Re:Russia Repression Index: The level of state repression in 2025 shows an even sharper increase than was observed at the beginning of the war

How to measure repression?

Repression has long been a constant backdrop to Russian political news and one of the main pillars of the political regime. If one imagines that political repression were to cease tomorrow, the political situation in Russia would begin to change rapidly. Yet we still lack a reliable tool for assessing the scale and overall dynamics of repression: is it increasing or decreasing, and how is its profile shifting, if at all? These indicators are crucial for understanding the condition and trajectory of an authoritarian regime.

Measuring repression in autocracies is a systemic problem. There is no reliable statistical data that would allow for comparison across authoritarian regimes, which makes comparative study of dictatorships difficult. Both the design and practical implementation of repression vary widely, as does the degree to which repression is public or verifiable. Autocracies typically balance between two contradictory aims: to conceal repression and to display it. Displaying repression has an instructive purpose: citizens must know what will be punished. But showing too wide a scope of repression signals that the regime does not enjoy the broad support it claims, and instead faces resistance which it punishes among 'ordinary people like us'. Therefore, autocracies strive for a deliberate ambiguity: repression should be clearly visible to one part of society while remaining seemingly invisible to another.

At the same time, one defining feature of today’s Russia as an 'atypical dictatorship' (→ Traisman: Somewhere between Mauritania and Portugal) is the extremely high level of documentation and public awareness of state repression. This is done by organisations such as Mediazona and OVD-Info, which have a large media audience, the Memorial Society and a number of other strong projects and organisations (First Department, Sova Centre, etc.). Information about repression is abundant and forms a significant part of the independent and opposition media agenda. As a result, among the politically aware and educated part of society, repression in Russia often appears even more extensive than it actually is. But the key question remains: what is 'actually'?

Several projects and institutions aggregate data on political repression in Russia. Memorial maintains a list of political prisoners. However, this is primarily a legal procedure rather than an informational monitoring tool: human rights experts examine individual cases against specific criteria to establish whether political persecution is present, and then include individuals in the register. Given today’s mass conveyor of politically motivated prosecutions, only a limited number of cases reach the list. At present, the list includes 689 people (excluding those prosecuted for religious reasons), a figure clearly insufficient to reflect the full scale of repression.

OVD-Info also maintains statistics on politically motivated prosecutions. In essence it tracks much the same phenomenon, but the procedure is significantly simpler: the project records cases where signs of a political motive are present, based on open sources, including other human-rights projects. The starting point is usually the opening of a criminal case (for more details, see the description of the OVD-Info methodology). Even if it does not go to trial, the fact of state pressure is recorded.

The OVD-Info database, maintained since 2012, is extensive and highly informative. As of early November 2025, it contains more than 6,100 entries. According to its data, around 4,100 people in Russia are currently subject to politically motivated prosecution (at some stage between the opening of a case and the end of a sentence), nearly half of whom (1,830 people) are in custody. The database shows the dynamics of repression. As shown in Chart 1, the number of such cases began to grow noticeably in the second half of the 2010s, increasing from around 120 a year in 2013–2014 to about 400 a year in 2019–2020. In the first year of the war, the number of cases recorded by the project jumped by 60% compared to 2021, while in 2023 and 2024 it began to decline slightly from that peak, by 5% and 8% respectively. The 2025 data should not yet be taken as final, as it will continue to be supplemented and the number of recorded repressive cases will rise.

Graph 1. Politically motivated prosecutions, according to OVD-Info, 2012–2025

However, this is only one statistical snapshot, derived through manual collection and therefore subject to omissions and errors. Moreover, in recent years OVD-Info staff have encountered another systemic problem. Whereas earlier, during a period of lower levels of repression, public attention could help those targeted by drawing greater scrutiny to their cases, today Russian law enforcement operates as a punitive system and, in particular, tends to show greater harshness towards defendants whose cases become more visible. As a result, those facing prosecution and their lawyers increasingly seek to keep information about their cases out of the public sphere.

Despite the importance of the OVD-Info database, another tool is needed which, in conjunction with it, can provide a more comprehensive picture of the scale and dynamics of repression. In our opinion, such a tool is court statistics, namely, data on the number of criminal convictions published every six months by the Judicial Department of the Supreme Court, for a defined range of politically sensitive articles of the Criminal Code.

The ‘repressive basket’ of judicial statistics

Political repression in Russia today is systematic and to a large extent institutionalised. This means that there is now a relatively stable set of Criminal Code articles that are used for such purposes, and a well-established body of court practice that allows repression to be scaled up. Some of these articles are explicitly 'political', for example, those targeting anti-war statements, such as the articles on 'discrediting' the armed forces and on spreading 'deliberately false' information about them (Articles 207.3 and 280.3). This group also includes various articles criminalising certain kinds of public 'calls' or 'appeals', for example to undermine territorial integrity, state security, or to engage in extremism (Articles 280, 280.1, 280.4). In all such cases, what is being criminalised is a spoken or written position, that is a 'speech crime'.

Also belonging to this 'political' group are the article on vandalism motivated by 'ideological hatred' (Article 214, Part 2), and the article on the 'rehabilitation of Nazism' (Article 354.1; for an overview of how it is used in repressive practice, see → OVD-Info: Unusual Fascism). Finally, this category includes articles punishing failure to comply with the requirements imposed on 'foreign agents' (Article 330.1), and articles criminalising involvement with organisations declared or deemed 'undesirable' by the Russian authorities (Articles 284.1 and 330.3).

Yet even more frequently, Russian authoritarianism employs an expansive, deliberately distorted interpretation of criminal concepts in order to prosecute political opponents. This is the tactic of repressive populism, where the authorities attempt to ascribe to the words and actions of opponents a supposed public danger which they do not in fact pose (→ Rogov: The Art of Coercion). The clearest example here is the so-called anti-extremism legislation (Articles 282, 282.1, 282.2, 282.3, 282.4). According to the OVD-Info database, these articles were used in 142 political cases in 2022 (16% of all prosecutions), 209 in 2023 (26%), and 128 in 2024 (17%). Initially used to target far-right nationalist groups, by the late 2010s these articles had become an instrument for suppressing opposition speech and activity more broadly, as well as values and social identities stigmatised by the regime. Under this logic, in 2021 the Anti-Corruption Foundation was declared an 'extremist organisation', and in 2023 the (non-existent) 'international LGBT movement' received the same designation. According to Human Rights Watch, in 2024 and the first half of 2025, at least 20 people were criminally charged for alleged involvement in this 'movement'.

Within the same strategy of repressive populism, 'mass disorder' is the term used by repressive justice to describe unauthorised demonstrations and gatherings, that is, attempts by citizens to exercise their right to freedom of assembly (Articles 212 and 212.1).

The situation is somewhat different with 'terrorism' articles. In the early 2000s, Russia experienced a high level of terrorist activity, largely linked to the Chechen wars. According to the International Terrorism Database, in 2000–2005 around 600 people per year were killed in terrorist attacks in Russia. This figure subsequently declined steadily and by the late 2010s was at a minimal level. Yet while in 2005 there were 28 convictions under three forms of Article 205 (Parts 1–3), by 2010 there were 10 convictions under seven forms of the article; by 2015, 69 under 14 forms; by 2020, 471 under 16 forms; and in 2024, 868 convictions under those same articles. The elastic label of 'terrorism' has become a tool of mass repression.

Throughout the 2010s, most of those convicted under 'terrorism' articles were Muslims, with the key proof of guilt being 'membership in a terrorist organisation'. Within this designation, 'terrorist' referred not to organisations that had committed terrorist acts, but to those designated as such by the Russian state. During the war, the term 'terrorism' began to be applied to Ukrainian servicemen and to expressions of sympathy towards them. More recently, the label has been extended to opposition and civil society groups. As a result, those now accused of terrorism include, for example, chess world champion and opposition figure Garry Kasparov (who resides in the United States), political scientist Ekaterina Schulmann, other members of the Anti-War Committee, and the entire Anti-Corruption Foundation (for more information on how ‘terrorism’ articles are used in repressive practices, see → Amnesty International review; Re: Russia: The Fictitious Anti-terrorism Industry; First Department: Five sentences every day).

Alongside these, espionage and sabotage articles (Articles 275, 275.1, 276, 276.1, 281, 281.1, 281.2, 281.3, 281.4) must also now be included in the repressive basket. The number of cases under them has grown exponentially; proceedings are typically classified, and trials are held behind closed doors. This makes it impossible to verify the charges, though sporadic leaks often raise serious doubts.

Analyst Kirill Parubets, working with First Department, identified 1,000 cases brought under 'espionage' Articles 275, 275.1, and 276 between 1997 and 2024, of which 792 occurred during the war period. In other words, every 10 days of the war (excluding holidays and weekends), 11 new cases of treason, 'confidential cooperation' or espionage were initiated, making these charges another mass repressive channel. Parubets also concludes that the Judicial Department intentionally understates the number of such cases (its statistics list only 256 convictions for 2022–2024). A substantial share of these prosecutions punish assistance or even sympathy towards Ukraine; the mass use of espionage articles evokes a 'Stalinist' pattern designed to amplify public fear and reflects the security services’ institutional interest in maintaining high 'clearance rates' for such 'offences'.

Finally, sabotage cases (Articles 281, 281.1, 281.2, 281.3) also fall within the repressive basket, as First Department analysts show that their rise is linked to the reclassification of certain acts (for example, arson of military enlistment offices) from 'property damage' to terrorism or sabotage (→ First Department: How ‘terrorism’ articles work in modern Russia; First Department: Sabotage and terrorist attacks). In most such cases, individuals did commit unlawful acts, but the severity of punishment for burning a recruitment office door or a railway relay cabinet is not determined by the actual danger posed. Rather, the shocking harshness of the sentence serves an intimidating purpose.

The widespread use of terrorism, espionage and sabotage charges is intended to instill in the public a sense of wartime emergency, of a world infiltrated by 'enemies', and thus to legitimise the high level of state repression and severity of punishment. Extremely long sentences, from 10 to 20 years or more, become normalised through constant public exposure; accordingly, sentences of five to seven years for a social media post appear less shocking. In this sense, the expanding use of these articles is a key structural element of high-repression policy, even though some cases may relate to actual criminal acts. The strategy of repressive populism largely consists in creating a unified grey zone of both real and invented crimes.

The repression index and its meaning

A total of 33 articles of the Criminal Code are included in our 'repressive basket' (see detailed list in Table C in the Appendix). They can be grouped into six thematic categories: 1) espionage, treason, sabotage; 2) terrorism; 3) calls for terrorism; 4) anti-war statements; 5) extremism; 6) other ‘political’ articles. We treat the article on the justification and propaganda of terrorism (Article 205.2) as a separate category, since in substance it represents another form of 'prosecution for speech', similar to certain 'extremist' and 'political' articles. Its particular status and role are evident, among other things, in the fact that sentencing practice under this article differs significantly from sentencing practice under other 'terrorism' articles.

In our view, the dynamics of convictions under this basket of Criminal Code articles should be considered as a Repression Index, indicating the regime’s inclination to increase pressure on society through intimidation. Unlike the OVD-Info database, which is based on manual collection and evaluation of individual cases, the data on convictions within the repressive basket provide a statistically more reliable, but strictly categorical, snapshot. That is, we cannot claim that the number of people convicted under these articles equals the number of politically prosecuted individuals. The Index measures not the quantity of repressions, but the level of repressive intensity.

The relevance of the basket is confirmed by cross-checking with the OVD-Info persecution database. The articles included in the basket accounted for 77% of politically motivated prosecutions recorded in 2022, and 89% in 2023 and 2024. Such a high overlap, together with the considerations mentioned above, indicates that although some of those convicted under these articles may indeed have been involved in terrorism, sabotage, or espionage, these articles function predominantly as political instruments. This is also reflected in their continuous fine-tuning: the selected group of articles is the subject of ongoing legislative 'creativity' – new compositions are added, circumstances clarified, and penalties made harsher.

Detailed data on the dynamics of convictions under this basket and its thematic sub-categories from 2021 to the first half of 2025 are presented in Table A of the Appendix. The statistics reflect the number of individuals convicted under the primary article. What matters to us is not how many times an article was applied in total, but how many times it served as the main instrument of repression. Moreover, statistical observation shows that when articles from the basket are used as supplementary charges, the primary charge is typically also drawn from within the same basket.

In total, 1,120 people were convicted under the selected articles in 2021, 1,514 in 2022, 1,621 in 2023, and 2,120 in 2024. In the first half of 2025 alone, 1,611 people were convicted, almost as many as in the whole of 2023.

The dynamics of repression during the war

The first and most important result that the analysis of the level of repression, i.e. court statistics on the ‘repressive basket’, gives us at the moment is that, contrary to the data from OVD-Info, the repression index rose throughout the three years of the war and, alarmingly, jumped sharply in the second half of 2024 and the first half of 2025.

In the first year of the war, the increase in sentences was 35% compared to 2021, while in 2023 the increase was insignificant — 7% compared to the previous year (see graph 2). In the second half of 2023 and early 2024, judging by the dynamics of convictions, there was a relative stabilisation in the level of repression. However, in the second half of 2024, the total number of convictions increased by almost 40% compared to the previous six months, and in the first half of 2025, by another 30% compared to the end of 2024. As a result, in annual terms, the total number of convictions in the first half of 2025 increased by almost 80% compared to the first half of 2024. This jump significantly exceeds the jump in repression observed in the first year of the war.

Graph 2. Dynamics of sentences by category of charges, 2021–2025

Table 1. Value and dynamics of the repression index and its components, 2021–2025

As can be seen from Table 1, in 2022 the increase in convictions within the repressive basket was driven by the group of 'anti-extremist' articles (a rise of 40%, from 448 to 630 people), which to some extent reflected the conclusion of prosecutions initiated back in 2021, as well as by convictions for incitement to terrorism or justification of terrorism (from 199 to 274 convictions). At the same time, the first 16 convictions under the new 'anti-war' articles (Articles 207.3 and 280.3) appeared, and the number of convictions for 'mass riots' (88 people) and ideologically motivated vandalism (74 people) more than doubled compared with 2021.

In 2023, the total number of convictions within the repressive basket increased by only just over one hundred. Prosecutions for anti-war statements took shape in full (105 convictions), and growth began in the cluster of charges relating to treason, espionage, and sabotage. The number of convictions under 'terrorism' articles continued to grow moderately, while the intensity of enforcement under the 'extremist' and the specifically 'political' articles showed a decline (particularly charges for incitement to extremism and participation in mass riots). Finally, in 2024 (with a shift towards the second half of the year) there began a strong increase across almost all categories. The number of convictions in the treason, espionage, sabotage group quadrupled (from 60 to 240), convictions for propaganda and justification of terrorism grew by one and a half times, convictions for anti-war statements rose to 145 (a 1.4-fold increase), and charges under extremism articles grew steadily by around 20%.

This trend continued in the first half of 2025. Once again, the number of convictions in the 'Stalin-era style' espionage–sabotage cluster more than doubled compared to the first half of 2024. After relative stabilisation in 2023 and early 2024, the terrorism-related articles surged again, and the number of people convicted for carrying out a terrorist act rose almost eightfold, from 27 to 210 (this is linked to the reclassification of arson attacks on military enlistment offices and railway relay cabinets as terrorist acts).

The number of convictions for justification of terrorism again increased by one and a half times, from 167 to 243 (these are generally convictions for expressing sympathy towards individuals previously convicted under 'terrorism' articles, or towards the Ukrainian side). Likewise, convictions for extremist activity, i.e. not under the main 'extremism' article 280, but under variations of article 282, increased by more than one and a half times, from 220 to 363. Mediazona has established that more than 100 cases in 2025 related to donations to the Anti-Corruption Foundation after it was designated an extremist organisation, of which 79 had already led to convictions. This group appears to account for most of the increase; however, the figures of the Judicial Department may indicate that there were even more such cases and convictions.

Finally, the number of convictions under the 'political' articles increased three and a half times compared with early 2024. Convictions for failure to comply with 'foreign agent' obligations and participation in 'undesirable organisations' rose from two to 33, and the number of trials for 'rehabilitation of Nazism' doubled, but the main increase came from the article on mass riots. The rise in convictions under this article, from 12 in the first half of 2024 to 121 in the first half of 2025, was driven primarily by prosecutions in the 'Baymak case', following the mass protests in Bashkortostan in January 2024(for a detailed description of the case, see → Memorial: The Baymak case; OVD-Info: Baymak, one year later).

Integrated repression index

However, the dynamics of repression are determined not only by quantitative indicators in judicial statistics but also by qualitative parameters, the main one being the proportion of convictions involving actual imprisonment.

The share of such convictions increased steadily from 2021 to 2025, although initially at a fairly moderate pace (see Table B in the Appendix for detailed figures). In 2021, imprisonment was imposed in 40% of convictions in the repressive basket (447 of 1,120). In the first year of the war, this proportion rose to 44%, alongside an overall increase in convictions of 60% (665 of 1,514). By 2023, imprisonment was imposed in more than half of convictions (54%); in 2024 the figure approached 60%, and in the first half of 2025 it reached 67%. This again indicates that the rise in the level of repression in 2024–2025 significantly exceeds the increase observed at the start of the war.

Graph 3. Share of convictions involving imprisonment as a proportion of all convictions, 2021–2025

Across the different categories of the 'repressive basket', the situation with sentences involving imprisonment varies significantly. In the 'treason, espionage, sabotage' cluster, this is almost always close to 100% real prison terms. Under the 'terrorism' articles, real terms accounted for 70% of convictions in 2021 and rose to around 83% during the war. Under the article on justification of terrorism, the share of prison sentences has remained stable at around 40% of convictions. For the articles concerning anti-war statements, sentences were initially fairly lenient, with only 13% of convictions in 2022 resulting in real terms. In 2023, this share increased to 40%, and in 2024 it exceeded 50%. Under the 'extremism' articles, the share of convictions involving imprisonment rose from about 20% in 2021–2022 to 40–45% in 2023 and the first half of 2025. Finally, the share of such sentences under the 'political' articles was normally around 30%, but it depends heavily on the weight of the most severe article in this group, that is 'mass riots'. As soon as the number of those convicted under this article increases, the proportion of real prison sentences rises sharply.

Average prison terms imposed under the various article groups are also, of course, significant. Our analysis shows that there has been no substantial increase in sentence lengths during the war years (see Table 3). The impression that sentences have become longer is mainly due to the growing share of convictions under the 'heavier' articles (terrorism, treason, sabotage). In reality, we see an increase in sentence length only under the 'terrorism' articles in the first half of 2025 (12.5 years compared to the average of nine years for 2021–2024). Otherwise, sentence lengths have remained stable: convictions for incitement to terrorism and under the 'extremism' articles result in an average of three years, while convictions under the 'anti-war' articles result in around five years. The average terms under the 'political' articles depend heavily on how many cases of 'mass riots' are included in this category.

Table 2. Average prison sentences by category of article, 2021–2025, years

It is clear that the qualitative parameter of imprisonment is extremely important for assessing the level of repression. Therefore, the final, integrated Repression Index of the regime must reflect this variable alongside the variable of the total number of convictions. As a result, the integrated index can be represented as the sum of the total number of convictions under the articles in the repressive basket and the number of convictions resulting in real prison terms. In such an index, those sentenced to actual imprisonment are counted twice, i.e. with a coefficient of 2, while all other defendants are counted with a coefficient of 1. At the same time, the natural (non-normalised) value of the index is useful, as it indicates the approximate scale of the total number of people subjected to the state’s repressive efforts. The values of the integrated index are presented in Table 3, with the figures for the first half of 2025 multiplied by two, in order to show the comparative intensity of repression on an annualised basis.

Table 3. Integrated repression index, 2021–2025

The second wave of repression and its profile

As already noted, the Repression Index based on the analysis of judicial conviction statistics under the basket of politically sensitive Criminal Code articles shows a sharp upward rise beginning roughly in the second half of 2024. It is clear that we are witnessing a second wave of intensifying repression, following the initial surge in 2022 and after a period of relative pause in 2023. This wave remained largely unnoticed partly because in 2022–2023 the main focus of human rights organisations and the wider public was directed towards the new articles concerning anti-war statements. Meanwhile, as our data shows, such cases occupy a relatively small place in the overall picture of repression (around 6% of all convictions in the basket), and although the number of convictions under these articles is not decreasing, it is rising only at a moderate pace (by 16% in the first half of 2025 compared with the first half of 2024).

A key feature of this new rise in repression is that, alongside a striking increase in convictions under the usual 'political' articles, we also observe a disturbing expansion of repression within the cluster of the most severe charges: terrorism, treason, and sabotage. If we divide the conviction basket into two clusters – the most 'severe' (terrorism, sabotage, treason, espionage) and the remaining, strictly 'political' articles – we find that in the first half of 2025, convictions under the political articles grew by 1.5 times compared with the first half of 2024, while convictions under the severe articles grew by 2.1 times. In 2021–2023, the share of these severe articles among all convictions in the repressive basket fluctuated between 26% and 32% and averaged 30%. In the second half of 2024 it increased to 34%, and in the first half of 2025 it reached 37%.

At the same time, as already mentioned, another defining feature of the repression profile in the first half of 2025 was the sharp rise in convictions under the article on 'mass riots' (Art. 212), from 12 in the first half of 2024 to 121 in the first half of 2025. This category of cases significantly increased both the weight of convictions involving real prison terms and the proportion of convictions resulting in long sentences. The number of convictions under the 'severe' articles, including Article 212, grew 2.4 times year-on-year in the first half of 2025, while convictions under the other 'political' articles increased by one and a half times. As a result, the expanded group of severe articles accounted for 44% of the entire repressive basket.

Graph 4. Dynamics of the number of convictions for the most ‘serious’ and other ‘political’ offences, 2021–2025

It should be remembered that convictions represent the final stage of the criminal process, meaning that their dynamics must lag by at least six months or more behind the political decisions that shape them. Therefore, for the rise in repressive convictions observed in late 2024 and the first half of 2025 to occur, the decision to expand repression must have been made in early 2024.

On the one hand, one could argue that the rise in repression under the treason and terrorism cluster is a response to numerous and relatively successful Ukrainian sabotage operations inside Russia. These operations, as is known, sometimes relied on recruitment through commercial inducement or deception (this is how many arsonists of military recruitment offices and railway relay cabinets were recruited), and at other times on the recruitment of Russian citizens sympathetic to Ukraine. However, in our view, the sharp increase in repression in late 2024 and the first half of 2025 was also influenced by two major domestic political events: Prigozhin’s mutiny (June 2023) and the mass protests in Bashkortostan in defence of Fail Alsynov in January 2024 (the 'Baymak case'). A key feature of both episodes was that resistance to the regime (and sympathy for that resistance) came not from the liberal milieu of 'educated metropolitan citizens', which are the primary target of the 'political' articles in the repressive basket, but from entirely different social groups.

The response, to some extent, was the extension of a 'Stalinist pattern' of repression, whose defining features are very long prison terms and a relatively 'random' distribution across different social groups outside the 'liberal ghetto'.

Overall, as this review shows, the statistics on convictions under the selected basket of Criminal Code articles allow us to assess the dynamics of repression and to track changes in its profile. They also allow us to supplement and clarify the picture of repression beyond what is reflected in the OVD-Info database. In particular, we have grounds to believe that the number of cases in the 'Baymak case' and in some other categories was higher than what came to the attention of human rights organisations. Finally, this tool allows us to monitor the dynamics within the 'grey zone' of repression, where assessing the justification of charges and the proportionality of punishment to the act committed is extremely difficult.

Appendix

Table A. Dynamics of convictions under articles in the ‘repressive basket’, 2021–2025, number of persons convicted under the main article

Table B. Summary table of the number of convictions, convictions with imprisonment, their share in the total number of convictions, and the integrated repression index, 2021–2025

Table C. List of politically sensitive articles of the Criminal Code included in the ‘repressive basket’ of the Re:Russia repressiveness index

@ Re:Russia / Evgeny Antonov, Kirill Rogov