1Population is one of the oldest demography journals in the world. Its birth immediately after the Second World War, a decade after Population Index and Genus, and a year before Population Studies, was the result of two processes. The more general one was the expansion and establishment of demography as a discipline during the interwar period. There was an earlier Population in the 1930s, the journal of the International Union for the Scientific Investigation of Population Problems (IUSIPP), the forerunner of the IUSSP, published in English, French and German. It did not survive the war and its title became available for the French journal [1]. The name was Population and not Demography, an important distinction. It recalled the fact that between the wars international researchers had focused not so much on a discipline, demography, as on a topic, population, which belonged to more than one rival speciality, beginning at that time with biology and economics (Kingsland, 1985; Ramsden, 2003; Rosental, 2003b).

2The other more specific reason for the title Population was that it was chosen by Alfred Sauvy, INED’s first director. He too wanted to focus more on a topic than a method. His other concern was to make the journal an instrument for disseminating knowledge rather than a collection of specialist research reports. “Without going so far as to become a journal of ‘popular’ science, Population will comment on demographic data for a wide audience of senior civil servants, scholars, parliamentarians, doctors, lawyers, captains of industry, etc.” Sauvy established from the outset a clear division of tasks: Population was to target strategic groups such as scholars, administrative elites and leaders of opinion, while Travaux et Documents, the Institute’s other publication, was addressed to “the more restricted audience of specialists” [2]. This division was modified over time: one sign of Population’s move towards the rank of scholarly journal was the creation of the Population et Sociétés bulletin in the 1960s in order precisely to recover the function of scientific communication. Another indicator of this shift was the gradual removal of indications of the journal’s official government origin from its cover page, clearly described by Clerc (1995).

3Alfred Sauvy and INED, research and dissemination: this from the outset was the institutional framework for the journal’s development. Nowadays, INED is seen primarily as a research institute and its journal as a scholarly one, but this image is the result of a quest for autonomy extending over decades, involving a continually renegotiated “transaction” between the Institute and its supervisory ministries. Population was founded immediately after the creation of INED on 24 October 1945. France was caught up in yet another regime change: the peace-time Third Republic, the war-time Third Republic with its emergency laws (September 1939–June 1940), the Occupation and the collaborationist Vichy regime (July 1940–August 1944), were followed by the Liberation and the provisional government headed by General de Gaulle. When he stepped down in January 1946, the parliamentary regime of the Fourth Republic was formed.

4For the whole of this period, the public research community as we know it today in France did not exist. No central administration was devoted to it: the first would be the General Delegation for Scientific and Technical Research (Délégation générale à la recherche scientifique et technique, DGRST), created by General de Gaulle at the start of the Fifth Republic in 1958. Research was not a profession; at best an initial phase before joining one of the canonical university faculties: for population sciences this meant economics, medicine, history or sociology. The first public research institutes had only just been founded. In this field as in many others, Vichy initiated a more pro-active government role, combining the ideology of its “National Revolution” with the technocratic activism of the 1930s (Dard, 2002). It set up the French Foundation for the Study of Human Problems (Fondation française pour l’étude des problèmes humains, FFEPH), also known as the Carrel Foundation, and such specialist bodies as the Institut national d’hygiène and the Office de la recherche scientifique coloniale. The Liberation authorities followed on, founding the Office national d’études et de recherches aérospatiales, the Commissariat à l’énergie atomique and INED, itself partly a successor to the FFEPH [3]. Apart from this latter foundation, relatively general in its aims, since it was designed as a sort of anti-Republican CNRS [4], these other bodies, as their names suggest, worked on sectors targeted for government action. They undertook research but had little autonomy from their administrations and the Ministry of Finance. What was the right balance between pure and applied research, as the issue was formulated at the time? Under the Fourth Republic, this caused tensions between the emergent research community and its various ministerial overseers (Duclert, 1998).

The long march to autonomy

5This was the background to the identity of the new-born Population. From one side, some of the Institute’s researchers, particularly in the Demography section under Paul Vincent, submitted articles on the latest advances in analysis and methodology (Henry and Vincent, 1947). From the other, the Institute’s influential Technical Committee, whose members were listed on the journal’s cover page until 1973 (Clerc, 1995), demanded material accessible to a wider educated readership [5]. The resulting tension in the Institute and its journal reflected the ambiguity of public policy at the time [6]. This caused difficulties for Alfred Sauvy, obliged to turn articles down because they were “too demographic” (sic). How could he resist pressure from the senior administration personalities he had brought into the Technical Committee to protect the Institute, such as Pierre Laroque, first director-general of the French social security system, while maintaining the motivation of young untenured researchers, many in their early thirties? With the lowly job title of chargé de mission, their contracts were renewed annually, causing a continual loss of staff, when recruitment was already difficult. Demography was an “unrecognized science” that did not exist at universities, and students in established disciplines were reluctant to commit themselves to a small institute that might not even survive, and preferred to stay at university (Rosental, 2003a).

6The dilemma was resolved by developing “demographic intelligence” (Rosental, 2003b) at INED: the Institute’s young demographers did their best to link the “topical” work they were given to the general methodological questions of the discipline. Jean Bourgeois-Pichat, Sully Ledermann and later Louis Henry, working with Paul Vincent, benefited from the expansion of their discipline after 1945. At the United Nations, a Population Division was developing, which, because the Cold War prevented any advance in population policy, offset this immobility by over-investing in research. Until around 1960, the Division was a key forum for demographic exchanges and, with support from American foundations and major institutes in other countries, was able to award recognition for scientific quality, which INED could display to its ministerial overseers. The new statutes of the IUSSP – which had been nationally based before the Second World War and in 1947 became a fully-fledged international scientific body – also helped to structure the discipline on the basis of international scholarly exchange and competition.

7Although this reconciliation of the scientific and the administrative was not originally planned, it did not contradict the role intended for the Institute by its two founders, Alfred Sauvy and Robert Debré, the influential medical luminary (Simonin 1997), who managed to persuade General de Gaulle to direct part of the Carrel Foundation towards population. The two men had been friends since the Occupation and shared rationalist and “natalist” convictions. They viewed an increase in the birth rate as a scientifically justified and demonstrable necessity: if it was properly explained to couples, they would modify their behaviour in their own and society’s interest and enable the French population to reproduce itself. This conviction explains the complex relationship of agreement and hostility maintained by Sauvy and Debré with the National Alliance against Depopulation (Alliance nationale contre la dépopulation). In general, they shared the objectives of this natalist lobby which had been operating under various names since 1896. But they criticized its propaganda campaigns, based on dramatization and an appeal to the emotions. They expected an institute devoted to population to provide detailed backing for demographic reasoning, in order to educate French people “dispassionately” about the seriousness of their country’s position, in the manner of an office of scientific information. Alfred Sauvy’s catch-phrase “Prévoir pour ne pas voir” (predict to prevent) took institutional shape with the foundation of INED, supported in that aim by its journal Population, targeted at a readership assumed to be influential.

Psychology as demography’s next-door neighbour

8Reproductive behaviour and demographic knowledge formed the link in the middle decades of the twentieth century between demography and social psychology, and they underpin the article by Jean Stoetzel (1910-1987) reprinted in this issue of Population. This link was naturally established by the “propaganda” in favour of high fertility between the wars, at a time when the manipulation of crowds was an obsession (Tchakhotine, 1939). But it was also supported by scientific considerations. One of the priorities of demographic research right up until the 1930s was to determine, on the basis of an organic concept, an indicator of the rates of reproduction of populations that would be independent of what we would now call “structural effects”. The introduction of the fertility rate rather than the birth rate, the development of crude and net reproduction rates came essentially from statistical work on census data. In the 1930s, further considerations appeared. Closer examination of the marriage rate raised questions about the timing and spacing of childbearing: couples’ intentions became an increasingly important component in the interpretation of changes in the fertility rate (Wicksell, 1937; Hajnal, 1947; Clark, 1949). This approach was supported by the development of large-scale fertility surveys in the United States, which for the first time attempted to establish statistical quantities on the basis of survey samples and interviews. These surveys were motivated by the old eugenicist theme of differential reproduction between the social classes and displayed an increasing interest in the sociological bases for contraceptive practice: they were part of a more general shift of Foundations and funding in the United States away from eugenics and towards birth control (Allen, 1991; Borell, 1987).

9After the Second World War, this interest in fertility was revived by debates on contraception that divided Catholic intellectuals and attracted the attention of demographers such as Frank Lorimer in the United States, and Louis Henry and Jean Bourgeois-Pichat in France [7]. Surprisingly from the viewpoint of an ahistorical classification of the social sciences, until the late 1950s at least, “psychology” became the discipline closest to demography. The quotes signal that it is important to distinguish between two uses of the word. The first was fairly imprecise: demographers sought to find “explanations” for the behaviour observed, often in a common-sense way [8]. But the second more scientific use took seriously the whole question of social psychology. This speciality had an unhappy history in France, where it was engulfed by Durkheimian sociology at the start of the twentieth century. And yet Jean Stoetzel took it as his banner, and made “psychosociology” the purpose of the section entrusted to him when INED was created.

A Janus of sociology: Jean Stoetzel

10At the time, the young sociologist Jean Stoetzel was one of the pioneers of empirical and quantitative sociology in France, and had imported survey techniques he had learnt in the United States (Blondiaux, 1998). When he graduated from the Ecole normale supérieure in the late 1930s, his priority, like so many of his contemporaries, was to escape from secondary school teaching [9]. Like his fellow graduate Louis Chevalier, he readily took advantage of the new resources provided by the institutions that the Vichy government had created [10]. Shortly before Chevalier joined the Délégation générale à l’équipement national, dedicated to what would later be called regional development policy, Stoetzel became chargé de mission at the Carrel Foundation. His position was justified primarily by his theoretical and methodological knowledge. The competences he was held to possess in survey sampling put him in charge of a “Survey and Statistics” department, which was one of the Foundation’s “common teams and departments” (Drouard, 1992). This central position at the Carrel Foundation was strengthened by the official relations Stoetzel had with the national statistical service (Service national de la statistique, SNS) [11]. But his legitimacy came also from his career. As a founder of the Institut français d’opinion publique (IFOP), he had run surveys as early as 1939 on French attitudes to demographic issues, and continued this work during the Occupation (Paoletti, 1939a and 1939b; Huss, 1990). After the Liberation, Stoetzel was a central figure among the fifteen or so chargés de mission that Alfred Sauvy took from the Carrel Foundation to create INED [12]. On this much smaller scale, he kept the cross-cutting role he had held at the FFEPH. His “psychosociology section” was only one speciality among many, such as demography (headed by Paul Vincent), economics (Georges Létinier), history (Louis Chevalier), the study of “the relationship between population number and quality” (Jean Sutter), or “factors of heredity and environment” (Robert Gessain). But at the same time, Stoetzel’s team, the direct forerunner of INED’s present surveys department, included technicians (male and female calculation clerks) who also worked for other sections [13], and maintained a network of 400 interviewer-correspondents throughout France.

11Although INED’s demography section rapidly asserted itself as the Institute’s flagship, Jean Stoetzel and his colleague Alain Girard remained central. When Stoetzel was appointed Professor of Sociology at the University of Bordeaux and then the Sorbonne, he remained as technical advisor (i.e. associate researcher) at INED, and was elected by his peers to succeed Alfred Sauvy, who retired in 1962, but declined the post in favour of Jean Bourgeois-Pichat (Girard, 1986).

12This strategic position is essential for understanding the article Jean Stoetzel published in the first issue of Population. He not only represented a discipline related to demography, but also advocated a social theory and a method crucial for conceptualizing population phenomena, at a period when such phenomena were not seen as belonging to demography alone. What social theory was it? The text that is republished, for all its academic format and air of consensus, reveals the tension that was to mark the development of French quantitative sociology in the post-war years. It combines two schools of thought whose theoretical constructs were radically different and would be opposed during the following decades, not least by Jean Stoetzel himself. On the one hand, there was North American empirical sociology, whence the brilliant young graduate imported not only the concepts and methods but also the practices, mid-way between university research and marketing. At a time when the scholarly ethos required a certain distance from the commercial world, Stoetzel cut a strikingly novel figure, as both a businessman of the polling trade who had founded IFOP [14], and a scholar at the outset of a long cursus honorum that would take him to the heights of the Académie des sciences morales et politiques [15]. On the other hand, the article republished here contains a long discussion of the work of Maurice Halbwachs, mention of whom leads straight back to the sociology of Durkheim. And yet Stoetzel was to remain an indefatigable opponent of that school throughout his career, seeking to throw off philosophy’s sway over the social sciences: indeed, as a young sociologist he described as a conversion his painful initiation into the American empirical tradition, which radically contradicted the teaching he had received at the Ecole normale supérieure (Blondiaux 1998) [16].

13If the classifying approach to the history of ideas made any sense, Stoetzel should not have been Halbwachsian, a dilemma sharpened by the fact that Halbwachs had been his thesis supervisor (Stoetzel, 1943). This reference may well be the result of the emotion he felt at his mentor’s death in the Buchenwald concentration camp only months earlier, and perhaps, in passing, an opportunity to stake out an ideological position contrasting with the Carrel Foundation, widely perceived as the Vichy research institute par excellence [17]. But not least, the circulation and transmission of ideas involve a reappropriation and flexible use of scientific constructs.

14Stoetzel’s repeated reference to Halbwachs’s work is based on similarities of purpose and theoretical developments, reinterpreted and redirected by the younger man in terms of his own priorities: population as a major structure in social morphology, the effect of sociological factors in determining apparently more biological demographic phenomena, such as the sex ratio at birth (Halbwachs et al., 2005), and the importance given to social representations (Marcel 1998), which did not exclude a certain culturalism, slipping at times into national psychology [18]. Since as a professional pollster he presented public opinion as an object, while as a scholar he critiqued the concept sociologically, Stoetzel was no stranger to paradoxical positions (Gingras, 2002). He chose to see in Halbwachs not so much a faithful heir of Durkheimian sociology as a necessary intermediary converting that sociology and setting the stage for a French version of social psychology, in line with what had long been developed in America.