The Rise of the Conjuncturists: Building Economic Expertise on the Fall of the Popular Front

Abstract

This article traces the emergence of a new type of economist in interwar France—the conjuncturist—through a study focusing on Alfred Sauvy and Robert Marjolin. We argue that these neglected figures helped to shape a new, autonomous, field of expertise that consisted of diagnosing and forecasting the economic situation to guide public decision making. As we show, the history of the conjuncturists is closely linked to that of the Popular Front in general, and to its emblematic law on the forty-hour week in particular. By becoming the most vocal opponents of this law, the conjuncturists fomented an open mutiny against the very government that had given them their first prominent position, in order to obtain the repeal of the forty-hour week, which Sauvy achieved in November 1938. Although the Popular Front was by then a thing of the past, and a future war with Germany had become the most likely outcome, the figure of the conjuncturist had succeeded in firmly occupying the institutional landscape of 1940s France—a form of economic expertise that was henceforth inseparable from political activity itself.

Introduction: a new type of economist, a new style of expertise

In November 1938, after years of persistent requests from economists, the French government created the Institut de conjoncture under the auspices of the fledgling Ministry of National Economy. This institute was headed by Alfred Sauvy (1898–1990), an established economist–statistician and a trusted adviser to Paul Reynaud, then minister of finance. That same year, Sauvy published his Essai sur la conjuncture et la prévision économique, the first introduction to the study of economic observation and forecasting in France.¹ When the young Robert Marjolin (1911–86), another promising figure in the field of economic observation, reviewed Sauvy’s 1938 essay, he not only praised its “high scientific value” but also stressed its “topicality.”²It was obvious to Marjolin that Sauvy wrote his book “under the influence of the economic events of which France has been the theatre for the last two years.” This influence had to do with the social laws passed by the Popular Front in June 1936, in particular a law that was highly symbolic of the workers’ desire for profound change: the reduction of the working week to a maximum of forty hours.

Sauvy’s Institut de conjoncture was arguably the most vivid embodiment of a deeper process at work since the early 1930s, that of the installation of an entirely new variant of economic expertise in French politics. Documenting how this expertise became institutionalized and autonomous is the central aim of our article. As we shall see, this specific style of public expertise was in principle carried out by a new type of economist, called at the time a conjoncturiste.³The role of the conjuncturist included, but went beyond, that of the statistician, in that it aimed to capture quantified and unquantified information in order to diagnose and, ultimately, forecast the macroeconomic situation by providing relevant empirical knowledge. Thus, if Sauvy and Marjolin happened to contribute to the fabrication of data, their essential characteristic as conjuncturists was to interpret such material. In this way, conjuncturists acted as mediators between the economic and statistical information produced by professionals on the one hand, and the decisions taken by political leaders on the other hand. The conjuncturist style was essentially expertise-oriented, but how did it gain such momentum in mid-1930s France?

As the literature in history and sociology of expertise has amply shown, the entry of economists into politics was clearly a gradual phenomenon throughout the Third Republic (1870–1940), with the Popular Front experiment (1936–8) being the most decisive moment of all.⁴ Indeed, it was the first Blum government that made the conjuncturists enter the state apparatus. From this position of influence, the conjuncturists paradoxically singled out the forty-hour week not only as the “dominant factor in the whole experiment” of the Popular Front,⁵but also as the “main reason for the failure of the 1936 recovery plan.”⁶For them, the forty-hour week was much more than a chronic economic problem; it had become a true obsession, especially given the arms race with Germany that was becoming more and more urgent by the day.

In this article, we trace the close links between the history of the forty-hour-week debate and the installation of conjuncturists such as Sauvy and Marjolin as leading economic experts. More precisely, we argue that the way in which they weighed in on the debate about the forty-hour week sheds light on the importance of their role as experts to the political authorities. They were the most vocal opponents of the forty-hour week and they united in a mutiny to get the law repealed by the very government that had for the first time given them a prominent place. This role can be understood through the lens of the emergence of a new field—in the Bourdieusian sense—of economic expertise situated in relation to the political field. In other words, the mutiny of the conjuncturists is indicative of the growing autonomy of the field of economic expertise at the time.

We approach “field autonomy” through three key aspects: the evolution of economic expertise into a profession with distinct epistemological and normative standards (compared to other types of expertise); the increasing capacity of experts to influence economic policy through the generation of diagnostic knowledge recognized as scientific in the political realm; and the institutionalization of this expertise in various structures, notably exemplified by the creation of the Institut de conjoncture in 1938, coinciding with the decree-law abolishing the forty-hour workweek. This autonomy explains how the experts established for themselves a lasting position within the state structure of the Ministry of the Economy, thereby acquiring the capacity to retain their position through political changes. By combining intellectual history with the Bourdieusian sociology of ideas (mainly through the concept of the field), our article is in line with the methodology outlined by Stefanos Geroulanos and Gisèle Sapiro.⁷

By documenting the rise of the conjuncturists, this article contributes to two main bodies of literature. First, the article documents the diversity of economic expertise in interwar France by shedding light on the institutionalization of a new type, hitherto neglected in the literature. The emerging conjuncturists differed from other competing—and at the time already established—styles of economic expertise, such as the statistical expertise conducted by the officials at the Statistique générale de la France,⁸ the financial expertise of civil servants (mainly from the corps of inspecteurs des finances, less often from law professors) at the Ministry of Finance and the Banque de France,⁹ or even the microeconomic expertise of state engineers in charge of assessing the utility of public infrastructure and transport.¹⁰ Each of these styles of expertise will be discussed in this article in order to paint a clearer portrait of the conjuncturists.

Second, while the rise of the conjuncturists was new in the French context, it echoed in fact a broad and vibrant international dynamic. Indeed, after the First World War and especially during the Great Depression, the industrialized countries experienced a boom in empirical and quantified studies of the world economy.¹¹ The role of leading interwar economists in research centers responsible for improving the quantity and quality of economic data and for conducting business forecast, as well as the expertise they provided to policy makers (and business leaders), has been the subject of several studies in recent years.

For instance, Walter A. Friedman’s Fortune Tellers told the story of the first economic forecasters in the United States, with particular emphasis on the difference between the work of Charles J. Bullock and Warren Persons in a private agency like the Harvard Economic Service, and that of Wesley Mitchell in government agencies and at the National Bureau of Economic Research (NBER).¹² In Europe, the work of Jan Tinbergen at the Dutch Centraal Planbureau and the League of Nations has been carefully considered by historians because of the pioneering way in which he introduced macroeconometric models into the expert’s toolbox.¹³ From a more transatlantic perspective, Laetitia Lenel retraced the dissemination and methodological adaptation of the Harvard index (the famous ABC curves) throughout Europe. Lenel showed how the apparently mechanical approach to the barometers (given by the respective variation of each curve) was soon replaced by a more informal analysis linked to the personal exchanges that the forecasters had with the decision makers.¹⁴

There is no doubt that French conjuncturists resembled their international counterparts in many respects, particularly in terms of their object of study and the tools they used: a national or global economy objectified by quantitative indicators. From a methodological point of view, Sauvy and Marjolin adopted an approach reminiscent of Mitchell’s “educated intuition,” advocating constant observation rather than recourse to a well-articulated theory.¹⁵ And yet there were also certain peculiarities to the French conjuncturist style, the most striking of which was its lack of interest in forecasting as an exercise in futurology, which was standard practice in the international research centres mentioned above.¹⁶ On the contrary, we could use the term “nowcasting” to qualify the practice of the conjuncturists, understood as a kind of public-policy assessment that focused on monitoring the current situation informed by the recent past and occasionally by (very) short-term forecasts. Overall, French conjuncturists were less interested in charting the anatomy of the business cycle, a central focus of forecasters like Mitchell, than in understanding the implications for the domestic economy of this or that recent reform (in our case, the forty-hour week) in the context of current government policy.

The Popular Front brings in the conjuncturist vanguard

Robert Marjolin, a young autodidact in Blum’s service

When Léon Blum came to power in June 1936, he created a “General Secretariat” attached to the government. This sort of brains trust was chaired by the leftist politician and engineer Jules Moch, and involved about fifteen socialist technical experts, including economists such as Étienne Antonelli and the young Robert Marjolin (who was twenty-four at the time). The political use of a professor of economics like Antonelli (former SFIO deputy, from 1924 to 1932) was neither new nor surprising. Generally speaking, the professionalization of political staff in economic matters had been underway since 1919.¹⁷ The most expert parliamentarians of all were precisely those who already held a chair in economics, as in the case of Antonelli, and of André Philip (a professor in Lyon, elected SFIO deputy in 1936), who was a passionate advocate of the forty-hour law and acted as its spokesperson (rapporteur) in parliament. Thus, in the cases of Antonelli and Philip, the economist and the parliamentarian were one and the same person, and their economic expertise was mobilized to the extent that they were already political insiders. The figure of Marjolin was far more symptomatic of the growing demand for a new kind of economic expertise from people outside the political and administrative establishment itself.

When Marjolin was appointed Blum’s expert adviser, he was by no means an academic economist (he neither followed the standard curricula nor had a degree). Political activist, journalist, data maker, a perpetual student but already a respected researcher, Marjolin wore many hats at once. As a result, his intellectual and professional career took turns. In 1925, at the age of fourteen, Marjolin began working in menial positions at the Paris Stock Exchange and he joined the SFIO in 1929, a dual commitment that would lead to his interest in economic, financial and monetary issues. After resuming his studies in philosophy in 1930, he turned to economic sociology and spent a year at Yale University from 1932 to 1933 as a Rockefeller fellow. There, on the other side of the Atlantic, Marjolin grew more and more concerned with social mobilization and set himself the goal of writing a short memoir about US labor movements, which was eventually published as L’évolution du syndicalisme aux États-Unis. ¹⁸ Returning to France after his American experience, Marjolin felt that he needed to master statistical techniques in order to take his research to the next level.¹⁹

In the US, Marjolin had witnessed the first phase (the “Hundred Days”) of the New Deal. He developed a keen interest in Roosevelt’s policies, on which he published a short volume for the SFIO press entitled Les experiences Roosevelt, becoming an authority on recent US economic policy.²⁰ This book attracted the attention of Blum, who contacted Marjolin in early 1935. Blum entrusted Marjolin with a column in the party newspaper Le Populaire. In a series of articles published at the 1935–6 turn of the year, Marjolin gave a very positive assessment of the ongoing “Roosevelt experiments,” and used them to dismiss the French proponents of financial orthodoxy. According to Marjolin, the spectacular recovery experienced by the US economy had clearly demonstrated the value of increasing purchasing power. In particular, he emphasized three key measures of the National Industrial Recovery Act of 1933: the policy of public works, the implementation of collective agreements and the reduction of working hours with no reduction in wages.²¹

At the same time as he was becoming involved in economic journalism, Marjolin joined the Institut scientifique de recherches économiques et sociales (ISRES) as a research assistant specializing in economic conjuncture. The ISRES had only recently been founded—in October 1933—by Charles Rist, a liberal economist who was an expert on monetary issues, together with Gaëtan Pirou, the editor of France’s premier economic journal, the Revue d’économie politique.²² Rist was a highly respected figure in the profession, both nationally and internationally. Sixty years of age in 1934, he was also probably the academic economist of his generation most involved in practical and political issues, serving on various boards of private companies and in the civil service (notably as deputy governor of the Bank of France), which explains his interest in economic information and forecasting.

Rist created the ISRES thanks to Rockefeller funding, in line with the US foundation’s policy of developing and standardizing the social sciences in Europe.²³ The ISRES became de facto one of the leading organizations dedicated to economic expertise in France.²⁴ Its main goal was to “promote the use of scientific methods in the study of economic and social phenomena,” which meant collecting, producing and publishing observations on the evolution of key indicators such as prices, wages and production.²⁵

The ISRES also sought to develop partnerships with other similar institutes abroad. Links were established with more than twenty institutes around the world, and the first direct exchanges included Lionel Robbins (the London and Cambridge Economic Service²⁶) and Ernst Wagemann (the Institut für Konjunkturforschung in Berlin²⁷), who were invited to give lectures on their working methods. Other organizations contacted included leading centres of business cycle research, such as the Cologne and Kiel Institutes in Germany,²⁸ the NBER in the US,²⁹ and the economic services of the League of Nations in Geneva.³⁰

Located in a bourgeois apartment near the Luxembourg Gardens in Paris, the ISRES employed a dozen permanent members (up to sixteen in 1937), equally divided between “technical” and “administrative” staff. The first group was all male, the second (typists, illustrators, librarians) all female. In addition to Marjolin, there was another research assistant learning the profession of conjuncturist at the ISRES, Philippe Schwob (two other young men completed the research team: Henri Lemaître and John H. Herberts). Like Marjolin, Schwob was trained in philosophy, but had graduated in law and had completed his PhD.³¹

Despite their young age, both Marjolin and Schwob made a name for themselves in the economics profession and emerged as the faces of the new career of conjuncturist. Thanks to their proximity to Rist and their command of the English language, Marjolin and Schwob were able to cultivate a fairly extensive network of contacts abroad. They also enjoyed regular study trips abroad (especially in the UK) and made contributions to leading international journals such as Economica and the Economic Journal. In London, Marjolin recalled, he often discussed economic and political matters with colleagues at the London School of Economics, such as Lionel Robbins and Friedrich Hayek, both of whom were involved in the forecasting centre called the London and Cambridge Economic Service, as was John Maynard Keynes (also editor of the the Economic Journal).³² Marjolin’s contacts with Keynes at this time are difficult to assess, although he was among the early French readers of the General Theory. Thanks in part to Marjolin’s international connections, the ISRES had in a short time acquired a “considerable reputation” in the field of economic information.³³

Unfortunately, we have no direct account of the discussions within Blum’s Secretariat. In his memoirs, Marjolin recalled lively exchanges with Blum and Moch, in which he often appeared as an “opponent from within” who ultimately endorsed the position of a “liberal economist.”³⁴ Marjolin had been calling for devaluation since August 1936, at a time when the government wanted to avoid this remedy, before moving on to denounce the error of the forty-hour week. In both these respects, Marjolin was very much in line with his friend Raymond Aron,³⁵ who, although convinced of the virtue of state organization of the economy, had denounced the whole redistributive logic of the Popular Front’s economic policy as doomed to failure.³⁶

After the fall of the first Blum government in June 1937, Marjolin formally remained in the Secretariat, but in a very distant manner, and he finally abandoned this position in March 1938, at which point he broke off all contact with his socialist friends.³⁷ Of all the individuals studied here, Marjolin is the one whose intellectual trajectory varied the most. Admittedly, he was very young at the time, but there is something more fundamentally elusive about his political stance, so much so that commentators have sometimes called him a socialist, a (Keynesian) liberal or even a neoliberal.³⁸

The X enter the new Ministry of National Economy

The economic policy of the Popular Front was a combination of different and sometimes conflicting orientations. It is not very surprising that a political coalition ranging from communists to radicals would lead at best to compromise, at worst to discord.³⁹ However, the lack of a clear and coherent economic policy was due not only to the coalition of heterogeneous parties, but also to the emergence of a new kind of economic expertise. Although the first Blum government broke with financial orthodoxy to some extent, tensions over spending and the reduction of working hours arose between the various ministries with an economic purview, in particular between the old and mighty Ministry of Finance on the one hand, and the new but ambitious Ministry of National Economy (MEN)on the other.

When Charles Spinasse took office at the MEN, he brought into his service state engineers such as Jean Coutrot, Jacques Branger and Alfred Sauvy—often referred to as “the X” after the nickname of the prestigious school from which they graduated, the École polytechnique. All three were members of X-Crise, a think tank considered to be the “main laboratory of economic ideas” in 1930s France.⁴⁰ Founded in 1931 by polytechnicians to find scientific and moral answers to the Great Depression that was sweeping the world, X-Crise was open to intellectuals, politicians and businessmen.⁴¹ In bringing the X into the MEN, Spinasse departed from the long-standing tradition of seeking economic expertise mainly from senior civil servants attached to the corps of inspecteurs des finances. Although an ambitious experiment, the MEN was too short-lived (it returned to being a sub-ministry under the Ministry of Finance with the fall of the Blum government in June 1937) to stimulate the deep structural changes in the public management of the economy that it was intended to bring about.⁴² Nevertheless, this episode marked a decisive first step towards the permanent incorporation of the conjuncturists into the state apparatus.

From the outset, the question of the impact of the forty-hour week on cost prices arose among the MEN technocrats. Spinasse was anxious to encourage increased production, and in March 1937 he stated, “if the forty-hour law leads to an increase in cost prices, it will be a fatal blow to the government’s experiment.”⁴³ But Spinasse seemed confident that the upward pressure on cost prices could be offset by a downward effect caused by investment in new equipment and better work organization. In other words, if productivity gains could counterweigh the increase in wages due to the forty-hour week, cost prices would remain stable.

Spinasse’s warning and the remedy he envisaged were not just wishful thinking. On the contrary, as soon as he took up his post at the MEN, the minister gave himself the means to realize his ambitions. He wanted to benefit from the economic expertise of the X in two complementary ways. First, on the solution side—that is, facilitating productivity gains—in November 1936 Spinasse created a National Centre for the Scientific Organization of Work for the Lowering of French Cost Prices (Centre national d’organisation scientifique du travail pour l’abaissement des prix de revient français), known as COST. Officially directed by Jacques Branger and Claude Bourdet, COST was in fact supervised by Jean Coutrot, one of the main French advocates of the scientific organization of work.⁴⁴ Second, on the evaluation side—that is, to measure the real impact of the social legislation on prices—in September 1936 Spinasse set up a National Committee on Price Surveillance (Comité national de surveillance des prix), chaired by himself and composed of twenty members, including Sauvy as representative of the Statistique générale de la France (SGF).

In addition to this function, Sauvy’s part-time job at the MEN consisted of writing notes de conjoncture for the ministry’s staff. The purpose of these notes was to select, digest and interpret the data provided by the SGF. While the material existence of these notes remained elusive to commentators for many years, we found examples of them in Reynaud’s archives. These notes dealt with both the general economic situation and the impact of the policies pursued by the minister.⁴⁵ Sauvy’s notes were used by Reynaud as ingredients (mixed with notes from other advisers) for his many speeches to the parliament and on the radio, as well as to prepare responses to anticipated criticism from political opponents.⁴⁶

The content of Sauvy’s original notes to the ministry was very similar to what he wrote in “Le point économique,” a monthly chronicle that appeared at the beginning of each X-Crise bulletin from 1937 to 1939 (more on this below in this article). These short pieces were written by Sauvy alone, but they were based on the collective work of the équipe de conjoncture (a subgroup within X-Crise). The aim of this team was to provide a regular update on the economic situation. On the basis of the relevant information, the team had to provide an “interpretation rooted in a few fundamental curves,” and, incidentally, give an opinion on future “trends.”⁴⁷Sauvy’s method of analysis was to observe and interpret the various indices (of production, unemployment, prices, etc.) in order to produce descriptive analyses of the current situation. Rather than a proper forecast in the Anglo-Saxon sense of the term, “Le point économique” was essentially an exercise in nowcasting, as defined in the general introduction to this article.

This way of studying the economic situation—doing, strictly speaking, conjoncture—was quite new in the French landscape. It was the result of a common impulse from X-Crise members who were increasingly dissatisfied with the work of the SGF. The first signs of such frustration were already evident in 1935, in the discussion that followed an X-Crise lecture by Léopold Dugé de Bernonville, deputy director of the SGF. In his lecture, Bernonville presented various statistical indices produced by his institution. He stressed that the “conjuncture” tracked by the SGF had nothing to do with “drawing up and publishing forecasts,” and even less with assessing the impact of governmental policies, for the good reason that this was simply “not part of its remit.”⁴⁸Sauvy, of course, knew this position only too well, having worked at the SGF himself since 1922. And yet he argued for an active line of data interpretation. In essence, this disagreement between the SGF and emerging conjuncturists embodied divergent approaches to objectivity. While the SGF advocated the production of data alone in the name of neutrality, the conjuncturists stressed that it was the interpretation of data using scientific methods that provided a neutral picture of the economy. The aim was to show the economic world as it really was, thus emphasizing the disinterested, and therefore autonomous, nature of this “new science, called conjuncture.”⁴⁹

There were, however, other voices in the discussion following Bernonville’s lecture, voices that took a different approach from Sauvy’s. Indeed, Robert Gibrat and the Guillaume brothers (Georges and Édouard) stressed the need to establish systematic and logical relationships between the quantities measured—in short, they urged the use of mathematical models to provide automatic economic forecasts. In X-Crise, these advocates of economic modelling in the name of “rational economics” were gathered in the econometrics team.⁵⁰ Extremely attentive to the developments of the fledgling Econometrics Society, this team was far more enthusiastic about Jan Tinbergen’s models of the business cycle, which he was invited to lecture on,⁵¹ than about the Mitchell-type research developed at the NBER.⁵²

A conjuncturist like Sauvy and the econometric team did share a central focus on facts and quantification: empirical economics had to go beyond what was then practiced at the SGF, namely the mere collection of index tables. It also had to go beyond the synthetic economic barometers à la Harvard that had fallen into disuse after the 1929 crisis.⁵³But while the econometrics team called for more robust modeling by using “at least some theoretical notions of the business cycle,”⁵⁴Sauvy did not believe that such abstract models were the appropriate solution for an accurate and useful short-term study of the economy. Essentially, the two teams diverged in their approaches to economic fluctuations: Sauvy advocated a more intuitive approach based on induction from the widest possible variety of empirical sources, while the proto-econometricians advocated theoretical modeling on an essentially deductive basis. After the Second World War, this kind of methodological disagreement was at the heart of a battle in the US between the NBER and the Cowles Commission, which came to be known as the “measurement without theory” controversy.⁵⁵

As a matter of fact, the two types of X-Crise commitment to the empirical method in economics (econometrics versus conjuncture) did not cooperate very well. While Gibrat and Georges Guillaume were indeed part of the équipe de conjoncture, Sauvy, to the regret of the former, never used or even mentioned the economic models developed in the context of X-Crise.⁵⁶ In his memoirs, Sauvy was particularly dismissive of the work of the Guillaume brothers. He recalled that when Minister Spinasse had asked them to give their opinion on the economic situation in France from the first quarter of 1937 onwards, their model-based forecasts proved to be completely wrong in relation to the observed course of the economy, and therefore of no practical use.⁵⁷

The debates about the best way to provide policy makers with tools for analysing the national economy were clear signs of the emergence of a new field of economic expertise. Within the SGF, the production of statistics was perceived as a straightforward response to a government order. Conversely, as part of an initiative aimed at developing economic expertise, the conjuncturists’ goal was to support the need for a scientific interpretation of the data to serve as a framework for government decision making. The legitimacy of this new expertise was established on the basis that the conjuncture was a depoliticized empirical science. In their desire to introduce scientific thinking into bureaucratic routines, the conjuncturists endorsed a position very similar to that of the X engineers of the Ponts-et-Chausséesas analysed by Theodor Porter.⁵⁸ However, in terms of tools (cost–benefit analysis), scale and subject matter (microeconomics related to specific engineering structures), their expertise was completely different from that of the conjuncturists.

Turning the economy into a “glass house”

Transparency: overcoming the chronic lack of economic information

Although different characters with different backgrounds and interests, Marjolin and Schwob at the ISRES as well as Sauvy and part of his team at X-Crise shared the essence of the conjuncturist’s mission: to provide and interpret information about the day-to-day functioning of the national economy. In order to make an informed diagnosis of the economic situation, conjuncturists had to rely on both technical skills and personal judgment. From this perspective, conjuncture was more of an art than a science.

The variety of data and information the conjuncturist had to master was immense: the many “quantifiable and even unquantifiable facts” had to be combined with global politics and individual psychological factors. Interpretation, therefore, required a “flexible mind” and a “broad general culture,” far removed from any hermetic hyper-specialization.⁵⁹ Still, access to quality and accurate information was the conjuncturists’ basic concern, as it is the very bedrock of interpretation. Perhaps Sauvy put it best in 1936 when he linked such a broad and crucial issue as better income distribution to a basic question of information:

The social problem is currently unresolved because it is poorly formulated. On many questions, agreement would be easy if the parties were sufficiently informed. An abundance of objective information (documentation) appears even more necessary [in organized capitalism] than for any other system. This information … should receive all the necessary diffusion so that all classes are widely penetrated by it. To be healthy, the house of tomorrow must be a glass house.⁶⁰

Sauvy’s image of a “glass house” (maison de verre) was very personal to him, but the underlying idea was far from unique among economists. In fact, the quest for better and more widespread economic information embodied by the X in the 1930s had been increasingly shared by others in the economics profession since the First World War. Indeed, they found that, just as in many other areas related to economic and statistical sciences, France was seriously lagging behind other developed countries.

This observation reflected a more general sense of “backwardness” that would affect both the French economy, which performed disastrously by any international standard throughout the 1930s,⁶¹ and French economics, which many practitioners, including university professors (e.g. Gaëtan Pirou and François Perroux), felt was still too much mired in legal reasoning.⁶² To confine ourselves to the field of statistics, there were indeed significant differences between France and, say, Germany: in terms of manpower, resources and public authority, the Weimar Republic provided the Reich Statistical Office with a “scale of support” for economic research and information that was unparalleled in Europe.⁶³

The demand for more information took a new and more acute turn in the second half of 1936, when the discourse on the lack of data became inextricably linked with the forty-hour week. While this association became the spearhead of the opponents of the reform, it can also be found among some of its early supporters. During the parliamentary debates, the SFIO economist Philip had called for the government and the SGF to “carry out a much more thorough study than has so far been possible of all the elements of cost prices in the various industries.”⁶⁴ Just as the Popular Front was more concerned with economic knowledge than Sauvy would have us believe, the movement in favor of economic information and conjuncture studies was deeper and more diverse.⁶⁵

For the conjuncturists who became the face of the struggle against the forty-hour reform, the origin and the error of its introduction, as well as its maintenance, were one and the same: the lack of assessment and anticipation based on thorough empirical economic knowledge. On a question as crucial as the skills of the unemployed, the équipe de conjoncture noted that, “just as in all other areas, the lack of information is very much felt.”⁶⁶A couple of years earlier, Marjolin had complained that France, unlike other European countries, had no proper way of measuring unemployment.⁶⁷ In fact, Sauvy declared, “no one in France can boast of knowing the country’s economy as well as one should, because the necessary information has not been produced.”⁶⁸Unsurprisingly, then, the forty-hour week was introduced without the “critical studies,” even the “most elementary,” that should “always accompany reforms of this magnitude.”⁶⁹And the choice of “forty” hours, to the exclusion of any other number, would not have been based on a rational and “precise calculation,” but essentially on the ridiculous “appeal of round numbers.”⁷⁰ The conjuncturists, as we shall see later, produced their own empirical studies claiming that this policy was hampering economic activity.

According to the conjuncturists, the lack of data and forecasts was not just bad luck, but had a very deliberate political origin. It stemmed from the general “contempt for statistics and economic forecasting in France”⁷¹—a contempt that was widespread in society, from public opinion to the ruling elite. A vivid illustration of this can be found in a lecture given by René Belin, leader of the Confédération générale du travail (CGT) workers’ union, on 5 February 1937, at X-Crise. In his speech, Belin expressed skepticism about gross numerical evaluations of wage fluctuations and thus rejected the indicators of the SGF (a fortiori the calculations of the para-official teams), while proposing other figures linked to this or that specific profession. Belin went on to undermine the conjuncturist spirit that was developing at the time by stressing how “really comfortable it must be to indulge in calculations and assessments of averages in the tepidness of a study room.”⁷²

In a note to Belin’s lecture, Sauvy engaged in a battle of numbers, claiming that “a correct calculation leads to very different results” regarding indices of production, employment and wages.⁷³Adjusting series, eliminating seasonal or random variations and constructing indices—all these tasks of cleansing and formatting the data (or at least the knowledge of how to perform them) were part of the conjuncturist’s job and in fact the first steps towards proper interpretations. Later in his 1938 essay, Sauvy somehow replied to Belin’s attack on the armchair conjuncturist by stressing the importance of “maintaining constant contact with life,” and of not being confined to an ivory tower. Indeed, in addition to his scientific research, the conjuncturist had to make “daily observations as the man in the street perceives them” in order to avoid creating a “gap between a small number of insiders and public opinion.”⁷⁴

In any case, the general blindness of economic actors (in particular of the government) was all the more damaging, the conjuncturists argued, because information on the economy and the performance of the economy were inseparable. This view was based on a double argument, political and economic.

On the one hand, the conjuncturists believed that if citizens shared the same picture of the overall economic situation—by looking through the glass house—they would somehow agree on economic policy irrespective of class and doctrinal differences. The immediate impression might be that the conjuncturists were advocating enlightened and open political deliberation. Rather, they expected the views of citizens to eventually converge with those of experts (i.e., in the case of the forty-hour week, with their own views).

On the other hand, economic knowledge was essential for agents to behave more efficiently in terms of consumption, investment decisions and the like. While the need for rationality applied to both private and public actors, the problem was even more pressing in the case of the latter, since the government’s management of the economy seemed destined for rapid and inexorable expansion, making its mistakes and failures damaging nationwide. In early 1938, X-Crise conjuncturists took this argument a step further, suggesting that the poor state of French information was one of the main factors in France’s inability to benefit from a global recovery that virtually all other countries had enjoyed “regardless of the regime and the skill of their governments.”⁷⁵

Sauvy’s assessment of the economic amateurism of the Popular Front was later strongly criticized by the historian Michel Margairaz.⁷⁶ The parliamentary debates of 1936 show that the law on the forty-hour week was supported by various studies. Philip and Blum defended their reform using a comparative argument: the USA, Russia, Italy, Czechoslovakia and Bulgaria all reduced working hours without this causing an increase in cost prices that would hamper their economies.⁷⁷ In addition, Blum and Philip identified a number of economic challenges that needed to be considered to ensure that the measure did not have a negative impact on the economic performance of French businesses (such as the costs of reorganizing production). Thus Sauvy’s judgment of total unpreparedness was therefore questionable. It was also highly questionable that Sauvy and the conjuncturists completely overlooked the Popular Front’s arguments about the long-term dynamics of improving the living conditions of the working class by redistributing productivity gains to it, which were just as important as the purely economic arguments (reducing unemployment and increasing purchasing power).

Efficiency: measuring the performance of the national economy

To build the glass house of tomorrow, Marjolin and Sauvy worked together in two complementary directions: on the one hand, they aimed at taking the pulse of the economy on the finest possible time scale, and on the other hand, they promoted medium-term research to shed light on some crucial—yet poorly understood—current features of the national economy. As Thomas Angeletti recently showed, the concept of a French economy was not an immediate and naturally given entity, but had to be constructed.⁷⁸ The designation of a specific area of activity (the national economy), its relative autonomy, and the designation of a group of individuals responsible for the production of knowledge about that area served to make it an objectified and even naturalized entity. By participating in this process, the conjuncturists delineated their own object of study and endowed it with specific laws and mechanisms.

Regarding the two points stressed by Sauvy and Marjolin—taking the pulse of the economy and supporting medium-term surveys—not everything had to be done from scratch; in other words, the house was not completely bathed in darkness. At the beginning of the twentieth century, the SGF developed and became professionalized under Lucien March’s leadership.⁷⁹It had set up an ambitious professional census in 1896, as well as new surveys on consumption in 1907, on industrial unemployment in 1908, and on retail prices in 1911.⁸⁰ It is therefore not surprising to find two leading conjuncturists at the SGF, polytechnicians Jean Dessirier and Alfred Sauvy; the two men, although from different generations, shared in the interwar period a very interconnected path. Sauvy had joined the SGF in 1922 and replaced Dessirier in 1929. Dessirier had tried to establish a set of forecasting tools, such as price indices, at the SGF in order to give France its own economic barometer.⁸¹ He was following in the footsteps of March, who had established close connection with the Harvard Economic Service in the 1920s.⁸²

However, Dessirier grew increasingly dissatisfied with the SGF’s reluctance to carry out predictive analysis and set up his own “artisanal observatory” devoted to publishing La conjoncture économique et financière.⁸³ This monthly bulletin later became the basis for Dessirier’s “Point économique,” a two-page rubric that sporadically opened the X-Crise bulletins from June 1933. The équipe de conjoncture was then formed to reflect on this theme in 1935, initially directed by Coutrot, and Sauvy took over leadership at the turn of 1936–7. It was also at this time that Sauvy took the helm of “Le point économique,” once again continuing an effort initiated by Dessirier. Like Dessirier, Sauvy was deeply convinced of the need to make economic information widely available, in contrast to the habit of confidentiality that characterized the SGF.⁸⁴

The arrival of the Popular Front in power coincided with a renewed and more sophisticated form of “Le point économique,” which became a regular—monthly—chronicle in the X-Crise bulletins. From the first (January) issue of 1937, the team set itself the double task of establishing a “rational terminology” and developing “deductive predictions.”⁸⁵Basically, the point focused on discussing indices (unemployment, wholesale prices, industrial production, financial and monetary values) often displayed in tables but interpreted in a descriptive way (and without any underlying model). Also, the domestic situation of the French economy was always connected to the international context. Progressively, during the year 1938, factual findings (constatations) were explicitly separated from a general and conclusive part called interpretation. Throughout its existence, “Le point économique” continued to grow, going from three pages in 1937 to six in 1938, and reaching ten pages in the final couple of issues in mid-1939.

Parallel to X-Crise and its “Point économique,” Rist’s ISRES had developed its own economic information, L’activité économique, published in association with the Institute of Statistics of the University of Paris. Marjolin and Schwob were among the main creators of this quarterly, first issued in April 1935, which quickly became a leading source of economic information next to the SGF. Again, the Popular Front contributed to the success of ISRES, as Charles Spinasse ensured and even strengthened its development by securing a 600,000-franc credit from the parliament for personnel and equipment.

Contrary to what is often assumed in the literature, which strictly separates the culture of economic engineers from that of university professors, academic circles were not entirely aloof from the progress of economic information and conjuncture. In particular, the Revue d’économie politique opened its pages to conjuncturists. From 1928 to 1939, Dessirier published an annual analysis of stock market movements (entitled “La bourse des valeurs”). Schwob authored several articles with Léonard Rist (Charles’s son) on the current monetary situation, but above all he contributed a regular section entitled “Jugements sur la conjoncture française” from 1937 to 1938. Schwob wrote eleven articles under this heading over two years, to which we can add Sauvy’s contributions.⁸⁶ Hence the conjuncturists often portrayed themselves as doctors at the bedside of the national economy, trying to record its daily palpitations. But the conjuncturists were also interested in longer-term studies on structural issues. The studies of unemployment and productivity deserve our attention here because both were instrumental in conjuncturists’ struggle with the forty-hour week.⁸⁷

The enquête sur le chômage, supervised by the ISRES, intended to shed light on the nature of unemployment in 1930s France. More specifically, the enquête aimed to document the material and psychological impact of unemployment on workers’ families through field studies and questionnaires on household management and general living conditions. Although the methodology of the survey brought together the considerations of sociologists and economists, it essentially resulted in a formidable production of statistics coupled with a meagre sociological analysis.⁸⁸ Marjolin played a central role in this survey, both as data producer and as analyst. The bulk of the survey was carried out and processed between 1935 and 1939, and the first results were available in the course of 1937 and circulated in the conjuncturist community.

In June 1937, after the fall of Léon Blum’s first government and the installation of a new one led by Camille Chautemps, the latter sponsored a major survey of French production and productivity supervised by a Comité d’enquête sur la production.⁸⁹ From August to December 1937, this tripartite structure (state, employers and CGT) aimed to document the situation of the national economy broken down into thirteen different sectors. The technical assistance for this survey was provided by the Service d’observation économique (later the Service d’observation et de conjoncture économiques) of the MEN, a department that had been headed by Sauvy since July 1937 and produced confidential notes on the “economic situation” for the MEN.⁹⁰

Clearly, these two major surveys could not remedy the entire lack of economic information at once, but they were clearly moving in the right direction according to the conjuncturists. The multiplication of data would help to solve the fundamental problem that had been the source of “social discord” over the recent reforms.⁹¹This tension would have more to do with errors of facts than of doctrines. If only the facts were better known to all, the conjuncturists seemed to believe, then decision makers (as well as public opinion) could reach at least partial agreement on the remedy to be implemented. It was the role of the conjuncturists to provide this information, which was shaped and interpreted by the new science of conjuncture.

Of course, this image of neutral economic expertise was an opportunity to defend normative positions. Presenting economic conjuncture as fundamentally disinterested (what Bourdieu called “interest in disinterestedness”⁹²) was a way for experts to increase their political influence through epistemic superiority. However, Marjolin and Sauvy did view the social order through a specific prism. They placed technological progress and economic growth as the prerequisite for social progress. Without going into too much detail here,⁹³ they were essentially fighting “economic Malthusianism,” a term (popularized by Sauvy) used to describe any policy that consisted in slowing down growth, leading to a reduction in the resources available to society, which was particularly damaging in the context of a lower birth rate. Therefore such measures had to be avoided at all costs, including production quotas, increases in retirement pensions or, as it were, the “most costly manifestation” of economic Malthusianism in the form of a reduction in the working week to forty hours.⁹⁴Hence, at a deeper level, the conjuncturists helped to establish the criterion of economic efficiency (itself reduced to productivity and growth) as the most relevant way to assess socioeconomic policies.

The conjuncturists’ mutiny

Early diagnosis of the forty-hour week

The bill, voted on on 12 and 21 June 1936, was debated in a context of intense social struggle. Following the victory of the Popular Front in May, strikes broke out and factories were occupied throughout the country. These social movements found a political outlet in the Matignon Agreements, signed on 8 June by the Blum government and the workers’ and employers’ unions, which provided for two-week paid holidays (congés payés), collective agreements and the forty-hour week (paid forty-eight hours). Forty hours were not really in the Popular Front program, but were obtained by the demands of the quasi-revolutionary situation of June 1936.⁹⁵From April 1937, when the forty-hour week came into full effect, until April 1938, when the Daladier government took office, the conjuncturists staged an open mutiny to get rid of the reform introduced by “their” government. Sauvy and Marjolin were the two main architects of this campaign, working essentially in the same direction, occasionally even together.

For Sauvy, a reform as crucial as the forty-hour week had to “remain on a purely technical and objective level,” the level of expertise, and not be polluted by “special interests,” the political level in the broadest sense of the term.⁹⁶ In other words, the conjuncturists called for a dispassionate debate on working time as a vehicle to promote measures in the general interest. Thus the conjuncturists wanted to reverse the prevailing opinion that the debate on the forty-hour week embodied the class struggle between workers and employers.⁹⁷On the contrary, all classes would benefit from economic prosperity: employers certainly, but also workers, since they would end up with far more material benefits than any form of social legislation could possibly give them.⁹⁸In the same vein, the conjuncturists rejected the hypothesis of technological unemployment that was still very much at the heart of workers’ union doctrine, for instance reaffirmed by CGT leader Belin in a lecture he gave at X-Crise.⁹⁹ Even before the forty-hour reform, Sauvy had consistently argued that technological progress was not the enemy of labor but would help it to grow in the interests of society as a whole.¹⁰⁰

Essentially, the conjuncturists’ ethos was in many points similar to that of any economic expert trying to embody the qualities of clarity, rigor, competence, consistency and impartiality. These were the conditions to obtain what Sauvy called a “rational determination of working time.”¹⁰¹ Hence, by contrast, Sauvy was suggesting that irrationality prevailed in the design and adoption of the reform. In addition to the lack of time for consultation and evaluation, conjuncturists also deplored the haste in adopting the forty-hour week in the last “Point économique” of 1936—incidentally it was the last one Dessirier authored before the équipe de conjoncture, and Sauvy, took over. More precisely, Dessirier stressed what became one of the leitmotivs guiding the utter condemnation of the reform—the idea that it was far too hasty and abrupt, highlighted by the associated lexical field: the “sudden application” was characterized by “too sharp” an increase in social charges, a transformation “neglectful” of working conditions, which would result in “too rapid” an increase in the cost of living.¹⁰²

A related point that conjuncturists also immediately raised was the inconsistency of the stimulus policy set up by the Popular Front. Two reforms seemed to be in serious conflict: the forty-hour law and the devaluation of the franc. Performed in September 1936, the devaluation brought immediate effects (in terms of rapid growth in capital goods). While in the last months of 1936, the French economy experienced a strong economic recovery—yet one limited to certain sectors—Dessirier anticipated that this upturn would be jeopardized by the progressive implementation of the forty-hour week—a forecast later confirmed by Sauvy and his team.¹⁰³ By early 1937, the conjuncturists’ concerns were growing but still relatively limited, and the open question was how long it would take for the shock to production to be absorbed—“digested”—by the French economy. At this point, however, it was already clear that the reduction in working hours was “preventing the incipient economic recovery” fuelled by the monetary devaluation.¹⁰⁴

Marjolin’s article “La vérité sur le chômage,” published in May 1937, shows that he was in a similar frame of mind to his X-Crise counterparts. In particular, Marjolin feared that the initial results of the Popular Front, which were already so meagre, would now be completely wiped out by the depressed state of the French economy.¹⁰⁵ Specifically, the problem was that the reduction in unemployment seemed to have been entirely due to state intervention (arms contracts, public works and the forty-hour week) without any recovery in the private economy, and that this state action seemed to have reached the end of what it could achieve.¹⁰⁶ When the new devaluation in June 1937 failed to produce the expected benefits of such a policy, it became clear to the équipe de conjoncture ¹⁰⁷ that the economic recovery (generated by devaluation or otherwise) would in future come up against the “technical obstacle” of a generalized forty-hour week¹⁰⁸—the conjuncturists made the same argument in the Revue d’économie politique.¹⁰⁹

For the conjuncturists, speaking truth to power meant speaking data, figures and statistics. Economic information, however, was not tantamount to economic expertise per se; it had to be interpreted and made usable for the layperson. In fact, it was the role of conjuncturists to start from empirical data compiled according to scientific canons and arrive at ready-to-use results intended for the ruling elite. A good example of this effort can be seen in the special issue of L’Europe nouvelle on “The Social Laws in Practice,” published in October 1937. Indeed, on the front page before listing the topics of the articles, the journal announces in capital letters “studies with graphs and statistics.” In truth, the number of numerical or graphic elements was rather limited, but it was still something very unusual in the press of the time. L’Europe nouvelle was a weekly magazine with high profile among the ruling elite, penned by technocrats as well as intellectuals and political figures.¹¹⁰ The special issue, thought of as a progress report on the effect of the social reform, included contributions by the conjuncturists Marjolin, Sauvy and Schwob.¹¹¹

The article that presented the most data and offered a real analysis of the conjuncture was that of Schwob. He was interested in assessing the impact of the social reforms on wages, cost prices and consumer prices.¹¹² Building on data from his own (ISRES) publication L’activité économique, Schwob estimated that the hourly wage of a worker rose by 60 percent between June 1936 and January 1937. Meanwhile, cost prices increased by 40 percent on average. In their contributions to L’Europe nouvelle, both Marjolin and Sauvy expressed deep concern about the government’s reliance on the theory of purchasing power (the discourse of the Popular Front was here imbued with US Keynesian-style considerations about aggregate demand management). Anticipating beneficial effects would be a fundamental error in the diagnosis of the “factual situation,” concerning both unemployment and “production elasticity.”¹¹³ And if technological progress (productivity gains) was supposed to compensate for rising wages, it was simply not there yet. Marjolin, Sauvy and Schwob all pointed to economic discrepancies with official interpretations—which they presented as plain diagnostic errors—associated with the forty-hour week.

Speaking truth to the ruling class: unemployment and productivity

As a conjuncturist, Marjolin perceived his main task as that of “clarifying the thousand ambiguities” that surrounded the question of working time in France.¹¹⁴By monitoring the economic situation in depth, conjuncturists would discover “unsuspected” mechanisms and compatibilities, or incompatibilities, which they could communicate to the public.¹¹⁵ Yet as the conjuncturists soon realized, abolishing the forty-hour week was not just a matter of showing its negative impact on economic performance. It was also about debunking some persistent economic views—which the conjuncturists presented as myths—and replacing them with a new set of explanations. From that standpoint, their criticism was essentially twofold. First, the forty-hour week was judged to be ineffective in dealing with the phenomenon of unemployment in France. Second, the forty-hour week would be dangerous because it generated a vicious circle of low production, public deficit and financial instability, as well as low availability of capital for investment.

For the conjuncturists, 1937 was an eye-opener as to the cause and nature of unemployment in France. By the time the forty-hour week was fully implemented in industries in the spring of 1937, partial unemployment had disappeared, and the economy was close to full employment. This result may seem impressive, but it obscured a fundamental point according to the conjuncturists: France had never had an unemployment problem in the first place. Figures in the 1936 census, which became available later, estimated the number of unemployed at 823,000, of whom only 487,000 were receiving assistance (i.e. chômeurs secourus). According to Marjolin, this total represented about 4 percent of the working population.¹¹⁶

In early 1937, all this information was unavailable, and yet it quickly became clear to the conjuncturists that unemployment had been “overestimated in quantity and quality.”¹¹⁷ Indeed, by February 1937, the équipe de conjoncture was reporting that while employment figures continued to improve, the rate of increase was showing the first signs of slowing down.¹¹⁸The problem was twofold: not only had the real level of unemployment been “overestimated,” but so had the ability of the unemployed to meet the needs of the economy by getting back to work. As a result, the French economy experienced a paradoxical situation in which labor shortages and unemployment existed side by side.¹¹⁹ On the one hand, the reform was limiting the productive capacity of the French economy by introducing bottlenecks. The forty-hour week created an “artificial scarcity of workers” because France had no reserve of skilled workers in several key industries, such as aluminum, mining and mechanical and motor engineering.¹²⁰ On the other hand, the remaining unemployed were unable to take a job because they were mainly “old, disabled or unskilled.”¹²¹ By mid-1937, the enquête sur le chômage carried out by the ISRES had produced its first results, thus “confirming that a large proportion of the unemployed are barely able to hold a normal job.”¹²²

According to the équipe de conjoncture, the government had made the “great mistake” of being obsessed with unemployment, whereas, below a certain level, a level never exceeded in France, unemployment was not a relevant economic issue.¹²³ In other words, in the current state of affairs, unemployment was less an economic than a social problem, and therefore must be addressed by a completely different set of measures than adjusting working hours (for instance by reskilling workers, as suggested by the members of X-Crise). From October to November 1937, the X-Crise group undertook its own survey on unemployment to better understand the reallocation of labor and how to promote its qualitative improvement.

In order to make public opinion aware of the inefficiency of the forty-hour week, the équipe de conjoncture measured the disparity between effort and result.¹²⁴The government had reduced unemployment by a maximum of 100,000 (over 400,000 unemployed) but at the cost of a 20 percent reduction in the attainable output; that is, a loss of 30 billion francs in production corresponding to a shortfall of 8 billion francs in annual taxes (respectively about 20 billion and 5 billion euros in 2023 values). The results were particularly unflattering when France was compared with other countries. The team emphasized how much less the fall in unemployment was in France (18 percent over the 1935–7 period) than in other countries (ranging from 30 percent in the UK to 70 percent in Germany), despite a generally greater devaluation and a reduction in working hours in France.¹²⁵

If unemployment was not the real problem that the French government had to tackle, then what was? And with what consequences for the evaluation of the forty-hour week? The conjuncturists wanted to show that there was no point in looking at the question of reducing (or extending) working hours in isolation. Rather, the issue was the “compatibility” of working time with the other economic objectives of the government, such as increasing production in general, encouraging rearmament in particular, and stabilizing the franc.¹²⁶ The coordination required by the conjuncturists was central to their style of expertise, which aimed to overcome the political isolationism that existed from one ministerial department to another, leading to inconsistent and sometimes even contradictory policies.

Contrary to the government’s claims, none of the expected results in terms of productive increase were observed by the enquête sur la production carried out from August to December 1937. But the results of the enquête were often ambiguous, even contradictory, depending on the sector of the economy concerned. In particular, the impact of the forty-hour week on hourly output was inconclusive. The conjuncturists could have dealt with these inconclusive results by updating their narrative in a more nuanced and complex way. Instead, they chose to ignore the survey’s ambiguity and stick to their initial assessment—after all, while they had expected that this survey would “undoubtedly fill in some gaps,” they also claimed that it would by no means “be able to make up for the lack of quantitative information” in general.¹²⁷ In any case, most economists, like the conjuncturists, were very keen to oppose the forty-hour week, claiming that it led to a “catastrophic fall in labor productivity.”¹²⁸In that regard, the conjuncturists expressed the majority view of the profession, and their specificity had less to do with what they said than with how they tried to demonstrate and enforce such a claim.

Framed in terms of productive capacity, the need for rearmament presented the French people with what the conjuncturists saw as a clear trade-off: either keep the forty-hour week and suffer a “lower standard of living,” or “work more” to maintain private consumption and at the same time increase military expenditure.¹²⁹ For the conjuncturists, therefore, the parallel drawn between 1919 and 1936 by those such as CGT leader Belin, to argue that the forty-hour week would not hamper industrial production, was completely misleading.¹³⁰

According to Sauvy, the crucial difference between the two situations was that in 1919, France was in a phase of disarmament, whereas since 1936 the government had been aiming at rearmament.¹³¹ Thus the introduction of the forty-hour week was particularly damaging because it added an extra burden to the existing effort to divert resources from useful production (consumer goods) towards useless production (military goods). Sauvy concluded that while the reduction of working hours in 1919 facilitated the structural transition underway, it had, in contrast, complicated it since 1936. Sauvy’s observation was even more relevant to the circumstances of 1938, when the international situation became dramatically more tense, and France’s need for rearmament even more urgent. In this context, the conjuncturists stressed that the forty-hour week was seriously out of step with the efforts being made abroad, especially in Nazi Germany where the average workweek was forty-seven hours, and overtime hours were increasing by 50 percent.¹³²

Fundamentally, the forty-hour week was presented by conjuncturists as the root of both the productive and the financial issues of France.¹³³ The government was facing a vicious circle: insufficient production meant loss of potential tax revenue for the state, which fed the public deficit, thus aggravating the instability of the franc. In this context, available capital flowed out of the country to be invested abroad, and little foreign capital was invested in France, depriving companies of the resources needed for productive investment. Even in the best-case scenario of cheap capital and full utilization of productive resources, the margin of progress of French production under the current forty-hour regulation was a maximum of 10 percent. However, a spectacular leap of about 50 percent was mandatory if the government did not want France to be irretrievably overtaken by Germany in the arms race.¹³⁴ Accordingly, for the conjuncturists, the only informed decision possible for the government to get the productivity boost it urgently needed was to return to the forty-eight-hour week.

The Reynaud–Sauvy synergy

The Third Republic was notoriously plagued by governmental instability. In the four years between the victory of the Popular Front and the defeat of France in June 1940, there were eight different governments. Édouard Daladier’s assumption of the presidency of the Council in April 1938 marked the end of the Popular Front experiment. Indeed, a series of decrees reversed the social measures of June 1936, culminating in the abolition of the forty-hour week in November 1938 under the leadership of the newly appointed finance minister, Paul Reynaud. Between April and November 1938, when the conjuncturists’ goal seemed closer than ever, they kept on voicing their contempt for the forty-hour week. For Marjolin, the urgency was clear: the new government must withdraw the forty-hour week—this “law of national treason”—for nothing less than the “national salvation” of France would depend on it.¹³⁵

Although the forty-hour week was gradually dismantled by the Daladier government, it was Reynaud who struck the final blow. Reynaud was an iconoclastic politician, an early advocate of the devaluation of the franc and an opponent of the forty-hour week from the start.¹³⁶ Furthermore, he vehemently opposed the Munich Agreement (September 1938) and called for a very active preparation for war. But Reynaud’s singularity also stemmed from the fascination he exerted over the economic–technocratic milieu; indeed, his arrival at the Ministry of Finance was greeted with a quasi-messianic fervor by Marjolin and Sauvy (but also by economists with more experience in public action, such as the financial experts Rist and Rueff).¹³⁷ Thanks to Reynaud’s diary, we know that he was in close but informal contact with Sauvy for advice from April 1934, before officially taking Sauvy into his Cabinet in November 1938. Of the future members of Reynaud’s Cabinet, Sauvy seems to have had the most contact with him, and their meetings became more frequent when Sauvy joined the Popular Front.¹³⁸ Overall, Marjolin emphasized the happy “homogeneity” of Reynaud’s entourage at the Ministry of Finance, a close circle of “young, intelligent, capable and courageous men.”¹³⁹

According to Coutrot, the campaign led by Sauvy was instrumental in “triggering or facilitating” the flexibilization of the forty-hour week, as many legislative measures and bills would have been based on X-Crise contributions.¹⁴⁰ More than that, Sauvy was charged with drafting, with Michel Debré, a series of decree-laws designed to overhaul France’s economic policy. These included the decree-law of 12 November on “improving the efficiency and effectiveness of work.” This put an end to the forty-hour week, notably by introducing the six-day week and authorizing virtually unlimited overtime (forty-eight hours without authorization) with a very low premium. Years later, Sauvy wrote at length in his memoirs about the “carte blanche” that Reynaud had given him to draft what he considered to be “his” decrees.¹⁴¹ If Sauvy was able to make himself the main protagonist of this story, it was also because conjuncturists’ economic expertise had assumed a new and critical role in the state apparatus.

For Marjolin, the November decree-law had finally settled the question of working hours.¹⁴² In a first progress report, written a month after Reynaud began as minister of finance, Marjolin highlighted the “very substantial inflow of capital” into France, which had allowed a fall in the short-term interest rate (ensuring more financing for the private sector) and an increase in subscriptions to treasury bonds (ensuring more financing for the public sector).¹⁴³Shortly afterwards, in early 1939, both Marjolin¹⁴⁴and Sauvy’s team¹⁴⁵stressed that France was experiencing a “very clear economic recovery” visible in the rise of the indicators of general activity. The average length of the working week, in particular, was increasing quickly, rising from 39.2 hours (in October 1938) to forty hours (in January 1939) then to forty-one hours (in May 1939), and with it the level of production and taxes.¹⁴⁶According to Marjolin, Reynaud’s record was clear and “leaves littles room for criticism,” for he had been the architect of a “considerable improvement in the economic situation.”¹⁴⁷

One difficulty, however, was still the lack of statistics and economic information, making it impossible to know whether this recovery had mainly benefited the defense sector (as most public opinion seemed to believe) or whether the civilian sector had also benefited.¹⁴⁸In any case, the équipe de conjoncture was keen to show people the tangible effects of Reynaud’s policy. The two last “Points économique” (before publication ceased due to the war) contained two graphs, which was rather unusual, as they normally only offered index tables.

The first graph documented output fluctuations over the previous ten years. By taking the year 1928 as reference, it shows that the output of 1939 was still below that level, but also that it reached a point not attained since 1931. In order to suggest to the reader explanations for the trends, some sociopolitical events are reported on the graph (the last one being Reynaud’s decree-law). The second graph follows the evolution of the number of workers occupied for more versus less than forty hours per week in recent months, showing a strong inversion of the lines. We found a similar hand-drawn version of this graphic in the archives of Reynaud: it was part of a set of preparatory notes for a series of speeches given in 1939 to demonstrate the benefits of relaxing the forty-hour working week.¹⁴⁹

By intensifying his contributions to economic journalism in 1938–9, Marjolin developed a clear strategy for gaining access to public opinion. The manner and frequency with which he combined his role as a researcher and data analyst with economic and political assessments of government policy in the press seemed unparalleled among his fellow French economists. In many ways, though on a much smaller scale, Marjolin’s profile is reminiscent of that of the American journalist Walter Lippman, who, though a Keynesian, became increasingly dissatisfied with Roosevelt’s New Deal.¹⁵⁰ But while the year 1939 coincided with Sauvy’s heyday in the political–administrative sphere, it contrasted with Marjolin’s relative decline. Cut off from the decision-making centers, Marjolin concentrated on research and completed his doctorate.¹⁵¹The opposite was true of Sauvy, who established himself as the conjuncturist in chief. In Reynaud, Sauvy had found the perfect political vehicle for his ideas and a strong supporter of the development of the field of conjuncture studies in France.

At that time, Sauvy did not limit himself to imposing an intellectual magisterium, but also received very concrete material rewards. In fact, the new Institut de conjoncture (IdC) was created in the same set of decree-laws (12 November 1938) that ended the forty-hour week. At its head, Sauvy was given his own institution for economic information and expertise independent of the SGF but placed under the authority of the MEN.¹⁵² Not surprisingly, the IdC’s tasks as set out in the decree were closely aligned with Sauvy’s prerogatives in recent years: the monitoring of economic trends was to be pursued through both statistics and various types of information, both quantified and nonquantified. The formulation of “forecasts of likely future developments” was also mentioned, although we have seen that this part remained marginal. The IdC was also to make itself available at the request of the government, public administrations and even private bodies, and should ultimately contribute to the “formation of an enlightened public opinion.”¹⁵³

This episode provides a compelling illustration of the emergence of economic expertise as a distinct and autonomous field. Sauvy’s contributions not only stabilized the role of economic nowcasters in the political sphere, but also gave their discourse a distinctive strength. This ability enabled the conjuncturists to delineate the range of possible options for economic policy. Nevertheless, their institutional position was still precarious. Although announced with great fanfare, the IdC was never fully set up during the Third Republic. It was opposed by the Finance Committee, which was reluctant to introduce an additional tax to finance this new institution (the IdC), which tended to duplicate the existing services, creating confusion rather than rationalization.¹⁵⁴ The so-called financial experts took a dim view of the rise of the conjuncturists, competitors who were draining resources and stepping on their toes in terms of the economic management of the state. But despite this initial setback, the story of Sauvy’s IdC did not end there.

Under Vichy France, the institute was finally attached to the SGF (29 October 1940) and then reestablished with a specific budget and about ten employees (law of 1 September 1941). At that period, Sauvy helped train new conjuncturists, such as René Froment, Jean Romeuf and André Vincent. Sauvy’s new team at the IdC produced analyses and forecasts of the moribund French economy, which were published in the form of the famous “brick-red” reports (due to the color of the cover). These confidential volumes were intended for the Vichy authorities to help them manage the corporatist organization they were trying to set up. Nonetheless, these reports were also distributed to the Free French abroad (in London and Algeria) to help them prepare the postwar economic order and the reconstruction of the country. On 21 October 1944, after the Liberation of France, Sauvy’s IdC was asked by Pierre Mendès-France (head of the MEN) to assess the cost of the German and Italian occupations.

Conclusion: autonomy without objectivity

The first Blum government, notably in the person of Prime Minister Léon Blum himself and the minister of the national economy, Charles Spinasse, brought young economic experts into office. The figurehead of this technocratic vanguard was the conjuncturist. From their positions of influence, conjuncturists like Sauvy and Marjolin gradually legitimized their empirical knowledge and expertise on the daily evolution of macroeconomic aggregates such as unemployment, production and prices. They wanted to establish themselves as the indispensable guides for public decision making. Remarkably, the conjuncturists achieved this goal not so much by constructing this or that economic policy, but mainly by undoing a social law passed in June 1936.

This expertise had a specific object, the national economy; several new instruments, embodied in the various versions of the note de conjoncture; and a set of dedicated institutions, attached in particular to the MEN. This new field of expertise gradually gained autonomy, which was tantamount to the ability to produce expert knowledge that had an impact on policy making. In the context of the economic and social policies conducted by the Popular Front, the conjuncturists’s autonomy was quickly demonstrated by their ability, as new experts led by Sauvy, to take a stand against these policies, and in particular against the law on the forty-hour week. Proof of their newly gained autonomy, these experts survived the political changeover of 1938.

Of course, winning the battle of the forty-hour week did not silence the conjuncturists in terms of economic guidance. If the abolition of the forty-hour workweek was essential to solving the production problem, it was by no means sufficient. According to them, it was also necessary, in the short term, to bring in foreign labor and, in the longer term, to encourage an increase in the birth rate and a more complete rationalization of production.¹⁵⁵Finally, by shedding light on the long-term objective set by Sauvy and Marjolin—in a nutshell, combating economic Malthusianism—we can take a critical look at the ethos of objectivity and defence of the general interest that they sought to embody. In their campaign against the forty-hour week, the conjuncturists did not simply offer a neutral and depoliticized assessment of the reform. More importantly, they also framed the debate in a way that was compatible with some specific objectives, to the detriment of others that might have been at least as legitimate, such as increasing the leisure time of workers and a strong distributive effect in their favor.

In other words, the conjuncturists helped to impose the sole criterion of economic efficiency (and from a narrow supply side perspective), to the exclusion of the criterion of social progress, which was an essential point in the defense of the working-time reform in the first place. In this respect, it is perhaps too reductive to assume that the conjuncturists’ discourse was only about the means to an end imposed externally by political power—a narrative often endorsed by both experts and politicians. Rather, we have seen that through specific tools and a unique style of economic expertise, the conjuncturists simultaneously reshaped both ends and means, as well as their overall articulation.