Do respondents' answers reflect the public opinion?

When P. Bourdieu in his famous report "Public Opinion Does Not Exist" substantiated the thesis presented in its title, he meant, among other things (and perhaps primarily) that "The opinion poll is, at the present time, an instrument of political action; its most important function is perhaps to impose the illusion that a public opinion exists, and that it is simply the sum of a number of individual opinions. It imposes the idea for instance that in any given assembly of people there can be found a public opinion, which would be something like the average of all the opinions or the average opinion. The "public opinion" which is stated on the front page of the newspapers in terms of percent (60% of the French are in favour of .... ) is a pure and simple artefact whose function is to conceal the fact that the state of opinion at any given moment is a system of forces, of tensions, and that there is nothing more inadequate than a percentage to represent the state of opinion... Today, the equivalent of 'God is with us' is 'public opinion is with us.' This is the fundamental effect of the opinion poll: it creates the idea that a unanimous public opinion exists in order to legitimate a policy, and strengthen the relations of force upon which it is based or make it possible" [Bourdieu, 1993: pp. 163–164].

As W. Ossowski notes, the idea of public opinion, which P. Bourdieu writes about, is characteristic not of sociologists, but of pollsters: "For pollsters, according to their vision of society as a mass society in a nominalist spirit, the object of public opinion is a certain set of individuals or their categories that were covered by the sample, and public opinion itself is the sum of individual opinions... For a pollster, the respondent's answers are components and elements of public opinion" [Ossowski, 2005:pp. 12, 17].

Therefore, respondents' answers to the questions of a questionnaire are interpreted as "individual opinions" of respondents. Meanwhile, it is obvious to researchers of public opinion that the respondents' answers, for various reasons, do not always reflect their true position (opinion, attitude, stand, motives) on a specific issue. And the key question is how often this happens in a particular situation. As O. Myagkov writes, depending on how sincere the verbal messages ("self-reports") of respondents turn out to be, "a lot depends for the reliability and quality of the final data" [Myagkov, 2012: p. 5]. Otherwise, the desire to answer not the way the respondent actually thinks but "as required" can reach such a scale that it may cast doubt on the reliability of survey results.

Most often, conditions for this arise in monocentric (authoritarian, totalitarian) societies. W. Ossowski notes that "in a monocentric society, the public sphere is controlled by the centre in three aspects:

• with the official semantic code, within which the communication process takes place, using ritual symbols;
• with the law, more precisely, subordination to the ruling centre, which, thanks to its dominant position, can almost arbitrarily legalize every decision and term any social action independent of the centre illegal;
• coercively, by enforcing loyalty of individuals and communities, which lasts as long as a threat of coercion exists.

It is quite obvious that direct and "legal" coercion create extremely negative conditions for the manifestation of public opinion on the problems of public life. No less negative is the impact on public opinion of the fact that the centre is dominated by a supra-semantic code (language) used in public life. In particular, "the symbolism of mandatory support for power structures is built into the official semantic code, which, as long as the centre dominates over law and has sufficient coercive power, successfully fulfils the function of maintaining the existing social system. True public opinion is blocked due to the lack of freedom and pluralism in the sphere of information exchange and assessment, and attempts to break through the boundaries of the official semantic code by publicly expressing opinions or information about facts not represented in the official code are interpreted by the authorities as deviant behaviour, social pathology, even a state crime" [Ossowski, 2005: pp. 57–59].

If one point of view dominates the information field when covering political events, then it becomes the "generally accepted norm" of answering questions about these events, that is, it is reproduced by respondents as a "normative answer". This, in particular, was observed in the times of the USSR, when polls on political topics did not make sense, since their results mainly reflected the rhetoric of the Vremya news program.

In a monocentric society, a respondent often sees researchers as representatives of the government: "In case of a respondent's distrust of the institution represented by the researcher, when the issues of his interests are related to public life, the answers can be determined according to the principle: "What do you want to hear?" [Ossowski, 2005: pp. 69–70].

According to the Levada Center poll conducted in April 2022 [Levada Center, 2022], only a bit more than half (54%) of the polled Russians trusted the data of sociological surveys on political topics in Russia. Confidence in the results of such polls strongly correlates with the political position of the respondents. Say, among those who approve of Putin's activities, 60% trust the data of polls on political topics, among those who do not approve of his activities, only 30% trust the poll results.

Of course, even in the most democratic countries the public opinion depends on mass media, as there are people who do not dare express an opinion during polls, as they feel that it may be condemned by the authorities or the people around them.

In addition, in a situation of military conflict (like the one is Ukraine now), intolerance to the enemy and any support for it is expected to grow. No matter how democratic Great Britain was, during World War II, few citizens dared openly support Hitler's Germany or, let's say, the Nazi ideology.

Distortion of survey results caused by respondents’ insincerity or fear of possible condemnation of their answers largely depends on the subject of the research. As methodological experiments show, answering sensitive questions, actual information obtained from respondents, as compared to unbiased data, may differ from the latter by 15–45% [Davydov, Davydova, 1992: p. 5].

The insincerity of the respondents is also facilitated by the doubts of a significant part of them in the anonymity of the poll. Although the interviewer assures the respondent that his answers will be used only in a generalized form together with the answers of other respondents and will not become known to other people and the authorities, not all respondents believe these promises. And the less the respondent's "subjective feeling of anonymity" is ensured, the more insincere answers should be expected.

This is confirmed by the results of a methodical experiment carried out by Razumkov Centre in October 2012, when the electoral attitudes of Ukrainian citizens were simultaneously studied by street and home polls. According to the latter, the electoral support of respondents for the ruling Party of Regions turned out to be overestimated compared to the official election results, and support for the opposition Svoboda party, on the contrary, was understated. Instead, the results of the street poll made it possible to predict the election results much more accurately. The difference between the survey data and the official election results (for all parties) in the street survey appeared 9.2% smaller than in the home survey.

After the beginning of Russia's full-scale aggression against Ukraine, polls in Ukraine show a significant increase in patriotic sentiments and a negative attitude towards Russia and its political leadership (which was quite bad even before this aggression). However, both journalists and sociologists suggest that polls may slightly overestimate these sentiments, as "hidden" sympathizers of Putin may not dare openly express their opinion. As V. Paniotto points out, "sincerity of answers decreases, because people who, for example, support Russia, will more often refuse to communicate" (quoted from: [Chernysh, 2022]). Therefore, there may be systemic bias in the sample structure.

Expert poll results

In this regard, an expert survey was conducted in July 2022 on the Facebook page of the group "Academic Sociology: Observations, Measurements and Conceptualizations", uniting many Ukrainian professional sociologists. Its participants were asked to what extent the results of sociological surveys (polls) can be influenced by the tendency of respondents to give "conformist" answers due to the fear of expressing views.

The group of experts who took part in the survey was formed on an initiative of the Facebook group participants. Different questions were answered by 29 to 43 experts. The author is grateful to the Facebook group "Academic Sociology: Observations, Measurements and Conceptualizations" moderator, Doctor of Sociological Sciences Serhii Dembitsky for his help with the expert survey.

Experts were asked to assess to what extent survey results are influenced in various situations:

• surveys in the regions of Ukraine controlled by the Ukrainian authorities;
• surveys conducted by Ukrainian research institutions in the territories of Ukraine temporarily occupied by Russia;
• surveys on the territories of Ukraine temporarily occupied by Russia, conducted by Russian research institutions;
• and finally, surveys conducted on the territory of Russia.

Of all these situations, according to experts, conformism is less frequent in the case of polls in Ukrainian regions controlled by the government of Ukraine: 14% of experts believe that there is no bias in poll results due to the fear to express one's views, 30 % of experts say that "conformist" answers shift the poll results by no more than 10%, 28% — by no more than 20%. 16% of experts believe that such a shift in results may exceed 20% (see Table 1).

Table 1

Distribution of experts' answers to the question: "Respondents' answers regarding politically sensitive topics may be influenced by their willingness to give "conformist" answers, due to the fear of expressing views that may be punished by the authorities, which may force them to express opinions that do not reflect their true views. We ask you to assess to what extent this affects public opinion poll results in different situations.

If we talk about the regions of Ukraine that are controlled by the Government of Ukraine, estimate by how many percent the above factor is likely to shift the results of sociological surveys (among all respondents) towards a "conformist position" with the authorities?", n* = 43

* Number of experts who answered the question.

The situation with polls in the occupied territories and in Russia looks much more dramatic. If we talk about surveys in the territories seized by Russia, when these polls are conducted by Ukrainian research institutions, the majority (56%) of experts believe that "conformist" responses caused by the fear of expressing views that may be punished by the occupation authorities shift the poll results by up to 30% or more (and only 13% believe that such a deviation will not exceed 10%, see table 2).

Table 2

Distribution of experts' answers to the question: "If we talk about the occupation of Ukraine’s regions by Russia, estimate, by how many percent the above factor is likely to shift the results of sociological surveys towards a "conformist position" with the Russian occupational authorities, if the survey is conducted by Ukrainian research institutions? , n = 39

If polls in the occupied territories of Ukraine are conducted by Russian sociologists, then, according to 61% of experts, “conformist” responses caused by fear of expressing views that may be punished by the occupiers shift the poll results by 50% or more (see table 3). This totally calls into question the reliability of the results of any polls in the occupied territories, especially if they are conducted by Russian sociologists, since these results may reflect the real situation "just the opposite".

As the correspondent of the Russian media resource "Proekt" K. Arenina writes, "last spring, the Russian Public Opinion Research Center twice, in late April and late May, conducted polls for the Kremlin in the cities and villages of Donetsk region recently seized by Russia, one of the organizers of these polls and confirmed an interlocutor close to the Kremlin leadership said. People were asked how they feel about the "special operation", Vladimir Putin and Volodymyr Zelenskyy, whether they want their city to become part of Russia or the “DPR”, remain part of Ukraine or become independent. The results have not been published, but according to the survey organizer, most people said they wanted to be with Russia. It is clear why this happened — in Volnovakha, Mangush, Volodarskyi and Bezymenny, the interviewers worked in temporary accommodation centres for displaced persons. In Mariupol, they were interviewed at the end of April only at the centre of delivery of humanitarian aid from "United Russia" [Arenina, Rubin, Badanin, 2022].

Table 3

Distribution of experts' answers to the question: "If we talk about the regions of Ukraine occupied by Russia, estimate, by how many percent the above factor is likely to shift the results of sociological surveys towards the "conformist position" with the Russian occupational authorities in a situation where the survey is conducted by Russian research institutions?", n = 33

Surveys cannot be conducted in the occupied territories, Russian sociologist O. Konyeva says: "Combination of demonstrative humanitarian actions with brutal repression causes a feeling of total unpredictability. Frightened citizens will give "safe" answers, opponents of the war will give "correct" answers, and at night they will paint graffiti against the occupiers." The answers received are simply unreliable, but, O. Konyeva notes, the Russian authorities do not care — the results will be used primarily for reports to Moscow and for propaganda purposes: "We are welcome here" (cited by: [Arenina, Rubin, Badanin, 2022]).

And when we talk about surveys of Russian sociologists on the territory of Russia and estimate by how many percent the survey results can be shifted towards the "conformist position" with the Russian authorities, expert are divided as follows: 41% of experts believe that the results will change by 20–30%, 45% — up to 40% or more (see Table 4). That is, the situation is slightly better than when Russian sociologists conduct surveys in the occupied territories of Ukraine. 

Table 4

Distribution of experts' answers to the question: "If we talk about the territory of Russia, estimate by how many percent this factor may shift the results of sociological surveys towards a "conformist position" with the Russian authorities? » , n = 29

This is not surprising, considering the socio-psychological atmosphere in Russian society, whose social practices are increasingly reminiscent of Stalin's times. Open expression of views, especially about the "special operation", can result in imprisonment. For example, as the Moskovsky Komsomolets newspaper reported, "a resident of the village of Gorki-10, Odintsovsky District, Moscow Region, filed a complaint against his wife with the police. ... A 37-year-old man asked to take measures against his wife. He reported that the woman expressed anti-Russian sentiments related to the SVO and set up their common child against the government. An investigation is being conducted" [Zhitel Podmoskovya, 2022]. As Meduza reports, "schoolchildren in Penza reported on a teacher who spoke about the war in Ukraine. A criminal case was brought against the teacher based on the article about "fakes" [V Penze, 2022].

As the Russian sociologist O. Myagkov writes, "recently, our country has seen a clear tendency to a decrease in the honesty of people in interpersonal and social communication, as well as in sociological surveys. The degree of normative support for lies has also grown significantly... Our studies show that almost 64% of respondents are ready to lie in order to protect themselves from obsessive and excessive curiosity on the part of other people, about 54% — to keep their thoughts and plans secret, 34% — to hide their flaws, mistakes and miscalculations from others, 11% of respondents would, according to their words, lie and cheat if they were paid well for it. On the other hand, almost every fourth respondent (24%) is sure that it is quite acceptable and permissible to answer a sociologist's question insincerely (in a questionnaire or interview)" [Myagkov, 2012: pp. 5–6].

According to the Levada Center survey results, the majority of Russians are now afraid to express their opinion during these surveys [Levada Center, 2022]. When respondents assess the position of the "majority of citizens", they, as a rule, attribute themselves to that "majority" (as a result of the "psychological projection", used by sociologists as a methodical technique to identify the opinions of respondents, which they try to hide from others or even from themselves).

Quite expectedly, the fear of the "majority of citizens" (and actually their own) is more often expressed by opposition-minded respondents — among those who do not approve of the actions of the Russian state leadership this share reaches 75%. Even among those who approve of the state leadership actions, the majority (52%) also suggested that the majority of Russians are afraid to express their position during public opinion polls.

As the Russian sociologist V. Zvonovsky writes, "the measurement of the attitude to the war is undoubtedly influenced by the fear of its opponents to speak out clearly and consistently. The influence exists, is empirically proven, and its indicators are statistically significant. However, it is not possible for responsible researchers to directly calculate the "real" share of supporters of the war on this basis" [Zvonovsky, 2022].

Conclusions

Therefore, the results of the analysis of both pollster studies and of the above expert survey indicate that during public opinion polls, there is always a problem of insincerity of respondents' answers, and a systemic error caused by the bias of the sample towards respondents "loyal to the government" or those who attribute themselves to the "ideological majority" is significantly aggravated in the conditions of growing authoritarian trends in society or the atmosphere of intolerance, in particular, caused by the war. Under certain circumstances, these factors can significantly reduce the reliability of survey data.

At the same time, the analysis makes it impossible to quantify the shifts in survey results caused by the mentioned factors. One of the reasons is that the magnitude of such a shift will vary significantly, dependent on changes in the social and psychological atmosphere in society, the subject of research, the questions and their sensitivity.

https://razumkov.org.ua/statti/gromadska-dumka-isnuie-ale-chy-zavzhdy-ii-mozhna-vyiavyty