Turning Science into Fiction? Censoring Population Research in the Soviet Union, 1964–1982

Population statistics reflect a nation's quality of life and accordingly have the potential to be highly politically charged, with implications for a government's legitimacy. In the Brezhnev era Soviet Union, emerging negative trends regarding life expectancy, fertility and mortality had the potential to de-legitimise the Soviet regime just at the moment when population issues were taking the spotlight through the United Nations. For this reason, population statistics were subject to significant censorship. The article examines how this censorship worked for domestic and international audiences. I show the main form of censorship was an editorial review by trusted experts in the Party and argue that the process was defined by uncertainty and negotiation, with personal networks mediating the result. In general, the period was characterised by tension between the need to expand demographic research and leaders' desire to suppress knowledge of unfortunate demographic truths.

feedback. Other kinds of valuable information are provided via interviews and written accounts by Soviet demographers produced since the fall of the Soviet Union. Interviews have either been published, orin one case cited hereconducted by myself. Memoirs can be problematic for historians, providing a self-conscious form of communication, which tends to omit details showing the author in a poor light. 10 This can sometimes lead them to be discounted completely; yet memoirs do have value, particularly in the context of the Soviet Union, where the general distrust of speaking freely or writing down personal thoughts means that it can be hard to understand a person's real motivations or feelings from archival documents. 11 As a result (and as many demographers of this generation have now passed away), memoirs offer one of the only windows into the experiences of people who, for political reasons, were unable to speak freely. In many cases, we only know about the falsification (or excessive manipulation) of data because Soviet experts later wrote that they did this. People who were there and took part in these events thus have something valuable to say about this history, provided their writings are approached with a degree of caution.
Censorship of population statistics must be seen in its broader context and was not unique to the Soviet Union. In post-war Czechoslovakia the government kept population data so secret that demographers likened themselves to astronomers without telescopes. 12 In Western European democracies research funding shaped science, but governments did not intervene to control demographic information to the same extent as in the Eastern Bloc. 13 In Poland, for example, Rosset explains that demographic research could only take place with the express permission of the Polish Central Statistical Office. 14 Literature highlights that governments censored population statistics most heavily in the Eastern Bloc while Stalin was alive. 15 After his death, the thaw affected the collection of population statistics across Europe, and governments relaxed the restrictions to the point where demographic research was possible. In the Soviet Union demographers gradually started to assert themselves, proposing new initiatives to expand the discipline. However, in the wake of the Prague Spring 1968, censors again sought to exert greater control over social scientists, both in the Soviet Union and the Eastern Bloc. 16 Much has been written about different kinds of Soviet-type censorship, and researchers have established some general principles. While some scholars have conceived of Soviet censorship as offering writers almost no freedom, the majority of the literature does not view censorship in this way. 17 In contrast to the early Soviet period and purges of the 1930s, Dewhirst and Farrell demonstrated that censorship after Stalin was a flexible process for writers, where self-censorship, control by cadres, and denunciations from colleagues functioned as the main form of control over publication. 18 Formal censorship remained in place, but writers and journalists were unlikely any longer to face prison if their works were rejected, and this emboldened authors to seek to circumvent censors. 19 The goal of abolishing all ambiguity, highlighted by scholars of the Stalinist period, no longer existed in the same way. 20 In intellectual life, Sandle argues that scholars were more likely to be stifled by 'dogmatism, conformism, ideological hairdressing and intellectual mediocrity' within the bureaucracy than outright censorship. 21 Denunciations of colleagues meant censorship thus became less about the state and more about powerful people 'hiding behind the state' in order to silence the less powerful. 22 However, to suppress embarrassing scientific evidence, the state needed loyal censors who were also knowledgeable experts, making scientific censorship distinct from other forms of censorship. A number of studies have investigated the manipulation of population data, but the majority have focused on correcting statistical distortions rather than censorship itself, seeking to establish accurate figures through retrospective calculations. 23 By doing this scholars have uncovered numerical evidence of abuses by the Soviet regime. 24 Literature that does explore the censorship process mostly focuses on the Stalinist period, when demographers were executed or sent to gulags for revealing negative trends. 25 As the dynamics of government and political control changed after Stalin's death, so did scientific censorship. Some studies have also engaged with the social or political implications of academic censorship, such as Oates-Indruchova's work on this type of censorship in Hungary and Czechoslovakia, in which she describes how academic careers were structured around Party loyalty as a form of censorship and control. 26 This article will build on this literature by arguing that the Soviet government's goal was to suppress demographic bad news because it undermined the legitimacy of the regime, but that tensions between leaders' desire for expert knowledge on the one hand, and secrecy on the other, made this a highly variable and unpredictable process. This article will begin with an overview of Soviet demographic trends and the development of Soviet population statistics in the twentieth century. I then examine how the state mediated demographic knowledge for three key audiences in the Brezhnev era: spectators abroad, the Soviet public, and domestic specialists. I show how population statistics aimed at an international audience faced the strictest controls because they had the potential to undermine the Soviet Union on the world stage. In this period, obvious bad news was self-censored, but where statistics could be interpreted in multiple ways there were few clear rules; both authors and censors were often uncertain. As a result, it was the interpretation of data that was most highly scrutinised; experts would try to position raw data positively, spinning trends as good news to reassure censors. Where the information was not clearly positive or negative, censorship became a flexible negotiation between bureaucrats, experts and the Party, and personal networks mediated the process. This article focuses on the everyday censorship of demographic publications. However, it was also very difficult for experts to access raw data from censuses or surveys to use in their work, and this formed part of the system of control over demographic knowledge, which is also discussed.

Context and Background
During the nineteenth and early twentieth century, the Russian empire had traditional patterns of mortality and fertility: very high birth rates coupled with high death rates. An extremely high number of children died in infancy from infectious diseases and lack of modern medicine. 27 As greater understanding of disease and hygiene began to develop globally, mortality rates sharply declined in the Soviet Union, spurred on by vaccines and antibiotics. Soviet healthcare, though far from perfect, was provided free to the population, and propaganda campaigns encouraged hygiene. However, describing a trend of downward mortality obscures enormous population shocks from upheaval, famine, collectivisation and, most significantly, the Second World War, which wiped out almost an entire generation of young men. 28 As mortality decreased, fertility fell sharply. The Russian empire had one of the highest fertility rates in the worldon average more than seven births per womandue to early marriage and lack of contraceptive knowledge. By 1940 fertility in Russia had fallen to an average of 4.25 births per woman. It continued to decrease; in 1955 the average was 2.83 and by the 1970s the fertility rate for the Russian part of the Soviet Union had fallen below the replacement level of 2.1. Fertility rates for the Soviet Union as a whole stood at 2.46, boosted by high fertility in the Central Asian Republics. 29 Life expectancy had also increased enormously from less than forty years in the Russian Empire to just over seventy in 1964. 30 If demographic good news was more people, longer life and improved health, then the early twentieth century provided much to be positive about. Communist countries across Europe were rapidly catching up with established democracies in terms of public health and life expectancy. 31 Any increase in mortality was, of course, seen as negative, but mortality increases only worried Soviet leaders from a censorship point of view when they reflected badly on the regime. Horrendous as the extreme mortality of the Second World War was, this spike in mortality was not deemed to reflect badly on the government; rather it was a sacrifice for saving the nation from fascism. 'Naturally', Khrushchev wrote in 1944, 'this led to significant population loss'. 32 Demographic bad news, then, was an increase in death and a shortening of life that could not easily be attributed to circumstances beyond the regime's control. It implied government action, or more often lack of it, was to blame. fertility was also felt to be bad news, though it was discussed more openly than mortality in an attempt to persuade women to have larger families. Where negative demographic trends could be attributed to government policy, leaders were keen to suppress this information.
Historically, the suppression of population statistics had involved violence against demographers, and this meant statisticians of the Brezhnev era were constantly aware they could be caught in the crossfire. In pre-revolutionary times military and local government statisticians began the process of population data collection and established the first statistical organisations and census. 33 Development continued in the twentieth century with the establishment of the Kiev Demographic Institute in 1919 under M. Putkha. Lenin encouraged the development of statistics as a means of learning about the population and managing the economy. Under his rule, statisticians completed the 1926 census, providing the first comprehensive data set about Soviet society. 34 Under Stalin statistics continued to develop as institutes were established and more data collected, but there were tragic consequences. When the 1937 census revealed the outcome of purges and the 1932-33 famine, the NKVD began a witch hunt. Figures showed nineteen million fewer people than the projections of the second five-year plan. 35 Secret police shot many demographers or sent them to camps, and the Party closed the Demographic Institutes in Leningrad and Kiev.
Following the purge, the Central Statistical Administration completed a new census in 1939, but from 1939 until the end of Stalin's rule, demography essentially disappeared as a discipline. It had proved too dangerous to the regime to be allowed to continue. Some survivors of the period went on to work in demography after Stalin's death. They laid the foundations for the revival of the discipline in the late 1950s and early 1960s. 36 The revival began under Khrushchev, who agreed to a Central Statistical Administration proposal for a new population census in 1959. The Party considered the census a success, and it opened the door to an expansion of the discipline. Statistical agencies did not re-evaluate the 1939 census, but nor did statisticians attempt to falsify new census data. 37 No independent institute was ever re-established during the Soviet period, but larger institutions created population departments and study centres. 38 Leykin has shown that scholars strove to breathe new life into their discipline and expand it, hoping to correct some of the Stalin-era failures. 39  Union. 40 But despite this expansion, the Party often suppressed information about mortality and health, and expanding research meant a greater need to monitor and control information to avoid embarrassing bad news being revealed. From the mid-1970s the level of censorship clearly increased. The primary reason for tighter state control was that demographic trends worsened almost as soon as Brezhnev assumed leadership. Life expectancy reduced every year from 1964 to 1981 driven by poor male health. By the end of Brezhnev's premiership, men in the Soviet Union lived three and a half years less on average than when he had taken power. 41 Until the 1970s infant mortality had followed a consistent downward trend: from 1960 to 1971 it fell from 35.3 to 22.9 (deaths per 1000 live births). From 1971, however, this trend reversed, peaking at 31.1 in 1976, but remaining higher than the 1971 level until the late 1980s. 42 As a result, infant mortality data was banned from publication. 43 That fertility had also fallen below replacement level for certain republics increased the perception of an impending demographic disaster. This raised a new issue for censors because the government's main priority in this area was to grow the population and improve health. Doing this would benefit the economy and population, and this in turn would legitimise the regime and demonstrate the effectiveness of the Soviet system to its own people and the outside world. As population problems intensified, the Party expanded research, while suppressing dissemination. The dual expansion and suppression required a complex system of censorship by experts. There were some written rules, but censorship as practised extended far beyond these. Both elements will now be examined.

Written Censorship Rules
The Index of State Secrets contained the basic rules on censorship of demographic statistics for the open press. 44 It functioned as a pre-censorship rulebook for authors. Officially all materials for publication passed through the General Directorate for the Protection of State Secrets in the Press (Glavlit), but in the case of demographic data it was difficult for a non-specialist to reliably censor it, and so practically, censorship fell to the Central Statistical Administration (TsSU). The Index makes it clear the agency was the final censor, a typical entry listing forbidden data read as follows: 'Statistical data about the mortality of the population by reason of death in the Soviet Union as a whole by republic, krai, oblast' and capitals of the union republicswithout the permission of the Central Statistical Administration of the Soviet Union or of the Union Republic'. 45 The Index prohibited most detailed population statistics from publication without the express permission of the TsSU. Many featured under the military section, supposedly as information that could be used by the enemy to calculate the number of potential conscripts. In fact, the Department of Defence encouraged censorship of demographic indicators so as not to give any advantage to foreign powers seeking intelligence about the Soviet military The primary censor mentioned here, the Central Statistical Administration of the USSR (TsSU), was the government agency responsible for collecting population statistics, running censuses, and supplying the government with demographic forecasts. There were republic, regional and local offices of 40  the TsSU, but in practice the TsSU was highly centralised. Templates for data were established in Moscow. 46 Anderson, Katus and Silver note that one of the purposes of the TsSU was actually to restrict access to data, and research demonstrates that TsSU leaders generally strove to acquire control over all statistics for their institution. 47 This tendency to restrict access to data, especially data showing abuses of the regime, meant specialists outside the agency were hostile to it. 48 When the agency's director, Vladimir Nikonovich Starovskii, was nominated to be a member of the Academy of Sciences in the 1980s, Academician Nikolai Trofimovich Fedorenko claims there was sharp opposition from the Academy because, 'in general, everyone at that time knew that the TsSU was nothing other than the "Ministry of Lies"'. 49 Material for open publication was censored at several levels. There was a hierarchy of secret information, from those data available only to party leaders and senior figures within the TsSU, down to data widely possessed by experts in universities and organisations outside of the TsSU, but which were nonetheless secret from the public (such datasets were stamped 'for professional use only' (dlya sluzhebnovo polzovaniya)). Data given the status of a state secret and therefore restricted totally from circulation included information about the following: homicide, suicide, internal migration and deaths caused by infectious diseases (plague, cholera, etc.). The TsSU allowed information about age-sex structure of the population and internal migration patterns to circulate between specialists but repressed it totally from publication. Mortality by age and infant mortality were published openly, but only rarely and with express permission of the TsSU. 50 The information contained under each of these categories had the potential to reflect poorly on the Soviet government. Indeed, suppressing these categories of information was not uncommon in the Eastern Bloc where such statistics were possible sources of embarrassment and accordingly out of bounds for publication. 51 The official rules on the publication of demographic data became stricter from the mid-1970s. The 1976 Index of State Secrets lists the additional following categories of statistics as forbidden for publication for the year 1975 onwards: 52 • The size of the population by economic region or by town with fewer than 50,000 inhabitants • Information about the numbers of children at school or numbers of workers in a given area • Data about the distribution of newborns by sex • Statistics about accidents at work An increasing number of citizens were dying from accidents at work, mainly as a result of increased alcohol consumption. Depopulation of rural areas due to migration and falling fertility was causing concern about labour resources, and the Party was also keen to obscure the potential numbers of conscripts given the Cold War environment. As demographic indicators worsened in the 1970s, censorship became stricter, and the new version of the Index reflected this.
In addition to these written rules, the Party tightly controlled who could access unpublished data, and this, along with the control of cadres, functioned as a precursor to censorship of publications discussed in the next section. Some indicators were particularly vulnerable to manipulation, such as 46 Anderson, Katus, and Silver, 'Developments and Prospects', 5; Tolts, 'The Failure of Demographic Statistics'. 47  infant mortality, which saw an entire network of hospitals involved in obscuring data before the figures ever reached the statistical offices. 53 Once datasets existed, many were circulated rather than published. Most census tables, for example, were kept from the public in this period and provided only to a select few specialists. 54 Copies of data were created with the stamp 'for professional use only', and lists were kept by the Central Committee detailing who was allowed to use these publications. 55 Sample surveys, forecasts and detailed census data provided by the TsSU fell into the 'professional use only' category. 56 In addition, use of demographic literature, especially foreign publications, was tightly limited and demographic works often ended up in the restricted sections of libraries. As one former TsSU worker recalled, 'simply to get into the Lenin Library for work you needed a signed letter from the head of the institution'. 57 Despite this, experts would sometimes break the rules by secretly sharing data with trusted colleagues and students. 58 In general, specialists found these controls irksome, and archival records show numerous complaints about them. A 1968 letter to Kosygin from Professor Dimitri Valentei requested that the government 'disclose certain portions of population and labour resource data currently designated for "professional use only" making them available to a wider scientific audience'. 59 His letter warned that this secrecy was stifling progress on population issues and noted that the Soviet Union was one of the few countries where population data was secret. In the Presidium meeting of the Academy of Sciences Social Sciences Section that year, several demographers criticised the TsSU for its secrecy, and conference resolutions from the period also called for the expansion of publications. 60 These senior demographers were hardly outsiders: most, including Valentei, were Party members and worked closely with the Central Committee. Nevertheless, they often opposed censorship because they desired professional autonomy and censorship limited this. As experts, they wanted their work to be recognised internationally, to have creative freedom and independent advocacy. In considering the meaning of censorship for scholars' lives it is therefore important to note that benefitting from the status quo was perfectly compatible with trying to negotiate ways round its constraints, and both featured heavily in the late-Soviet period.
These rules discussed abovesuch as the Index and systems of classificationset down certain categories of information that were out of bounds, but other information could be censored too depending on the view of the censor. In all cases, clearly bad news was censored even if it did not appear in the Index. Most of the time, however, exactly what constituted bad news was unclear. A grey zone existed where statistics could be interpreted in multiple ways, making the process variable and flexible. The following sections explore censorship as practised, considering how the state controlled demographic information for international and domestic audiences, where the limits to this control were, and what the government aimed to achieve. 53

Censorship as Practised: The Global Stage
The censorship process worked in different ways depending on the kind of publication and audience; and information about the Soviet population intended for an international audience faced the strictest censorship of all. International experts were difficult to mislead and not constrained by the Soviet state. Like governments across Europe, the Soviet leaders wanted to appear prosperous and righteous on the world stage by showing successes in the area of health and population. But with the ongoing Cold War, the Soviet Union also aimed to draw developing nations into its sphere of influence by demonstrating communist superiority over capitalism. 61 In this respect, population politics was an important arena for the exercise of soft power. Documents for international events were intended as diplomatic tools and were heard by specialist foreign audiences, which meant they were picked over by censors to ensure anything that could reflect negatively was removed. The publishers and editors discussed in later sections of this article were not involved here; drafts were simply reviewed by senior leaders at the TsSU. 62 Examples of censored demographic texts showing the precise process are rare in the archives, but some examples are accessible to researchers. This section uses one such paper as a case study in the censorship and politics of the time. The paper was that written by Arkadii Mikhailovich Merkov, a medical demographer working at the Semashko Institute of Public Health, for the Second World Population Conference. The conference took place in 1965, right at the beginning of the Brezhnev era explored in this article. Spurred on by the successful 1959 census, demographers were gradually recovering from the horrors of the Stalin era and finding space to assert their professional opinions; the tightening of scientific censorship following worsening mortality and the Prague Spring was yet to occur. The conference was organised by the United Nations and took place in Belgrade. Like all World Population Conferences, speakers met to discuss global population challenges and share research. In preparation for the conference, leaders in the field were chosen by TsSU directors and invited to write specific papers in their area of expertise for presentation at the conference.
The papers were translated into English and other languages by the UN and circulated for international scholars to read. Before each Soviet paper was approved, it had to pass through the TsSU for review by two trusted Party demographers. 63 This process was disguised as peer reviewboth reviewers were, in a sense, Merkov's peers. They were experts themselves who knew Merkov well and frequently worked with him on state projects and on the conferences circuit. Nevertheless, the function of this review was ideological censorship. The possibility of demographers sharing secret and revealing materials containing bad news to foreign colleagues was a constant fear for Party leaders. 64 In this role as an additional reviewer because he specialised in medical demography and had worked at the Semashko Institute himself as a doctor before becoming a population specialist. 66 Merkov wrote a long report exploring fertility and mortality in the Soviet Union, using England, France and the United States as comparators. 67 His report used standardised tables of fertility and mortality, which accounted for differing age structures of the population, as is customary in demographic literature. It demonstrated the Soviet Union had made similar levels of progress to Western countries in reducing mortality, but that it lagged behind in certain areas. For example, the standardised mortality rate per 1000 of the population showed the Soviet Union at 7.3. This was better than the United States at 7.5, but worse than England and France, which both had rates of 6.9. 68 Infant mortality was also shown in tables to be higher than other developed countries. Merkov repeatedly claimed that these effects, along with reduced population growth, were hangovers of the Second World War and its devastating impact on the population, which other countries had not suffered. In general, it showed the Soviet Union in a positive light, claiming much progress had been made. It did not, however, demonstrate that the Soviet Union was actively better than the West in this field, instead highlighting that the Soviet Union was doing similarly, despite facing more difficult challenges. The response it drew is interesting because it shows how the process of censorship moved the report away from a neutral statement and towards a declaration of nationalism intended to promote the Soviet Union on the world stage.
Both Smulevich and Podyachikh wrote back to Merkov with damning feedback. Smulevich began his letter by addressing the report's sense that the Soviet Union was faring no better or worse than other developed countries. This implied, to Smulevich's mind, that, 'socialism does not create advantages in solving the problems of public health'. 69 His feedback letter asked, 'from where did he get such a scientifically incorrect and politically harmful conclusion?', and continued, stating, 'only socialism creates the conditions for the provision of health in its widest sociological form'. Smulevich's feedback then turned to Merkov's choice of statistics. He claimed Merkov had used too few indicators, all of which were demographic rather than medical. His letter reads, 'all this cannot be reflected only in mortality indicators, which are widely used by bourgeois scholars with the purpose of furthering capitalism'. Nor should the medical health of the nation 'be reduced to an analysis of the process of reproduction'. 70 Podyachikh's response was equally critical, and he claimed the report required much work. Podyachikh particularly stressed the role that Merkov's conference report was to play for the Soviet Union: In the report it is necessary to include the main aim of the conferenceproviding assistance to developing countries in studying the interrelation of demographic indicators with economic development. In connection with this, the report should demonstrate the historical side of reproduction and show that in the early years, pre-revolutionary Russia and the Soviet Union had a low economic and cultural level, and that after a short period of time it caught up with advanced capitalist countries, not only in economic and cultural terms, but in indicators showing reproduction. You need to show the role of healthcare in this and show that the standard of healthcare in the USSR is higher than in capitalist countries. 71 In short, the report needed to show developing nations that the Soviet route to health and wellbeing was faster than the capitalist route. This highlights the perennial difficulty for Soviet health and 66 'Iz zhizni sovetskikh demografov. K  population statistics: how could they demonstrate the standard of health was higher, when in fact the standard of health was similar if not lower? Focusing on progress and change over time was one such way; though the Soviet Union might still be behind, it had modernised faster.
What is particularly notable about this instance of censorship is that Merkov had already attempted to include change over time as a mitigating factor. Much of the criticism focused on the fact Merkov had not stated this uncompromisingly enough. Both reviewers particularly objected to the following sentence by Merkov, for example, which they crossed out multiple times: Despite the fact that child mortality in the USSR has reduced by more than 8.5 times in comparison with the previous period, it is still higher than comparative countries. The future reduction of child mortality is an important task for Soviet healthcare. 72 Podyachikh suggested the following formulation to reframe the narrative: You should have given a comparison of what child mortality was in 1911-1913 in the USSR and in other countries, and by how much it has reduced in these countries in the same time frame as in which the USSR has reduced it by more than 8.5 times.
Any statistic that didn't show the Soviet Union as better needed to be looked at in a different way. Methods had to be found to present the data such that the Soviet Union appeared as a model for the entire world.
Despite the overwhelmingly negative responses by censors explored above, most of the draft report text actually remained in the final version. Merkov had described the Soviet Union incredibly positively wherever it was possible but had stopped short of artificially manipulating dataturning science into fiction. On 14 March 1964 he replied to Podyachikh about the feedback, agreeing to make the changes, but he wrote that Smulevich's comments about the report not having a medical focus and overly relying on mortality and fertility were of little help. He wrote that comparing medical statistics with capitalist countries 'is not always in our interest, moreover, the Soviet materials on this issue are not available for open publication'. 73 The exchange demonstrates the extent to which demographers were forced to tie themselves in knots, in this case excluding nearly all public health data from a public health report so as not to embarrass the Soviet government (and in doing so risk their career). Evidently, Merkov understood broadly what was required because he had excluded politically damaging data on diseases in the first place, but he was reluctant to include deliberately misleading statistics until forced to by those in charge. For a domestic audience, comparator countries could be carefully selected, but an international audience had the potential to access other data of their own accord. Censors therefore tried to take charge of the narrative and frame it much more aggressively.
The final report shows the requested changes were made and standardised mortality rates were replaced with unadjusted ones (Table 1). 74 This meant that the mortality rate per 1000 appeared as 7.5 for the Soviet Union, 9.5 for the United States and 11.9 for England. It concealed the true situation by relying on the fact that the Soviet Union had a much younger population and so the statistics did not compare like for like. Though it superficially provided better numbers, the problem would have been immediately obvious to specialists. Ironically, Blum points out that this tendency to conceal and manipulate actually increased speculation in Western literature, sometimes causing scholars to presume that Soviet demographic indicators were worse than they actually were. 75 The issue of data not being cleared for publication reoccurs often in sources from the time and is highlighted by this case. It shows how refusing to publish data was a strongly ingrained default with 72 Ibid., l. 14. 73 Ibid., l. 6. 74 Ibid., ll. 135-41ob. 75 Blum, 'Rodit'sya, zhit', i umeret', 39. government agencies. Head of the TsSU Department of Demography, Andrei Volkov, recalls how at this time they used to joke that TsSU leaders could say no to suggestions even before they'd been uttered. 76 At times this produced farcical situations. As a side note in his feedback to Merkov, Podyachikh informed him that he would need to take out the population pyramid based on the 1959 census data, which showed age groups as five-year blocks. He said that this data had not been cleared for publication and ordered Merkov to use the openly published data instead. 77 Merkov replied to say that the statistics 'not cleared for publication' were actually from page three of Podyachikh's own book, Population of the USSR (Naselenie SSSR), published three years previously. In fact, he argued, he couldn't change this data as there was no other openly published data to use. 78 There is no further comment from Podyachikh preserved in this file so his precise response to this revelation is unclear, but the final report shows a population pyramid with those aged under twenty aggregated into ten-year age groups to display less detailed information. Despite the data being in open publication by the censor himself, Merkov was ultimately still made to remove it. Podyachikh had deemed the pyramid safe for public scrutiny at one time, but as the head of the Census Bureau he would have been constantly looking at various different pyramidsmost of which were secretand evidently he did not instantly recognise this version of the data. Upon seeing revealing statistics, his impulse was automatically to refuse to publish it in the United Nations report translated into English. The vast majority of data was never published, so that it was not actually unreasonable of him to assume such data would not have been cleared for publication. Hiding data was a strongly ingrained default for TsSU leaders, one that was exacerbated when the information was for international publication.
This censorship exchange reflects the tension between the expansion and suppression of demographic knowledge and tells us much about the nature of Soviet censorship and population research. Global challenges were driving the dissemination of population research at an international level, such that Soviet leaders were forced to choose between publishing or abstaining from UN population activities completely. The desire to take part in global debates ultimately defeated the desire to conceal information, but the reviewer's true task was to marry these joint aims. As this case study shows, this task was impossible to fully complete. It should be noted that in this example, the editorial review transformed the paper significantly beyond the level of self-censorship. Population specialists understood the politics of the time, but it is clear in many cases they did not want to present inaccurate data. Merkov's response indicates he saw many of the revisions as a source of embarrassment and attempted not to include them in his work. It indicates the continuing tension late-Soviet scholars faced between their desire for professional autonomy and international respect on the one hand, and their privileged position as Party members able to attend international events and lead departments on the other.
Most of all, censorship of papers for the Second World Population Conference shows why population statistics mattered. Leaders wanted to demonstrate that their national community was populous, healthy and virile. Proving this, they believed, would show others that the Soviet project had been a success. Poorer nations were encouraged to visualise their own future through the success of the Soviet Union and choose socialism. For this reason, it was not good enough just to be equal to  78 Ibid., l. 5. capitalism for Soviet leaders. Only clear superiority would influence other nations to follow their path to socialism, affording the Soviet Union hegemonic power through strategic alliances. The less conference reports reflected the reality for the Soviet population, the more they therefore became imperialistic expressions of nationalism, carefully crafted to persuade, deceive and promote.

Censorship as Practised: The Soviet Popular Press
Many of the themes explored in the previous section can also be seen in the censorship of demography for domestic audiences, but the process, actors involved and, crucially, the importance placed on these publications differed quite significantly. The popular press here refers to any publications aimed at the general Soviet public. For such publications, censorship decisions rested with the editor of the publication. Editors invited specific demographers to write on the topic of population problems for their newspapers and magazines. The relationship between expert and editor therefore became the primary source of censorship; in the Brezhnev era, trusted editors could print without prior approval, though decisions would sometime come back to haunt them when bureaucrats or organisational leaders complained to the Central Committee. 79 As such, editors needed to be able to rely on experts to provide interesting material but not cause trouble, and so they tended to turn to the same individuals time and again. Viktor Perevedentsev, for example, wrote articles for Literaturnaya Gazeta every year throughout the Brezhnev period, having a strong working relationship with its editor for social and domestic problems, Anatoly Rubinov. 80 Ultimately, all demographic censorship was a fallible social process. It was shaped by the tension between the Party's declaration that demographic policy was an 'important task' for natural and social science, and leaders' desire to avoid any demographic bad news. 81 However, in search of interesting stories to publish, editors sometimes prioritised these goals differently to other Party figures. In 1968 demographer Boris Urlanis's famous article 'Take care of the men!' (Beregite Muzhin!) was published in Literaturnaya Gazeta, highlighting some male health problems in the Soviet Union. First deputy editor in chief Vitaly Syrokomsky recalls it was opposed by both those 'above' (verkhi) and by some other editors. 82 Nevertheless it was published, suggesting, perhaps surprisingly, that the final word on the matter rested with the editor. The article went on to be reprinted in many other newspapers.
It was not uncommon for organisations to complain about editorial decisions to the Central Committee. 83 State actors at all levels negotiated acceptable practice between themselves as the desire to reveal new information conflicted with the fear of political scandal. In 1970 for example, after the demographer Boris Urlanis had written an article entitled 'Fatherlessness' in Literaturnaya Gazeta, the director of the TsSU, Vladimir Starovksii, wrote to the Communist Party Central Committee to complain that figures relating to the number of children born out of wedlock had been published. The letter warned that foreign governments could use such figures in anti-Soviet propaganda campaigns. 84 He requested that in the future the Central Committee stop the editors of Literaturnaya Gazeta from publishing statistics that the TsSU had deemed secret. The potential for embarrassment through comparison by governments across Europe and North America loomed over any individual editor's censorship decision.
Such complaints were actually quite common; leaders at the TsSU complained to ministers whenever they felt publications had strayed too far from their own line on a demographic issue. Peter Gavrilovich Podyachikh, head of the Census Bureau at the TsSU, complained to ministers across government about articles supporting contraception as a demographic tool that were published in Literaturnaya Gazeta and Sputnik between 1966 and 1968. 85 'Inserting this tendentious compilation [of statistics showing the benefits of contraception],' Podyachikh wrote, 'the editorial office of the journal knew that the representative of the Soviet Council of Ministers, Comrade Kosygin, sent D. Rockefeller a negative answer to the request to collaborate on birth control policies in September 1967'. However, Podyachikh's letters on this topic went unanswered, and indeed by 1969 statements by Soviet delegates indicated that the Party had changed its mind on the acceptability of contraception as a means of population control abroad. 86 This highlights the somewhat unpredictable nature of censorship in the area of population issues. Policy changes from leaders, unforeseen by experts, could rapidly change dynamics. What constituted bad news was constantly being negotiated by censors and authors, and personal connections mediated the process.

Censorship as Practised: The Specialist Press
There were many types of publications for the specialist press including books, journal articles, TsSU reports and census data. The process differed depending on both the type of publication and the author. A university academic trying to publish a monograph faced different hurdles to a TsSU bureaucrat attempting to publish a demographic conference report. Often the process was unknown even to the author. For example, according to Russian demographer Valery Elizarov, who began his career at Moscow State University in the 1970s, academic articles had to be approved for publication by the Ministry for Education as well as the journal editor. 87 When permission to publish his first article was denied, colleagues told him it was pointless trying to find out who had made the decision or why. This makes establishing the stages of censorship in the specialist press difficult for historians, firstly because there was so much variation, and secondly because evidence suggests that clear and consistent procedures were not necessarily followed. Archival evidence and accounts do, however, provide snapshots of censorship decisions from which some general principles of censorship in the specialist press can be established.
The first principle is that specialist publications were able to contain more detailed statistical information than those for the general Soviet public, but they generally faced a more burdensome process for publication as a result, particularly because the need to publish quickly that characterised news reporting did not apply to academic literature. Glavlit was required to review all works for open publication, but for specialist scientific publications this was not the main form of censorship. 88 Instead, works faced a multi-layered editorial review, which depending on the type of publication could include the publisher or journal editor, a senior leader of the The second principle was that works should cast Soviet health and wellbeing in a broadly flattering light. The manipulation of life expectancy provides a useful case study of what could happen when figures revealed bad news. After Brezhnev publicly declared in 1967 that average life expectancy had reached seventy, staff at the TsSU were forced to manipulate the published statistics to try and keep life expectancy at seventy, even though it was already starting to decline. Evgenii Mikhailovich Andreev was part of a team of TsSU statisticians given the task of calculating life expectancy from the 1970 census data. To the statisticians' dismay, life expectancy for 1969-70 had fallen to 69.34, despite the fact that the TsSU and Soviet leaders had publicly announced they expected it to be at least seventy. His account describes how his bosses, Aron Ya. Boyarskii and Gerogii A. Pavlov, finally decided to publish data for the four-year period 1968-71 instead. As 1968 was a better year, its inclusion brought the average to 69.5. This figure could then be appropriately rounded to seventy. He reports that after this time, mortality tables stopped being published, saying TsSU statisticians did not even discuss their publication until perestroika. 90 That no discussion on the matter even took place is important. It demonstrates that specialists had an unspoken knowledge of censorship rules. A grey zone existed when statistics could be interpreted in many ways, but data that clearly showed a decline in health was not to be published. Because statisticians knew this, they censored it themselves, rather than be reprimanded by a censor before the data was inevitably excluded.
A third principle that applied to the censorship of specialist publications is that much depended on the interpretation of numbers and the positioning of the research within the political context. Even when statistics themselves were approvedand demographers would rarely attempt to publish statistics clearly prohibitedthe written text was subjected to harsh scrutiny. For the most part, quantitative analysis was affected little by Marxism in this period. Instead, ideology played a large role when interpreting and explaining demographic changes and their consequences. Writing after the collapse of communism, Anatoli Vishnevskii, who worked at the TsSU Scientific Research Institute as a demographer in the 1970s and 1980s, claimed that 'distorted and dogmatized Marxism was bizarrely mixed with fragments of randomly introduced modern Western ideas, which gave rise to a strange, sciencelike mixture that made it possible to see real life, as if through poorly pointed binoculars'. 91 This is supported by archival evidence. A manuscript entitled 'Contemporary Socio-economic Issues of Population Reproduction of the USSR' sent for publication to the journal The Health of the Russian Federal Republic (Zdravookhraneniie Rossiiskoy Federativnoy Respubliki) shows extensive sections crossed out by the editor of the journal. 92 The redacted sections were all explanatory text; the editor retained the numerical portion of the document untouched. In particular, the editor removed text that might touch on politically sensitive events. For example, the author wrote that studies on natality, mortality and life expectancy were completely 'stopped' (prekrashchennye) in the 1930s, which the editor replaced with the word 'aggregated' (agregirovannye), a particularly vague term chosen to gloss over the repression of demography in those years.
Beyond these principles the process was characterised by conflict between authors and censors. As one trainee at the Moscow State University Population Centre explained: Essentially there was no statistical data available that could be used as the basis of an interesting article for publication in the open press. Like detectives, we extracted this data in snippets from scientific literature, and then we'd recalculate the figures so the censors would agree to print them. 93 90  Negotiating censorship successfully was a major part of scholars' working lives. As the account above highlights, if such tactics were not used, publication became impossible, making a career hard to sustain. Often experts wanted to use their knowledge to improve society and found it difficult to remain silent. Urlanis was well known for such conflicts with the authorities. His daughter wrote of her father working in this difficult atmosphere: I remember how hard it was for him to overcome battles with censors, when every figure was checked, not just the secret ones. To publish this or that data, characterizing the unfavourable situation in the country (high mortality and morbidity, a huge number of abortions, etc.), my father had to present the material so that secret figures did not appear, but at the same time a true picture of the demographic phenomena in the country was given. How many times he spoke in the TsSU, the Ministry of Health, and in other ministries with alarm for our country! He would say something like: 'you write the figures in red ink, but you should be writing them in blood'. 94 Whether or not Urlanis ever really expressed his concerns in quite such provocative wording is, of course, impossible to verify, though stenogram records from the Academy of Science and TsSU do show several other passionate outbursts by Urlanis, denouncing the way population problems were disregarded and kept secret by those in charge. 95 Conflict over the censoring of certain figures appears to have been common. Much to his dismay, censors removed large chunks of Urlanis's 1974 book Problems of the Dynamics of the Population of the USSR (Problemy dinamiki naseleniya SSSR), including an entire chapter detailing the demographic forecast for the Soviet Unionsomething that by the mid-1970s was looking far from rosy. The data used in that chapter was already in open publication, but censors banned it anyway. 96 Though scholars could negotiate, they frequently lost the negotiation, and this became a bigger risk as trends worsened.
Accounts of specialists successfully negotiating with censors show why uncertainty was key to the way this type of censorship operated. 97 Censors saw thousands of potential publications and no two were the same. They applied a set of general rules using their judgement, and for this reason experts needed to sell their trustworthiness to the censor. Small adjustments to style or citations to demonstrate loyalty could be enough to reassure a censor when they were unsure whether or not something constituted demographic bad news. For example, Vishnevskii's 1980 article was published only after he agreed to reduce the number of foreign demographers cited throughout. 98 Publishers tended to be risk averse. Raisa Sergeevna Rotova was a former publisher turned demographer working under Professor Dimitri Ignatovich Valentei, director of the Moscow State University Population Centre. She recalls preparations for the publication of his 1971 monograph, Marxist-Leninist Theory of Population (Marksistskaya-leninskaya teoriya narodonaseleniya) (by publisher Mycl'): DI.
[Valentei] felt that the book might not be allowed for ideological reasons. He asked me to consult with the editor of our monograph from the publishing house V.I. Budarina (whom I knew well) to find out what needed to be done so that the book saw the light of day. The editor recommended giving an approving assessment of Academician Strumilin at the beginning of the book. DI.
[Valentei] followed this advice, and the monograph came out. 99 This example shows Valentei successfully negotiating with a publisher to achieve his desired result: the publication of his book. In such an environment contacts were key, and specialists tried to maintain good working relationships with publishers to ease this task. Where the science was not clearly good or bad news for the regime, demonstrating loyalty through other means could tip the scales in an author's favour.
The purpose of all this tinkering was to ensure that demographic works cast Soviet socialism in a flattering light. It served to make censors feel secure in the publication of certain information. By the very nature of their job editors and publishers needed something to publish. Unlike censors at the TsSU, if they simply rejected every book or article sent their way they would very quickly find their journal or publishing house ceased to exist. They were, therefore, open to negotiation within limits. Reassuring the publisher by editing the text, citations and quoting the right Marxist authorities were central to the process.

Conclusion
When governments are faced with statistics showing their population is dying earlier and in larger numbers than before, their instinct is often to conceal this information. What is distinctive about the Soviet context is the extent to which leaders intervened in science to do this. 100 Ultimately, where trends could reflect badly on the government, politics triumphed over science. But in censoring demography, the regime faced a group of scholars and experts who did not want their work concealed and spent much of their energy trying to influence censorship decisions in their favour. These experts were not dissidents; like the majority of intellectuals under communism they occupied the space 'between conformity and dissent', benefiting from the status quo in some ways but disliking many of the restrictions on their professional freedom. 101 They internalised the rules of the game. Anatolii Vishnevskii, Russian demographer who began his career in the late 1960s, wrote it was hard to believe that by the end of the Brezhnev years there were people sincerely convinced that the most important task for demography was 'to confront the methodological line of bourgeois science'. 102 Yet, he noted, demographers wrote and even lived as if this was true. Yurchak has observed this phenomenon as central to late-Soviet society. 103 People did not need to accept communism, 'so long as they accepted it as the socio-political framework of reality within which to further their personal interests and goals'. 104 In this environment, censorship of demographic work became a process of negotiation between scholars, bureaucrats, agencies and formal censors. Uncertainty was central to enabling this negotiation. As case studies explored here have shown, specialists could and did push back against censors, arguing their case and explaining why their chosen work should be published. Provided their work was within the zone of uncertainty and not clearly bad news, censors could be persuaded, or a compromise solution found. Merkov's paper was not made to include infectious disease data or remove information about infant mortality because he successfully argued against it. Nevertheless, censors retained a veto on the process. Having good relationships with potential censors was vital because it allowed scholars to discover what kind of inclusions would reassure the reviewer so that their work saw publication. As has been shown, small inclusions and adjustments, usually to the accompanying text, were incredibly significant, and scholars became adept at playing the game and adjusting their work accordingly.
The purpose of demographic censorship was to contain bad news and promote good news. Though it is difficult to gain a comprehensive picture of demographic censorship because of the scattered nature of sources, evidence provided by memoirs, archival documents and testimonies indicates the process followed no precisely established logic because it was unclear what constituted good or bad news in many cases. The large number of potential censors meant works went round in circles waiting to be approved. Data deemed acceptable for one audience was unacceptable for another, adding to the confusion. However, this paper has also underlined editorial review by specialists chosen for their loyalty to the Party as the primary form of academic censorship. If an official in one part of the system was unhappy with a censorship decision made by someone else they could and did complain to central government.
After the horrors of Stalinism, demographers gradually reasserted themselves, a process beginning with the 1959 census and the establishment of new departments and study groups in the early 1960s. By the early 1970s demography was well accepted as a discipline. Nevertheless, the worsening state of health and mortality encouraged censorship, even as it provided experts with a further raison d'être. Only in the era of glasnost' were the most stringent restrictions on publication lifted, paving the way to a reappraisal of the period and its demographic past. Newspaper articles suddenly revealed negative trends and asked why the government had shown an 'unjustified complacency and proved totally unable to combat them'. 105 The Soviet healthcare system was criticised as it emerged a quarter of district hospitals lacked plumbing and over half had no hot water. 106 The Minister of Health, Yevgenii Ivanovich Chazov, even told the Soviet Party Conference in 1988 that the Soviet Union had been silent for too long on its poor international rankings in infant mortality and life expectancy. 107 This had an effect as opinion surveys showed young people were less satisfied with Soviet healthcare than the older generation. 108 As the Soviet authorities had earlier feared, revelations about demographic trends both past and present contributed to undermining the regime's authority and the legitimacy of the socialist developmental model and political system.
Finally, this research confirms the importance of population politics and statistical knowledge to modern statecraft. German physician Rudolf Virchow once famously claimed that 'politics is nothing but medicine at a larger scale'. 109 Through demographic data governments compare themselves and stake claims to superiority. New research shows that Soviet legacies have shaped Russian obfuscation of morality data from the COVID-19 pandemic. 110 Nor does this tendency to manipulate mortality apply to Russia alone. From the censorship of suicide statistics in East Germany, to the suppression of Brazilian ethnicity data, examples can be found everywhere. 111 As such, population statistics remain an important platform for competition and influence for nation states across Europe and the wider world. If history is any guide, demographic statistics will continue to be important to governments and their populations alike long into the future, making demographic censorship a topic of continuing relevance for researchers.