Second Demobilisation: How public opinion changed during the second year of the war

During the second year of the war, public opinion polls show a slight decrease in the proportion of respondents who declared support for military action, and a much more significant decrease in the core of supporters of the war — the pro-war segment of society. As a result, this core has shrunk by one and a half times and has become proportionate to those who generally lean towards anti-war views. The share of those who would not be willing to support a decision by Putin to immediately withdraw troops also decreased, and by the end of the year, it became less than the share of those who would be willing to support such a decision. However, all this did not lead to an increase in the proportion of those who oppose the war, but rather to the expansion of a blurred and alienated attitude towards it.

In general, four periods can be observed in attitudes towards the war over the past two years. The highest levels of support were observed in the spring and early summer of 2022, this was followed by a phase of partial demobilisation of supporters of the war due to the realisation of its protracted nature and the higher costs associated with it, which peaked in September 2022. However, after the autumn retreat of Russian troops in the winter of 2022/23, there was a second phase of mobilisation among supporters of the war, stimulated by the fear of defeat and the information campaign of Prigozhin's turbo-militarism. Finally, from the spring of 2023 we can observe a second phase of demobilisation and a marked decline in pro-war groups. However, the absence of a simultaneous increase in anti-war sentiment is due not only to the fear of repression against the backdrop of relatively low costs of the war, but also because the scenario for ending the war and withdrawing troops appears unclear, potentially conflict-ridden, and worries respondents.

Supporters, opponents and evaders

After the first year of the war, we tried to summarise some observations on the dynamics of public opinion, based on polling from the 'Chronicles' and ExtremeScan projects. The main conclusion at that time was the, at first glance, paradoxical idea that in the data of mass polls we can identify not one, but two majorities simultaneously.

In general, the Chronicles and ExtremeScan surveys, thanks to the presence of two prompts offered to respondents from the beginning (the answer options 'I find it difficult to answer' and 'I do not want to answer') allowing them to avoid expressing a certain opinion, we see two large groups and one small group. The first are those who respond affirmatively to the direct question 'Do you support the actions of Russian troops in Ukraine?' They constitute the majority of 'declaratory support' for the war. In the spring of 2022, at the start of the war, this was 63% of those surveyed. This then declined and averaged 53% for most of 2023 (from April 2023, across four measurements). The second group are those who said they did not support the war: their share (as measured by the Chronicle and ExtremeScan) throughout almost the entire two years of the war was 10%. In addition to these, however, there is another large group, the group of 'evaders,' that is, those who found it difficult or refused to answer. In the spring of 2022 this was 28% of those surveyed, and for most of 2023 they formed 36%. At the same time, the 'hesitant' group grew by only 1 percentage point, and the 'refusers' group grew by 7.

It is important to note that support for the war in today's Russia is a strictly normative position (a normative requirement), the absence of which can lead to social pressure and persecution. Therefore, avoiding expressing this opinion in the survey largely looks like a meaningful gesture, and in any case, refusing to answer appears significant. If in response to an ordinary (non-sensitive) question the answer 'I find it difficult to answer' is simply an evasion of 'choice', here it becomes an evasion of confirmation of loyalty to a normative position supported by an effective repressive apparatus. In this sense, it is significant that at the beginning of 2023, 40% of respondents ('evaders' plus open opponents of the war) and at the end of the year, 48% did not express loyalty to the normative position of approving of the war..

Figure 1. Dynamics of answers to the direct question about support for the military activity, % of those surveyed

In general, according to the polls of 'Chronicles' and ExtremeScan, four periods can be distinguished in the dynamics of attitudes to the war in terms of a direct question about support for it. The highest figures of declarative support were observed in the spring and early summer of 2022. Then, for several months, there was a decline to its lowest levels in September (after the announcement of 'partial mobilisation'), with a simultaneous increase in the share of 'evaders'. There was a new rise in the winter of 2022/23, followed by a new decline from spring 2023, with support levels fluctuating between 51-55%.

These trends are also discernible in polling from the Levada Centre. The peculiarity of the latter results are, first, that they are conducted face-to-face (interviewers come to people's homes) and, second, that the respondent is offered a scale of possible answers: 'I definitely support', 'I somewhat support', 'I somewhat do not support' and 'I definitely do not support'. This tactic (often and quite legitimately used in surveys) encourages respondents to give a meaningful response. As a result, the Levada Centre's aggregate support figures ('Definitely support' + 'Rather support') are higher, ranging from 70-81% with an average of 75%, and show virtually no dynamics. However, if we look at the levels of support, we see a similar picture to that painted by the results of 'Chronicles' and ExtremeScan. The share of those who definitely support military action has decreased compared to the first six months of the war by 6 percentage points. The share of those who are uncertain in their support and uncertain if they do not support the war ('More likely yes' + 'More likely no') has increased from 37% to 42%; the proportion of those who strongly disagree has remained unchanged at around 10%.

Table 1. Support groups for the war according to the polls of 'Chronicles', ExtremeScan and 'Levada Centre', 2022-2023, % of those surveyed

It is also important to highlight that over the course of two years, the majority of declarative support for the war fluctuated within a very significant range (51-66%), while the share of open opponents of military activities remained unchanged. This is a rather uncommon situation for polls, when a decrease in support for one position is not converted into an increase in support for its alternative, indicating some special 'blocking' circumstances, in this case, presumably, the influence of the climate of public opinion and the repressive environment.

Why two majorities?

Two additional questions posed to those surveyed by Chronicles and ExtremeScan asked about support for 1) a hypothetical decision by Putin to withdraw troops, even if the goals of the military operation were not achieved, and 2) directing budget funds to military needs as a priority. The answer to the first question demonstrates whether the goals of the war are really essential for the respondent and whether a return to the starting point ('before the war') is deemed acceptable. The answer to the second question is whether these goals are deemed significant enough to 'tighten their belt' by restricting other budget expenditures. At the same time, the wording of both questions is not as sensitive — state officialdom does not require a single correct answer — and therefore leaves room for the safe expression of one's opinion.

The combination of the three questions allows for a more fine-grained segmentation and the identification of a broad and narrow core of supporters of the war. At the beginning of 2023, the broad core was 39% - this is those who not only declaratively supported the 'military operation', but would not be ready to support the decision to withdraw troops until its goals were achieved. The narrow core of support — turbo-militarists who declaratively support the war, would not agree to end it before its goals are achieved and are ready to 'tighten their belts' for the sake of victory made up 22% of respondents.

The fact that out of approximately 55% of respondents expressing declaratory support for the war, only about 39% would not support the decision to withdraw troops means that about a third of those who joined the normative majority of declaratory support would probably not have done so if the 'climate of opinion' had been more favourable. This allowed us to assert that, in addition to the majority of declarative support for the war, there is also a majority of ‘non-resistance to war’ — those who either shy away from a normative gesture of support or, while expressing declarative support, do not share pro-war views.

The situation where two majorities are identified in society may initially seem paradoxical. In fact, the direct question about support for the 'military operation', which sociologists are constantly posing to respondents, as well as other highly sensitive (polarising) questions, changes its actual content within the specific political context. Similarly, for example, when respondents in Russia are asked about their approval of Vladimir Putin's activities, within the context of the widespread belief that the absolute majority of Russians support him, the question essentially sounds to them as follows: 'Do you approve of President Putin in the same way that the absolute majority of Russians approve of him (as is well known), or do you belong to the minority that does not approve of him for some reason?' As a result, the respondent is not really answering the question about their attitude to Putin, but rather the question about their relation to the 'imaginary majority' of Russians, and in their response, they must declare either their belonging to it or their opposition to it.

The question about support for the 'military operation' within the specific repressive and propagandistic climate of today's Russia also has a specific meaning for the respondent. The answer 'I support' is strictly normative, while the answer 'I do not support' is openly confronting the regime. That is, in answering this question, the respondent first of all decides whether they are ready to define themselves as being in confrontation with the regime and in opposition to it — as well as to the 'imaginary majority' of Russians who support the war. The 10% of respondents who say they do not support the 'special military operation' are those who recognise that they are in a state of confrontation with the regime and the 'imaginary majority' (even if it does not involve public action).

Thus, in the distribution of responses to these three war-related questions, we are dealing with three main groups: 1) those who recognise themselves as being in a state of confrontation with the regime and the 'imaginary majority'; 2) those who actually support the war; and 3) those who are not ready for a confrontation with the regime even at the level of personal positioning, and are therefore forced to seek niches for declarative support or avoid answering. The last group forms another majority — the majority of 'non-opposition', the contours of which emerge in the shadow of the majority of declarative support.

Gradient of support/non-support for war

Results from qualitative sociological research, based on in-depth interviews, suggest that support/non-support for the war in Russian society is not a set of segments structured around articulated positions but rather a continuum of arguments and narratives situated between two poles: staunch support for the war and its non-support. To some extent, quantitative research based on mass surveys confirms these findings. The most significant changes in attitudes towards the war over the past year have occurred not in the distribution of responses to the direct question about support for the ‘special military operation’ but in responses to additional questions about the war.

Figure 2. Dynamics of indicators of attitude to war, 2022-2023, % of those surveyed

In the second half of 2022, as can be seen in Figure 2, the question of the immediate withdrawal of troops as a result of a decision by Putin stumped respondents, and the groups declaring support for this decision, disagreement with it, and those having difficulty answering were roughly equal. At the beginning of 2023, the share of those who supported this decision grew, but simultaneously the share of those who did not support it saw an even greater increase. Such simultaneous growth of two groups with opposing views is a sign of the polarisation of society. Moreover, those who did not support the withdrawal of troops were more numerous last winter than those who supported it. At that time, the share of those who did not support the decision to withdraw troops decreased twice — in spring and in the autumn of 2023. As a result, the proportion of this group shrank by almost a third, from 47% to 33%. The proportion of supporters remained stable — 40%, while the proportion of those having difficulty or refusing to answer increased (from 13% to 28%).

In other words, support for the idea of withdrawing troops (without achieving the set goals) is still difficult for a significant portion of respondents, but the share of real supporters of continuing the ‘operation’ decreased quite dynamically. The process of partial demobilisation of staunch supporters of the war was also evident in the distribution of answers to the second additional question about budget expenditures: in February 2023, 37% supported their priority allocation to military needs, in October 2023 that had fallen to 26%.

Accordingly, the narrow and broad cores of support for the war have undergone significant changes. In the February 2023 survey, the narrow core comprised 22% of respondents, while in the October survey, it decreased to 12%. The broad core (those who support the war and those who do not support a hypothetical troop withdrawal) was 39% in February 2023, but dropped to 27% by the end of the year. At the other end of the spectrum, in addition to the open opponents of the military operation (those who say they do not support the war in response to a direct question), are those who do not express declarative support for the war and yet support withdrawal of troops, as well as prioritising budget spending on non-military purposes. This group shares similar views with opponents of the war. Avoiding declarative opposition to the regime, they do not support pro-war policies and do not consider the goals of the war important enough.

Moreover, there is another group of those who do not express support for the military operation and support the withdrawal of troops without achieving the objectives of the war — these are the 'neither-siders' of the war. They are not willing to identify themselves as opposition, engage in confrontation with the regime, but at the same time, they would prefer the absence of war or a quicker resolution. They are not necessarily ‘secret liberals’; most of them are quite loyal to ‘Putinism’ but ‘Putinism without war’. Finally, there is a group of those who express declarative support for the ‘special military operation’, but support the withdrawal of troops and do not support the prioritisation of military spending — this is the group of purely declarative support for the war.

Thus, we can construct a kind of spectrum of attitudes to the war as it is presented in the data of mass surveys. Accordingly, we can observe how this spectrum has changed between the surveys conducted at the beginning and end of 2023.

Table 2. Spectrum of attitudes to the war

Figure 3. Spectrum of attitudes towards the war and its dynamics in 2023, % of those surveyed

As can be seen in Figure 3, the pro-war segment of the sample (turbo-partisans and supporters of the war) has significantly decreased, the share of those who are anti-war has not changed in size (20-25%), and the average zone of ambiguous, unclear or hidden attitudes towards the war has expanded from 41% to 54%.

How and why attitudes toward the war changed in Russian society in 2023

Figure 2 clearly shows that the beginning of 2023 became a sort of peak of mobilisation for war supporters and polarisation in society. Both the group supporting the decision to withdraw troops and the group not supporting such a scenario grew simultaneously. This seems quite logical given concurrent events.

After the autumn retreat of Russian troops from the Kharkiv region and from near Kherson, the 'smell of defeat' greatly unsettled the Russian public and instilled a sense of dangerous uncertainty. Moreover, these emotions were supported by quite specific changes in the information landscape. Yevgeny Prigozhin, with his furious tirades against Shoigu and Gerasimov for allegedly not giving him shells, and the war-supporting bloggers' tribe brought into the federal media scene by Prigozhin, radiated an atmosphere of 'the Motherland is in danger' (Interestingly, the version of the pro-war discourse was supported by a powerful anti-elite message).

This led both to an increase in declarative support for the military action after its decline in the autumn of 2022, and to an expansion of the militarist core of support. Conversely, by the end of 2023, the failure of the Ukrainian counteroffensive, the futile outcome of the coup attempt, the disappearance of Prigozhin, and the diminishing significance of pro-war bloggers, who were under pressure from Kremlin censorship, decreased the level of mobilisation. The pro-war frenzy lost ground and fuel, and the war became routine, shifting to the periphery and into the zone of what in late Soviet times was called 'doublethink' - formal loyalty in the absence of any conviction.

Speaking fairly about the routinisation of the war, however, experts overlook an important aspect: routinisation is associated with the shrinking of the group of those who support the war. This effect of general demobilisation is clearly visible in the 'petal' diagram (Figure 4): in positioning in relation to the war, the sample has shifted from the zone of real support towards the zone of declarative support and the 'grey zone', where the attitude of those surveyed towards the war is difficult to define.

Figure 4. Spectrum of attitudes towards war and its dynamics in 2023, % those surveyed

It is worth noting that even the increase in the share of families in which someone has participated or is participating in military activities (24% of respondents stated this in the last surveys) does not significantly change the profile of attitudes towards the war: in this group, 58% expressed their support for the war, while the average for the sample as a whole was 52%. Thus, among close relatives of direct participants in the war, we do not observe either increased criticism of the war or increased solidarity with its goals. In general, we observe a double phenomenon: on the one hand, the indifference of society to the losses of the war (which reached unimaginable scales not so long ago), on the other hand, the lack of increased involvement and sympathy with its goals even among those whose relatives are risking or have risked their lives and health at the front.

The signs of public demobilisation that we have highlighted align with the phenomenon discussed in Vladimir Zvonovsky and Alexander Khodykin’s article, and perhaps provide a clue to its understanding. While the share of those who said they support the 'military operation' in 2023 decreased slightly, the share of those who said that the majority of people around them support the war has noticeably contracted. Conversely, the share of those who say that they are equally surrounded (50/50) by opponents and supporters of the 'military operation' has grown. As we can see, the share of the broad pro-war core (39%) was one and a half times higher than the share of those who held more sceptical views about the war (25%) at the beginning of 2023. In the second half of the year, however, the sizes of these cores became almost equal. The group of convinced supporters of the war has shrunk by almost one and a half times, with unconvinced or declarative supporters likely to have less incentive to speak out about the war. For the average Russian person, the result of this could be that the 'audibility' of the pro-war position in their environment has noticeably decreased.

In our previous article, we described several phases of change in attitudes towards the military activities and focused on the phase of 'immersion in war', when, last winter, the public became aware of its duration and inevitability. In 2023, the main trend was the routinisation or displacement of the war. An important part of this phenomenon, as noted above, is the reduction in the share of real supporters of the war and the decline in the 'audibility' of the pro-war position in society, a kind of demobilisation of society in the spring and autumn of 2023.

At the same time, the fact that, with a sharp decrease in the share of real war supporters, the share of its opponents did not increase in both narrow and broad definitions, apparently indicates the effectiveness of authoritarian institutions: respondents perceive the costs of confrontation with the regime to be high, while the costs of the war are quite low. Some decrease in declarative support in 2023 was observed in most age groups, but in all of them, it led not to the growth of the group with an openly anti-war position but to an increase in the share of 'evaders'. As a result, in the youngest ages (18-34 years old) their share reached 50-60%, with 30-35% expressing a 'normative' position of support for the war. It is noteworthy that it is young people, where even declarative support for the war is a minority, who are now the main target of the authorities' propaganda pressure.

Thus, over the two years of the war, we have observed two periods of mobilisation and two periods of demobilisation of public opinion regarding the war. The first demobilisation in July-September 2022 was probably associated with the protracted nature of the war and its considerable costs. At the same time, the second, 'Prigozhin's' period of mobilisation last winter ('Motherland in danger') brought the pro-war coalition to the brink of acute internal conflict as a result (criticism by pro-war bloggers, Prigozhin's mutiny).

It is difficult to say whether the trend of the second demobilisation described here (from spring to the end of 2023) is sustainable and irreversible. The decline in the share of opponents of the hypothetical decision to withdraw troops during this period certainly indicates a growing demand for an end to the war. However, not only did the share of open war opponents not increase, but also the share of those in favour of its immediate end by a decision by Putin. In a September 2023 ExtremeScan poll, when asked whether Russians' lives would improve or worsen if Russia withdrew its troops from Ukraine now, 12% said it would improve, 34% said it would not change, and 37% said it would worsen. Among those who said they supported the war, when asked directly, 50% chose the last option. This indicates that respondents view the scenario of troop withdrawal as unclear, potentially conflict-ridden and alarming. This circumstance, along with the fear of repression, also hinders the consolidation of anti-war sentiments.