About 60% of Paper Votes for United Russia Might be Fabricated

According to the preliminary results from Russia’s Central Election Commission, the ruling United Russia party is winning 58% of the vote on party lists and is also winning in most single-member districts. It is about to win 355 of the 450 seats in the State Duma, which will be the largest parliamentary faction in the party’s history: in the 2021 elections, it held 324 seats. According to the Election Commission, the Communist Party of the Russian Federation (CPRF) is running second (with 14% of the vote), followed by the Liberal Democratic Party of Russia (LDPR) with their 9%, followed by New People (8%) and A Just Russia (5%). The final results will be announced on Friday, September 25.

The election was accompanied by numerous violations. Observers were denied access to polling stations, beaten, detained, and prohibited from recording videos. There were reports of ballot stuffing, voter coercion, ballot tampering, and other violations.

The authorities had created the conditions for an uncontested election from the start. A month and a half before the vote, Russia’s Supreme Court barred Yabloko, the only anti-war party, from running in the election. The authorities launched repressions against the party members. Twelve of them have been charged with criminal offenses, and eight are in custody. The party’s deputy chairman, Lev Shlosberg, was sentenced to 11 years in prison. Two Yabloko members, Konstantin Smirnov and Elena Perepelitsa, remain in pretrial detention.

Voting took place on September 18–20 using the new GAS Vybory 2.0 (‘Elections 2.0’) system, which was used for federal elections for the first time. IStories had published a detailed investigation into the system and discovered that it allows for results tampering, voter rolls fraud and uploading unverified results from remote electronic voting systems.

In this article, we report on the anomalies we were able to identify in the voting results using statistical methods.

United Russia got no more than 35% of the real vote share

According to IStories calculations, instead of the claimed 58% of the vote, United Russia may actually have no more than 35%. Novaya Gazeta Europe reached a similar conclusion. Boris Ovchinnikov, an election analyst we consulted, also agrees with this assessment.

We arrived at the 35% estimate based on the voting results analyst Petr Zhizhin compiled from the Central Election Commission’s website data by the evening of September 21. At that time, 94% of the vote count had been entered into the database. The commission had not yet published the complete data for Bashkortostan with its 3 million voters, for the Leningrad and Smolensk Oblasts and the occupied territories of Ukraine (apart from Crimea).

We excluded Moscow from the analysis: approximately 80% of the votes there were cast via electronic voting, and the results were published without a breakdown by polling station. We also did not include the results of electronic voting in 32 regions. Based on this calculation, it appears that approximately 18.4 million of the 30 million paper ballots cast for United Russia — according to the data as of September 21 — may have been falsified. This means that more than 60% of the “paper” votes for the ruling party may be fabricated.

The method we used to analyze the data was first applied to Russian elections by Sergei Shpilkin, a physicist and election analyst. The method is based on the premise that the ratio of votes cast for different candidates and parties should not change depending on voter turnout. When commission members stuff the ballot boxes, they thereby increase voter turnout numbers. And since they stuff the boxes in favor of pro-government candidates, United Russia’s result would be rising alongside voter turnout.

The polling stations where the voting was fair should be forming a dense “cloud” on the chart, meanwhile the polling stations where ballot stuffing occurred are located at the top right and look like a comet’s tail. In the 2026 State Duma elections, the cluster of fraudless polling stations is located around the 35% vote share for United Russia. Most polling stations are in the “comet’s tail” zone, which indicates widespread ballot stuffing. You can read more about Sergey Shpilkin’s method in the explainer.

We defined the real turnout range — at which United Russia’s result does not yet correlate with the turnout numbers — at 20–29 percent. Based on the vote ratio for different candidates at this turnout level, we calculated the potential number of fraudulent ballots.

In fair elections, the graph showing the relationship between the number of votes and voter turnout resembles a bell curve. For example, this is what the turnout distribution looked like in the 2000 presidential election. Experts call those elections the fairest in the history of modern Russia. A different picture emerges in the 2016 State Duma elections and further on, where “peaks” appear on the right side of the graph. Those peaks are what votes rigged for the ruling party look like, as its share of the vote increases along with voter turnout.

Turnout fabrications

According to the Central Election Commission, turnout stood at 59.32 percent. This is a record for State Duma elections since 2011 (60.21 percent). “The Kremlin went too far with its another wartime election,” Mikhail Komin, a political analyst, notes. “In 2024, Putin achieved a result more typical of Latin American and Central Asian dictatorships (87% with a 77% turnout) than of his own previous term's results. Now, in the 2026 State Duma elections, the results are no less historic.” Analysts Roman Udot and Boris Ovchinnikov believe that the actual turnout was around 40%.

Election commissions began falsifying turnout figures from day one, analyst Ivan Shukshin pointed out. For example, by 8:00 p.m. on September 18, an unusually high number of commissions reported a turnout of 30%. The same thing happened on the second day of voting, when the most common reported turnout figure was precisely 40% — and 85% on the third day. In normal circumstances, people across decide to go to the polls randomly and independently of each other, so the turnout distribution among law-abiding commissions should follow a simple bell curve,” Shukshin explains. “Peaks at round numbers (40%, 60%, 65%, etc) appear when the figures are fabricated.”

Virtually all voters at unusually high turnout polling stations — those where the turnout neared 100% — voted for United Russia members. Mikhail Matveev, a State Duma candidate from CPRF who had been temporarily barred from participating in the election, drew attention to polling station 0618 in the Borsky District in Samara Oblast. The commission reported a turnout of 98.53%; according to the same data, Matveev received 14 votes, while Alexandr Zhivaykin, his opponent from United Russia, received 704 votes. The Yabloko initiative group also reported similar results favoring candidates from the ruling party in New Moscow .

During previous elections, election commissions artificially increased turnout numbers in order to be able to falsely add votes for the ruling party's candidates. According to sources cited by Meduza, ahead of this election, officials from the presidential administration demanded that regional administrations ensure turnout numbers of more than 50 percent.

On the first day of the election, neighboring polling stations under the “Sovetskaya” Territorial Election Commission in Dagestan submitted reports showing identical voter turnout figures—another sign of vote rigging also noted by Shushkin. Similar unnatural turnout figures were also detected at election commissions in the Ivanovo and Leningrad Oblasts, in Kuzbass, in the Primorsky and Krasnodar krais and in St. Petersburg. In Kabardino-Balkaria, virtually identical turnout figures were recorded by every single polling station in the republic.

Anomalous turnout and United Russia’s share in the final hours of the voting

According to the official data, at some polling stations, election commissions may have added votes for United Russia right before the polls closed. For example, at two neighboring polling stations 322 and 338 in district 48 in the Kamchatka Krai, two hours before polls closed (from 6 to 8 p.m. on September 20), turnout suddenly jumped from their respective 7% and 17% to 100%, as noted by IStories. Irina Yarovaya, co-author of the law on internet surveillance, was running for the State Duma in this district. During the last two hours of voting, she identically received a record-breaking 97% of the vote at both polling stations and was re-elected to the State Duma. Yarovaya has held a seat in the State Duma for 19 years, and she now has five more years to go.

Identical results at neighboring polling stations

Another sign of fraud is when neighboring polling stations report virtually identical results. This may indicate that they are following a directive handed down from the authorities. In a fair election, votes at neighboring polling stations cannot be distributed identically down to hundredths of a percent. The probability of this is extremely low, but we have identified numerous regions with this anomaly.

For example, in Dagestan, the results at 11 out of 53 territorial commissions are practically identical. At 45 polling stations in Derbent, United Russia “won” with their results ranging from 93.84% to 95.03%. This is statistically impossible in a genuine voting process.

Vote manipulation

Observers recorded commission members transferring votes for other candidates to United Russia by literally putting ballots in a different pile. “The chairperson takes a ballot with a vote for candidate No. 4, New People's Zharkova, but reads out the surname of the United Russia candidate, Romanenko. The same thing happens with the next ballot,” Daria Bagina, an observer, reports. A Yabloko observer at polling station 440 in Moscow noticed that ballots used for voting for the Yabloko candidate Viktor Kogan-Yasny, as well as for other candidates, were being placed in the United Russia pile.

But most of the vote tampering takes place without observers present. “At Moscow polling stations, they did indeed shift ballots in favor of United Russia candidates during the vote count, but this is apparently only the first level of fraud,” Boris Ovchinnikov notes. “I am certain that virtually all result reports showing a turnout of more than 75–85% (but less than 100%) were filled out with no connection to the actual count of completed ballots whatsoever. This explains the recurring vote shares across different polling stations as well as other anomalies. Where there were no actual observers present, they could have entered any numbers into the protocol and then further alter the results when entering them into the digital elections system.”

Vote transfer is evident in the data. It occurs at polling stations with a turnout of more than 82%. Until then, United Russia’s results are rigged through ballot stuffing: its share increases along with turnout, while the results of the other parties remain unchanged in absolute numbers. But starting at a turnout of 83%, votes for all parties except “United Russia” begin to disappear. Their results drop by more than half. At the same time, the ruling party’s results rise in a corresponding manner. This may indicate that at such polling stations, ballots for candidates from other parties are being counted as votes for United Russia.

On the graph, this pattern of votes resembles a pair of tongs. Some of the 18.4 million falsified votes may have been obtained through such vote transfers.

At the same time, when the turnout approaches 100%, United Russia’s share of the vote declines, while that of all other parties, on the contrary, rises. Boris Ovchinnikov explains this by the fact that there is usually no fraud at such polling stations: “A turnout of exactly 100% automatically occurs at ‘special polling stations’: in hospitals, detention centers and on ships. Anywhere where there are no permanent residents, everyone votes on a temporary voter list. Those who do not vote simply do not make it onto the list.”

In addition to United Russia, votes for A Just Russia party may also have been rigged in these elections. Election analyst Stanislav Andreychuk noted that on September 21, A Just Russia suddenly gained 60,000 votes, which allowed the party to reach the 5 percent threshold needed to enter the State Duma. “Just one phone call, a decision was made, and the numbers were rewritten right away,” Andreychuk suggests.

Results abroad: emigrants against United Russia

The New People party received 43% of the vote and came first at polling stations abroad, according to independent exit polls conducted by the Vote Abroad project. Surveyors interviewed those voting in 22 countries and collected 15,700 responses.

United Russia only got 5.5% of the vote. The New People party also leads according to the exit polls conducted by Vesna and Progress movements.

At the same time, according to preliminary data from the Central Election Commission, United Russia won at polling stations abroad, with 45.49% of the vote. New People came in second, getting 22.42% of the vote. A total of 212,000 Russians cast their ballots abroad.

United Russia lost the elections in at least 40 out of 152 countries. In 39 countries, New People came first; they got their highest vote share in Slovenia, Serbia, and the Netherlands. United Russia also lost in the anti-war emigration hubs, such as Armenia, Kazakhstan, and Montenegro. Previously, Russians living in these countries would mostly vote for pro-government candidates and parties. The ruling party also lost to New People in those countries in the Middle East and Asia where thousands of Russians have also moved since 2022: the UAE, Israel, Vietnam, Thailand, and China.

United Russia received its highest share of the vote abroad (79%) in the Central African Republic. Its vote share was also high in Moldova (77%), Russian-occupied South Ossetia (73%), Lebanon (70%), and Tajikistan (70%). United Russia also won in a number of Western countries: Greece (56%), Estonia (50%), Italy (40%), Switzerland (36%), and the United States (34%).

Unverifiable remote voting results and the occupied territories of Ukraine

Anomalies in the remote electronic voting system at the elections were detected as early as on the first day of voting. The Belgorod Territorial Election Commission closed remote voting in District No. 80 just three hours after it began, Ivan Shukshin noted. Approximately one-third of the voters who had registered for electronic voting cast their ballots in those three hours. This was later attributed to a “technical error.”

In addition, Petr Losev detected three simultaneous spikes in the electronic ballots generation within the remote voting system — one on the afternoon of September 18 and two on the morning of September 19. For two to three minutes, the number of ballots allegedly issued to online voters increased by 2.5 to 4 times in several regions at once.

Electronic voting has long been difficult to subject to public oversight. But in the 2026 election, the blockchain data on electronic voting in Moscow was completely removed from public access. “For the first time since electronic voting began in Moscow, the ability to view or download records was disabled. In previous years, we would be able to create graphs showing how a candidate’s vote share depended on the time,” says Petr Zhizhfin, a remote voting expert. As a result, this voting method opens up numerous opportunities for manipulation. United Russia received 3.5 million electronic votes in Moscow and elsewhere, but it's not clear how many of those are genuine.

Voting results in the occupied territories of Ukraine also remain completely unverifiable. There is no independent observation there, and the results are only published as overall totals not broken down by individual polling stations or election commissions.