Stability in Numbers: Central Banks, Expertise and the Use of Statistics in Interwar Europe

Abstract

This research examines the central banks of interwar Europe through the lens of statistics. It focuses particularly on how the rise of economic and statistical expertise simultaneously supported the existing goals of central banks to retain national autonomy and the tenets of liberal internationalism espoused by the League of Nations. The institutionalised efforts to improve quantitative research culminated in the 1928 Conference of Central Bank Statisticians, where delegates envisioned creating new channels of cooperation based on standardised terminology and a centralised information bureau. By framing central banks within the historiographies of statistics and interwar internationalism, this article details the confluence of factors that shaped a new dependence on expertise.

In the spring of 1928, Paris hosted a gathering of twenty-four central banks for a conference on the collection of monetary statistics. ‘For the first time’, reported one French civil servant, ‘the representatives of almost all the central banks [banques d’émission] of Europe and the United States found themselves assembled to work together and organise a lasting cooperation.’¹ Economists and statisticians convened at the Banque de France's ornate Galerie Dorée, where they discussed potential advancements to the standardisation of statistical reports and terminology. According to one official at the Bank of England, the priority was to establish ‘personal contact between people doing the same work in different central banks on ‘purely statistical’ matters.² The meeting itself bore many similarities to more prominent statistical conferences hosted by the League of Nations in Geneva. Yet, as the only conference of central bank research departments in the interwar period, it also represented an alignment of interests among national and international institutions in the realm of statistics. By examining efforts to standardise and centralise data, this article argues that the convergent goals of central banks and the League were able to sustain a new form of interwar internationalism.

After the First World War, national governments, central banks and intergovernmental organisations initially relied on economic advisers, not because of an underlying faith in the academic discipline, but rather as a means of restoring international financial stability.³ High wartime debts exacerbated exchange-rate volatility and inflationary pressures in many countries. For central banks, which had hitherto employed bankers and industrialists, the outside counsel of trained economists helped to address these concerns. Officials then turned to experts whose use of quantitative tools provided an enhanced view of the global economy. If the classical gold standard, free trade and balanced budgets presented an opportunity to return to the supposed stability of the prewar years, then the work of experts was viewed as the means to achieve such ends. This evolution not only reflected the changing priorities of central banks themselves, but also redefined the structure of interwar economic governance, now dependent on the procurement of expert knowledge.

An extensive historiography has looked at the contributions of experts in the realms of economic development, public health, warfare and welfare.⁴ The field of science and technology studies (STS) has shown how specialised knowledge was integral to the production and transmission of new forms of policy.⁵ These works have further historicised the concept of expertise both as a political tool for enacting policy and as a reflection of shifting societal perceptions of ‘apolitical’ specialists. Meanwhile, other historians have traced the role of expertise in the history of statistics. They have shown how nineteenth-century statistical measurements and methodologies – from the fields of cartography and demography to the study of production and trade – aimed to quantify societal changes and bring order to, what contemporaries viewed as, uncertain times.⁶ The application of statistics was particularly important for emerging forms of technocratic governance. Through the language of efficiency, rationality or neutrality, experts were able to gain authority through their use of quantitative tools and measurements.⁷ By the twentieth century, government officials and economists supported the development of more complex economic methods and indicators, such as business-cycle forecasts, national income accounting and calculations of gross domestic product.⁸ As a result, the use of statistics to measure trends and influence policy had become integral to the formation of the modern administrative state.⁹

Why, then, did central banks turn to expertise? This article examines the circumstances in which statistical expertise – from the hiring of economists and statisticians to the expansion of research departments – proliferated in the aftermath of the First World War. Central bankers sought to obtain (or retain) autonomy over the publication of national monetary statistics, thereby strengthening the institution's own reputation. The establishment or expansion of an in-house research department provided one means for achieving these goals, since statistics gave greater insights into the ongoing changes in the international economy.¹⁰ As Charles Maier has shown, the work of technocrats to preserve and justify existing social hierarchies was fundamental to the restoration of a bourgeois society.¹¹ More importantly, the making of economic expertise concerned what was researched as much as who was doing the research. Experts with technical qualifications conferred legitimacy because they provided an intellectual defence of monetary policy based on a language of rationality. Raising credibility involved not only having a stable currency tied to gold, but also employing economists to make decisions appear more impartial. For many central banks, the regular publication of economic statistics proved to be an additional avenue for gaining prestige. New research resulted in an increasingly globalised network of scholars, notably centred at the Bank of England and the Banque de France.

Expertise also aligned with the internationalist goals of cooperation supported by the League of Nations. According to Patricia Clavin, interwar institutions – as well as their postwar counterparts, including the Bretton Woods institutions of the International Monetary Fund (IMF) and the International Bank for Reconstruction and Development (later, the World Bank) – had a strong proclivity towards in-house expertise, often along orthodox lines.¹² Indeed, the League's Economic and Financial Organisation employed leading technical specialists from member countries to support the compilation and distribution of statistical research.¹³ Projects to harmonise data collection, as Roser Cussó has shown, were vital in reshaping global economic governance in the interwar years.¹⁴

Building on these works, this article demonstrates the role of experts in creating a new type of ‘interwar internationalism’ through global cooperation over statistical standardisation.¹⁵ This variant drew on characteristics of liberal–internationalist rhetoric that sought to bind together self-governing, independent states through a common set of intergovernmental norms.¹⁶ Increased transparency and coordination of monetary policies helped to sustain many tenets of economic liberalism, such as free trade and open markets. In this process, in-house expertise was an attractive model because it allowed central banks to retain a degree of autonomy from international organisations, while simultaneously fostering cooperation among European states. It arose amidst a favourable political environment made possible by the Locarno Treaties, as well as the stability brought by the widespread adoption of the gold-exchange standard in Western Europe: Germany in 1924, the United Kingdom in 1925, France and Belgium in 1926 and Italy in 1927. The new international monetary standard forced the central banks of Europe to concede some amount of operational autonomy, as they had to exercise greater fiscal constraint to maintain the parity to gold.¹⁷ Yet it also permitted them to retain authority over the publication of national statistics, an area in which they could have claimed to have expert knowledge.

Thus, the following analysis examines how the proliferation of in-house expertise helped to reconcile the League's interests in sustaining liberal internationalism with the nationally oriented goals of central banks. Section I traces the establishment of central bank research departments before and after the First World War; Section II places these research departments within the wider geopolitical context of the 1920s, primarily engendered by the ‘spirit of Locarno’; Section III looks at the first and only interwar conference of research departments held in 1928; and a brief conclusion offers some broader views on expertise.

The Making of Central Bank Research

Perhaps the oldest central bank research department in Europe was that of the Banque de France. In 1872, at the behest of Governor Georges Pallain, Pierre des Essars established the Department of Economic Studies (Service des études économiques), which engaged in intellectual debates and discussions with the Statistical Society of Paris, in addition to providing intellectual opposition to the adoption of bimetallism.¹⁸ A member of the International Statistical Institute and a professor at Sciences Po, des Essars advocated for the use of statistical analysis in policy making, mainly by collecting data from issuing banks in foreign countries and assessing monetary conditions.¹⁹ While the Service occasionally exchanged information with other banks, it was primarily responsible for advising the directors of the Banque de France.²⁰

A similar responsibility was bestowed upon the German central bank. As with its French counterpart, the Reichsbank conducted significant work on domestic financial policy. In the 1890s, Reichsbank President Richard Koch commissioned a banking expert, Karl von Lumm, to help create a statistical research department that offered assessments ‘from a scientific point of view’ (vom wissenschaftlichen Standpunkt).²¹ The goal was to make technical (and seemingly apolitical) arguments in favour of the gold standard amidst a populist backlash against it.²² Further supported by another economist, Karl Helfferich, the new Economics and Statistics Department (Volkswirtschaftliche und Statistische Abteilung) thereafter aimed to employ the ‘best-qualified civil servants’ to advance the Reichsbank's agenda.²³

Yet the Banque de France and the Reichsbank were the exception, not the norm, as statistical research factored relatively little in the work of other central banks. At the Bank of England, prewar economic research had been rare, usually only encompassing the compilation of ‘about a dozen graphs, mostly of the Bank's own figures’.²⁴ It was only after the First World War that economic research gained broader appeal at the Bank of England. Under the direction of Governor Montagu Norman, a research department called the Economic Section was established in August 1921 to help officials understand domestic economic conditions. The Section began collecting statistical data from various government agencies and newspapers, including The Economist. By the mid-1920s, the department sought to hire additional staff members, especially those with foreign-language abilities that allowed them to read European publications.²⁵ There were desires not only to aggregate existing data, but also to pioneer original research in the field of statistics. Although one in-house statistician, J. A. C. Osborne, admitted in 1925 that statistical research had not yet been fully institutionalised at the central bank, it seemed ‘that a service of this kind might possess real value’.²⁶

In response to Osborne's suggestion, the Bank began weighing the possibility of inviting an external economist, possibly on a temporary basis, to support the expansion of statistical work.²⁷ As Norman admitted, ‘we are in need of a system for the collection and preparation of Statistics and other information on economic questions which may guide us in our financial policy.’²⁸ Whereas Osborne suggested finding ‘a trained Economist [who] could be consulted whenever the views of an “expert” were considered likely to be of assistance in clearing up obscure points of economics’, Norman hoped to procure the advice of an outside specialist who already had experience in the practice of central banking.²⁹ In March 1927, Norman received a recommendation from Governor Benjamin Strong of the Federal Reserve Bank of New York to invite the former Director of Research and Statistics at the Federal Reserve Board, Walter W. Stewart, ‘to build up a statistical and information office’.³⁰ Although Norman recognised the unprecedented nature of having a ‘non-British’ employee, he felt that the decision was justified given that Stewart was able to provide novel insights from an outsider's perspective and from a country where central bank research had already been better established.³¹

Expertise was also vital for statistical work at the Belgian central bank. There, new forms of research came from the publication of economic studies and the construction of an in-house library.³² Paul van Zeeland, after spending time at the research department of the Federal Reserve Bank of New York and studying at Princeton University under Edwin Kemmerer, returned to Belgium to establish the Department of Economic Analysis (Service d'analyse économique and, later, the Service des études économiques) in 1921.³³ Similar to its French counterpart, the Belgian Service aimed to streamline statistics into a single publication by aggregating the weekly balance sheets of other central banks in the process.³⁴ One of its reports discussed the overall importance of research as a basis for cooperation among central banks:

In the current state of the economic sciences and of statistics, while our information on the affairs of [Britain] is still so vague, so questionable, due to errors and fraud, direct contact with trade and industry still appears to be, as empirical as it is, the best way to audit an economic nation. The situation would be greatly modified if the central bank had precise, sincere and abundant documentation on the movement of business.³⁵

Until statistical standardisation across countries had been achieved, the Belgian central bank aimed to employ monetary specialists to interpret data published by other central banks. To this end, Van Zeeland and director Albert-Édouard Janssen hired a cohort of economists in the 1920s and 1930s.³⁶

The growth of interwar economic expertise was even more pronounced at the French and German central banks. At the latter, Reichsbank President Hjalmar Schacht searched for ways of improving policy through the research department (Figure 1).³⁷ While often presenting himself as a pragmatist without an interest in abstract economic theories, Schacht was part of the changing institutional landscape of Weimar Germany where statistical calculations and expertise had increasingly begun to shape economic policy.³⁸ Between 1913 and 1926, the number of staff members in the Reichsbank's research department increased from twenty to ninety-two.³⁹ Schacht assigned a leading economic expert, Karl Nordhoff, to lead the group. Nordhoff, who had worked at a Hamburg-based bank prior to obtaining a doctorate from the University of Halle, understood the importance of statistical research for policy. He later ascribed great importance to the compilation of data from private German banks, given that existing calculations were insufficient in delineating detailed information on the assets held by commercial banks, such as the quality of bills or maturity date.⁴⁰ Because of this new work, Schacht described Nordhoff as ‘the conscience of the Reichsbank’ (das Gewissen der Reichsbank) for his technical knowledge and ‘careful thoughtfulness’.⁴¹

Figure 1. The Reichsbank's Economics and Statistics Department (1928). These figures do not include the number of administrators or secretaries employed; Hermann Schneebeli, ‘Die Statistische Abteilung der Reichsbank’, Studienbericht, Zurich, 1928, 2; ‘Aufbau und Aufgaben der Statistischen Abteilung der Reichsbank’, Feb. 1931, BArch R2501/6346.

The research department of the Banca d'Italia had a similar technical function. Following the reorganisation of the old Economic and Financial Studies Office (Ufficio Studi Economici e Finanziari) by Governor Bonaldo Stringher in 1926, the new Department of Economic Studies and Statistics (Servizio studi economici e statistica) expanded its operations by aggregating data from both national and international sources, while also publishing articles in academic journals, such as the Giornale degli economisti.⁴² Its director, Giovanni Santoponte, was tasked with the responsibility of ‘present[ing] data, overviews, information on the current situation or on special questions’.⁴³ At the same time, the Servizio was able to gain greater standing at the central bank by advising on matters that helped to support existing goals.⁴⁴ Its reorganisation occurred the same year that the central bank obtained a monopoly on note issue.⁴⁵ Subsequent reports on the benefits of the gold standard conferred legitimacy to the decisions made by the Italian central bank, as well as to the government as a whole.⁴⁶

Meanwhile, the Banque de France's Service evolved from a largely technical department to one focused, more broadly, on foreign relations under the leadership of Jules Décamps. Having joined the central bank in 1910 and becoming head of the Service in 1920, Décamps represented the Bank at several postwar conferences, where he emphasised the importance of sound fiscal policy and cooperation for supporting France's return to gold.⁴⁷ Through publications and lectures, he defended monetary orthodoxy and frequently advocated a policy of balanced budgets.⁴⁸ He was further interested in achieving political stability in Europe by economic means, namely through the correction of wartime capital movements, which he believed had brought financial troubles to Europe. In 1918, for instance, he wrote that the US Treasury had imported gold at a rate that exacerbated economic instability.⁴⁹ The structural dislocations of capital during the war – namely an estimated 800 billion francs spent by the belligerents – left France with considerable debts owed to the United States and the United Kingdom.⁵⁰

After Décamps’ death in 1926, the Banque de France expanded its research department under director Pierre Quesnay, who had been a student of Deputy Governor Charles Rist at the Sorbonne.⁵¹ Similar to Décamps, Quesnay took part in numerous internationalist projects, such as those at the League and on the Allied Reparations Commission. At the Banque de France, he oversaw the reorganisation of the research department by specialisation: the Division of Monetary and Economic Studies was the main research group, the Division of Foreign Relations facilitated communication with other central banks and studied monetary conditions overseas and the Division of Information and Translation was responsible for collecting information from foreign publications and newspapers (Figure 2).

Figure 2. The Banque de France's Department of Economic Studies (1928). Source: Quesnay, ‘Composition, organisation et fonctionnement du Service’.

Due to their line of work, research departments depended on a particular type of economic expertise, one centred around hiring ‘practical’ specialists who helped ‘in the formation of Central Banking opinion’.⁵² To be sure, Norman believed that statistics offered value in testing his hypotheses, but they could not have been used for making policy decisions alone.⁵³ Nevertheless, he recognised that proprietary statistics offered a multitude of benefits for the central bank: to justify existing policies, to retain autonomy over the national monetary system and to increase prestige by hiring those with technical skills. Strong from the Federal Reserve noticed how there had been a general movement to hire ‘practical-minded central banking men’ who ‘had both an acquaintance with the theoretical side of monetary matters and with the practical aspects of central banking requirements’.⁵⁴ Indeed, technical expertise and university-level training in economics or statistics were seen as mutually inclusive. According to one statistician at the Bank of England, ‘a trained Economist could be consulted. . . . He should be fully “qualified” as far as Degrees, &c., indicate; must not be a crank, and must have the gift of applying economics to practical affairs’.⁵⁵

Conducting original research conferred prestige in itself, but also offered other strategic benefits. For the Bank of England, the City of London's status as an international financial centre was a pressing concern for central bankers and commercial bankers alike.⁵⁶ The Bank claimed that the goal of its research department ‘has always been to study the London Market’ in an effort to be the leading source of information on the domestic economy.⁵⁷ Similarly, the Banque de France invested resources in assessing the financial positions of foreign countries to achieve its own goals. Indirectly, its research department intended to address the highly politicised question of inter-Allied debts and German reparations through publications that assessed ‘the situation of foreign central banks and of foreign markets from a French point of view’.⁵⁸ A favourable assessment of Germany's economic recovery based on the reports of economic experts, for instance, might have proven the country's capabilities in being able to make its reparation payments. Because statistical work aligned with wider geopolitical goals, central banks planned on supporting the development of in-house research.

A Cooperative Environment

The rise of in-house expertise was as much a product of direct decisions made by central bankers as it was an outcome of European geopolitics in the interwar years. Throughout the 1920s, the League was at the forefront of new projects to refashion the political and economic landscape of Europe.⁵⁹ In contrast to its prewar variants, the ‘new internationalism’ of interwar Europe, according to Glenda Sluga, worked in tandem with the rhetoric of national sovereignty and self-determination.⁶⁰ This new form of liberal internationalism largely helped to support the movement towards in-house expertise and statistical standardisation among the central banks of Europe.

Initially, divergent views on economic policy made the possibility of a conference of central bankers seem unlikely. Strong noted the challenges of finding common ground given the different languages, political ideologies and domestic goals of each institution.⁶¹ Yet the Bank of England imagined that a conference might have provided the direct means of finding common ground, notably over the project of European reconstruction. Central banks, alongside their respective national governments, had affirmed the orthodox doctrines of balanced budgets and the gold-exchange standard at the Brussels Conference of 1920 and the Genoa Conference of 1922.⁶² After the latter, Sir Charles Addis proposed that the Bank hold another meeting to address ‘the question of Central Banks urging upon their respective Governments the summoning of an International Monetary Conference to consider the Economic Reconstruction of the World’.⁶³ Although the Bank's directors briefly discussed which central banks to invite, they ultimately decided to postpone a meeting until there had been a more favourable agreement with the United States over the repayment of wartime debts.⁶⁴

Yet interwar internationalism rested as much on agreement over economic disputes as an alignment of political motives for cooperation. Even with the stabilisation of the Reichsmark in 1923 – followed by the renegotiation of German reparation payments and an international loan under the terms of the Dawes Plan – the future stability of European politics remained unknown.⁶⁵ Planning a conference of central banks, Norman reported to Strong in April 1925, still appeared unlikely to contemporaries.⁶⁶ The Bank of England certainly wanted to host an ‘informal and private Meeting of representatives of the Central Banks of those Countries’ that had re-established the gold standard.⁶⁷ While such a meeting may have been possible given the convergence in orthodox policies among the United States, United Kingdom and Germany, uncertainty followed the 1924 election of the Cartel des gauches in France and general civil unrest in Italy. Because of these ‘political stumbling-blocks’, Norman again contended that ‘a formal Central Bank Conference, even if it could now be arranged, would probably prove unsuccessful’.⁶⁸

Yet central bank cooperation rested on agreements in both the geopolitical and the economic spheres. The signing of the Locarno Treaties in December 1925 provided a temporary resolution to the territorial disputes in (Western) Europe following the Franco–Belgian occupation of the Ruhr.⁶⁹ To its onlookers, the ‘spirit of Locarno’ created a temporary peace, particularly after Germany joined the League in September 1926.⁷⁰ It signalled the potential for future collaboration between France and Germany, as well as for European security when foreign ministers Aristide Briand and Gustav Stresemann were jointly awarded the Nobel Peace Prize for their work on the treaties.⁷¹ In the financial realm too, central banks led reconstruction projects in various European countries through international loans and currency-stabilisation programmes.⁷² This work largely stemmed from the high degree of coordination among financial experts at the League and the central banks of Europe.⁷³

In October 1926, Quesnay from the Banque de France travelled to London to explore the possibility of fostering new channels for communication between the two central banks.⁷⁴ Both Norman and Quesnay now believed there to be a greater chance of cooperation in the favourable political atmosphere in Europe.⁷⁵ The de facto stabilisation of the devalued French franc (franc Poincaré) in December, while viewed as a temporary expedient, necessitated communication among central banks for managing the movement of gold.⁷⁶ During the negotiations over a French loan, Strong noted that there had been a remarkable degree of consensus reached in his meetings with private bankers, the Bank of England and the Banque de France.⁷⁷ Officials began to believe that cooperation was the foundation for proper central bank policy.

The stabilisation of the Belgian franc further confirmed such views. In the months prior to the reform, the Bank of England's research department took part in negotiations with the Belgian central bank. After Osborne compiled a report on Belgian public finance that emphasised the need for fiscal reform, Norman intended to meet with Janssen and Governor Fernand Hautain to discuss the practicalities of a loan to balance the state's budget.⁷⁸ Coupled with an ongoing exchange-rate crisis, the inability to negotiate a foreign loan culminated in the resignation of the Poullet cabinet in May 1926.⁷⁹ The new Belgian government entrusted Émile Francqui, a banker and minister without portfolio, to oversee the stabilisation of the franc in October.⁸⁰ Officials, in spite of their disagreements over the specific terms of the plan, were able to negotiate an international loan, leading Norman to refer to the entire project as nothing short of a ‘miracle’.⁸¹

Although central banks adhered to new stabilisation programmes and the fiscal orthodoxy of the gold-exchange standard, they sought to exert influence through other channels. Expanding statistical work through in-house research proved to be one such avenue. Adopting the British, French and German research models, central banks across Europe established new statistical departments: Czechoslovakia (1926), Denmark (1928), Greece (1928), Italy (1926), Lithuania (1928), Norway (1924), Poland (1924), Portugal (1932), Romania (1927) and Spain (1931).⁸² They employed technical experts, who were responsible for aggregating data from other institutions, advising directors and publishing regular reports on national economic conditions. As Béla Imrédy, a director at the National Bank of Hungary (and later Minister of Finance and Prime Minister), stated:

While the statistical studies of Central Banks should have a basis in theoretical economics they must not be allowed to become doctrinaire and in the pursuit of theory to neglect the observation of economic phenomena which may perhaps appear of little importance from the theoretical point of view. The policy of Central Banks is a practical matter having the widest possible general interest. . . . Statistical departments must then have as their characteristics a practical outlook, quick methods and a power of observing economic phenomena as a whole without theoretical prejudices. . . . Statistics are of use only when they give life to the figures studied.⁸³

Cohorts of new staff members were employed to support the compilation of monetary statistics: Austria (7), Bulgaria (22), Czechoslovakia (11), Finland (6), Hungary (6), Norway (4), Poland (9), Romania (10) and Sweden (4).⁸⁴ Others, such as the Banque de France and the Reichsbank, made concerted efforts to hire statisticians with foreign-language capabilities and dedicated a whole division to translation.⁸⁵

Officials also seemed to agree that one of the prerequisites for financial stability was the coordination of statistics. If other central banks were willing to provide more transparency over their purchases of gold, for instance, greater stability could have been secured for the maintenance of the gold-exchange standard. Along these lines, the Bank of England was again willing to support ‘a conference of Central Bankers’, as had been recommended at Genoa.⁸⁶ Harry Siepmann, one of the Bank's experts, argued that the meeting needed to have been solely statistically focused, with no direct discussion of monetary policies.⁸⁷ He considered inviting the leading economist or statistician of each central bank to address the possibility of standardising data: Quesnay (France), Imrédy (Hungary), Antonín Basch (Czechoslovakia) and Alf Eriksen (Norway).

The Conference of Central Bank Statisticians

It was with the acceptance of economic experts and the geopolitical conditions of interwar Europe that the prospect of a conference materialised. As Emanuel Goldenweiser from the Federal Reserve remarked, such a meeting would not have been possible even a few years prior.⁸⁸ Relatively little has been written on the 1928 conference, even amidst the plethora of research on interwar central banking.⁸⁹ Michel Margairaz has contextualised the conference's proceedings in relation to the broader efforts of statistical collaboration facilitated by the League.⁹⁰ Yet fundamental to understanding the conference is not only what it aimed to achieve, but also how its proceedings reflected the emerging dependence on expertise.⁹¹

Indeed, many projects to address the issue of statistical standardisation had already been underway.⁹² Government statisticians, including Alfred Flux in the United Kingdom and Ernst Wagemann in Germany, sought to improve the usability of ‘economic barometers’ for shaping government policy.⁹³ The League had also undertaken its own work to support uniformity in economic statistics, such as through the Industry Committee of the International Economic Conference (1927).⁹⁴ At the International Conference on Economic Statistics (1928), delegates again discussed a vast array of statistical work, including research on industrial output, tariff nomenclature and international trade.⁹⁵ With the League’s support, these projects aimed to standardise statistical terminology and, by extension, centralise information in a systematic way. For instance, efforts since March 1927 to compile and translate the financial legislation of different countries later culminated in the publication of the Monetary and Central Bank Laws (1932).⁹⁶

While the benefits of standardisation had long been discussed in these fora, the League endeavoured to work on an arrangement ‘dealing directly with the question of the comparability of monetary statistics’.⁹⁷ In August 1927, the League's Financial Committee concluded that there was ‘a considerable number of persons directly interested in monetary statistics’ that necessitated ‘the convening of a quite informal meeting of intelligence and statistical officers of central banks’.⁹⁸ It was at this recommendation that Alexander Loveday, a British economist working in the League's Secretariat, agreed to arrange such a conference.⁹⁹ Loveday believed there was a need to establish uniformity among the disparate modes of economic statistics gathered in each country, particularly with regard to the varying statistical terminologies employed.¹⁰⁰ He was also keen to highlight the benefits that central banks would have gained from the conference. ‘A meeting of this type’, he told the Banque de France governor, Émile Moreau, ‘would constitute not only the most effective means of achieving the particular goal of the Financial Committee itself, but could still be . . . of interest to the central banks themselves.’¹⁰¹ To another official, Loveday explained that ‘although . . . the League will benefit from such exchange of information it is primarily a matter between the central banks themselves.’¹⁰²

Central bank governors, in turn, sent their head statisticians to represent their institutions abroad.¹⁰³ The Banque de France was willing to host the representatives for five days at its main building in Paris.¹⁰⁴ During the first few days, the press reported that the conference was ‘without precedent’ in bringing together representatives from twenty-four central banks for matters of an ‘entirely practical’ nature.¹⁰⁵ The Economist wrote that the meeting centred around ‘close international co-operation in the field of intelligence and statistics, including that obscure but important question of the movement of capital from market to market’.¹⁰⁶ Continuing the cooperative spirit created by the stabilisation of European currencies and the settlement of the border question through the Locarno Treaties, the conveners of the conference were now able to assemble statisticians from each central bank: Goldenweiser, who had, by then, succeeded Stewart as the Director of Research and Statistics at the Fed; Osborne, now a principal at the Bank of England; and Nordhoff and Quesnay from the German and French central banks, respectively. Finally, the League's representatives included Loveday, Jacques Rueff and Ansgar Rosenborg (Figure 3).

Figure 3. The Conference of Central Bank Statisticians (1928). The delegates shown are: Moreau (1), Rist (2), Santoponte (3), Pratsicas (4), Goldenweiser (5), Rosenborg (6), Nordhoff (7), Loveday (8), Quesnay (9), Rossi (10), Vincent (11), Karpinski (12), Szawlezki (13), Basch (14), Tchakaloff (15), Rueff (16), Radesco (17) and Stanesco (18). Unpictured or unlabelled delegates include: Allport, Burgess, Cohn, Eriksen, Imrédy, Kerschagl, Osborne, Poom, Protitch, Ricard, Rimka, Schmidt, Schneebeli, Stalbovs, Stanesco, Tudeer, Trygger, Van Buttingha Wichers and Yovanovitch; ‘Hier matin s'est ouverte la conférence des banques d’émission’, Excelsior, 19, 6331 (1928), 1.

During the conference, the delegates discussed the possibility of establishing greater uniformity in their statistical reports.¹⁰⁷ They began by providing an overview of the current state of monetary research at each central bank, along with details of what they hoped to do in the future. One of the topics was the inadequacy of existing economic data. Rather than relying on the League's Monthly Bulletin, which some criticised, central banks wanted to establish their authority over the collection of national statistics.¹⁰⁸ In light of the inadequate state of existing reports, Nordhoff noted the importance of research departments: they were not purely for collecting statistics as an end unto themselves; instead, they could have been used to influence policies on both theoretical and practical questions of money, banking and credit.¹⁰⁹ The degree of consensus over the prevailing vision for reorganising economic statistics was striking. For the Bank of Finland, research captured ‘the economic life of the country’, as their reports sought to provide an all-encompassing view of the industrial, financial and agricultural changes in each European country.¹¹⁰ The delegates believed that statistics had an illuminating character. Imrédy emphasised how research provided a ‘practical outlook, quick methods, and a power of observing economic phenomena as a whole without theoretical prejudices’.¹¹¹

Moreover, the economists understood the major challenges to the lofty objective of standardisation across national boundaries. Each central bank compiled weekly statements in a different manner such that establishing uniformity seemed nearly impossible owing to the differences in national accounting practices.¹¹² Research departments often confronted the problem of unreliable sources since many of them did not compile their own information, but instead relied on government reports and newspapers, both foreign and domestic.¹¹³ Along with monetary statistics, a few economists proposed improving statistical work in other areas as well. Einar Cohn, an economist at the National Bank of Denmark, feared a further decline in agricultural prices would have led to an unfavourable trade balance that, in turn, forced the central bank to suffer reserve losses.¹¹⁴ Statistics on agriculture, however, were in short supply. Due to the lack of standardisation, the National Bank of Bulgaria took a leading role in centralising data from the various municipalities of the country.¹¹⁵

As a result, a key part of the discussion revolved around whether a centralised ‘information bureau’ was necessary or desirable.¹¹⁶ Such was the suggestion of Hermann Schneebeli from the Swiss National Bank, who thought it might have been worthwhile to compile the banking and monetary laws of each individual country for comparison purposes.¹¹⁷ Although Schneebeli seemed unaware of the concurrent efforts by the League to initiate such a project, Loveday offered the support of the Economic and Financial Organisation for ‘the centralisation and publication of information’ among central banks.¹¹⁸ These and other proposals, however, were less well received by the larger central banks. Osborne from the Bank of England noted that most monetary statistics were intended for internal use only.¹¹⁹ Others feared the loss of autonomy or of competitive edge once a document was published. Quesnay, who chaired the proceedings, suggested that central banks should not have been allowed to publish information from other central banks.¹²⁰

Transparency over statistical data made sense from a practical standpoint when it benefited the institution’s existing goals. Many officials believed that publishing original research could have been a means of garnering prestige, a view that was particularly important for the newly established central banks. Jean-Jacques Vincent from the National Bank of Belgium boasted how ‘the heads of the department possess very serious university degrees, notably doctorates in law, in political science and diplomacy, in physical sciences and mathematics, etc. . . . in addition to knowing foreign languages.’¹²¹ As in-house experts were able to interpret both quantitative and qualitative information, they aimed to be the definitive source of financial information, thereby legitimating their own usefulness. According to Richard Kerschagl, an economist from the Austrian National Bank, central banks filled the gap between the theoretical assertions made by academics and political decisions by policy makers.¹²²

More concretely, central banks were able to elevate their own prestige based on the publications and the number of visitors whom a central bank was willing to host. For the Bank of England, the publication of accurate and comprehensive reports was a mark of status. As one economist, Henry Clay, later remarked, the Bank's Statistical Summary was renowned for its accuracy and presentation of material: ‘the work of the Department seems to me as good of its kind as anything that is being done by any Government Department or University; the Summary of Statistics is already recognised as the best publication of its transfer, which is as it should be.’¹²³ Clay imagined the Bank's Statistical Section adopting an even more prominent role abroad by providing and sharing economic information with other central banks.¹²⁴ Sending and receiving economic advisers to and from other central banks had helped to create a transatlantic network of economic advisers. While in 1928 the Bank had hosted several foreign visitors, including two from the Federal Reserve, two from the Reichsbank and one from the Hungarian National Bank, by the 1930s it welcomed additional economists from the United States, continental Europe, India and Japan.¹²⁵

Despite moments of cooperation, the proceedings of the 1928 conference were dominated by the priorities of the larger central banks. Not entirely surprising was the fact that the conference's working languages were the interwar linguae francae of English and French, both official languages of the League.¹²⁶ Proposals to standardise statistical terminology rested on the assumption that central banks would have considered publishing their reports in a single language. The Bank of Norway published many of its reports in English, which were then shared with ‘the Bank's foreign correspondents and central banks . . . [as well as] other corporations and persons abroad’.¹²⁷ Other delegates, however, were unable to find agreement on the matter. ‘Some common language should be chosen’, posited the Swiss delegate, ‘such as English, French or German’, while Quesnay thought it best to ‘leave [this matter] out of the discussion’.¹²⁸

Alongside the question of foreign languages were the differences in technical terminology. The delegates from the central banks of Austria, Czechoslovakia and Hungary used price levels as an economic barometer of their country's position, but admitted they were not satisfied with the non-standardised processes for compiling this information.¹²⁹ Prices were even more difficult to compare with countries that used different baskets of goods to measure change over time. An economist from the Federal Reserve, W. Randolph Burgess, recognised that this problem, in part, stemmed from the multiple indices used for various purposes: some allowed for seasonal variations, others focused solely on the activity of domestic businesses and all frequently changed the ‘base year’, thereby making comparisons more difficult.¹³⁰ In response, Vincent argued that all indices should have been updated after the war since 1913, a year frequently used as a point of comparison, was no longer suitable for practical purposes.¹³¹

Unresolved issues notwithstanding, the discussions over statistical standardisation as a whole exemplified a convergence in views on the importance of in-house expertise. At a dinner on 14 April, Governor Moreau gave a toast to the delegates and ‘this spirit of collaboration’:

we are delighted to see the manifestation [of this spirit], so numerous and so effective, in your Conference. We have experienced what happy effects can result from it, and we think we are the interpreter of all desires by forming the wish that it will continue to develop between Central Banks in the years to come.¹³²

The Conference of Central Bank Statisticians appeared to be one of mutual benefit, as reported by business journals. According to the Journal of Commerce, central banks had laid the foundation for ‘a more adequate system of gathering statistics in each important European country for use in guiding their monetary and credit policies’.¹³³ Meanwhile, a journalist writing in the Berliner Tageblatt und Handels-Zeitung anticipated that the recognition of ‘the commonality of the problems at hand’ would have soon led to ‘closer and permanent relations between the central banks’ in the realm of monetary policy.¹³⁴

After the meeting, economists and statisticians agreed to share information with one another on the import and export of gold reserves, a concern mentioned in the collective report.¹³⁵ They wanted to calculate and manage the flow of gold that stabilised the purchasing power of each of their currencies. These talks echoed what had already been discussed at the 1927 meeting of central bankers on Long Island, which aimed to promote financial stability by controlling the volatility of gold flows to New York.¹³⁶ Yet the conference differed from prior attempts at reform because of its specific promotion of statistics-based cooperation. Each institution took a great deal of interest in the findings of one another's reports, forwarding copies to one another and looking for ways of improving their own data sets.¹³⁷ As the Bank of England reflected, ‘what has taken place has been useful, and the approach of one Bank to another on points of interest to economists has been made infinitely easier.’¹³⁸

At the same time, the conference's shortcomings were also apparent. On the first day of the proceedings, an Austrian newspaper acknowledged the fact that ‘the conference itself constitutes only a non-binding discussion’.¹³⁹ Indeed, the collective report presented only a general outline of the discussion points and main findings. Even before the conference, maintaining relations between central banks seemed fraught with challenges. The delegates had been unable to establish closer ties with the Soviet Union's new state bank (the Gosbank), created in 1922 to support Lenin's New Economic Policy.¹⁴⁰ While its president, Aron Sheinman, initially accepted an invitation to the conference, he later cancelled due to suspicions of unfair gold restrictions conducted by France.¹⁴¹ Between 1915 and 1917, the Banque de France had maintained 52 million francs in gold at the State Bank of the Russian Empire.¹⁴² Although the Banque had continually affirmed its ownership of this gold, the Bolshevik seizure of power had led to its confiscation in 1918.¹⁴³ Seeing an opportunity in March 1928, the French government laid claim to $5.2 million worth of Soviet gold held by private commercial banks as collateral for orders from US firms.¹⁴⁴ The Russian central bank denounced the action and asserted that it was responsible neither for the assets nor for the liabilities of the old central bank, leading to its decision not to attend the conference.¹⁴⁵

Experts later undertook efforts to promote statistical collaboration through the Bank for International Settlements (BIS).¹⁴⁶ Its Monetary and Economic Department (formerly the Central Banking Department) helped to share information among central banks as a means of improving the prospect of collaboration.¹⁴⁷ Quesnay noted the success of the BIS in restoring monetary confidence by giving the various banks access to lines of credit during the 1931 crisis.¹⁴⁸ In a similar endeavour, Sprague from the Bank of England hoped to calculate the volume of short-term capital flows as part of a new study. At the General Assembly in May 1932, he gave a presentation on his plans to estimate the movement of foreign loans between New York and London, even suggesting a credit bureau could have helped to collect the information from individual companies.¹⁴⁹ There were, however, great difficulties in institutionalising new avenues for statistical research. Sprague stated, for instance, that ‘it would be misleading to attach importance to the net balance’ of capital movements in his estimates.¹⁵⁰ Even Quesnay regarded his project as overly ambitious: ‘[Sprague] thinks that measures taken according to the principles indicated by him can, to a certain extent, represent a guarantee against imprudent investments and thus constitute an indirect insurance against the serious depressions that we have recently known.’¹⁵¹

Yet it would perhaps be more insightful to understand the conference based on the criteria of what its delegates envisioned rather than through the dichotomy of success or failure. Experts imagined a way of fostering ties between central banks through a shared vision of standardised monetary statistics. They looked to statistics as a basis for future economic relations and as a tool compatible with their own national goals. As the vice-president of the Bank of Poland described, ‘a more regular co-operation of brains should precede the official co-operation of central banks . . . and better co-ordination of the Gold Bullion Standard’.¹⁵² Technical experts particularly helped central banks establish authority over the collection of national data, while simultaneously improving relations with others through exchange programmes. Although the 1930s may be characterised as an era of diminished prestige and influence of central banks, the 1920s offered a potential moment of enhancing credibility through in-house research. With the common language of statistics, central banks thus accepted the view that expertise conferred institutional benefits.

Conclusion

The proliferation of new statistical tools rested on a confluence of factors specific to interwar Europe: the attempts by the League to promote cooperation, the ambitions of central banks to retain national autonomy and bolster legitimacy, the problems of economic instability wrought by the First World War and the prevalence of experts in the early decades of the twentieth century. As a result, the 1928 Conference of Central Bank Statisticians stood as a testament to the possibility of standardisation across national and linguistic boundaries. It embodied a particular vision of liberal internationalism among central banks, one predicated on the standardisation of data. As central banks sought new avenues for filling an apparent gap in the collection of monetary statistics, they were also able to conceptualise a new basis for collaboration that relied on neither the views of politicians nor the direct control of international institutions. Rather, they believed that, by focusing solely on the technical aspects of monetary policy, one could have divorced economic concerns from political ones. This view, however problematic, was a notable feature of expert-based knowledge.

The rise of expertise in international governance further depended on the political conditions of interwar Europe. Both economic and geopolitical agreements, from the gold-exchange standard to the Locarno Treaties, compelled central banks to identify new avenues for cooperation over statistical matters. Among the recently established (or reformed) central banks of Central and Eastern Europe, hiring experts who published statistical reports and attended international conferences helped to bolster credibility. Meanwhile, the Bank of England and Banque de France observed the strategic benefits of expertise to advance their own geopolitical goals, notably the reestablishment of London or Paris as international financial centres. As central bank officials sought to address their own needs, made possible by the hiring of economists and statisticians, they became increasingly interested in the standardisation of statistics, both for its quantitative insights and for its institutional prestige. The recruitment of specialists, analogous in many ways to the concurrent practices of the League or BIS, had the added benefit of being internally sourced: central bankers, often reluctant to divulge their proprietary data, were able to retain a degree of control over what information they shared.

The broader implication of this article is to show how the developments of the interwar period provide insights into understanding the dominance of expertise in postwar organisations, such as the IMF or World Bank.¹⁵³ New statistical tools and models created in the interwar years have become entrenched in a wide range of institutions, from government agencies to think tanks.¹⁵⁴ Through the use of quantitative data, central banks have increasingly seen the dominance of economic analysis in the upper echelons of their organisations. The rise of in-house expertise as a model of economic governance in interwar Europe thus set an important precedent for postwar variants of liberal internationalism.