The Social Reach of War: The Family Networks of Russians Fighting in the War and Their Characteristics

1. The proximity of the war and the situational activation of family networks

This study centres on the question of the social reach of war. Since spring 2022, the proportion of those surveyed by ExtremeScan reporting that they or their relatives had participated in the war has risen steadily, from 15% in May 2022 to 23% in February 2023, following the ‘partial’ mobilisation, and to 38.2% in the July survey (Figure 1). Of these, 1.4% reported having participated in hostilities themselves, 22.6% reported that one relative had taken part, and 14.7% that more than one relative had done so. According to the July figures, 37.1% of those surveyed have relatives who have participated in the war. Some respondents who have themselves taken part in the war also have relatives who have done so, meaning that this group cannot simply be calculated by adding together the three subgroups. A substantial proportion of those surveyed reported the death of a relative (14.2%), while smaller proportions reported that a relative had been wounded, gone missing or returned from the war. Thus, two-thirds of those with relatives at war, representing more than 25% of the entire sample, provided additional information about them beyond the fact of their participation. It is the penetration of war into family networks that transforms it from an external political and social event into a matter of private life for a surprisingly large proportion of Russians, making the war relationally close to society as a whole.

In the July 2026 survey, another question asked about events affecting them and their families over the previous two years. One of the prompted response options was ‘the mobilisation of a relative’. Some 20% selected it. Formally, Russia’s ‘partial’ mobilisation took place four years ago. However, mini-focus groups conducted to clarify how respondents understood the concepts and wording used in the questionnaire revealed two things. First, the phrase ‘over the previous two years’ could encompass significant events that had begun earlier but continued to affect the family. Second, respondents interpreted mobilisation broadly. Participants in the discussions used the term not only to describe the autumn 2022 campaign, but also conscription for compulsory military service and even cases in which relatives had been coerced into signing military contracts. Thus, reports of mobilised relatives generally reflect families’ experiences of state coercion in sending a relative to war (and in this sense may contain a certain element of judgement).

Figure 1. Proportion of respondents reporting their own participation or that of relatives in military operations in Ukraine, 2022–2026, % of those surveyed

Table 1. Forms of involvement in the war and their prevalence, July 2026, % of those surveyed

In some surveys, sociologists ask whether respondents have friends or acquaintances who are participating in the war. In our view, this wording expands the social reach of war too broadly. The concept of an ‘acquaintance’ is rather vague and encompasses people with whom emotional ties may be minimal. As a result, the question conflates two distinct degrees of social proximity. We deliberately restricted our wording to ‘relative’, which indicates a higher level of personal involvement.

At first glance, the figures appear unexpectedly high: almost 40% of those surveyed have at least one person who has participated in the war somewhere within their family information network. However, it became clear in the very first wave of the survey that respondents understood the circle of relatives much more broadly than the immediate family, including cousins, relatives by marriage and other members of the extended family. This understanding was formalised in the survey instructions: family members were defined as all relatives whom respondents themselves considered part of their family, including distant relatives and relatives on their spouse’s side. All of those surveyed who reported having relatives who had participated in the war were also asked an open-ended follow-up question about how they were related to them.

The fact that those surveyed expand their family and kinship networks when answering this question is itself an important sociological finding. Every respondent has a network of family ties and a hierarchy of information about the people within it. When particularly sensitive and emotionally charged events occur, such as participation in war, and especially death or disappearance in action, information is exchanged beyond the bounds of regular communication. The ‘family grapevine’ spreads the news more widely. This effect can be defined as a situationally activated family information network: a circle of relatives about whom a person acquires information in connection with a socially significant event. As responses to the open-ended questions show, in this case we can speak of the activation of a network comprising around 20–30 relatives across different generations, whose fate may become a subject of discussion when they are involved in such events.

In the July survey, 14% of all of those surveyed reported having a relative who had been killed, while 4.5% reported a relative who had been wounded. This is a particularly revealing finding. It is generally assumed that the number of wounded in war is two to three times higher than the number killed. Yet the survey data reverse this pattern: a fact with greater social and emotional significance helps activate a wider family network. This departure from the ratios found in conventional statistics is a striking indication of how information spreads across a broader circle of social ties. Overall, 20% of those surveyed reported having a relative who had been killed, wounded or gone missing. This represents one in five respondents in the survey sample and half (52%) of those with relatives who had participated in the war. These proportions reflect not so much the actual distribution of outcomes as the relative salience of the information.

Some 23% of those surveyed reported having one relative who had participated in the war, while 15% reported having several. Open-ended responses occasionally mentioned three, four or more relatives who were currently fighting or had fought in the war. The interviewer clarified that ‘participation’ referred specifically to involvement in hostilities. However, relatives may not have complete information, and those counted as fighting may also include personnel serving or working in logistical support units.

Among those surveyed aged 18–29, 13% reported that an uncle or aunt had participated in the war, 11% cited a brother or sister, and 7% a parent. Among older respondents, the collateral line of the next generation plays a notable role: 17% reported the participation of nieces or nephews, while 5% each reported that a child or a grandchild or great-grandchild had participated. Among all women surveyed, 5% reported the participation of a current or former husband. A further 8% cited other relatives by marriage, 10% nieces or nephews, and 3% children. Interestingly, respondents were twice as likely to report the participation of a parent as that of a child. This, too, is a significant characteristic of the current war: for pragmatic reasons, it is middle-aged people rather than young people who are more likely to be sent to fight.

Thus, our findings indicate that familial involvement in the war has become a widespread phenomenon in Russia, one that warrants dedicated sociological investigation. For Russians, the war has long ceased to be an external ‘event on the television’ or in news bulletins, lying beyond the bounds of their immediate social experience. Through a broad, situationally activated network of family ties, it has penetrated the fabric of everyday life in Russia and its emotional landscape.

2. Uneven family experiences of the war: areas of deprivation and the redistribution of costs

Measuring the social reach of war provides important insights not only into how widely it has penetrated family networks and everyday life, but also into how this experience is distributed across social groups and how it correlates with social sentiment and political attitudes.

Table 2. Prevalence and forms of relatives’ participation in the war by respondent income, July 2026, % of those surveyed by income group

As Table 2 shows, the proportion of those surveyed with a relative who participated in the war and was killed is higher in lower-income groups. In the lowest-income group, 48% of those surveyed have no relative at war, compared with 67% in the highest-income group. The income divide is most pronounced in the experience of mobilisation: respondents in the least affluent group are approximately twice as likely to report having a mobilised relative as those in the most affluent group, at 35% versus 17%. Both patterns appear to reflect the role of family social capital. A substantial proportion of the relatives of less affluent respondents living in provincial towns reside in the same or similar towns and are themselves unlikely to enjoy high social status or wealth. The relationship with income is therefore not mechanically linear, but is most evident at the two ends of the income spectrum. It is more accurate to speak of particularly high involvement among the least affluent group and relatively low involvement among the most affluent. The difference in the proportion of relatives who died also appears significant: among the more affluent, 9–10% give this response, compared with 17% among less affluent respondents, who are also more likely to report relatives who have gone missing. However, this effect may be a consequence of the greater number of people who have participated in the war within the family networks of these groups.

Table 3. Relatives’ participation in the war, mobilisation and casualties by type of settlement, July 2026, % of those surveyed by group

The prevalence of participation in the war is also clearly skewed towards family networks rooted in villages and urban-type settlements. Those surveyed living in these areas are substantially more likely to have relatives who are fighting or have fought in the war. The most important and pronounced difference, however, concerns not whether a respondent has a relative who has participated in the war, but the degree of involvement across the family network. Some 23% of urban residents and 26% of rural residents report having one relative who has participated, a relatively small difference. However, 12% of urban residents, compared with 24% of rural residents, report having several relatives involved. The social reach of war in rural areas is greater both in terms of mobilisation and severe losses, with particularly marked differences in reports of relatives killed or missing. Urban-type settlements occupy an intermediate position: in terms of mobilisation, they are closer to cities, but in multiple participation and reports of relatives killed, they resemble rural areas. For residents of major cities and those of small settlements, both the war itself and the density of its presence in civilian life look substantially different.

The mere presence of someone who has participated in hostilities within a respondent’s family network does not affect their assessment of their material circumstances: on this measure, they remain at the sample average (Table 4). However, respondents with relatives who have participated in the war but no relatives who have been killed or wounded rate their material circumstances slightly more positively than average. This should not be taken as evidence of ‘mass enrichment’ within this group: the majority do not report any material gain from the war (it is worth remembering that the activated family network extends far beyond those eligible for compensation payments or those who benefit directly from signing a military contract). The difference appears instead to stem from the presence of groups with more negative assessments of their circumstances. Such assessments are more common among respondents whose family networks include mobilised personnel, meaning relatives sent to war through mobilisation, or several people who have participated in hostilities.

Table 4. Assessment of the past year and expectations for the future, July 2026, % of those surveyed by group

Among those who reported having a mobilised relative, 51% said their financial circumstances had worsened over the past year, while 44% expected them to deteriorate further. Those with relatives who had been killed displayed a similar pattern of more negative social assessments. Here, the same mechanism of concentrated disadvantage discussed above appears to be at work. The point is not that having a mobilised or deceased relative necessarily worsens a respondent’s sense of social well-being, but that the human costs of war accumulate within the same family networks.

Within the kinship networks of those who have been mobilised, assessments of the impact of the ‘special military operation’ on their lives and their expectations regarding its continuation are significantly more negative (Table 5). It is important to note that negative assessments predominate in all groups, although to varying degrees. Among relatives of contract servicemen, 55% assess the war’s impact negatively, compared with 60% of those without relatives at war and 70% or more among those with relatives who have been mobilised, killed, wounded or gone missing, as well as those with several relatives participating in the war.

Table 5. Negative consequences of the ‘special military operation’: lived experiences and expectations, July 2026, % of those surveyed by group

The war does not affect society uniformly, but rather creates concentrated social clusters. In our sample, an area of hardship is clearly visible, comprising the extended families of those who reported the forced deployment of a relative to the front, losses within their family network, or several relatives who had fought or were fighting in the war. These groups are concentrated predominantly in lower-income segments of society and smaller settlements. The manpower for this war is drawn from the social periphery, and this occurs through both channels – compulsory and commercial. As a result, the social costs of war accumulate within this cluster, to some extent shielding major cities and more socially influential groups, particularly those on middle and higher incomes, from those costs. However, while close relatives of contract servicemen and those killed in the war may receive financial compensation, the wider social cluster does not benefit as a whole.

3. The social reach of the war and political attitudes

In a telephone survey conducted in July this year, ExtremeScan abandoned its previous practice of allowing those surveyed to avoid giving a direct answer to the question about their attitude towards the war (the earlier wording was: ‘Do you personally support or oppose Russia’s “military operation” on the territory of Ukraine, find it difficult to give a definite answer, or prefer not to answer?’ The question was now posed more directly: ‘Please tell me, do you personally support or oppose Russia’s “military operation” in Ukraine?’) In the March survey, which used the less direct wording, 54% said they supported the operation, 14% said they opposed it, and 32% avoided giving a clear answer, with 16% undecided and 16% unwilling to answer. In the July survey, the proportion saying they opposed the ‘military operation’ rose to 35%, while 18% still gave no definite answer and 47% expressed support. The decline in support in July may have been influenced by Ukraine’s intensive strikes on Russian infrastructure, which contributed, among other effects, to an acute petrol crisis. However, an even more important factor may be that most of the ‘silent’ respondents in earlier surveys declared their opposition to the military operation when given the more direct wording. In this section, differences in attitudes between groups are examined in relation to this overall distribution.

Two substantially different political profiles emerge within the broader group of relatives of people who have participated in the war. If we examine these respondents according to the level of involvement in the war across their family networks, differences in political attitudes appear relatively modest. Among those with one relative who has participated in the war, 51% express support for the ‘special military operation’, whilst 37 per cent do not. Among those with several relatives involved, the corresponding figures are 46% and 38%. However, a much sharper contrast emerges when we compare two mutually exclusive, similarly sized subgroups defined by the reported mode of involvement in the war, coercion or voluntary enlistment. In other words, we distinguish between relatives of ‘special military operation’ participants who reported mobilisation and those who did not mention it.

Defining the group of those who reported having a relative at war but did not mention mobilisation as ‘relatives of contract soldiers’ is, to a certain extent, arbitrary. The absence of a report of mobilisation does not in itself establish that the relative enlisted under contract or participated voluntarily. A respondent may not remember, or may not know, when or how their relative, who may sometimes be quite distant, went to war. Conversely, the group classified as ‘mobilised’ probably includes not only those mobilised in 2022, but also people subsequently sent to the front under coercion, as noted above. However, regardless of the actual circumstances in which the relative entered the war, reporting that they were mobilised reflects the respondent’s perception that a member of their family network was forced or compelled to participate in hostilities. The absence of such a report suggests that the respondent does not associate their relative’s participation with this particular emotional association of coercion.

Among relatives of contract soldiers, support for the ‘special military operation’ is significantly higher than among those who have no relatives at war, by 9 percentage points. Among relatives of mobilised personnel, support is only slightly lower (Table 6). However, the picture is reversed when we look at the groups stating that they do not support the war. This group is almost twice as large among relatives of mobilised personnel as among relatives of contract servicemen, and 14 percentage points larger than among those without relatives at war. In effect, almost one in two respondents with a mobilised relative said they opposed the war. Support and opposition are not symmetrical expressions of opinion. Support may reflect a conformist disposition or a tendency to give a socially approved response. Expressing opposition, by contrast, may require a stronger degree of personal conviction and even a willingness to register a form of protest in response to a pollster’s question. Thus, while support for the war is significantly more prevalent among relatives of contract servicemen, relatives of mobilised personnel are markedly more likely to express opposition.

Table 6. Personal support for the ‘special military operation’, July 2026, % of those surveyed by group

The higher level of support for the war among relatives of contract servicemen may have several explanations. It may reflect solidarity with a relative whose participation the family regards as a personal choice and which official narratives portray as a morally exemplary, heroic decision. This solidarity may extend to the army and the ‘special military operation’ as a whole. Furthermore, financial payments may further link the material well-being of close relatives to the military system. However, the causal relationship may also run in the opposite direction: members of families that were already more supportive of the war may be more likely to enlist voluntarily under contract. Overall, however, the increase in support for the war among relatives of those who joined voluntarily remains modest. Having a relative involved in the war is clearly not a key factor in shaping attitudes towards it.

Although the proportion of those surveyed with wounded or missing relatives is small, making reliable statistical conclusions difficult, the available data show that support for and opposition to the war are almost evenly balanced among relatives in these categories. Among the families of those who have gone missing, a group that is also too small to support reliable conclusions, opposition to the ‘special military operation’ is twice as prevalent as support. Unlike a confirmed death, disappearance leaves families facing a prolonged period of searching, delayed death-related compensation payments, administrative procedures and uncertainty. In the previous section, this group stood out for its social and material vulnerability. Here, that vulnerability is accompanied by the strongest degree of disengagement from the ‘special military operation’.

Attitudes towards a possible new mobilisation follow the same patterns as attitudes towards the war itself (Table 7). Compared with those surveyed who have no relatives at war, relatives of contract servicemen and relatives of mobilised personnel diverge in opposite directions by almost symmetrical margins of 5–6 percentage points. However, among relatives of mobilised personnel, support for halting hostilities and reaching a peace agreement is substantially higher, largely because fewer respondents in this group are undecided. Two-thirds want the fighting to stop, compared with half of the relatives of contract servicemen.

Table 7. Choice between a new mobilisation and a cessation of hostilities, July 2026, % of those surveyed by group

The two groups we have identified have significantly different views of Russian society (Table 8). Relatives of contract servicemen are more likely than respondents without relatives at war to believe that a majority of people around them support the war. It should be noted that the perceived prevalence of such a majority has declined recently across all groups. Among relatives of mobilised personnel, only one in five perceives a pro-war majority in their immediate surroundings, while more than 44% are convinced that they are surrounded by a majority who oppose the war. Russians generally tend to avoid discussing the war, both because the subject is potentially dangerous and because it is divisive. Consequently, answers to questions about the views of those around them often reflect respondents’ own attitudes projected onto their social environment.

Table 8. Immediate social environment, perceptions of society and discussion of political issues, July 2026, % of those surveyed by group

Nevertheless, relatives of mobilised personnel are significantly more likely to discuss political issues with people around them than those whose relatives chose to participate in the war. In this respect, as in their greater willingness to express negative views of the war, they again depart from the consensus of passive acquiescence that has prevailed in Russian society in recent years. Relatives of contract servicemen, by contrast, do not differ in this regard from those without relatives at war. Finally, relatives of mobilised personnel are more likely to perceive Russian society as divided. A similar pattern is evident in family networks with several relatives at war and among those who have lost relatives. In these groups, difficult family experiences coincide with a bleaker view of relations between people: a weaker sense of unity, and greater distrust and perceived injustice.

Nevertheless, this critical attitude has its limits. Negative assessments of Putin’s performance are twice as common among relatives of mobilised personnel as among relatives of war participants who did not mention mobilisation, at 26% versus 13%. Positive assessments of Putin’s performance are also less common in the former group, although they still represent a majority, at 55% compared with 66%. Consequently, the proportion who believe Putin should be replaced as soon as possible reaches 33% in this group, compared with 22% among relatives of contract servicemen. Yet 50% agree that Putin should remain in office. Thus, although almost half of those surveyed in this group say they oppose the war, only around a quarter to a third appear, judging by their answers, to extend their opposition to the president himself. Others, by contrast, may associate him with hopes that their relatives will return from the war and that its hardships will come to an end.

4. Some conclusions

The war has become deeply embedded in everyday life in Russia, albeit very unevenly, including through family networks. The extent of accumulated experience of this kind of involvement is considerably greater than one might have assumed in the absence of sociological data.

The material circumstances reported by relatives of contract servicemen are somewhat better than average: members of this group are slightly less likely to report a deterioration in their financial situation and slightly less likely to anticipate further hardship than those who have no relatives at war. At the same time, family networks are generally much broader than the circle of beneficiaries of financial payments to war participants, although this factor may have some influence. Mobilisation, the involvement of several relatives, and deaths are more likely to affect less affluent and rural family networks. Here, the experience of war is intertwined with social vulnerability, and the gap between these networks and families without a relative fighting in the war is substantial. This is because the involvement of socially disadvantaged clusters in the war is considerably higher than that of more advantaged ones. In other words, we are dealing less with the consequences of involvement than with the conditions that shape it.

When all respondents with relatives participating in the war are treated as a single group, their assessments of the war and the wider social situation are similar to those of respondents whose family networks have not been drawn into the war. However, this aggregate figure conceals pronounced differences associated with the circumstances in which their relatives became involved. These differences may reflect both the family’s experience of war and the attitudes of the respondent’s wider family circle that were already in place before their relative went to war, in other words, the prevailing attitudes within their social circle.

Among relatives of contract servicemen, where coercion as a reason for being sent to war is not reported, support for the ‘special military operation’ is somewhat higher, as is social optimism, understood as a perception of cohesion within the surrounding society. Relatives of mobilised personnel take a more critical view of the ‘special military operation’ and, crucially, are much more willing to speak openly about their views and discuss political issues. Their perceptions of society are markedly different: within their immediate circle, society appears predominantly anti-war, while in their wider circle it seems fragmented and permeated by distrust.

In any case, being sent to war under coercion is clearly associated with lower levels of support for it and a greater willingness to express opposition openly. At the same time, higher levels of support for the ‘special military operation’ among the relatives of contract soldiers do not imply a willingness to continue the war and bear its costs. The majority in the overall sample and in both groups being compared note the negative impact of the war on everyday life and expect a further deterioration should hostilities continue. Although the proportion of relatives of contract soldiers who support a new mobilisation is slightly higher than among those who do not yet have relatives fighting (25% versus 20%), it is still only half the proportion who oppose it. The desire to see the war brought to an end extends beyond the differences in attitudes identified here between the three groups.