Central banks' thinking at test: the unexpected birth and fate of the “Availability Doctrine” By Pierrick Clerc (Swiss National Bank) and Eric Monnet (Bank of France, PSE, CEPR) This draft. May 2019 Very Preliminary Bridging the gap between history of economics and history of economic thought is an important venture. It requires considering economic knowledge produced outside academia, in governments, international institutions and central banks. This leads us to explore what sociologists of science call the “boundaries of science” (Gieryn 1983, Cartwright 1999, Lamont and Molnar 2002) and to investigate the relationship between academic and non-academic production of economic discourses. Even in countries with a strong postwar academic community of economists like the US, policy institutions produced economic thinking which was integrated into policy discourses, public or parliamentary debates and, sometimes, then, in academia. Central banks hired some economists and researchers but they did not have to meet academic requirements. During the postwar period, we know at least three examples of economic doctrine shaped by economists in central banks and which eventually had a strong influence on national academic debates: the 1959 Radcliffe Report in England, the Économie d'endettement in France in the 1970s, and the “Availability Doctrine” in the United States in the 1940s-1950s. These doctrines were based on the work of central banks economists attempting to build their own theory so as to describe monetary policy implementation and formulate guidance for policymakers. They criticized mainstream monetary theory of their time for being too abstract and undocumented about the true process of money creation. They also blamed its incapacity to account for institutional differences in financial systems and monetary policy operations across countries. Most of these works were published in reports or documents rather than scientific journals. They however gained recognition in academia, by being reviewed and discussed by economists in scientific journals, textbooks and included in course syllabuses. As we will discuss in the last section of this paper, a similar story can be told today about the work that developed at the BIS in the early 2000s – and which became extremely influential after the 2008 financial crisis – on macroprudential policy and the risk taking channel of monetary policy 1 transmission (Borio & Zhu 2008). In the past and today, these approaches have in common that they criticize conventional monetary theory in academic circles because it is too disconnected from the implementation of monetary policy and it does not take into account the importance of the composition of financial institutions' balance sheets for both monetary policy transmission and financial stability. Our research focuses on the Availability Doctrine, which was notably developed by Robert Roosa at the New York Federal Reserve in the late 1940s. Roosa served as Acting Manager of the Research Department since 1950 and was then Vice President of the New York Fed, in the Research Department, when he left this institution for the Treasury in 1961. The Availability Doctrine gained wide recognition in academia in the early 1950s, particularly through hearings at the US Congress (on credit controls and monetary policy) where it was debated. The published Congress hearings were considered as such an important piece of economic discourse that they were reviewed by James Tobin in the Review of Economics and Statistics. The Availability Doctrine was a narrow theory about monetary policy implementation but – as recognized by Jaffee and Stiglitz (1990) – eventually gave birth to the now standard theory of credit rationing which emerged in the 1960 and 1970s (Acosta 2016). The distinctive feature of the Availability Doctrine is the change in emphasis, on the loanable funds market, from borrowers' to lenders' behavior. Its proponents argued that, due to changing institutional structures on this market and the large holding of public debt by banks after the war, lenders became strongly sensitive to variations in the yields on government securities. Therefore, contrary to a widespread belief at that time, monetary policy would have important effects on aggregate demand: even if the demand for funds were quite interest inelastic, monetary policy would nevertheless operate through the large response of the amount of funds made available by lenders to movements in yields on government securities. The main message of the Availability doctrine was that quantities (rationing of bank assets) were much more important than prices (interest rates) for the transmission of monetary policy. Hence the term “availability” denoted the emphasis on “quantities”. Such conclusion relied on a detailed analysis of the shift in the composition of the assets of banks after a monetary policy shock. This shift was driven by market imperfections which were missing from standard models. The “doctrine” went against standard theories which focused on the money base or interest rates. It reasoned with the main concerns of policymakers at that times, that is the interaction between monetary policy and public debt management. 2 The Availability Doctrine cannot be understood without the institutional and historical context of postwar monetary policy (which notably had to deal with a massive amount of government securities in bank balance sheets). This doctrine was not supposed to be a general and full fledge theory that could be applied in all situations. Instead, it was very context-specific and recognized as such. It was indeed a “doctrine”, whose main objective was to set rules for policy actions. In a 2002 speech, Paul Volcker remembers that when he started as a young economist at the Fed in the early 1950s, “the idea that was promoted by this particular institution at the time was something called the availability doctrine”. 1 Moreover, even though its theoretical foundations were weak, the Availability Doctrine played a key role in showing the limits of mainstream academic models and in formulating new areas for research. As such, the historical study of the Availability Doctrine we aim to provide here is not only interesting for historians of monetary policy in the US. It is also an important, but underestimated, milestone of postwar economic thought. It also shows the limits of the usual interpretation of postwar monetary thinking as a mere fight between monetarists and Keynesians. The Availability Doctrine, as other central bank’ doctrines, cannot be understood on the basis of these usual categories (Monnet 2018). Why was the Availability Doctrine developed at the New York Federal Reserve? We particularly stress two reasons. First, as acknowledged by Roosa, it was motivated by the perceived weaknesses of the leading monetary theories in the 1940s and 1950s as well as the evolution in the practice of central banking that occurred at that time in the US. Secondly, economists of the US central bank were looking at the experience of credit controls in European countries as a way to renew their thinking about the transmission of credit and monetary policy. This remains an hypothesis – as it was not explicit in Roosa’s writings – but the work of Roosa on French quantitative credit controls in the the late 1940s (at the time when the US Fed policies depended on open-market operations and reserve requirements only) might have led him to recognize the importance of the availability of credit for the transmission of monetary policy. Hence, international comparisons – which were an active task of the NY Fed research department in the 1940s – might have helped to design a theory especially adapted to the US context, but influenced by foreign experience with quantitative controls. Interestingly, Donald Hodgman - the University of Illinois Professor who made the connection between the Availability Doctrine and the modern theory of credit rationing in a 1960 article in the Quarterly Journal of Economics (Jaffee & Stiglitz 1990, Acosta 2016) – later became the most prominent US specialist of credit controls in Western Europe during the 1960s and 1970s (Hodgman 1971, 1972, 1973, 1974). 1 https://www.newyorkfed.org/research/economists/medialibrary/media/research/epr/02v08n1/0205volck.pdf 3 After 1945, economist became a profession in its own right, but through different institutions across countries (Fourcade 2002). Hence, institutions such as central banks played a key role of intermediary between economists and policymakers. In the US, the Federal Reserve of New York played such a key role as early as the 1940s. Some academic journals, like the Journal of Finance, devoted a large part of their publications to the description of monetary policy implementation and engaging debates on the means of intervention of the Federal Reserve in the money market. In times of major negotiated reforms of the international monetary system (Bretton Woods), central banks were also immediately exposed to the influence – if not of foreign economic thinking – of the diversity of monetary experiences and institutions in the Western world. What were the relationships between the Availability Doctrine and academic research? We first discuss how this doctrine became legitimate among academic circles in the 1950s. We believe it was the case since the Availability Doctrine was elaborated within an important central bank, and that the NY Fed and Roosa had strong academic connections. Then, we detail the successive steps which