Opinion polling behind and across the Iron Curtain

From bipolarity to multidimensionality

In 1957 the West German polling institute Infratest caused a stir with a new study on the life of the East German population behind the – at that time still pretty porous – Iron Curtain. From the start, this survey was quite controversial. Up until then, information from migrants was collected by the Western secret services and the West German Ministry for All-German Issues only in relation to the political and military situation in East Germany. Feeling uncomfortable about these obviously biased and one-sided reports, the press office of the ministry commissioned Infratest to carry out a survey based on interviews with migrants from East Germany in the refugee transit camps, thereby questioning the images produced through the official information channels.

This shift towards a systematized, systematic and empirical, or ‘scientific’, approach was part of the boom of polling as a ‘modern’ way of collecting information for political purposes like the design of election campaigns. Infratest was one of the major companies in the field, but not the most favored by the conservative government in West Germany (which was the Allensbach Institute). Infratest’s major field of business was polls on media usage for the public broadcasting services (see Kruke, 2007; Rüß, 1973).

The new survey program was not well received by the rank and file of the West German ministry, because this institution was predominantly staffed by Germans who had at some point fled from the East, and who felt that they could better rely on their first-hand experiences intuitively for evaluating East German life and mentalities (Rüß, 1973: 89; cf. Creuzberger, 2008).

The head of the Infratest institute, Wolfgang Ernst, responded to the lukewarm reception of ministry staff with the characteristic self-confidence of his young and booming profession, positioning opinion polls as a ‘policy advice science’ based on ‘reality’ instead of ‘ideology’. In the foreword to the first study on industry workers in East Germany he argued that the alleged familiarity of West Germans with the living conditions in the East was

… in most cases based on knowledge collected randomly and shaped by the personality of the reporting person … This study is the first attempt to draw a systematic, methodologically well-based and controllable image of a whole layer of the population under the rule of the zone regime. (Graf Blücher, 1959: v) ¹

The subsequent surveys on white-collar employees, the intelligentsia and youth arrived at equally unexpected and controversial results (Infratest, 1960; Schröter, 1959). The methods employed and findings, at this point, stood far from mainstream communist studies, which were dominated by militant anti-communism and an extremely polarized understanding that distinguished only between the rulers and the oppressed population and for which assessing potentials of resistance took center stage. Communist studies bit by bit began to question the presuppositions of US liberation policy (cf. Stöver, 2002; Hüttmann, 2008).

In a process over the following 10 to 15 years, this new approach, favoring society as an analytical subject (rather than the system of communist rule) and using in particular opinion polling, gained importance in redrawing the image of socialism in West German communist studies and in rethinking the paradigms of decision-makers on which policy towards the East was made. ² Up until the 1970s the Infratest surveys in particular accompanied and shaped the new policy of ‘change by rapprochement’ of the West German Social Democrats under the future chancellor Willy Brandt.

The discovery of ‘the middle’ of society

The identification of middle opinion went along with a series of ‘discoveries’ in several areas: first, the perception and reconstruction of sub-groups within East German society; second, the differentiation of levels of observation beyond the traditional issue of pro and con attitudes towards communist rule; and finally an idea of the constellations and tensions under which East Germans positioned themselves in terms of value orientations and everyday life.

The decisive impulse in this regard was the successive shift in terms of the interpretation of ‘middle’ positions in the survey results. This shift modified questioning techniques, the range of issues asked about, and analysis methods. The first survey brought the insight that the attitude of ‘accommodation’ could not necessarily be interpreted as a form of ‘standby readiness for an uprising’, as had been the common-sense perception in Western circles during the 1950s. The question of which reasons and with which underlying motivations East Germans ‘accommodated’ the regime became the central driving force of the survey programs that followed. For instance, in the intelligentsia survey the Infratest pollsters noted:

When we conclude our observations and interpretations on the general ideological attitude amongst members of the intelligentsia in the Soviet Zone, the scale – taking into account all given differences – inclines towards the moderate middle group of attitudes. (Infratest, 1960: 53)

Moreover they claimed:

One does not primarily reject communism for ideological but for practical reasons; at the same time practical reasons and not a ‘fascination’ with ideology allowed a minority to become full supporters of communism. For the broad majority of interviewees, the reality and the subjectively perceived opportunity or inopportunity to realize certain aspects of Marxism–Leninism was the key criterion. Their thinking is pragmatic and not based on ideology. (Infratest, 1960: 78)

Not by accident the title of the survey was ‘A General Overview on the Stage of Accommodation of the East Berlin Population under the SED Regime’ (Infratest, 1964: 21). ³ East Berlin residents were differentiated into 6 categories as regards their political position in favor of or against the communist system. Besides one group with ‘strong sympathy’ for the communist system (type A: 3%) and an opposite group of ‘outspoken enemies’ (type E: 4%), Infratest detected a large ‘middle’ ground. The middle ground covered three distinct categories. The first was generally in favor of the regime, but did voice criticism or reservations (type B: 12%). The second and largest overall group was dubbed as ideologically distant, but completely accommodated in its visible behavior (type C: 31%). The third group rejected the regime and tended not to hide its discontent, but had adopted a position of ‘inner emigration’ (type D: 15%). Finally, a sixth group was identified, which was politically totally uninterested and indifferent to the eastern as well as to the Western system. At 29 per cent (type F), this group in particular was surprisingly large, given the traditional search for friend and foe. A residual group of ‘not classified’ (6%) rounded off the survey (ibid.: 2).

Thus, Infratest had a number of messages for its clients in the West Berlin town hall. The pollsters showed that the number of explicit active supporters was pretty limited at around 15 per cent, but the declared opponents were in the minority as well. A majority accommodated the regime while keeping an inner distance or proved to be totally uninterested in the world of politics, irrespective of whether the eastern or Western version. With the differentiation of demographic and occupational groups, Infratest at the same time drew more precise profiles of the several attitude types. They showed that younger East Berliners in higher professional positions, in particular, inclined to more system-loyal attitudes, that workers took a more accommodated stance and those without employment, meaning pensioners and housewives, were the most uninterested in terms of politics.

Such findings unsurprisingly caused a certain level of dissonance at the interpretation stage. Infratest rhetorically followed the friend–foe pattern of totalitarianism, but at the same time had the challenging task of drawing general conclusions from its data, which diverged from this model. Somewhat inconsistently, it declared that in the ‘totalitarian state’ everyone was pressed into ‘societal-political collaboration’ and interpreted the ability of large parts of the population to position themselves as indifferent or uninterested in politics (which clearly was quite the opposite to this totalitarian view) as an act of hidden resistance, yet at the same time also as a success for the regime:

To exclude themselves and to stay apart is a demonstration against the absolutist claim of the state. Even amongst the younger and middle generations there were people who showed no interest in politics at all or any kind of reaction to political-societal issues. This attitude could be the result of giving up resistance. To immunize against all kinds of trouble, this could lead to a total encapsulation – to a kind of political ‘hibernation’ … This kind of opposition is not explicit, but the passivity and lack of interest has to be evaluated at least as an insufficient readiness for collaboration with the regime. (Infratest, 1964: 28–9)

OPEN IN VIEWER

Table 1.

Infratest grouping of the political attitudes of East Berliners, 1963 (in %).

	Supporters (A + B)	Accommodated (C)	Opponents (D + E)	Politically not interested (F)	Not classified (%)
Total	15	31	19	29	6
Under 34	30	12	25	18	15
35 to 59	14	40	23	20	3
Over 59	9	24	9	46	12
White collar/Intelligentsia	23	42	19	13	3
Workers	13	37	21	16	13
No profession	2	19	18	56	5

Infratest, 1964: 23

Nevertheless, Infratest had a sense of the brittleness of the alleged hidden pro-western attitude among the middle groups. The allure and the pressure of the communist grand ideology were perceived as the major risk for the further development of East Germany:

Under the coercion of totalitarianism a part of the majority pays ideological lip-service [to the regime] while maintaining inner reservations. The mental burden of such kind of double thinking could in the long run lead to inner crisis, causing a transformation from a mock attitude to real conviction. (Infratest, 1964: 112)

Beyond these attempts to frame the discovery of broad currents of opinion into traditional categories, the 1963 survey paved the way for a broader range of issues to be addressed in future research. Besides the classical questions about political positions, Infratest was interested in the material circumstances of life in East Berlin and opened the field to collect data on standards of consumption and living. The results showed some modest improvements compared with the 1950s, not only in terms of the objective quantity of goods available, but in respect of the level of satisfaction expressed by the population as well.

Another 5 years later Infratest was authorized to start a new survey program under the title ‘Attitudes and Patterns of Behavior of the Population in the GDR’, after the Social Democrats entered the federal government in 1966 and were allocated the Ministry for All-German Issues. The aim of this program was deliberately kept broad: to investigate ‘attitudes and opinions of GDR citizens on certain political, economic and social issues’ through annual surveys (Infratest, 1968–89[1968]: 2).

Indeed, the program had a far broader and more open design than the 1963 scheme. The very first presentation of results in 1968 started with a chapter evaluating material circumstances and the receptiveness of East Germans to practical ‘achievements’ of the socialist system (Infratest, 1968–89[1968]: 14). The political attitudes proved to be more or less similar to the earlier surveys, distributed around the ‘accommodated middle’, with somewhat shifted emphases. In the 1969 survey 7 per cent were dubbed as ‘absolute supporters’, 21 per cent as ‘critical supporters’, 54 per cent as ‘accommodated’, 7 per cent as ‘opponents’ and 11 per cent as ‘not interested in politics’ (ibid.[1969]: 66). While the fear in the West of an ‘ideological’ identification with the regime shifted into the background, Infratest now began to recognize that a determining factor for this broad middle of society was the perceived practical advantages and weaknesses afforded by each system to the individual citizen. Infratest developed a large set of questions that directly compared perceptions of East German opportunities with those in the Western system in various areas such as politics, consumption and educational opportunities.

Through this lens, Infratest now concluded that the communist system was indeed quite well received, especially by the younger population. The educational system and the promotion of youth were ranked highly in the survey. According to Infratest, 37 per cent of East Germans held the GDR education system to be better than that in the Federal Republic. For the category ‘promotion of youth’, the rate was 28 per cent (Infratest, 1968–89[1969]: 68–9). Such findings of course diametrically contradicted the presupposition that a ‘totalitarian’ system generated compliance by force alone. Accordingly Infratest was alarmed and explicitly noted:

Astoundingly, a comparison of the statements of ‘supporters’ and ‘opponents’ shows no major differences … Those areas of public life which lie in a more apolitical sphere may be regarded positively by opponents of the regime as well … In conclusion one can say that the interlocutors in the GDR expressed a certain pride in some issues. (Infratest, 1968–89[1969]: 68–9)

As the following results show, this conclusion somewhat exaggerated the real situation, because in other areas, such as the supply of consumer goods or intellectual and cultural freedom, the West still maintained a fairly hegemonic position. Moreover, there was still a strong desire for reunification and especially for an opening of the border to West Germany (which was practically closed to East Germans since the wall was built), as the answers to questions about the current policy of the West German government towards the GDR showed. But it was obvious too that the views of the wide ‘accommodated’ middle group and few active opponents needed further elaboration.

This was confirmed by the results on the information resources available to East Germans, which proved to be astonishingly diverse. Western broadcasting and television played major roles, but personal experiences in the workplace and at home proved to be even more important for the observed judgements of East Germans, while all kinds of propaganda were only of minor relevance (on media usage see Meyen, 2003).

With the launch of the annual survey waves Infratest began to systematize its interview program: it established a trend analysis over several years, tested topics for their relevance and thereby successively differentiated its images of East German society. This process finally culminated in a complex typology of attitudes and value orientations, which was used for the first time in the analysis of survey results in 1973. As the pollsters explicitly claimed, this novel analysis should serve ‘to critically re-evaluate the information value of the former “political” division of GDR residents (into supporters, non-supporters, politically interested, etc.)’ (Infratest, 1968–89[1973]: II, 2).

From now on the Infratest surveys provided a multivariate analysis of the situation in East Germany. A set of 8 questions, including the degree of identification with the GDR system, a general assessment of living conditions, and the degree of political activity, was combined with direct comparisons of the performance of the West German system and the East German system in terms of income, working conditions and personal outlook for the future (Infratest, 1968–89[1973]: II, 4).

Using a method of computerized type building, Infratest opted for a total of 5 types among East German society. This typology in some respects continued the distinctions developed in past surveys. Two types of attitude were again identified as the exponents of strong ‘ideological’ identification with the eastern system (type C: 14%) and the Western system (type B: 19%), both showing a high level of political interest. And the politically totally uninterested (type A: 19%) were found again. But the largest group was formed of type D (29%) that showed a moderate political interest with a pragmatic and selective, but rather pro-western, stance, and a slightly smaller type E (19%), also moderately interested in politics and pragmatic in its value orientations, but with rather pro-eastern or neutralist positions.

Infratest in turn assigned these types to specific political and social profiles: for example, type C was most frequently found among young and middle-aged East Germans with higher education and professional positions. It included – not very surprisingly – the largest proportion of members of the Communist Party. In type A (non-political), women, pensioners and people with low education and occupational jobs were overrepresented.

Infratest itself highlighted two findings in the 1973 survey, which stood for the final farewell to the ‘ideological’ Cold War perspective of the 1950s. First, the polling institute stressed explicitly that type B persons, i.e. the decidedly pro-western GDR citizens, were no longer to be considered as representatives of direct resistance: ‘A group of declared enemies of the system of the GDR, defined just as clearly as the group of loyal supporters of the GDR system, cannot be observed even among the persons of this type’ (Infratest, 1968–89[1973]: II, 26).

The second decisive finding was that despite a strong sympathy for Western system features there could be no talk of a ‘critical mass’ of potential protestors among the total population:

Overall, a larger number of GDR citizens – recruited mainly from people attributable to type D and part of type B – showed a significant affinity for the societal system of the Federal Republic, but the relationship to their own system is quite differentiated, i.e. it is at least not marked by a pronounced dissatisfaction or frustration. (Infratest, 1968–89[1973]: II, 27)

Thus, the assumption of ‘rebels in waiting’ was finally off the table and replaced by an image focused on the silent and pragmatic ‘middle’ of society, responsive to concrete tangible and intangible benefits.

The typology developed in 1973 was used continuously by Infratest until 1989 as an analytic tool. The long-term trends showed on the one hand some shifts between types: type C became smaller and smaller and type E disappeared completely. On the other hand Infratest used the typology for profiling through socio-demographic categories (gender, age, education, etc.) for the evaluation of other issues, which facilitated a relatively complex picture on special issues, such as attitudes towards the Soviet Union, or the judgements of political initiatives of the West German government towards East Germany (extracts of the annual reports were published in Holtmann and Köhler, 2015).

The history of the Infratest surveys from 1957 to 1973 shows very clearly how the initial establishment of the survey method represented a striking break in an anything but accessible political and public environment. With its programmatic orientation towards empirical sociology and concrete behavior analysis, it stood by definition for a departure from classic political science approaches and the political implications of communist studies, which focused overwhelmingly on the actions of the communist leadership, practices of repression and the potency of ideology in the Cold War. With the establishment of sociological and psychological approaches it prioritized a view ‘from below’, instead of emphasizing the state system of rule. On the level of research design and the production of images of society, the ‘discovery’ of the accommodated middle triggered diversification and differentiation and gradually led to a multi-dimensional analysis. This analysis included not only political identification in the rivalry between East and West, but also the general interest or lack of interest in politics, and the juxtaposition of major political ideas and pragmatic orientations in life. Moreover, with its image of socialist states as complex societies, it paved the way for a new view on the ‘ruling class’ as a subject of sociological analysis. Thus, it also allowed a more nuanced picture of the Communist Party and its supporters themselves.

The craft of opinion polling between politics and academic sociology

One can tell the history of the Infratest program in a second way – as the history of opinion polling as a ‘democracy science’ between academic sociology and policy advice (see Kruke, 2009).

As mentioned above, its success was triggered by a sense of unease about the prevailing images of communism as a social order and the need to design new strategies for dealing with the East beyond the deadlock of ‘liberation policy’. It was deemed necessary to get a more accurate and reliable feedback tool by which the degree of ‘alienation’ of East Germans from West Germany, and the success of communist efforts to gain legitimacy, could be assessed. This desire went along with the rise of opinion polling as a tool for politicians in general. It offered the promise of replacing intuition with ‘scientific’, quantified knowledge through a consistent focus on systematic and empirical methods. It was expected that this form of knowledge would make the overall uncertainty in the observation of whole societies manageable and thereby create robust data. The presupposition was that communist societies are so complex that they require such a kind of second-order observation as was already an inherent part of the working methods of opinion polling.

Such a role as a science-based instrument of policy advice became prevalent in the late 1950s and early 1960s in the federal republic as part of a ‘second wave’ of westernization (after the original institutionalization of democratic procedures by the Western occupation powers). Infratest was one of two major polling institutes which advised the Social Democrats, at that time the opposition party, on US-style election campaigning. The decisive factor here was a network of social scientists such as the American Harold Hurwitz, who established the systematic use of opinion polls while part of the mayor’s office in West Berlin, Willy Brandt, the former East German communist functionary, his political adviser Leo Bauer, and the director of Infratest Wolfgang Ernst, who offered his services to the Social Democrats as a consultant for electoral campaign strategies (Kruke, 2007; Hurwitz, 1999).

Due to the central role of the policy towards East Germany in Brandt’s overall election campaign, there was a particularly strong sense of the need for empirically based data about the GDR, and indeed for information that went beyond the stereotypes maintained in the established circles of militant anti-communists (a group that often possessed a more or less pronounced biographical background in the Third Reich; see Creuzberger, 2008). In this spirit, Infratest positioned itself from the outset as explicitly providing a ‘factual-critical’ perspective on the GDR as a modern industrial society with growing prosperity. Such an image gained importance at this time in the sphere of public support for the Ostpolitik of the federal republic as well as in the academic field with sociologist Peter Christian Ludz as the major advisor of the social-liberal administration under Chancellor Willy Brandt.

In this context Infratest emphasized the superior potential of polling methods:

Even such an outstanding expert as Hanns Werner Schwarze [a journalist who had written a well-known and highly contentious book on everyday life in East Germany] had no ability to give quantitative-representative information about the attitudes of the GDR population. For this reason, in our present and future research it should be possible to come to more accurate results. (Infratest, 1968–89[1969]: 6)

Infratest even emphasized the ‘avant-garde’ element of such a perspective by occasionally hinting at the backwardness of the anti-communist West German population, which would to a certain degree taint the results of questioning West Germans (Infratest, 1968–89[1969]: 17).

The cult of ‘representativeness’, i.e. the transformation of mathematical and statistical methods into valid statements on state socialist society, was, as elsewhere in opinion polls, at the real core of Infratest’s self-representation as a provider of scientific knowledge.

However, procedure-based and quantitative representativeness in this sense was in fact more a normative idea than a reality in the surveys. Infratest began in 1957 with 80 respondents whose statements were actually analysed mostly ‘monographically’ (as the authors of the survey dubbed it), meaning qualitatively. Only gradually was the sample size increased to several hundred respondents and finally more than 1,000. With each survey Infratest announced a validation promise that in the next survey the sample would be extended. From 1970 onwards, the data were also at least roughly weighted based on the criteria of age and gender in line with official East German population data.

From the perspective of empirical sociology, however, the survey methods themselves were the real bottle-neck in respect of the validity of the results, because both the refugee surveys and also the proxy method in the technical sense definitely did not meet the requirements of mathematical accuracy usually proclaimed by polling institutes. The Infratest surveys on the GDR always remained a number-based, but ultimately qualitative, method.

Infratest itself took an ambivalent position on the rigorousness and accuracy of its data-collecting in this area. On the one hand, the pollsters were honest and professional enough to identify the limitations of their method. Regularly their reports included a sentence stating that the results were not statistically representative and could only indicate trends and relative distributions. Especially in the first years of the annual program, Infratest paid some attention to the documentation of statistical cross-checks. At the same time it was clear that these tests were biased by the interest of the polling institute to continue the program and successfully promote the use of the proxy method. In this respect, the critical revision of methodology to a large extent was rhetorical. At the end of the day the pollsters always relied on the ‘reasonableness’ of their findings and did not get beyond the point of internal consistency and compliance with external assumptions.

For instance, the introduction to the first proxy survey of 1963 with 101 interviews stated:

We could not rely on experience for this experiment, but had to start incrementally from scratch to solve this problem. Thus, it was – besides the results on the respective issue of investigation – one of the main goals of the survey to develop interview methods which could allow us to get optimal results – under similar preconditions – in the future from interviews with West Berliners about attitudes of inhabitants of the ‘GDR’ … After a critical examination of the results and a detailed comparison with previous studies, we believe we can say that the experiment was successful in principle. (Infratest, 1964: 7)

In 1973 the ministerial client ordered an evaluation of the survey method by the Central Archive for Empirical Social Research at the University of Cologne, presumably at the recommendation of academic social scientists advising the ministry. The evaluators, who were trained sociologists, acknowledged that Infratest ‘had embarked on a basically interesting path’ to determine attitudes and value orientations in East Germany. But they questioned the method of retrospective interviews with ‘naive observers’ and came to the conclusion that possibly hard facts could be surveyed in this way, but not attitudes (Herz, 1974: 11). Furthermore, they highlighted some weaknesses in the selection of interviewees and other procedures. Nevertheless, the academic institute stated that the data could be quite valuable for trend comparisons (which indeed make them interesting as a historical source for the history of GDR society even today).

Both the critical remarks and the suggestions for improvement in the evaluation report did not impress Infratest very much. The pollsters particularly avoided complex and thus expensive selection and control procedures. Remarkably, it defended its methods by referring precisely to those factors that the survey method should overcome: it declared that the interview techniques and the interpretation of results were embedded in an environment of qualitative assessment and ‘empathy’ with the subject, based, among others, on the personal experiences of the program heads as former East Germans and visitors to East Germany, and group discussions among West German respondents. Program head Anne Köhler had left East Germany after having been imprisoned for political reasons in the late 1950s. She herself embodied the mixture of proximity to the East German population and pollster professionalism. In cooperation with the ministerial staff she suggested issues to be addressed and designed questionnaires. ⁴ Thus, ultimately the inherent plausibility and persuasiveness of their differentiation and trend descriptions remained the decisive criteria for validation, not sociological research standards (ibid.: personal communication).

In other words, Infratest benefited with its survey program from the suggestive power of precise numbers and the appeal of ‘representativeness’ that lies at the very core of the legitimacy claims of the polling method. But the actual gain in knowledge and the potential impact on decision-making processes of the client owed more to the interest in new and unexpected results by researchers and the client, and, second, to the method’s departure from binary patterns (pro or anti the communist system) to differentiated options, culminating in the creation of the multi-level typology.

While there was some exchange between pollsters and academic sociologists about the methodology of the survey techniques, the significance of Infratest’s findings for conceptual debates within communist studies played a surprisingly small role in the report volumes, although they would have had a lot to contribute. As long as the reports were to be published, as was the case with the refugee surveys of the 1950s, theoretical reflections and references to sociological theory, and (at that time) to psychology, were extensive, because the pollsters were eager to earn the respect of their academic colleagues. These position statements did not reach the detail of the methodological and theoretical parts of the Harvard project, but were obviously designed for scholarly exchange and to make an impact on public debates. As mentioned above, references to various terms and works of totalitarianism theory can be found until the end of the 1960s. This theory provided a quasi-unquestionable ‘given’ point of reference, even if it offered little in terms of operationalization and the empirical evidence clearly contradicted it.

Starting with the classified survey of 1963, no elaborate discussion about alternative concepts such as theories of convergence or modern industrial societies took place in the Infratest reports. Obviously, these survey reports were designed around the expectations of the client, i.e. the political administration and ministerial bureaucracy, which had no specific interest in theoretical considerations. Due to the classified nature of their work, the Infratest researchers were observing the academic field of GDR and communist studies by, for example, attending relevant conferences as silent observers, but they had no opportunity until 1990 to feed their findings into the scientific community.

In this respect the inner structure and self-representation of the reports reflect, on the one hand, the sphere of clandestine observation of the enemy and, on the other hand, the rules of market research and strategic election campaign advice. In the context of the Cold War, opinion polling as a method of observation stood for the shift from confrontation and clandestine preparations for a hot or cold civil war towards system competition in the fields of political and cultural attractiveness and economic capabilities. This shift led to some uncertainty on the side of the Western client, as the ministry had to realize that the attitudes of the eastern population tended to be less clearly ‘pro-western’ than they had expected. But in the long run, the new instruments proved to be carefully targeted and directed at those spheres of competition in the Cold War where the West excelled in order to show the ‘victory of the West’, namely its political and cultural liberality and technological and economic superiority. The relationship between opinion polling and social sciences, between pollsters and social scientists, remained somewhat blurred: the theoretical concepts of communist studies that framed the scope and direction of the questionnaires and the interpretations showed some references to the shift from totalitarianism to concepts of modern industrial societies, but the pollsters at the same time emphasized the exclusive mixture of their findings and maintained a distance, supporting some paradigms of social science research on communism while contradicting others.

Build-up: Intentions and preconditions

On the other side of the Iron Curtain, in East Germany, the impulse for the establishment of empirical sociology and public opinion research came in the early 1960s. After the building of the Berlin Wall, which finally provided the desired stabilization of GDR statehood, the East German party leadership began to search for a new base of legitimacy in place of the establishment of party rule through violence and mobilization. Following the Soviet (and in part Polish) model in the wake of de-Stalinization, the East German party leaders opened the field for a group of social scientists to establish this discipline. Sociology as an empirical ‘reality science’ embarked on the search for a new ‘equilibrium’ between the objective historical laws of Marxist philosophy and the subjective consciousness of the ‘masses’ (see Reinecke, 2010).

Among other steps the path-breaking volume on public opinion by Soviet social psychologist Alexander K. Uledov was translated into German. Uledov assigned public opinion an important role in socialist society:

The relation between public opinion and party program is the relation between a basic state of consciousness and the highest. This does not mean that the influence of the former should be held in low esteem. Public opinion is a necessary link in the chain of tightly connected states of consciousness. No other state of consciousness can cope without it, not even the party program. (Uledow, 1964: 41)

In the course of liberalization in the field of economy, culture and youth policy, East German party leader Walter Ulbricht agreed to establish the Institute for Opinion Polling [Institut für Meinungsforschung beim ZK der SED (IfM)] with the aim of generating a ‘scientific-exact’ image of the moods of the masses in the GDR, as it was framed in the relevant Politburo resolution draft (cited in Niemann, 1993: 17). The party leadership advocated objective investigations of state socialist society. These would supersede an understanding shaped by situation reports by the party bureaucracy and the secret police as well as by the intuitive ‘empathy’ of experienced party cadres.

The new social sciences were part of the scientific enthusiasm of the time. Reporting the first large-scale survey of its readers, the party newspaper Neues Deutschland stated in 1964:

This survey is of high significance. It shows for the first time detailed information about the size, composition and range of interests of the community of ND-readers: Now can be gauged with a ‘tape measure’ what before was estimated more or less globally by rule of thumb. (Reuter and Wittkowski, 1964: 3)

Through this search for a new empiricism the discovery and recognition of societal diversity were initiated. Nevertheless, as the quote by Uledov indicates, the framework for opinion polling remained theoretically and practically precarious. The sociologists wanted to distance themselves from positivist ‘bourgeois’ concepts, but had to clarify the epistemological preconditions of accepting the ‘subjective factor’ by elicitation of empirical data with all its contingencies and potential deviance from Marxism–Leninism as a ‘scientific’ method of detecting objective laws of societal development.

The party leadership authorized this experiment, but only under the utmost restrictive conditions. Opinion polling was fashioned as another channel of regular internal ‘party information’. The Institute for Opinion Polling was subordinated directly to the Politburo and the secretariat for agitation and propaganda of the Central Committee. Karl Maron, a long-time party veteran and former Minister of the Interior, was appointed head of the institute. He had no professional background whatsoever in the field. The Politburo was entitled to set the research agenda, and results were to be reported to the central committee apparatus exclusively. Initially, the idea of publishing results was taken into consideration. It was up to the Politburo alone to decide about publications, and indeed it never published any survey during the whole 15 years of existence of the institute. ⁵

But the most sensitive question was the relationship of the institute to the lower ranks of party bureaucracy and their reporting of public moods– a reporting to which opinion polling would build an alternative channel. The whole interviewing apparatus was incorporated into the local party organizations. The interviewers were functionaries of the district party secretariats. This double role as reporters and pollsters was compensated only to a certain extent by the regulation that they should send the filled-out questionnaires directly to the institute, avoiding the usual path through the party hierarchy. Insofar, one can read the establishment of party opinion polling as a half-hearted attempt by the party leadership to distance itself from its own regional and local apparatus, which was suspected of glossing over unpopular findings and sanctioning those who might voice them in the questionnaires.

For this reason it was stressed in all concept papers and resolutions that the questionnaires had to be filled out strictly anonymously and that ‘nobody should be disadvantaged by his or her answers’. ⁶ Moreover, ‘suggestive questions are to be avoided in all cases to guarantee the psychological unbiasedness of the respondents’. ⁷ In training material for interviewer party functionaries circulated in June 1966 it was reiterated, clearly based on negative experiences, that anonymity was

… of crucial importance to the scientific quality of the survey results, particularly in the field of political problems and given that the survey should not only reproduce what was to be read in the newspapers day by day. Political opinion polling has to find ways that show each and every respondent, even the most mistrusting, that I can give my answer freely, without any risk of getting into trouble with state institutions or my employer. ⁸

Thus, the basic idea was to create a new sphere of open, ingenuous and anonymous elicitation and transmission of opinions and thereby to lift the relations between the regime and the population to a new stage.

This logic proved to be quite alien to the minds of party cadres. In the following years, the pollsters repeatedly inculcated local party officers with these principles and tried to get across the advantages of the polling method, which was dubbed by the latter as a dubious import from the West. Moreover, it happened more than once that uninformed party functionaries called the state security service to arrest institute employees conducting on-the-spot surveys. The party men could not imagine that someone else other than a Western spy (or a Western polling institute, which in their view would have been the same) would be interested in conducting such kind of information collection (Erxleben, 2013). ⁹

Reservations within the population were similarly strong. The pollsters were fully aware of the differences from the working conditions of Western polling institutes in this respect. In the above-mentioned training material from 1966 it was stated:

In socialist society the private and thereby ‘neutral’ veneer [of the polling institute] does not apply. The respondent associates all societal institutions with the state administration and the leading party. This gives rise to the thought for many people not to answer each question frankly. [This applies] particularly to those [questions], to which they held an opinion differing from socialist ideology and the official opinion [sic!]. It would be wrong to assume nowadays that the fear of consequences has been totally overcome … Examples have shown that some respondents cut out the printing number of the questionnaire, gave no socio-statistical information about themselves, wrote responses to open questions on a typewriter, or noted in the questionnaire that [they were] aware of the fact that fingerprints will be taken from it. (Interview training material, cited in Niemann, 1995: 28)

The clearly widespread impression that the surveys were just another variety of staged public performance, such as the forced attendance at elections, the mass demonstrations on Labor Day, or the signing of protest resolutions against imperialism, was amplified by the fact that the institute decided to conduct part of the surveys in ‘classroom situations’. For this they assembled all workers of a certain factory section in a room to let them fill out the questionnaires. This method was even praised as particularly successful and promising due to the high return rate.

To reduce distrust of the process and of polling in general, the institute in the first years placed articles in the local newspapers to announce their surveys and stress the importance of citizens’ opinions for the government. ¹⁰ Indeed, open reservations over filling out the forms or suspecting Western spies or pollsters of being behind the surveys decreased in the following years (Erxleben, 2013). ¹¹ On the other hand, the survey results indicated that large numbers of the respondents took the apparent anonymity of the questionnaires with a pinch of salt. In the outcome, some predictable patterns emerge. First, the results fell more into line on politically sensitive issues. Second, in all cases where it was possible to sidestep a question (by answering ‘don’t know’ or ‘do not wish to express my opinion on this issue’), the rates for this option were significantly high, while options that explicitly flouted the official line were mostly avoided. Third, the rates show a certain socio-political differentiation: the more distant the occupational position was from the party system, the larger the share of less loyal views. For instance, responses by older workers in the factory surveys proved to be less friendly towards the party than the average of the population (‘Report’, 1969).

The shaping of images of society under the conditions of ‘socialist representativity’

Besides all the mentioned restrictions and structural distortions, the social scientists involved in survey work took their task quite seriously: they were keen to learn what East Germans ‘really think’ and what they as researchers and loyal party members could do to improve communist policies and increase the legitimacy of the claim to leadership. The preconditions for such an attempt were actually not too bad, given that the data did not come from the usual ‘99 per cent in favor’ results of the election rituals and were also at least in part so differentiated that they invited further interpretations.

However, this open-mindedness toward complex and critical findings and the willingness to touch taboo subjects was significantly less pronounced among the party leadership than on the Western side. Indeed it exhibited some similarities to the reaction of the Roman Catholic Church to the Kinsey reports on the sexual behavior of Americans, namely an indignant insistence on the validity of a quasi supra-empirical truth (Ziemann, 2007: 147). It therefore remains unclear whether the pollsters believed their own analysis or just anticipated the expectations of the party leadership. In any case, they pursued a peculiar ‘middle way’ in interpretations.

By documenting more complex patterns of opinion-building and unfolding diverse subgroups with respective differences (a rudimentary mapping of social attitudes), they – as with their colleagues in West Germany – also aimed to overcome the bi-polar image of society in friend–foe categories. The existence of a broad ‘middle’ spread of differentiated positions between the polarized edges was stated in the institute’s reports.

However, this tentative openness of the party pollsters was negated by highly normative interpretations of the findings. Approval ratings of around 80 per cent for official SED policies were considered appropriate, while lower values were taken to indicate a deficit in party propaganda and influence on the ‘masses’. For instance, the IfM revealed in 1965 in one of its first surveys that 70 per cent of the East German population held the current eastern borders of Germany to be permanent, while 22 per cent expressed the view that the borders of 1937, including the lost Eastern Territories, should be restored, and 8 per cent chose not to answer the question. This was a broad majority for the official position, but the pollsters nevertheless commented:

The results of the survey on this issue are a serious signal in our assessment, especially when one considers that only two-thirds of all surveyed workers or only 61 percent of the rural population held the current borders to be permanent. (Survey Report, 1965, in Niemann, 1993: 77)

It was always the idealized workers who enjoyed special attention in opinion polling, since they consistently showed below-average approval rates of the official party line. Party boss Erich Honecker followed closely this pattern in his commentaries and annotations on the survey reports.

The explicit reference to such deficits in popular support and the problems with the ‘working class’ as a population group that was the least susceptible to ideological speech and the most distanced from party politics in particular, was offset by a general upward trend identified by the pollsters in their studies (and which found its parallel in the Infratest surveys). In a speech to party officials in 1973, IfM deputy director and actual head pollster Dieter Rückmann not only emphasized that the population supposedly showed a strong interest in politics (nearly 90 per cent), but that it was able ‘for the most part to deliver clear verdicts on political, economic and cultural issues’. Even in the direct competition with the West German system he saw East Germany on a triumphant path:

We have been asking for years and years the question which societal system the respondents approve of – the social conditions in West Germany, [or] the social conditions in East Germany. Since 1965 we ask this question every year anew. [In] 1965, we had an approximate value of around 68%. We have now reached a level of 81%. (Institut für Meinungsforschung, 1973: 6)

This upward trend in the results began to reverse around the mid-1970s. The polling institute recorded in detail the reactions among the population to the announcements of the IXth Party Congress of the SED in 1976 and highlighted the responsiveness of East Germans to social benefits, better housing, higher pensions and a good education system. The report on the institute’s survey for General Secretary Erich Honecker showed at the same time, however, that economic performance, opportunities for democratic participation and above all the lack of freedom to travel to the West were considered as definite weaknesses of the East German system (Niemann, 1993: 401). Despite all the differences in working conditions from Western opinion polling, the East German institute equally showed East German citizens to be selective and differentiated in their judgements and more receptive to tangible material or immaterial benefits than mass ideological campaigns.

The fate of opinion polling

In a sense, the party leadership under Honecker (as in other socialist states) proved to be quite responsive to such findings. His policy relied heavily on broad-scale improvements to social benefits and in consumption, including an extensive housing program. However, in the second half of the 1970s it became obvious, first, that this program was economically unsustainable, and, second, that the population’s expectations were heavily influenced by West German standards and could never be fulfilled.

Honecker finally responded to this inconvenient truth in January 1979 with the closure of the institute and the (fairly successful) order to destroy all its documents, with the lapidary justification that

… party information, complex investigations of the departments of the Central Committee and the results of social science research today provide essential information and insights for the scientific leadership of the party and the state … In respect of this the activities of a special Institute for Opinion Polling of the Central Committee of the SED prove no longer necessary. (Politburo, 1979)

During the previous years, working conditions had been increasingly restricted: the proportion of surveys on political issues had decreased, instead surveys about banal subjects such as holiday accommodation or cultural work in factories gained more space. In addition, the pollsters were not allowed to repeat questions over the years for trend diagnosis nor even provide substantial interpretative analysis of the collected data. To write such reports was the prerogative of the respective Central Committee departments to which the raw data were delivered. Shortly before its closure, the institute had in vain demanded greater leeway and highlighted the potential of a direct information channel to the people for the party leadership and the advantages of a quantifying method based on anonymous statements. Honecker’s radical decision in 1979 was symptomatic of his dead-end course. For example, he refused as well the distribution of critical mood reports by the Ministry of State Security. In his decision-making process he mostly relied on the regular channels of internal party information which always painted a rosy picture of the situation in East Germany.