-2026